ASSESSING QUALITY OF LIFE AND LIVING CONDITIONS TO GUIDE NATIONAL POLICY
Social Indicators Research Series Volume 11 General Editor: ALEX C. MICHALOS University of Northern British Columbia, Prince George, Canada Editors: ED DIENER University of Illinois, Champaign, U.S.A. WOLFGANG GLATZER J. W. Goethe University, Frankfurt am Main, Germany TORBJORN MOUM University of Oslo, Norway JOACHIM VOGEL Central Bureau of Statistics, Stockholm, Sweden RUUT VEENHOVEN Erasmus University, Rotterdam, The Netherlands.
This new series aims to provide a public forum for single treatises and collections of papers on social indicators research that are too long to be published in our journal Social Indicators Research. Like the journal, the book series deals with statistical assessments of the quality of life from a broad perspective. It welcomes the research on a wide variety of substantive areas, including health, crime, housing, education, family life, leisure activities, transportation, mobility, economics, work, religion and environmental issues. These areas of research will focus on the impact of key issues such as health on the overall quality of life and vice versa. An international review board, consisting of Ruut Veenhoven, Joachim Vogel, Ed Diener, Torbjorn Moum and Wolfgang Glatzer, will ensure the high quality of the series as a whole.
The titles published in this series are listed at the end of this volume.
ASSESSING QUALITY OF LIFE AND LIVING CONDITIONS TO GUIDE NATIONAL POLICY The State of the Art
Edited by MICHAEL R. HAGERTY University of California, Davis CA
JOACHIM VOGEL Statistics Sweden, Welfare Analysis Program, and University of Umeå, Sweden
and
VALERIE MØLLER Rhodes University, South Africa
Reprinted from Social Indicators Research, Volume 58, Nos. 1–3 (June 2002)
KLUWER ACADEMIC PUBLISHERS NEW YORK, BOSTON, DORDRECHT, LONDON, MOSCOW
eBook ISBN: Print ISBN:
0-306-47513-8 1-4020-0727-2
©2002 Kluwer Academic Publishers New York, Boston, Dordrecht, London, Moscow Print ©2002 Kluwer Academic Publishers Dordrecht All rights reserved No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher Created in the United States of America Visit Kluwer Online at: and Kluwer's eBookstore at:
http://kluweronline.com http://ebooks.kluweronline.com
CONTENTS
Introduction Michael Hagerty, Joachim Vogel, Valerie Møller PART I. CONCEPTS
AND
1
THEORY
Conceptualizing and Measuring Quality of Life for National Policy Sten Johansson
13
Why Social Policy Needs Subjective Indicators Ruut Veenhoven
33
Towards a European System of Social Indicators: Theoretical Framework and System Architecture Heinz-Herbert Noll
47
PART II. C URRENT SOCIAL INDICATOR AND SOCIAL REPORTING PROGRAMS: NATIONAL AND COMPARATIVE EXPERIENCE Strategies and Traditions in Swedish Social Reporting: A 30-Year Experience Joachim Vogel
89
Quality of Life and Living Conditions in the Netherlands Jeroen Boelhouwer
115
“Quality of Life” Research at the Israel Central Bureau of Statistics: Social Indicators and Social Surveys Charles S. Kamen
141
Quality of life in a European Perspective: The EUROMODULE as a New Instrument for Comparative Welfare Research Jan Delhey, Petra Böhnke, Roland Habich, Wolfgang Zapf
163
The NORBALT Project: Comparative Studies of Living Conditions in the three Baltic Countries Aadne Aasland, Guri Tyldum
177
Living Conditions in the Arctic Thomas Andersen, Birger Poppel
191
Cross-national Comparison of the Quality of Life in Europe: Inventory of Surveys and Methods Michaela Hudler, Rudolf Richter PART III. APPLYING SOCIAL I NDICATORS EFFECT S OCIAL C HANGE
217
TO
TELESIS: The Uses of Indicators to Set Goals and Develop Programs to Change Conditions Abbott L. Ferriss
229
The Role of Quality of Life Surveys in Managing Change in Democratic Transitions: The South African Case Valerie Møller, Helga Dickow
267
Monitoring the Impact of Land Reform on Quality of Life: A South African Case Study Julian May, Thildé Stevens, Annareth Stols
293
Toward a Social Development Index for Hong Kong: The Process of Community Engagement Richard J. Estes
313
International Comparisons of Trends in Economic Well-Being Lars Osberg, Andrew Sharpe
349
Declining Quality of Life Costs Governments Elections: Review of 13 OECD Countries Michael R. Hagerty
383
Considering Social Cohesion in Quality of Life Assessments: Concept and Measurement Regina Berger-Schmitt
403
Do Income Surveys Overestimate Poverty in Western Europe? Evidence from a Comparison with Institutional Frameworks Christina Behrendt
429
Index
441
INTRODUCTION
ASSESSING NATIONAL QUALITY OF LIFE AND LIVING CONDITIONS: THE STATE OF THE ART
In February, 2000, a call for papers went out to address “crossroads” issues in “Assessing National Quality of Life and Living Conditions”. Many of the preliminary papers were also presented at a special track at the annual conference of the International Society for Quality of Life Studies in Girona, Spain, 19–22 July 2000, and received further discussion by prominent researchers active in social indicators research. The papers addressed theory, current measurements and social reporting, as well as applied studies in various directions. Speakers were specially invited to address the current crossroads of the social indicator movement. Papers then underwent the review and selection process. The final contributions published here display an impressive diversity of social indicator work, from large scale survey and reporting systems providing systematic information for social planning, and large comparative projects, to targeted research focused on strategic issues. This volume presents these contributions along the three headings concepts and theory, current information systems, and applying social indicators to effect social change. Altogether they represent a good survey of current focus and practise. Social indicators, general social surveys and social reporting have a history of some 40 years, in the sense of a common movement united around a few simple principles: expanding the focus of social monitoring beyond traditional economic indicators, looking at output in terms of individual living conditions, and providing systematic facts on core issues of political debate and social planning. This was indeed a new movement; it attracted researchers as they recognised old dreams of social research; it responded to the administration’s needs for informed social monitoring, and it gained support by citizen’s movements waiting for ‘scientific’ support for Social Indicators Research 58: 1–11, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
2
INTRODUCTION
their cause. In effect, this kind of research interest was brought forward in all periods of qualitative jumps, such as the New Deal, the Great Society, or, in Sweden, the People’s Home. Nevertheless, as with earlier attempts in this direction, the social indicator movement was not to become an immediate success story. The first decades showed a rapid diversification of approaches, as well as considerable informal coordination of methods and analyses, but only a few institutionalised systems, many discontinued programs, and fragmentation into competing sub schools. About a decade ago the Journal of Public Policy invited a number of active researchers to give their assessment on the movement’s recent history.1 The initial enthusiasm and evolution, followed by decline and even failure of the movement to establish itself, was a general theme in these assessments. However, in retrospect, we should note that the research community went ahead with its institutionalisation towards a self-recognised movement, including attributes such as a journal, Social Indicators Research, newsletter, international working parties, and regular conferences. Public recognition and political influence, however, was less impressive. Early Attempts
The institutionalisation of social indicators was slow and eventually discontinued in many countries. Social reporting tended to stop at statistical compendia, avoiding the next step towards analysis and policy evaluation. It turned out that the strength of the movement in providing scientific facts to the political arena also mobilised its counter forces. The major drawback of the first two decades was the closing of OECD’s social indicator program. OECD had by 1980 recruited a large international network of researchers and national statisticians, joining OECD in developing its program over several years. This group developed a comparative model comprehensive survey, as well as a social indicator list intended to be incorporated in national regular statistics, to provide systematic indicators for national as well as comparative use. The OECD program was terminated 20 years ago, in parallel with the U.S. regression during the Reagan administration. Denis Johnston and Frank Andrews underlined that the American demise was ideological, as it was
INTRODUCTION
3
for the OECD Social Indicator Program. Johnston’s reflection on the poor resources given to social indicator research in those days deviates much from the European, in particular the Scandinavian experience.2 This comparison points at the difficulties, and even failure, to penetrate the administration, social planning, and general public debate. Institutionalisation of social indicators is still unusual a decade later. Economic and Social Accounting
In the initial stage of development in the 1960s, the founders of the social indicator movement were greatly influenced by the development of the System of National Accounts (SNA). The unity, integration, coverage, routine data collection, and standardised compilation and publication of economic SNA statistics appeared as a prototype for the social dimension as well. Similar systems were discussed under headings such as social accounting and a system of social and demographic statistics (SSDS), linking elements of social statistics to produce detailed as well as aggregate statistics of individual and societal well-being. The successful GDP concept demanded a social counterpart to monitor the social consequences of modernisation and economic change. The writers in this latest assessment of the social indicator movement provided self-critique on the early optimistic ambitions of the social indicator movement to produce a ‘social’ counterpart to economic accounting. Most writers shared this disappointment, but also pointed at the complexity of social reality, the overly optimistic early belief in a measurable ‘system’. However, by the early 1990s, most developed countries had developed information systems to monitor general living conditions along the lines of the Social Indicator Tradition. In the absence of a unifying concept (such as money), and already established statistical systems, the answer was a more or less standardised system of social indicators, general social surveys (data collection systems), and routine social reporting (descriptive statistics; advanced analysis of living conditions). These are still the three major pillars of the social indicator movement, along with a broad spectrum of targeted research.
4
INTRODUCTION
Diversification
No doubt, the first decades saw expansive conceptual development, new methods and analytical perspectives. Today we have an impressive diversification of concepts, research methods, analyses, and types of reports; but still of an ad hoc nature, and the general audience is often unaware of the concepts and reports. Diversification is indeed the key word around the turn of the century. It is fair to speak of sub schools within the movement, the repertoire of data has expanded much, reporting has been diversified and standardised towards sector reports and systematic overviews of marginalized categories. Actors and Current Crossroads
But, as already mentioned, while most developed countries carried out data collection and reporting, institutionalisation was slow, programs were terminated, and social indicators never played a major role in public debate, as did traditional economic and financial indicators. Social indicators of everyday living conditions had less influence in setting the political agenda than economic macro indicators and indicators of market performance. What should have been fuel to public debate, politics and social planning, tended to stay within the research community and within social planning, not finding its way into basic statistics, public recognition and political debate. During the last forty years, the original focus on individual as well as societal welfare was realised in most developed countries by researchers who not always define themselves as belonging to this tradition, but were definitely inspired by the early writings. Today, we have two mainstream actors. One tradition is official and semi-official social reporting located at public institutions (national statistical institutes, special research institutes, and sometimes ministries) with a public mandate and usually regular long-term public funding. The other stream is the independent research tradition, with temporary funding, ad hoc and temporary studies, sometimes less comprehensive in nature, and more focused on special problems. The tradition and focus of various actors are reflected in the current crossroads of the movement. The following list of eleven
INTRODUCTION
5
issues are currently being debated, and the papers in this volume contribute toward their solution: 1. conceptualisation: Quality of Life (QOL), Level of Living (LL) and Living Conditions (LC); 2. the feasibility of summary QOL indexes; 3. types of data collection (surveys vs administrative registers); 4. integrated surveys vs focused research (simultaneous measurement; integrated databases); 5. comprehensiveness (coverage of domains, topics and indicators); 6. units of analysis (individuals, households, regions, nations); 7. focus of analysis (cross-sectional vs longitudinal design); 8. how to communicate results to citizens; 9. producer–user relations: how to affect decision makers; 10. forms of social reporting; 11. improvement of comparative QOL research. PART I: CONCEPTS AND THEORY
Sten Johansson and Ruut Veenhoven deliver a plea for each of the two major competing traditions in the general welfare research: objective facts only, termed “level of living” (LL) versus reports that allow subjective perceptions of citizens as well, termed “quality of life” research (QOL). Johansson summarises the Scandinavian school, arguing for ‘objective’ facts being most relevant as target indicators in social planning as well as in the political debate in representative democracy. His ‘epistemology of the democratic process’ links social indicators to the original mission of the pioneers of the movement to add a social dimension to technological and economic change. His way of conceptualising the research field is oriented towards social planning. ‘Citizens’ discontent, wishes, demands and goals (in short: preferences) should not be revealed by a mechanical process such as opinion polls, but through a political process (representative democracy), based on scientific measurement of facts. Or, as formulated by Veenhoven in his interpretation of the objective tradition: ‘subjective indicators will distort the technocratic policy process and will give a voice to the irrationalities that have always hampered scientific management’.
6
INTRODUCTION
This has been a strong argument in Scandinavia, as well as in social surveys and social reporting run by official bodies elsewhere, as explicitly stated by Vogel (Sweden), Boelhouwer (Netherlands) and Kamen (Israel). Nevertheless, this tends to be not an absolute criterion, but a dominating one, which the reader will find in reading these contributions and their publications; see also Aasland and Tyldum (Norway). Ruut Veenhoven, whom we invited to formulate a critique of Johansson’s position and define the quality of life approach (QOL) does not reject objective indicators as such, but argues that objective indicators taken alone are inadequate: Social policy makers need both objective and subjective indicators. In principle, this position might not arouse even the hardest objectivists; their position would be that this is a matter of priority: additional subjective measurement would be nice to have, but only in addition and after the basic objective information is secured (Boelhouwer, Noll, Vogel). Veenhoven states subjective indicators are needed in selecting policy goals, since public support for policies is not altogether channelled through the political process. Furthermore, objective measurement often falls short of aggregating the whole, many items are difficult to assess objectively. When looking over the contributions in this volume, we see no orthodox statements, but mixed approaches. In official statistics and social planning there is a clear objective focus. Europe, with its diversity of cultures and levels of economic development, and the ongoing EC integration, is indeed the region where comparative studies are the most intensive. In his contribution, Heinz-Herbert Noll outlines the conceptual framework and major structural elements of a “European System of Social Indicators”, funded by the European Commission. His “European System of Social Indicators” is considered to be an instrument to continuously monitor the social change in Europe. This indicator system which covers 14 life domains is not restricted to individual quality of life, but also addresses social cohesion and sustainability as two major components of the “quality of society”. The indicator system will include 20 European countries, but also the U.S. and Japan as two important reference societies. As the final outcome of this work an electronic “European Social Indicators Information System” will
INTRODUCTION
7
be released. Noll’s contribution also summarises recent conceptual and theoretical development. PART 2: CURRENT SOCIAL INDICATOR AND SOCIAL REPORTING PROGRAMS: NATIONAL AND COMPARATIVE EXPERIENCE
Institutionalisation and Integration in National Programs
In this volume national social indicator programs are represented by two old and established national programs for Sweden, outlined by Joachim Vogel, and, for the Netherlands, presented by Jeroen Boelhouwer. A younger program still in its initial phase in Israel, is discussed by Charles Kamen. The Swedish and Dutch programs have found their institutionalisation in different ways. The Swedish program, reaching back some 30 years, is located at the national statistical institute, and is responsible for large-scale annual data collection forming a system of integrated social surveys, as well as a large program of integrated social reporting. The Dutch program is institutionalised at a special semigovermental institute focusing on social and cultural policy issues. This institute uses data from all sources, in particular the large Dutch statistical institute, and concentrates on policy analysis and social reporting. This appears to be a unique form of institutionalisation, which is close to, and focused on, social planning. Charles Kamen’s report reminds us of much of the early experience from the older programs. Many new programs start off from a wealth of already available information and analyses, but scattered between data collection systems, institutions and publications. Social indicator systems need to work towards integration in all three aspects, as well as towards institutionalisation. There is a long history of programs which never managed to achieve integration and institutionalisation, but were terminated, and are called in as ad hoc studies with long intervals, revived at the mercy of temporary fashion in politics. Among the positive trends in national social indicator work mention should be made of the increased use of register data linked to regular social surveys. This is an old tradition in the Nordic countries, but similar methods are on the way in other European countries (e.g. the Netherlands).
8
INTRODUCTION
The second major development is the series of panel surveys in USA and Europe (Germany, UK, Netherlands, Belgium, Luxembourg) adding a dynamic capacity to social indicator research, in particular with respect to employment, education and income. This is yet another tradition in its own right. In most countries it operates in parallel with traditional cross-sectional studies. One of the major questions concerns the usefulness of long-term panels in analyses of stocks and cross-sectional trends. Panel attrition, and problems of longitudinal comparability explain why most national actors tend to rely on cross-sectional design (Vogel, Boelhouwer, Kamen). Comparative Social Indicator Studies
The expansion of a variety of full-scale social surveys, with full domain coverage, coordinated between several European countries, is one of the most positive recent developments. After OECD’s Comprehensive Survey of the late 1970s, with participation of 7 OECD member states, comparative studies survived in the Nordic countries (Vogel). The next advancement followed with the European Community Household Panel (ECHP), with annual surveys in 12 member states, starting 1994. This was to be the largest survey ever conducted, with good coverage for income, employment, and education only, and poor information for remaining domains. ECHP is planned to be terminated after 2002, to be substituted by a cross-sectional survey as the major tool for comparative social indicator work in Europe (SILC = survey of income and living conditions). Neither the national panels nor ECHP are represented in this volume. Instead a collection of more limited comparative surveys are presented. The EUROMODULE project presented by Jan Delhey, Petra Böhnke, Roland Habich, and Wolfgang Zapf uses a negociated questionnaire module included in already existing surveys (first wave: 8 countries in West and East Europe, and Third World). The NORBALT project co-ordinated by Adne Aasland and Guri Tyldum covers the three Baltic countries, with extension towards the Nordic countries, with two waves. And the third is the new circumpolar survey of living conditions coming up in the Arctic region, SLICA, outlined by Thomas Andersen and Birger Poppel. A final paper in this section is by Michaela Hudler and
INTRODUCTION
9
Rudolf Richter, who outline the public database they are creating with EuroReporting to facilitate comparison of questions and results from dozens of previous surveys covering over twenty countries in Western as well as Eastern Europe. In national as well as in comparative social indicator research we can record important advancement in data availability, with respect to the three major reference points, i.e. comparisons between subgroups, over time, and between nations. Having access to systematic reference data is eminent in statistics, and in particular in social planning. Such data enhance the influence of social indicators in the political process. They stimulate causal thinking, and provoke discussions of responsibilities. Within the national context time series are built up by repeated surveys, which adds time as reference to social reporting of levels and distribution of living conditions. Comparative studies add yet another dimension. This double-sided perspective is central in economic accounting, and was also realised by the early writers of the social indicator movement.
PART 3: APPLYING SOCIAL INDICATORS TO EFFECT SOCIAL CHANGE
The papers above aim at establishing a system of regular descriptive and systematic statistics (which is stressed in the original early thinking of the social indicator movement.) In contrast, the papers in the current section aim at in-depth targeted and analytical research, providing direct answers to pertinent questions. Abbott Ferriss outlines a process he calls “Telesis” which a nation (or any interest group within a nation) can follow to achieve goals measured by social indicators. Valerie Møller and Helga Dickow describe the surveys designed to inform the major transition in South Africa toward full democracy and prosperity for all citizens. May, Stevens and Stols report on indicators applied to track progress in the domain of land reform in the new South Africa. The next three papers each give innovative solutions to creating a unified index of well-being, to allow an answer to the perennial question, “Are we better or worse off than before?” Richard Estes describes the development of an entire system of social indicators for Hong Kong, for which he creates an index composed of over 50 indi-
10
INTRODUCTION
cators, with weights selected by local experts. Andrew Sharpe and Lars Osberg utilize the rich economic and social reports of OECD nations to create their index of economic well-being (where economic is defined broadly to include leisure, poverty, inequality, household activities, and environment.) Also Michael Hagerty shows that voters in national elections reveal their preferences for components of living conditions, allowing a unified index of quality of life to be computed from the representative voter in national elections. Another political goal that is urgently needed in many parts of the world is what Regina Berger-Schmitt terms “social cohesion,” to increase unity among diverse nations or regions. She proposes dimensions and measurements for this important goal. Finally, Christina Behrendt examines the measurement of poverty in some developed nations, and presents a simulation that suggests that poverty is overstated in these nations because of underreporting of income during typical interview procedures. This special issue will be a valuable resource for four groups of readers. To researchers interested in best practices for wellestablished surveys of living conditions, the papers by Boelhouwer, Noll,Vogel, and Berger-Schmitt will be of special interest. To researchers and policy analysts interested in establishing a livingconditions report in their country, the papers by Kamen, Møller and Dickow, Estes, Andersen and Poppel, May, Stevens and Stols and Aasland and Tyldum give invaluable information about developing credibility, consensus-building, and survey design. For researchers interested in cross-national comparison, the papers by Hudler and Richter, and Delhey, Böhnke, Habich, and Zapf describe the rich resources already available, as well as problems of different wording, interpretation, etc. Finally, for citizens wishing to effect changes in public policy, and for researchers studying that process, the papers by Ferriss, Estes, Hagerty, and Behrendt outline how organizations should select goals, utilize social indicators, and develop programs that improve the Quality of Life in their nations.
NOTES 1
Whatever happened to Social Indicator? A Symposium, with contributions by Frank M. Andrews, Martin Bulmer, Abbott L. Ferriss, Jonathan Gershuny,
INTRODUCTION
11
Wolf-gang Glatzer, Heinz-Herbert Noll, Judith Eleanor Innes, Denis F. Johnston, Duncan MacRae, Joachim Vogel and Michael Ward; Journal of Public Policy, volume 9, number 4. 2 One of the two corresponding Swedish institutions (Statistics Sweden) surveying social indicators and publishing social reports appears to have had ten times larger staff in those years, and afforded six large reports every year, while U.S. reports appeared only once every three or four years.
Michael Hagerty Joachim Vogel Valerie Møller
This page intentionally left blank
STEN JOHANSSON
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE FOR NATIONAL POLICY From the Swedish Level of Living Survey to an Epistemology of the Democratic Process
ABSTRACT. The interests in social indicators and social reporting started in the 1960s with the new awareness of poverty in the midst of affluence. In this paper I first discuss the poverty concept and its implication for social policy strategy. The poverty concept should be but one in a system of concepts that throws light on the whole distribution of income and wealth and how income from labor as well as income from capital is generated. The central concepts in the system are income and economic standard, which I explain in different perspectives on command over resources. The command-over-resources concept is used to get from a narrow concept of material welfare that can be measured in money to a wider concept of welfare that includes the universal common social concerns. I confess to being intrigued by the fact that a list of social concerns can be agreed upon that seems to be relevant across cultures, political systems and times. I suggest that this surmised universality springs from the great “life projects” that all humans face over the life cycle. I then discuss the role of social indicators and social reporting as continuous information on these common concerns in the context of an epistemology of the democratic process. Social reporting would serve the democratic process best if it answers “how it is” and leaves the answers on “how it ought to be” and “what should be done” to come about through discussion among citizens. KEY WORDS: democratic process, income distribution, poverty, quality of life, welfare
INTRODUCTION
The Swedish Level of Living Surveys or more broadly the Nordic Living Conditions Surveys are but one approach among many in This paper is based on a lecture that I was invited to give for the opening session of the III Conference of the International Society for Quality of Life Studies, July 20–22, 2000, in Gerona, Spain. Social Indicators Research 58: 13–32, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
14
STEN JOHANSSON
the field quality of life studies internationally. The Nordic surveys may be of interest as an approach to social indicators and social reporting that has become institutionalized. They are a regular part of official statistics produced by the central bureaus of statistics of the Nordic countries and a distinct branch within sociology. This is most thoroughly the case in Sweden. In Sweden this line of research and official statistics started in the 1960s as an early indigenous branch of the international movement to develop social indicators and social reporting. The roots of this research in Sweden of the middle sixties as well as in other countries lie in a public concern – new at the time – with remnants of old-type poverty in the midst of affluence resulting from full employment and rapid economic growth over the two decades of the 1950s and 1960s. In the US, President Lyndon B. Johnson declared “war on poverty” and that there should be a set of statistical “yardsticks” in an annual social report with which the nation “can better measure the distance we have come and plan for the way ahead”. There was a lot of activity on social indicators and social reporting for some years. However, the one and only remaining social indicator from this period in the US is the official statistic on poverty based on the poverty line concept. In parallel – but independently – the Swedish government in 1965 set up a state committee of experts with a mandate to “consider what kind of information is needed to give the discussion on low income an adequate basis”. This committee was led by two economists from the LO, the central organization of the blue collar trade unions. Dr. Rudolf Meidner was the chairman of the committee that included also experts from the white-collar trade unions, the employers association and some independent experts. The late Per Holmberg was the very dynamic secretary of the committee. Meidner and Holmberg drastically widened the mandate of the committee in two steps; arguing initially that what is required is (1) information on the whole distribution of income and wealth rather than only on low income groups and (2) information on the distribution of welfare and well-being in a wider sense than just material standards. I was recruited as a sociologist for this second task in 1967 – in practice to design and execute a survey of the living conditions of the Swedish population as a basis for social policy.
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
15
In this paper I summarize some of the frame of reference for these and other surveys that were pioneered by the Low Income Committee in the 1960s and are now in developed form a regular part of official statistics. These can be seen as quality of life studies that have been specially designed to give information for national social policies. In the first part of the paper I will present a system of basic concepts for studies in the distribution of income, wealth and economic standard of individuals and households. It may come as a surprise that the poverty concept does not have a very central place in the system. Quality of life in this context is conceived of as individuals’ command over resources in terms of income and wealth that can be used for need satisfaction by consumption in the market. You will notice that the command-over-resources concept will be central throughout this paper. The next part is concerned with the wider concept of welfare that was developed for the special level of living survey that was first designed in 1968. The decision to structure the survey into components that were defined to coincide with the main areas of social concern and measure level of living in these areas “objectively” with indicators seems very natural – in retrospect. However, there was lots of economic and other theories standing in the way for the idea that welfare can be measured in an inter-subjectively valid way. As a young sociologist in 1967 – less than half my present age –I did not know that the variation in individual preferences makes objective measurement theoretically impossible. I will somewhat discuss the intriguing fact that these areas of social concern seem to be the same in all countries, cultures and may be also historically. The final part is supposed to present an “epistemology of the democratic process”, which may sound rather too pretentious. This “epistemology” is an attempt to formulate the information needs in the democratic process at the most general level and which of these needs should be satisfied by regular social reporting.
ECONOMIC CONCEPTS FOR QOL-INDICATORS
Meidner and Holmberg argued that an adequate information basis for discussing the low-income problem was not just statistics on
16
STEN JOHANSSON
various low-income groups. Low income and poverty are relative concepts, they claimed. Their concern was with equality generally rather than with low income specifically. Therefore they argued that the information must cover the total income distribution, not just income below some poverty line. The information must further show how the income distribution is formed as income from labor and as income from capital and finally how the distribution of factor income is changed by taxation and social benefits to set the level of disposable income for consumption of households of different types. Using the low-income concept rather than the much more loaded poverty concept is not an innocent thing in terms of national policy strategy. This shying away from the dramatic poverty concept in favor of the relative and rather neutral low-income concept is a characteristic of the Nordic welfare state in which universalistic policies for the whole people are preferred before selective policies for the poor. Why is this? In an economy under rather full employment the income distribution is shaped like an onion with most persons and households within a rather narrow income range. Rather few are below the big bulge and rather few are also in the thin stem (see Figure 1). With an anti-poverty strategy social policy becomes something that the majority above the poverty line do for the minority below the poverty line. This kind of strategy cannot generate sustained
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
17
political support from a big majority because it tends to split the working class in the middle. Many of those just below the poverty line will actually resent being called poor while many of those immediately above will not be enthusiastic about paying taxes to support neighbors whose economic circumstances are no different from theirs. Choosing the poverty concept as one of the central indicators of quality of life for national policy thus has some very strong implications for political strategy. However, the low-income and also the poverty concept, of course, have their place in the system but only as one among several in a system of concepts for material quality of life indicators for national policy. To what extent is low income, poverty or more generally low material standard caused by wages so low that even fulltime all year work does not give an adequate economic standard? To what extent is low income caused by inability to work full time all year because of unemployment, illness or other circumstances? To what extent is the material standard low because the income must support a large family? These are three basic questions that are interrelated. To answer them a system of basic concepts for material QOL-indicators has been developed as a frame of reference for a statistical system on income, wealth and economic standard (Figure 2). It was originally based on the experiences of the 1965 Low Income Committee but has been much developed since then. The many boxes and arrows in Figure 2 really illustrate the complexity of the system with the many factors that must be taken into account. The basic concept in a QOL-perspective is economic standard. However, to arrive at that concept in a way that is relevant for considerations of national policy we need several intermediary and auxiliary concepts and indicators. The many boxes and arrows are there to illustrate the many factors and policies that affect the economic standard of individuals and households. The point of departure at the top of the figure is individual resources (1) in terms of age, work experience, schooling, health, work impediments, etc, which are important for the individual’s ability to function on the labor market. Social policies can increase these resources in the population and somewhat influence their
18
STEN JOHANSSON
distribution, mainly through providing education and health care (2) in a broad sense. Policies to adapt labor market conditions (3) so that as many as possible can find gainful employment also come in early. Economic policies and the pricing of the factors of production are the most important determinants of employment. Various labor laws on union rights, working hours, labor protection, social insurances are also
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
19
important in shaping the quality of jobs that are available to the individuals on the labor market. At any point in time, e.g. a calendar week, there are close to five million jobs in the Swedish economy, which may be described in terms of wages per hour (4) but also in terms of other employment and working conditions. Annual income from labor (5) is determined by earnings per hour and the number of hours worked per year (6) in the context of general labor market conditions. Differences in hours worked provide much of the explanation for the variations in annual income from labor between individuals in Sweden as in other countries. Insurance for loss of earnings (7) in the event of unemployment, illness, childbirth, etc, does reduce the effects of number of work hours on annual income structure. Adding income from wealth to annual income from labor we get annual market income (8) or annual factor income. When taxes and rates (9) are deducted from annual income and allowances (10) added, we get disposable income (11), which needs to be summarized from the individual to the household level. In order to make adjustments for need, household disposable income must be related to household composition (12) by an equivalence scale that adjusts for household size and composition. We then arrive at economic standard, the most basic concept for income distribution studies. But even this concept must be interpreted with some consideration of two further factors. The household’s time margin (14) for own production and leisure as well as social consumption rights (15) needs to be pointed out as additional factors to take into account when one’s interest is really in final need satisfaction. In international comparisons it is especially important to note social consumption rights, which give the individual access to medical care, education, municipal home help, transportation services, etc, quite free of cost or at very low fees. Wealth in the form of financial and real assets (18) must also be included as a central concept in the system. In the diagram, the arrows to and from “Wealth” stand for saving and consumption of wealth respectively. The arrows to and from the box “Market prices” (17) represent inflation profits and inflation losses. A description of the economic situation of individuals and households using the concepts in the upper part of Figure 2 will detail
20
STEN JOHANSSON
how income is generated in the economy but also some of the main ways in which economic and social policies can affect the income generating processes particularly in respect to employment and productivity. Full employment is the key to equality of labor income also indirectly in that full employment affects labor productivity. Long spells of unemployment or long periods outside the labor force affects labor productivity negatively besides creating big dispersion in annual incomes from labor. Big differences in labor productivity between individuals will inevitably lead to big dispersion in wages and salaries in a market economy. Access to education and job training are the means to make all individuals employable at good wages and to make full employment possible without big wage differences. In this perspective the distribution of households’ annual labor income and total market income reflect the distribution of individuals’ command over other resources, both so-called human capital such as health, education and work experiences and wealth in financial and real assets, structural, that is, position in the production system, and collective, such as the efficiency of the production system. Under this assumption annual factor income is an aggregate indicator of all types of resources that individuals and households can command in their pursuits of income. With the factors in the upper part of Figure 2 statisticians can explain a very significant part of the distribution of labor income and total market income. In a QOL-perspective economic standard in the lower part of Figure 2 is normally regarded as the central concept because of its close relation to need satisfaction by command over resources for consumption. The smaller the dispersion in annual factor income the less must be done through taxes and social allowances in the middle part of Figure 2 to ensure a low dispersion in economic standard between households and individuals and that poverty is reduced as a problem. However, even a very egalitarian distribution of labor income of persons of active age must be adapted to differences in needs of households and families in terms of number of dependants. Income is an ex post indicator of command over (other) resources while economic standard is an ex ante indicator of (potential) need satisfaction through consumption of goods and services.
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
21
A system of social indicators like this and social reporting that serves to clarify the mechanisms behind the generation of the distribution of annual income and of economic standard is a very important instrument for national policy.
TOWARDS A WIDER CONCEPT OF WELFARE
The decision that the discussion on low incomes required information not just on the distribution of economic standard as potential command over goods and services on the market but also on the distribution of welfare and well being in a wider sense has been considered the really pioneering act of the 1965 Swedish Low Income Committee. Living standard or welfare “in a wider sense” was a popular concept at the time (quality of life actually came later). A big question for the committee then became; what is welfare or well-being in “a wider sense” than economic standard that can be measured by money? The committee actually got stuck for a time on the many issues that could be brought in to bear on the meaning of “welfare or well-being in a wider sense”. Particularly divisive was the idea that the committee needed a concrete specification of what is the good life in order to structure what should be included in “welfare or well-being in a wider sense”. Should living conditions be measured objectively according to a common standard or subjectively as well-being according to each individual’s evaluation or satisfaction? The committee heard experts from various disciplines and had questionnaires for a survey drafted but the discussions continued. The decision of the Swedish Low Income Committee in the end may seem almost trivial. The committee accepted the proposal that the survey of living conditions be structured by a level of living concept that was adapted from a UN technical report on International Definition and Measurement of Standards and Levels of Living. The first and foremost characteristic of that concept is that it both directs and restricts information to the areas where the political mechanism is by some degree of consensus used to affect living conditions through social policy. The concept also organizes the information into level of living components mainly by the sector
22
STEN JOHANSSON
divisions used in social policy. A corollary of this is that a unitary measure of welfare, a GWP, is rejected in favor of separate systems of indicators that are designed for each of the sector policies, health, education, housing, labor market and so on. The original nine components in the 1968 survey have been modified in later research in the sociology of welfare and in the practice of Statistics Sweden. The parliament in 1974 decided that Statistics Sweden should start an income distribution survey and an annual survey of living conditions based on the work carried out by the Low Income Committee. Both these surveys are still ongoing. Joachim Vogel designed the Living Conditions Survey with a core of indicators on each of the components that were included in the survey every year and then in-depth extensions of some set of components every third to fifth year. Living conditions surveys were carried out in the other Nordic countries. Statistics Sweden championed the idea that social indicators and social reporting should be based on a comprehensive social survey in the OECD working party on social indicators in the 1970s and nowadays in the European Union. The Norwegian FAFO-institute and Statistics Norway have spread their own version of the idea to a number of developing countries. The list of components of welfare that I use in presentations includes the following nine components. In practice it does not differ importantly from the official Swedish list that Statistics Sweden is using. 1. 2. 3. 4. 5. 6. 7. 8. 9.
Economic resources and consumers’ conditions Employment and working conditions Education and access to schooling Health and access to medical care Family and social relations Housing and amenities Culture and recreation Security for life and property Political resources and participation
In extending the welfare concept beyond the purely economic aspect, one is faced with a choice at the theoretical level that has been much discussed in Nordic QOL-studies. One can either define this wider welfare concept in terms of degree of need satis-
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
23
faction or in terms of command over resources (see Figure 3). In this context I just like to mention two of the Nordic sociologists who have developed more subjective approaches for quality of life studies in the 1970s and 80s; Erik Allardt in Finland who developed the concepts “having, loving and being” and Siri Naess in Norway who developed the concept “inner quality of life”. These concepts and similar subjective concepts give important insights in the psychology of the human conditions but cannot easily be translated into goals for national policy. I leave the discussion of such other approaches to QoL-studies to other sessions in this conference.
AN APPROACH TOWARDS A WIDER CONCEPT OF WELFARE
A very pragmatic approach to a wider concept of welfare is to get many people to agree on a list of components of welfare like the one I just presented that obviously capture common social concerns in most countries. I still like to report on my attempts to provide some theoretical reasoning as to why a system of social indicators and social reporting for national policy must go beyond indicators on economic standard. Figure 3 illustrates that first of all there are a number of intervening factors between command over economic resources in terms of disposable income and realized market consumption as need satisfaction of household members. There is no guarantee that the same amount of money gives the same amount of consumption, let alone the same amount of need satisfaction. If one includes also the fact that individuals have different preferences one ends up with the dogma in economic theory that inter-subjectively valid measures of welfare cannot be constructed. Figure 3 also tries to illustrate that there are needs which money cannot satisfy, but which are nevertheless of vital importance for the individual, for example the need for close relationships with other people. Such needs are satisfied via human relationships, which are “private” by nature, and in which exchanges are made and substitutions occur according to other than market rules. Figure 3 thirdly illustrates that there are other resources than economic ones which in the individual case compensate for or add to economic resources, e.g. health, education, work, family, social
24
STEN JOHANSSON
and civic rights, etc., which may be as important for the individual’s chances of achieving satisfaction as the economic ones. And fourth, individuals earn income at the expense of differing costs in other respects. Such “costs” cannot only or even primarily be expressed in terms of working hours. “Costs” should include all the conditions under which income is earned through work. For some people, for example, the character and conditions of their jobs are such that working time is positive and enriching. For others work is connected with considerable health hazards at the same time as it might be so physically and/or psychically demanding that a great deal of leisure time must be sacrificed for sleep or rest. In order to describe the situation of the individual from the point of view of welfare, therefore, one cannot content oneself with describing the individual’s command over economic resources and possibilities of consumption in the market alone. Neither is it enough to study the consumer’s utilization of purchasing power in the market as in household budget surveys. A wider concept of welfare than the economic one must be applied if one is to develop a comprehensive method for social reporting as a basis for national policies to promote quality of life. One way would be to proceed to
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
25
so-called household budget surveys that register how income is used for consumption (horizontally in Figure 3) and then widen such a survey to include other needs. But one can also proceed in the opposite direction in Figure 3, and solve the problem of definition in the same way as in economic theory with regard to economic resources as a measure of welfare, where income differences are relevant to welfare on the supposition that in a free market they reflect differences in access to goods and services. In the same way, one can in social policy theory assume that e.g. health, education, work, family, social and civic rights, etc., are resources with the help of which the individual can control and consciously direct his or her life. One would thereby have created a wider definition of welfare than the economic one. Welfare can thus be defined as individuals’ command over resources in terms of money, possessions, health, education, family, social and civic rights etc. with which the individual can lead his life. The difference between welfare as need satisfaction and welfare as command over resources is perhaps most clearly demonstrated when considering the behavior of the individual as deliberate actions in a time perspective. In the one case, it is the results of individual actions which are termed welfare and which are to be measured in an inter-subjectively comparable way. In the Nordic discussion this “result” may then be labeled “need satisfaction” as Erik Allardt or “inner quality of life” as Siri Naess have done, or leave it as “utility” as most economists do. In the second case, it is the individual’s having a choice of actions, which is termed welfare. It is not necessary to take a position on whether he actually does fulfill himself (e.g. by sacrificing satisfaction in many areas to gain perfection in one) or is productive of utilities. To borrow a term from economic literature, one can say that welfare is judged in the first case ex post, in the second ex ante.
ARE THE SOCIAL CONCERNS UNIVERSAL?
When I started to meet international colleagues in the field in the early 1970s, particularly in the OECD Working Party on Social
26
STEN JOHANSSON
Indicators, I was very intrigued by the fact that “my” list was very similar to the lists developed in other countries even if the political system and the cultures were very different. Indeed when I made an international overview of the varying definitions of level of living components – “subsystems” in a document clarifying common social statistical guidelines for the Eastern European countries, “goal areas” in the OECD document on social indicators, corresponding ones in the United Nations’ statistical system SSDS, “chapters” or corresponding grounds for categorization in the many different national reports of the Social Trends type – I found a surprising amount of similarity between the lists. The similarities reflect not only that social statisticians in different countries cooperate with and learn from one another. I think that the lists also reveal a high degree of universalism in what is considered as social concerns in all countries. The countries vary as to the form and degree of collective responsibility and in the relative importance ascribed to the different areas but the same areas are everywhere relevant. From a purely theoretical point of view, this surmised universality in common areas of concern is very interesting. Whether the collective values and needs of the individual are expressed by representatives who are accountable to the people for their decisions in recurrent elections, or whether they are arbitrarily defined by a class-bound, economic or military elite who cannot be removed by the will of the people, these areas comprising the political sphere remain the same. The underlying cause of this universality could be that the human condition is basically the same everywhere. Some of the problems and challenges facing people over the life cycle in every society must be solved collectively. The collective takes some form of responsibility for individuals who suffer loss of health or weakness in old age in every society. How work is organized and the results of production distributed must be collectively regulated in every society. In every society one must in one way or another collectively regulate how new generations are to be socially initiated into the society (educated). In all types of societies there are collective arrangements for recreation and cultural expression. The individual’s political rights and duties
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
27
are everywhere regulated, even if in very different ways. Everywhere, attempts are made to uphold order and – at least in principle – to protect individual life and property. Yet, I have been unable to find any structural principle for these collective interests. The list contains concepts at different levels of abstraction. They do not therefore form a general system of concepts, which might give a logically connected structure to the whole area of study. This is probably because the various areas of politics have not emerged according to any easily discernable logic and because citizens’ political ambitions cannot exclusively be classified in level of living terms, even using our wider definition of level of living as “the individual’s command over resources … with the aid of which he can control and consciously direct his own life.” There may be some infinite variation in the preferences of individuals as to the basket of goods and services that they want. However, when they gather to decide what should be the collective concerns they tend to arrive at the same answer in all countries, in all cultures, in all history because these concerns spring from the big life projects that all humans face over the life cycle: To be cared for, nurtured and fostered as a child To be trained or educated as a preparation for the adult roles To find a job in the system of production To find one’s own place to live and to form a family To maintain health over the life cycle To be protected against violence and crime To find a societal identity in culture and as a citizen
EPISTEMOLOGY OF THE DEMOCRATIC PROCESS
How should citizens in a democracy act to insure that their orientation in the world and their common decision-making in the public realm are informed by the best possible knowledge? What is the role of social reporting and how can it function in an epistemology of the democratic process? I do not know what the reader associates with the concept of “epistemology”.1 My own thinking originally went to lengthy
28
STEN JOHANSSON
hair-splitting arguments among abstract philosophers about what is knowledge and if knowledge is at all possible. This is not what I propose to do. I mean the lengthy hair-splitting. Epistemology as I will use the concept is the normative theory of knowledge, which I translate to mean how to secure good answers to important questions. In this context the concern is with good answers to important questions in the democratic political process, conceived of as a method for a citizenry to arrive at common decisions that are binding on all in the land. First we need to look for the important questions in the democratic process. Of course, we could stay there forever arguing over the many intriguing problems. Which are indeed the important questions in the democratic process? They cannot be detailed in substantive terms and set up in their order of priorities. Is disarmament more important than global warming or international trade? Is inflation more important than unemployment or the budget deficit? The important questions must be detailed along general rather than substantial lines. I take the important questions to be the three most general questions for any rational actor: (1) How is the situation, (2) How ought it to be, and (3) What should be done? These questions are important in that they are always there implicitly or explicitly whenever there is decision-making and whatever the substantive issue. We also need to see that these questions differ very much from each other as to how they should best be answered. How It Ought to Be
Embodied in how it ought to be are all conceptions of conditions desired or sought, as well as the individual’s aspirations for himself and his own group. Objective methods cannot determine how it ought to be. Therefore, citizens should not allow political decisions to be reduced to problems appropriate for experts only. At the very core of the concept of citizenship lies the notion that the view of every citizen on how it ought to be should be of equal weight. This does not mean that citizens collectively should decide how it ought to be by each and every one subjecting himself to ques-
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
29
tioning, as in an opinion poll. These different opinions on how it ought to be would then be calculated mechanically. The total would not represent a “common will” that is reasonably consistent as to economic, social and environmental developments and also fairly weighed as to the interests of strong and weak groups in society. Citizens’ discontent, wishes, demands and goals ought therefore not to be revealed by any mechanical process such as opinion polls, but through discussion in a political process. Discussion provides a means whereby individual personal interests can be weighed against each other. Discussions permit demands and wishes in different areas (schools. medical care, communications, defense, taxation, pensions, etc.) to be weighed against each other. Individuals and groups of citizens can be brought to see how demands in different areas may be incompatible or impossible to satisfy simultaneously. Discussions facilitate the dissemination and consolidation of opinions among the people. A view of how it ought to be only becomes a serious political factor when some group of citizens (a “party”) adopts it and takes responsibility for it. It is also by this means that elected representatives are linked to movements. Social reporting based on questioning the individuals as to how it ought to be does not facilitate the democratic process as outlined by the theory of representative democracy. Strictly speaking, such investigations merely simulate the democratic process in a nonconstructive way. Discussion within a political organization is the best way of formulating an answer to how it ought to be. What Should Be Done?
The question of what should be done is fundamentally of a different character than the question of how it ought to be. The difference may be expressed by saying that whereas citizens must agree on the goals on the basis of self-interest, they must choose the means on the basis of information concerning cause and effect. Nor can the choice of means be reduced to a problem to be left to experts. Experts can only give guidance as to what should be done on condition that they are given detailed instructions as to how it ought to be, viz. the goal of the measure in question.
30
STEN JOHANSSON
This does not mean that citizens’ views of what should be done can best be discovered by questioning each individual, as in an opinion poll. The reasons against this are partly the same as against investigating how it ought to be by solely mechanical listing. Measures in different areas must be coordinated, both with regard to costs and in order to avoid conflicts. Such compromising is essentially a question of evaluation, where the interests of each citizen should be considered equally. The question of what should be done must therefore in the final analysis be tested in a political process intimately connected with the choice of goals. Measures must be harmonized and crystallized into total programs. Social reporting based on questioning individuals about what should be done contributes nothing to the democratic process, since it is easy to confuse opinion polls with referendums, a situation which makes it easy for the elected representatives to be deprived of or to wriggle out of their responsibilities. A mechanical reckoning which combines informed opinion with spontaneous, uninformed or mainly emotional opinion is clearly unwise. An opinion poll on what should be done might be meaningful if respondents before answering the questions could consult with the best experts in the land on all that should be taken into account like elected representatives and responsible governments must do. As stated before, the citizens’ chances of collectively arriving at a conclusion as to what should be done, is best furthered by party participation, and by each individual studying the questions involved and ensuring access to competent expertise. Therefore, social reporting about what should be done ought not to have the form of questioning individuals about their private opinions. How It Is
The actual living conditions and how they change in those aspects that influence the citizens’ views of the welfare development are as important questions in the political discussion as the two analyzed above. Is the state of public health improving or deteriorating? Are working conditions becoming better or worse? Is the distribution of income and wealth among individuals and households widening or narrowing? Are human relationships within the family and in other
CONCEPTUALIZING AND MEASURING QUALITY OF LIFE
31
contexts becoming richer or more impoverished? Is crime in the forms of theft, vandalism, violence, etc increasing or decreasing? Is participation and involvement in politics increasing or decreasing? Citizens cannot get good questions to these answers by discussion, by voting or by experts deriving answers from some theory. Citizens can arrive at good answers only by organizing “counting” and by all citizens agreeing to contribute to the process in which each and everyone’s participation has equal weight. Social reporting based on comprehensive living conditions surveys can be conceived of as citizen reports on social and economic change. CONCLUSIONS
When we consider the democratic political process as a means of deriving, from many individual views, collective answers to the three questions of how it is, how it ought to be and what should be done, we can draw a basic conclusion: social reporting is not needed – according to the theory – in order to obtain answers to the questions of how it ought to be and what should be done. On the other hand, if social reporting is directed at answering the question of how it is, it can fill a gap in the theory and practice of the democratic political process. Epistemologically speaking, we would also want to understand the nature of “good” answers to these kinds of questions in the context of the democratic process. Truth is the most important
32
STEN JOHANSSON
concern when we think of “good” answers to the question on how it is, but truth is not the only concern. Authenticity might be one of the important characteristics of a “good” answer to the question on how it ought to be. Efficiency and effectiveness come to mind in relation to the third question on what should be done. But both answers to the latter questions must also be moored in the people through discussion informed by best available science. NOTE 1
According to Encyclopedia Americana, (New York 1971), vol. 10, pp. 430– 433, “we may say that the epistemologist is primarily concerned to analyze and understand certain philosophical concepts. These include the concept of knowledge itself together with many others which we employ when we characterize knowledge – the concept of meaning, belief, truth, proposition, faith, certainty, probability, evidence, confirmation, justification and rationality … the epistemologist hopes to discover and formulate the basic assumptions which underlie human knowledge, both common-sense knowledge of the world around us and the more abstruse knowledge of specialists”.
The Trade Union Institute for Economic Research Stockholm, Sweden E-mail:
[email protected] RUUT VEENHOVEN
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
ABSTRACT. There are many qualms about subjective indicators, and some believe that social policy would be better for not using them. This paper consists of a review of these objections. It is argued that policy makers need subjective indicators, the main reasons being: 1. Social policy is never limited to merely material matters; it is also aimed at matters of mentality. These substantially subjective goals require subjective indicators. 2. Progress in material goals can not always be measured objectively. Subjective measurement is often better. 3. Inclusive measurement is problematic with objective substance. Current sum-scores make little sense. Using subjective satisfaction better indicates comprehensive quality. 4. Objective indicators do little to inform policy makers about public preferences. Since the political process also does not reflect public preferences too well, policy makers need additional information from opinion polls. 5. Policy makers must distinguish between ‘wants’ and ‘needs’. Needs are not observable as such, but their gratification materialises in the length and happiness of peoples’ lives. This final output criterion requires assessment of subjective appreciation of life-as-a-whole
INTRODUCTION
There is a longstanding controversy in social indicators research between the ‘objective’ and the ‘subjective’ approach. In the objective approach the focus is on measuring ‘hard’ facts, such as income in dollars or living accommodation in square meters. The subjective approach in contrast considers ‘soft’ matters such as satisfaction with income and perceived adequacy of dwelling. The objective approach roots in the tradition of social statistics, which dates back to the 19th century. The subjective approach stems from survey research, which took off in the 1960’s. The objective approach is similar to mainstream economic indicators Social Indicators Research 58: 33–45, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
34
RUUT VEENHOVEN
research, though the topics differ, the method is the same. The subjective approach is akin to the psychological stream found in economic indicators research, which monitors things like consumer trust (Katona, 1975) and subjective poverty (VanPraag et al., 1980). The objective approach is currently central in the Scandinavian countries, in particular in the Swedish ‘level of living’ studies. From the beginning the basic tenet has been that social policy needs hard facts in the first place (Johansson, 2001). Hence Scandinavian social surveys do not involve items on matters like ‘trust’ and ‘happiness’. Similarly, the Dutch ‘welfare-index’ focuses on material living conditions (Boelhouwer, 1999). Though the Dutch welfare survey contains items about happiness and stress since the 1970’s, these data are seldom reported. The subjective approach originates from the US. Landmark studies have been published by Campbell et al. (1975) and by Andrews and Withey (1976). This approach is further refined in the German ‘welfare studies’ (Glatzer and Zapf, 1984). Specializations have been developed on subjects such as perceived poverty (VanPraag et al., 1980), values (Inglehart, 1990) and happiness (Veenhoven, 1997). Defenders of the objective approach hold that social indicators serve to guide social policy and that social policy makers need information about (1) the actual state of social problems, and (2) the effects of attempts to solve these problems. This information should be of an indisputable nature, in other words ‘objectively true’, and this scientific truth should enable rational social engineering. In this view, subjective indicators will distort the technocratic policy process and will give a voice to the irrationalities that have always hampered scientific management. In this paper I evaluate the objectivistic position. To do this, I first examine the difference between ‘objective’ and ‘subjective’ indicators in more detail. Then I take a closer look at the misgivings to be found about subjective indicators. Next I will explain why social policy still needs subjective indicators and why objective indicators taken alone are inadequate.
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
35
THE OBJECTIVE–SUBJECTIVE DISTINCTION
At first sight, the distinction between ‘objective’ and ‘subjective’ indicators is fairly clear. Yet when a closer look is taken, there are two dimensions of difference. Firstly there is a difference in substance matter measured. Objective indicators are concerned with things, which exist independent of subjective awareness. For instance: someone can be ill in an objective sense, because a tumor is spreading in the body, without that person knowing. Likewise, Marxists maintain that workers are objectively underclass people, even if they see themselves rather as middle class. Both the doctor and the Marxist give more weight to the objective condition and will press for treatment even if the patient protests. Secondly, there is a difference in assessment. Objective measurement is based on explicit criteria and performed by external observers. Illness can be measured objectively by the presence of antigens in the blood, and class membership by possession of means of production. Given these operational definitions, any impartial observer will come to the same conclusion. Yet subjective measurement involves self-reports based on implicit criteria. The ignorant cancer patient who reports to feel in good health may have based that appraisal on many cues and will not be really able to say how he came to that appraisal. The worker with false class-awareness fails to notice the whole point. The examples show that the differences in substance and measurement do not necessarily concur. The possible combinations are presented in Scheme 1. The two top quadrants concern extrinsic substance matters, such as physical condition, mental aptitudes and social position. The quadrant top left denotes the combination of objective substance and objective measurement. An example is the actual ‘wealth’ of a person when measured by her bank account. The top right quadrant also concerns objective substance, but now measured by self-estimate. An example is measuring wealth by perceived wealthyness. The two bottom quadrants concern subjective matters, such as identity, happiness and trust. The bottom left quadrant combines subjective substance with objective measurement. An example is
36
RUUT VEENHOVEN
Scheme I
measuring happiness by suicide. The bottom left quadrant measures subjective substance using subjective appraisal, for instance, measuring happiness by self-report.
QUALMS ABOUT SUBJECTIVE INDICATORS
Scheme 1 helps to chart the doubts about subjective indicators. Misgivings about mental matters must be distinguished from misgivings about measurement by self-reports. Since these reservations are independent, the criticism mounts most in the bottom right quadrant. The colours indicate the degree of scepticism. MISGIVINGS ABOUT MENTAL MATTERS
Pragmatic Objections
Part of the criticism is pragmatic. It is argued that subjective appraisals tend to be unstable, incomparable and unintelligible and are therefore of little use in social policy. Unstable. It is argued that attitudinal phenomena vary over time and that this variation has little link with reality conditions. For instance,
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
37
attitudes about safety in the streets could depend more on media hype than on actual incidence of robbery. In this view, subjective indicators cannot provide a steady policy compass and fail to protect policymakers against the whims of the day. Incomparable. It is also argued that the subjective appraisals cannot be compared between persons. One assertion is that different people use different criteria, so two persons stating they are ‘very happy’ could say so for different reasons. Another claim is that people have different scales in mind, and that people who report they are ‘very happy’ may in fact be equally as happy as someone who characterizes his life as ‘fairly happy’. In economy this reasoning is known as the theorem of ‘incomparable utilities’. If true, this would mean that subjective appraisals cannot show whether one person (or social group) is better off than another, and hence that this kind of indicator is of little help in selecting those most in need of policy support. Even if subjective appraisals are fairly comparable in a given language community, there is still the possibility that standards will shift over time. For instance, substantial improvement of living conditions might raise standards of comparison, and might thus result in rising dissatisfaction. A classic example is the dissatisfaction that soared among African Americans between 1946–66 in spite of their unprecedented emancipation in this era (Manning Gibbs, 1972). This would mean that success or failure can not be assessed in this way. Likewise it is argued that subjective appraisals could not be compared across cultures. The example of ‘poverty’ is often given in this context. Notions of poverty, and hence definitions of oneself as poor, will differ greatly between rich and poor nations, and within nations between upper and lower classes. This would mean for social policy that these kinds of indicators tell policy makers little about relative performance. Unintelligible. A related objection is that the criteria used for these subjective appraisals are largely implicit. Though people know fairly well how satisfied, anxious or trustful they are, they typically know less well why they think this is so. The appraisal process is quite complex and partly unconscious; this creates at least an interpretation problem for social policy. The declining trust in
38
RUUT VEENHOVEN
government (Vile, 1999) is an illustrative case. Though the trend is fairly clear, at least in the US, the causes are not and hence neither the remedy. Essential Objections
Next there is the more fundamental objection that subjective appraisals can be objectively wrong. This is the ‘doctor knows best’ argument. Unrelated to Objective Reality. One of the surprises of social indicators research is that correlations between objective conditions and subjective appraisals tend to be weak. For example actual income is only modestly related to income satisfaction and hardly related at all to overall happiness (Veenhoven and Saris, 1996). Likewise the incidence of mugging has little relationship to perceived safety (Noll, 1994). This all merges into the position that subjective valuation is in fact irrelevant. Satisfaction judgements in particular can depend too little on real quality of life and too much on fashionable beliefs and arbitrary comparison. In this view, policy makers can better ignore appraisals of citizens, just like some doctors disregard their patients complaints. Instead policy makers should look to objective statistical information, like doctors who believe only laboratory tests. Incorrect. Still there are some persistent patterns in subjective appraisals that cannot be so easily denounced as irrelevant. For instance: the above mentioned middle-class identification of blue collar workers and the rising distrust in government in the USA are clearly no inconsequential passing whims. In such cases subjective valuations are often debunked as ‘false consciousness’. Classic Marxists were quite explicit about this, but mostly condemnation is more covert. Disturbing findings about subjective experience are simply ignored. This seems to be happening to my studies that found no greater wellbeing in welfare states (Veenhoven and Ouweneel, 1995; Veenhoven, 2000b).
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
39
MISGIVINGS ABOUT MEASURING BY SELF-REPORTS
Several objections concern matters of validity. It is doubted that selfreports tap the things we want to access, even if the aim is inner matters. Next there are qualms about reliability. Self-reports are said to be imprecise and vulnerable to distortions. Though much of this criticism is overdone, there is some truth in it. Validity Doubts
When objective matters are measured by self-report, there is always the problem that survey questions may evoke responses to different matters than the investigator had in mind. Even with a seemingly clear cut matter such as ‘income’, there are problems: is it personal income or family income, gross or net, should capital revenues and non-monetary income be included, etc. This problem is particularly noticeable for ill-defined concepts such as ‘health’ and ‘social prestige’. When subjective substance is measured, a further problem is that people may not have thought much out in their mind. For instance, not everybody has a crystallized ‘self-concept’ or a clear ‘class-conscience’. Even when the person has some idea, this is not always fully consciously understood. For example, racists often fail to acknowledge their own opinions. Obviously these problems vary with subject matter. Elsewhere I have investigated the possible validity problems in the measurement of happiness in more detail but found no serious flaws (Veenhoven, 1997, 1998). Reliability Doubts
Even when self-reports fit the subject matter, there is still the problem of precision. Self-reports are typically made on fixed response options, the number of which is mostly no greater than 10. Not only are these scales rather crude, but the response on them are also fickle. The same amount of satisfaction may be rated by one person using the number 6 and by another person using the number 7. Such random error is no great problem for average scores, but it deflates correlations.
40
RUUT VEENHOVEN
Next there is the problem that responses may be distorted in a systematic way, such as by a tendency for respondents to conform to social desirability. There is some evidence that desirability bias inflates ratings of income, social prestige and happiness. Along side such cultural biases there may also be systematic distortions in interviewing, item sequence and response-formats. Schwartz and Strack (1999) have demonstrated several such effects in the measurement of happiness. Systematic measurement error is especially problematic if it works out differently across social categories. NEED FOR SUBJECTIVE INDICATORS
In spite of these weaknesses, subjective indicators are indispensable in social policy, both for selecting policy goals and for assessing policy success. Objective indicators alone do not provide sufficient information. Selecting Policy Goals
Political entrepreneurs must have an idea of what people want, to mobilize the necessary support. They must also get in view what people really need, to select the most meaningful objectives. Much of this information requires subjective indicators to be obtained. What People Want. When deciding on goals, policy makers meet time and again with the problem that the political process does not always reflect public preferences adequately. Representatives sometimes fail to pick up latent concerns and vested interests often keep appealing issues from the political agenda. Good political marketing therefore requires additional public opinion research, in particular polls on worries, aspirations and satisfactions.1 These indicators are subjective in both substance and measurement. This kind of research is common practice in all developed democracies. What People Need. Policymakers also operate in a more technocrat way and try to grasp what people really need. Here the problem is that expressed wants do not always reflect true needs. A good example is the case of materialist aspirations in affluent society. The western public wants ever more money and consumption, and this
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
41
demand is served well by politicians. Yet in spite of the stunning rise in the material level of living, average happiness has remained about at the same level. According to Frank (1999) this is because our material needs are already satiated. In his view, the constant craving for more luxury draws on an underlying need for supremacy, which could be equally well met in less wasteful ways. Lane (2000) likewise has observed in a decline in happiness in modern market economies, which he attributes to the institutional neglect of social needs. In this example the gratification of needs in a population is measured by happiness, that is at the very most a subjective indicator. Elsewhere I have argued that overall happiness is indeed the best available indicator of the degree to which true needs are met, especially if combined with the number of years lived (Veenhoven, 1996, 2000). Assessing Policy Success
Assessing policy success requires information about goal attainment, such as reduction of poverty or improvement of housing conditions, and information about public support. This demands both subjective measurement and assessment of attitudinal matters. Goal Attainment. Success in some goals can be measured objectively. Improvement of housing conditions can be measured using the gain in square meters per person or improvement in education using student/teacher ratio. Yet such measures have their limitations, and in some cases additional subjective indicators are required. This is for instance the case with public ‘health’. Considerable problems exist regarding assessment of average health based on medical consumption and registered incidence of disease. Longevity does not fully capture the phenomenon either, and the effect appears only in the long term (Veenhoven, 1998). Therefore all developed nations run health-surveys to gather data on subjective health complaints and reports of general feelings of health. Likewise, reduction of xenophobia manifests only partly in objective indicators such as racist attacks and interethnic marriage. Attitudinal data are needed to complete the picture.
42
RUUT VEENHOVEN
Progress towards some goals can hardly be measured objectively. This is for instance the case with restoring trust in government. Since the link between attitudes and behavior is typically weak, confidence manifests at best marginally in voting behavior or taxpaying. Assessing progress with regard to this attitudinal goal requires attitudinal data in the first place. Public Support. Success in social policy depends typically on public support. Without public backing most programs perish in the long run, even if planned goals are reached. Public opinion is not always fully expressed in the political process, hence polls are needed for additional information. Survey data are particularly needed for issues that are not on the political agenda and for groups that are ill represented.
WHY OBJECTIVE INDICATORS FALL SHORT
The need for subjective indicators must also be judged against the limitations of objective indicators. Objective indicators provide only a part of the required information and give generally a better view on details than on the whole. Hence categorical rejection of subjective indicators leaves the policy maker with an information deficit, which is inevitably replenished with private observations and hearsay. Limits to Observation
We have already noted above that social policy is not only concerned with objective matters such as ‘income’ and ‘sanitation’, but also with subjective things like ‘trust’ and ‘perceived safety’ in the streets. Such issues are typically intertwined, in the policy mix there is always a combination of material and mental matters. Hence objective indicators tell only half the story. We have also seen that objective measurement falls short on a lot of issues, not only in attitudinal matters but also in the assessment of objective substance. Remember that even the objective measurement of income is problematic. Objective measures also have limited validity and reliability. Joint use of objective and subjective measures is mostly helpful to get a complete picture,
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
43
while rigid restriction to objective indicators considerably narrows the perspective. Limits to Aggregation
Though objective counts are often quite useful for assessing detail, they are typically less helpful in charting the whole. For example, in assessing the quality of housing, objective indicators can help a great deal in quantifying aspects such as space, light and sanitation, but these aspect scores do not simply add into a meaningful overall estimate of dwelling quality. Social policy makers need indications for overall performance and hence there have been many attempts to combine piecemeal observations into a comprehensive index. Next to indexes of housing quality, there are counts for ‘livability’ of neighborhoods (Schoenmakers, 1999) and for ‘quality-of-life’ in nations (f.e. Kacapyr, 1996). These sum-scores fall short for the following reasons: Firstly they are typically selective and tend to focus on those aspects currently on the political agenda. Secondly, such indexes are incomplete, because they are limited to a few easily measurable aspects. Thirdly, all the items are mostly given equal weight, while it should be rather evident that the importance of aspects will vary. Fourthly, it is not acknowledged that weights also vary with satiation and that they are contingent to situations and personal capabilities. Together this greatly limits the comparability of such sum-scores across time and culture. I have analyzed these shortcomings in more detail elsewhere (Veenhoven, 1996, 2000a). Aggregation is less problematic with subjective indicators, because we can simply ask people about their overall judgement. Research has shown that people are quite able to strike a balance, both in life-domains such as housing and for their life-as-whole.2
CONCLUSION
Social policy makers need both objective and subjective indicators. The challenge of social reporting is to combine the strengths of these indicators and to make sense of the discrepancies they show. For some purposes objective indicators are best suited, for other uses
44
RUUT VEENHOVEN
subjective indicators are preferable. Assessments of overall lifesatisfaction is particularly needed to assess comprehensive policy success and to distinguish needs from want.
NOTES 1
Johansson (2001) warns that the democratic discourse should not be emptied by referendum like opinion polls, and stresses that social indicators research should focus on feeding the discussion with information on ‘how it is’, rather than counting opinions on ‘how it should be’. If this is an appeal to limit to objective indicators I don’t agree. Firstly because attitudes are typically part of ‘how it is’ and secondly because I don’t see harm in periodic counts of opinion on ‘how it should be’. 2 Subjective appraisals have sometimes been used to assign weights to items in objective sum-scores, mostly avowed value priorities and sometimes observed correlations with satisfaction. In fact that is a testimonial paupertatis. Rather than use subjective appraisals to construct a comprehensive index, one can better ask right away for an overall judgement.
REFERENCES
Andrews, F. and S. Withey: 1976, Social Indicators of Wellbeing: American Perceptions of Quality of Life (Plenum Press, New York). Boelhouwer, J. and I. Stoop: 1999, ‘Measuring wellbeing in the Netherlands’, Social Indicators Research. 48, pp. 51–75. Campbell, A, P.E. Converse and W.L. Rodgers: 1975, The Quality of American Life (Institute for Social Research, Ann Arbor). Frank, R.: 1999, Luxury Fever. Why Money Fails to Satisfy in an Era of Excess (The Free Press, New York). Glatzer, W. and W. Zapf: 1984, Lebensqualität in der Bundesrepublik (Quality of life in West Germany) (Campus, Frankfurt am Main, Germany) (English summary in Social Indicators Research, 1987 vol. 19, pp. 1–171). Inglehart, R.: 1990, Culture Shift in Advanced Industrial Society (Princeton University Press). Johansson, S.: 2001, ‘Conceptualizing and measuring Quality of life for national policy’, Social Indicators Research, this issue. Kacapyr, E.: 1996, Index of Wellbeing (American Demographics). Kanahan, D., E. Diener and N. Schwartz (eds.): 1999, Wellbeing, the Foundations of Hedonic Psychology (Russel Sage Foundation, New York). Katona, G.: 1975, Psychological Economics (Elsevier Scientific Publishers, Amsterdam, Netherlands).
WHY SOCIAL POLICY NEEDS SUBJECTIVE INDICATORS
45
Lane, R.: 2000, The Loss of Happiness in Marker Democracies (Yale University Press, USA). Manning-Gibbs, R.A.: 1972, Relative Deprivation and Self-reported Happiness of Blacks 1946–1966. PhD dissertation University of Texas, USA. Noll, H-H.: 1994, Zustand der offentliche Sicherkeit beintrachtegt Wohlbefinden der Bürger ISI-bulletin, nr. 12, pp. 5–8. Schoenmakers, A.: 1998, Leefbaarheidsmonitor Zeeburg (Bureau Onderzoek Op Maat (Boom), Erasmus Universiteit Rotterdam). Schwartz, N. And F. Starck: 1999, ‘Reports of subjective wellbeing: judgemental processes and their methodological implications’, in D. Kanahan, E. Diener and N. Schwartz (eds.), Wellneing, the Foundations of Hedonic Psychology (Russel Sage Foundation, New York, pp. 61–84). VanPraag, B.M., Th. Goedhart and A. Kapteyn: 1980, ‘The poverty line: A pilot survey in Europe’, Review of Economics and Statistics 63, pp. 461–465. Veenhoven, R.: 1996, ‘Happy life expectancy: a comprehensive measure of quality-of-life in nations’, Social Indicators Research 39, pp. 1–58. Veenhoven, R.: 1997, ‘Progrès dans la compréhension du bonheur’ (Advances in the understanding of happiness), Revue Québécoise de psychologie 18, pp. 29– 47. Veenhoven, R.: 1998, ‘Vergelijken van geluk in landen’ (Comparing happiness across nations) Sociale Wetenschappen 41, pp. 58–84. Veenhoven, R.: 1999, ‘World Database of Happiness: continuous register of research on subjective enjoyment of life’, Available: http://www.eur.nl/fsw/ research/happiness Veenhoven, R.: 2000a, ‘The four qualities of life: ordering concepts and measures of the good life’, Journal of happiness studies 1, pp. 1–39. Veenhoven, R.: 2000b, ‘Wellbeing in the welfare state: level not higher, distribution not more equitable’, Journal of comparative policy analysis 2, pp. 91–125. Veenhoven, R. and P. Ouweneel: 1995, ‘Livability of the welfare state, length of life and appreciation of life in nations varying in state welfare effort’, Social Indicators Research 36, pp. 1–48. Veenhoven, R. and W. Saris: 1996, ‘Satisfaction in 10 countries’, in Saris et al. (eds.), A Comparative Study of Satisfaction with Life in Europe (Eötvös University Press, Budapest, Hungary, pp. 223–231). Vile, M.J.: 1999, Politics in the USA (Routledge, London).
Erasmus University Rotterdam and University of Utrecht The Netherlands E-mail:
[email protected] This page intentionally left blank
HEINZ-HERBERT NOLL
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS: THEORETICAL FRAMEWORK AND SYSTEM ARCHITECTURE*
ABSTRACT. By developing a “European System of Social Indicators” the scientific community as well as policy makers are being provided with a theoretically well-grounded as well as methodologically sound tool of social measurement to be used to continuously monitor and analyse the development of welfare as well as general trends of social change in European societies. The article focuses on the development of a conceptual framework, to be used to guide and justify the selection of measurement dimensions and indicators. Based on an extensive review of theoretical concepts of welfare and an exploration of political goals of societal development at the European level, six major categories of goal dimensions, referring to the concepts of quality of life, social cohesion and sustainability have been identified as the backbones of the conceptual framework. In addition, the article presents the main features of the architecture of the European System of Social Indicators and lays out its main structural elements and characteristics.
INTRODUCTION 1
In recent years social reporting activities – efforts to monitor and systematically describe and analyse the current state of and changes in living conditions and the quality of life – have been given new priority. The process of European integration has obviously stimulated the development of such monitoring and reporting activities not only at the supranational, but also at national and sub-national levels. The improvement of living conditions and the quality of life in the member states are among the main goals of the European Union, as stated for example in the Maastricht treaty. Accordingly, the availability of appropriate knowledge and systematic information on social conditions within and across European societies as * This article has been written as part of the project “Towards a European System of Social Reporting and Welfare Measurement”, funded within the TSERProgramme of the European Commission. Social Indicators Research 58: 47–87, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
48
HEINZ-HERBERT NOLL
provided by social monitoring and reporting will be of crucial importance to enhance European integration and cohesion and to create the “Social Europe” of the 21st century. Establishing a science based European system of social reporting is the overall objective of the EuReporting project. As part of this project an “European System of Social Indicators” is going to be developed. As a result of our research the scientific community, policy makers as well as other potential users shall be provided with a theoretically as well as methodologically well-grounded selection of measurement dimensions and indicators to be used as an instrument to continuously monitor and analyse the development of welfare and quality of life as well as changes in the social structure of European societies and the European Union. To achieve these objectives, such an indicators system is supposed to meet certain requirements: coverage of the ‘European dimension’ (e.g. identity, cohesion) science based, concept driven approach incorporation of new dimensions of welfare and social change search for – with respect to validity and reliability – improved or new indicators making use of the best available data bases and ensuring comparability across national societies By constructing a system of social indicators one faces certain basic problems of measurement. First of all, there is the most basic question of what ought to be measured and monitored? In order to be able to provide a sound answer to this crucial question, a conceptual framework is needed, which specifies concerns and dimensions of measurement to be covered by the indicators system. Second, by developing a system of social indicators one has to determine structural elements and to define the procedures of measurement. This is the purpose of a systems architecture. And third, by constructing a system of social indicators, certain formal criteria need to be respected: The various parts of a system of social indicators need to be consistent, indicators shall be non-redundant, the system shall be comprehensive in terms of including all relevant dimensions of measurement and finally it shall also be parsimonious in the sense of using no more indicators than the number actually needed for appropriate measurement.
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
49
CONCEPTS OF WELFARE AND GOALS OF SOCIETAL DEVELOPMENT: ELEMENTS OF A CONCEPTUAL FRAMEWORK
As a first step in constructing such an indicators system, the development of a conceptual framework is of crucial importance. The purpose of such a framework is to guide and to justify the selection of measurement dimensions and indicators. As measures of welfare and social change social indicators are supposed to refer to societal goals and be relevant for the policy making process. Thus, the conceptual framework should determine the goal dimensions and political concerns to be covered by the indicators system. Since the indicators system is primarily supposed to serve the function of monitoring the attainment of welfare goals in Europe, the question arises, which concept of ‘welfare’ is going to be used, which components and dimensions are to be covered and to which goals of societal development it is related. In order to determine welfare components and the goals of societal development in Europe as the major elements of the theoretical framework, two kinds of analysis have been undertaken: an analysis of the goals of societal development as they are explicitly or implicitly considered by various concepts of welfare which have been developed and discussed within the social sciences as well as the political debate such as quality of life or more recent concepts like sustainability or social quality. an exploration of goals of welfare and societal development as they are expressed at the level of European politics. In order to identify these political goals an analysis of European social concerns, common objectives and goals of the EU member states as indicated in the European Treaties (Rome, 1957; Maastricht, 1992; Amsterdam, 1997) and official documents of the European Commission has been carried out. Concepts of Welfare
The main purpose of a European System of Social Indicators as being developed in our project is the measurement and monitoring of the level of and changes in the welfare of European citizens. There are different notions of what constitutes a ‘good life’ or a ‘good society’ and correspondingly different concepts of welfare
50
HEINZ-HERBERT NOLL
have been developed. Only a few decades ago, the notion of welfare still used to be synonymous with material wealth, and rates of economic growth turned out to be the main criteria for assessing social progress. Later on a broader conception of welfare emerged, which also included non-material and qualitative aspects of development, and thus quality of life became the leading welfare goal and perspective of societal development (see below). Among the welfare concepts included into our analysis, the concept of quality of life is probably the most widely recognised and the most frequently used framework for analysing changes of welfare across time and inequalities of welfare within a society. It has stimulated much research on empirical welfare measurement. Various approaches of operationalisation are to be distinguished. Each approach reveals a different notion of the concept and thus highlights different ideas on relevant components and dimensions of welfare. The more recent welfare concepts discussed here – liveability, social cohesion, social exclusion, social capital, human development, sustainability, social quality – are less approved so far. They are still characterised by deficiencies of empirical operationalisation and partially also of theoretical elaboration and clarification. Especially, the relationships among these “new” concepts as well as their relations to the quality of life concept have not been sufficiently clarified yet, although implicit linkages are obvious. Quality of Life The concept of quality of life arose at the end of the 1960s as an alternative to the by that time dominant societal goal of an increasing material level of living. Besides material dimensions of welfare, the concept encompasses immaterial aspects of the living situation like health, social relations or the quality of the natural environment. Moreover, quality of life was supposed to include objective features – the actual living conditions – as well as the subjective well-being of the individual citizens (Argyle, 1996).2 Among the various efforts to operationalise the quality of life concept, two rather contrary approaches are to be distinguished (Noll and Zapf, 1994): the Scandinavian level of living approach (Erikson, 1993; Uusitalo, 1994) and the American qualityof-life approach (Campbell, Converse and Rodgers, 1976). The
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
51
Scandinavian approach focuses almost exclusively on resources and objective living conditions, whereas the American approach emphasises the subjective well-being of people as a final outcome of conditions and processes. These distinctive views are the result of alternative conceptualisations of the idea of quality of life, welfare and not the least a good society. Besides the contrasting notions of quality of life as ‘individuals’ command over resources’ on the one side and as subjective wellbeing on the other, there have been broader conceptions and operationalisations of quality of life which include objective as well as subjective elements. The consideration of objective as well as subjective indicators is nowadays the prevailing research strategy. Such a broader concept of quality of life was taken as the basis of Erik Allardt’s “Comparative Scandinavian Welfare Study” as early as in 1972. This approach distinguishes between three basic needs of human beings – Having, Loving and Being (Allardt, 1973). Within each category, objective as well as subjective dimensions of need satisfaction are included. Another approach combining objective as well as subjective dimensions is based on the German notion of quality of life focussing on the constellation of objective living conditions and subjective well-being across different life domains (Zapf, 1984). Objective living conditions include the ascertainable living circumstances, such as living standards, working conditions or the state of health. Subjective well-being concerns general as well as domain-specific assessments and evaluations of living conditions and includes cognitive as well as affective components. The following typology of welfare positions distinguishes between four constellations of objective living conditions and subjective well-being:
The constellation of good living conditions and positive subjective well-being is called well-being. The combination of good
52
HEINZ-HERBERT NOLL
living conditions and negative subjective well-being is denoted as dissonance. Poor living conditions coinciding with low subjective well-being represents a situation of deprivation. And finally, poor living conditions but nevertheless high subjective well-being is described as adaptation (Zapf, 1984: pp. 25–26). In trying to determine the most distinctive features of the various conceptualisations of quality of life as presented above, one has to differentiate first between objective and subjective notions. Secondly, there are different conceptions of objective as well as of subjective approaches. Concerning the objective dimensions of quality of life, one can distinguish the idea of resources or capabilities (Sen, 1993) as means to enhance the quality of life from the focus on living conditions as the outcomes or end states of societal processes. Within the category of subjective assessments of quality of life, an important distinction is that between cognitive and affective components of well-being. A common feature of all approaches is the more or less implicit or explicit conceptualisation of quality of life as concerning individual characteristics. Dimensions of welfare related to societal characteristics and qualities such as equality, equity, freedom, or solidarity – which affect the welfare situation of individuals at least indirectly – have been rather neglected, at least as far as empirical measurement and research is concerned, although they have been part of the early notions of the concept of quality of life. In contrast to this, the more recent welfare concepts – as they are subject of the following section – put the focus more explicitly on aspects concerning the quality of societies, the distribution of welfare and social relations within societies. Concepts of the Quality of Societies The concepts referring to welfare related characteristics of societies, such as distributional and relational aspects, have become popular mainly during the second half of the 1980s and during the 1990s. Some of these theoretical approaches are rather comprehensive, such as the concept of Human Development and the most recent concept of Social Quality. Others, as for example the concepts of Social Exclusion and Social Capital, focus primarily on special welfare issues. There is a substantial overlap between these concepts
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
53
and some of the ideas dealt with are not really new but have also been part of early notions of the quality of life concept. Liveability and the Quality of Nations The concept of liveability has been introduced by Veenhoven (1996, 1997) as a performance criterion of societies. It is defined “as the degree to which its provisions and requirements fit with the needs and capacities of its citizens” (Veenhoven, 1996: p. 7). Two approaches to operationalise this concept have been discussed (Veenhoven, 1996: p. 17 ff). The first approach is measurement by so-called input indicators which refer to living conditions of a society and which are expected to match the citizens’ needs and capacities, such as wealth, political freedom, equality, access to education. Two problems have been identified with this approach: (1) the implicit assumption on human needs and capacities and (2) the assumption that the more of a condition the better the liveability. Thus, an alternative approach has been proposed, the measurement by so-called output indicators which are supposed to capture the degree to which people “flourish” in a society. Indicators of physical and mental health, overall satisfaction and happiness are proposed as appropriate measures of “flourishing” (Veenhoven, 1996: p. 12). It is assumed that a good health and a positive appraisal of life are outcomes of societal conditions which meet the citizens’ needs and capacities and thus can be regarded as indicators of the liveability of a society. As far as the relation between the concepts of liveability and quality of life is concerned, liveability on the one hand is considered as a characteristic of a good society, on the other it has been denoted as quality of life within a society and thus been defined with a clear reference to individual characteristics such as needs and capacities. Hence, the concept of liveability comes fairly close to the notion of quality of life as an end state, measured by indicators of subjective well-being. Social Cohesion, Social Exclusion, Social Capital The concepts of social cohesion, social exclusion and social capital are all closely related to each other, and there are further concepts such as social inclusion, social integration, and civil society which could be mentioned in this context as well. Referring to Emile
54
HEINZ-HERBERT NOLL
Durkheim, all these concepts can be seen as being primarily concerned with the possibilities and preconditions of societal integration and solidarity (Noll, 1999: p. 19). Common to all concepts is the concern with the interrelations between units of the society such as individuals, groups, associations, institutions as well as territorial units. Among these concepts social cohesion embodies the most comprehensive perspective, which includes aspects addressed by the concepts of social exclusion and social capital, too. Social Cohesion is considered to be a characteristic of a society dealing with the relations among members of that society and the bounding effect of these relations (McCracken, 1998). Among the notions mentioned in describing social cohesion are shared values and communities of interpretation, feelings of a common identity, a sense of belonging to the same community, trust among individuals as well as towards institutions and not the least the reduction of disparities (Woolley, 1998; Jenson, 1998b). The Social Cohesion Network of the Policy Research Initiative of the Canadian Government has promoted a definition of social cohesion as “the ongoing process of developing a community of shared values, shared challenges and equal opportunity within Canada, based on a sense of trust, hope and reciprocity among all Canadians” (PRI, 1999: p. 22). Emile Durkheim turns out to be the first scholar who discussed and made use of the concept of social cohesion. He considered social cohesion as an ordering feature of a society and defined it as the interdependence between the members of the society, shared loyalties and solidarity (Jenson, 1998b). In recent years the concept of social cohesion received great attention by policy circles at the national and supranational level. Besides the Canadian government, the French and the Dutch Government, the OECD, the Council of Europe, the European Commission, and the Club of Rome have Concerned themselves with issues of social cohesion (Jenson, 1998b; PRI, 1999). Moreover, the British Liberal Party established a “Commission of Wealth Creation and Social Cohesion” directed by Ralf Dahrendorf (Noll, 1999: p. 21). The increasing popularity of the concept is most likely due to various aspects of economic and social change which are currently considered to threaten the social cohesion of societies
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
55
such as rising income inequality, poverty, unemployment, and crime (Jenson, 1998b). As a detailed review of the literature reveals (Berger-Schmitt and Noll, 2000), the concept of social cohesion incorporates mainly two goal dimensions of societal development which may be related to each other but should be distinguished though analytically: The first dimension concerns the reduction of disparities, inequalities, fragmentations and cleavages which have also been denoted as fault lines of societies. The concept of social exclusion is covered by this notion too. The second dimension embraces all forces strengthening social connections, ties and commitments to and within a community. This dimension includes the concept of social capital. Towards the end of the 1980s the concept of social exclusion has become more and more popular. In recent years, it has represented one of the most widely used concepts in scientific and political debates on social issues and has inspired a large amount of literature. The popularity of the concept was particularly promoted through the growing interest in matters of social exclusion at the level of the European Union. The origin of the concept can be traced back to France, where the term has been used in the context of debates on a new poverty and defined as a rupture of the relationship between the individual and the society (Silver, 1994; Rodgers, Gore and Figueiredo, 1995; de Haan, 1999). In contrast to the concept of poverty, social exclusion refers not only to a situation, but focuses attention also to the processes and causes and thus represents a more analytical concept (Rodgers, 1995; Berghman, 1995; de Haan, 1999). According to Silver (1994), the different meanings and research perspectives derived from the concept of social exclusion are related to three basic paradigms: Within the framework of the first paradigm, ‘solidarity’, the term social exclusion is used in the sense of the French research tradition. It is defined as a disruption of the social ties between society and the individual due to the failure of institutions to integrate individuals into the society. Within the second paradigm – ‘specialisation’ – social exclusion has been defined from the perspective of the Anglo-Saxon
56
HEINZ-HERBERT NOLL
research tradition (de Haan, 1999). It is seen as a result of social differentiation and specialisation, of the individuals’ diversity of interests and capabilities. Contrary to the solidarity paradigm, social exclusion is considered to be caused by changes of the social structure as well as individual behaviour. Individuals may participate in some domains and be excluded from others due to their voluntary choices, the interests of other actors, contractual regulations, and notably also due to discrimination. The third paradigm, ‘monopoly’, highlights that society is ordered hierarchically with different groups controlling access to goods and services and protecting resources from outsiders. In this context, social exclusion is the result of processes of social closure by which more privileged groups protect their monopoly position. Contrary to the solidarity paradigm, the society is characterised by a hierarchy of inclusions and exclusions rather than a dualism of excluded and included (IILS, 1998). According to the theoretical approach of the ‘European Observatory on National Policies to Combat Social Exclusion’ and of the ‘European Poverty 3 Programme’ social exclusion is defined in terms of the denial of citizenship rights – civil, political and social rights – which major societal institutions should guarantee. Thus, social exclusion should be conceptualised as the failure of one or more of the following four systems: the democratic and legal system which promote civic integration the labour market which promotes economic integration the welfare state system promoting what may be called social integration the family and community system which promotes interpersonal integration (Berghman, 1998: pp. 258–259) The concept of social capital covers topics like the density and quality of relationships and interactions between individuals or groups, their mutual feelings of commitment and trust due to common values and norms, a sense of belonging and solidarity which are supposed to be the fundamentals of the internal social coherence of a society (McCracken, 1998; Woolley, 1998; Jenson,
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
57
1998b; O’ Connor, 1998). “The social capital of a society includes the institutions, the relationships, the attitudes and values that govern interactions among people and contribute to economic and social development. Social capital, however, is not simply the sum of the institutions which underpin society, it also makes up the glue that holds them together. It includes the shared values and rules for social conduct expressed in personal relationships, trust, and a common sense of “civic” responsibility, that makes society more than a collection of individuals. Without a degree of common identification with forms of governance, cultural norms, and social rules, it is difficult to imagine a functioning society” (Social Capital Initiative, 1998: p. 1). The concept of social capital has been defined in different ways by various scholars looking at it from different perspectives (Coleman, 1988; Putnam, 1995; North, 1990).3 But all of them have in common that they regard social capital as a property of a social entity and not of individuals. As a relational concept, social capital exists only as far as it is shared by several individuals. Thus, it cannot be regarded as an individual characteristic, but rather shows the character of a public good (Grootaert, 1998; Immerfall, 1999; Narayan, 1999). Corresponding to the different scopes of the concept, a distinction between three levels of manifestation of the concept has been made (Immerfall, 1999: pp. 121–122): the level of interpersonal relations, such as family, friends, neighbours the level of intermediary associations and organisations, such as clubs, firms, political parties the macro-level of societal institutions Regardless of the perspective taken one can conclude that social cohesion, social exclusion and social capital represent important welfare components which are merely being covered by the quality of life concept but rather need to be incorporated into the monitoring perspective of a European System of Social Indicators.
58
HEINZ-HERBERT NOLL
Sustainability During the 1990s the concept of sustainability has become the dominant model of societal development. There is a general consensus that the achievement of sustainable development ought to belong to the key priorities of local, regional, national and supranational policies. From a general point of view, the concept of sustainability can be seen as a new answer to the traditional concern with a balanced and harmonious societal development (Noll, 1999). The concept became popular in 1987 as the central message of the so-called Brundtland-Report “Our Common Future” of the World Commission on Environment and Development, where it was defined as “development that meets the needs of the present without compromising the ability of future generations to meet their own needs” (World Commission on Environment and Development, 1987: p. 43). In general, three dimensions of sustainable development have been distinguished: an environmental, a social dimension, and an economic dimension which are supposed to be linked to each other (OECD, 1998a,b; Wiman, 1999). Every policy guided by the principle of sustainability should take into account its impacts on the economic, the social and the environmental dimensions; it should ensure the continued preservation of the economy and the society without destroying the natural environment on which both depend. Actually, sustainable development proposes a new paradigm of decision making for all sectors of society. In order to achieve a sustainable development, environmental policies need to be socially and economically feasible, social policies need to be environmentally and economically feasible, and economic policies need to be socially and environmentally feasible (Bell, Halucha and Hopkins, 1999: p. 3). Thus, sustainability has been defined “as a continuous striving for the harmonious co-evolution of environmental, economic and socio-cultural goals” (Mega and Pedersen, 1998: p. 2). Among the various attempts to conceptualise and to operationalise sustainable development,4 the World Bank’s Multiple Capital Model (World Bank, 1997) is one of the most well known and widely recognised approaches. Within this approach sustainable development is conceptualised with reference to national wealth
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
59
and denotes the maintenance or enhancement of wealth for future generations. The World Bank approach distinguishes between four components of wealth: natural capital: the stock of environmental assets, such as land, water, wood, minerals, flora and fauna, which corresponds to the environmental dimension of sustainable development; produced/man-made capital: the stock of machinery, factories, buildings, and infrastructure, such as railways, roads which represents the economic dimension of sustainable development; human capital: people’s productive capacities based on skills, education, health which constitutes – together with the social capital – the social dimension of sustainable development; social capital: social networks, associations and institutions tied by common norms and trustful relationships that facilitate cooperation. From this perspective, the goal of passing on to the next generation at least as much natural, economic, human and social capital as the current generation has at its disposal is at the centre of the idea of a sustainable development (Hardi and Barg, 1997; OECD, 1998c). Concerning the relations between the various forms of capital, there is a debate about the possibilities of mutual substitutions between them. Advocates of a ‘weak sustainability’ concept claim, that all forms of capital can be substituted by each other, whereas supporters of a ‘strong sustainability’ concept deny that such a substitution is feasible, especially as far as the natural capital is concerned for which the possibilities of substitution are considered to be limited (Pearce, 1993; Pearce and Warford, 1993). It is argued, that natural capital to some extend fulfils life supporting functions which cannot be substituted at all by other forms of capital (Munasinghe and McNeely, 1995). It is obvious that the concept of sustainable development, especially in the specification of the World Bank’s four capital model, is clearly related to the concepts of social cohesion, social exclusion, and social capital. The notion of sustainability strongly emphasises the ideals of equal opportunities, equity and solidarity both within and between generations. These are also aspects addressed by the concept of social cohesion with the exception that it’s perspective
60
HEINZ-HERBERT NOLL
does not extend to the future. Likewise the creation and preservation of social capital is a goal dimension covered by the idea of social cohesion as well as of sustainability, but the focus of the sustainability concept is on the preservation of social capital for future generations. Thus, sustainability is a more comprehensive concept than social cohesion, because it includes a wider range of issues. As far as the relationship between sustainability and quality of life is concerned, one could consider as a major difference between the two concepts that quality of life is explicitly concerned with the individual welfare in actual life domains, whereas sustainability represents a general principle of acting which refers to collective or societal properties, such as equality, equity and the preservation of nature. This leads to the question, whether the idea of sustainability is compatible with the goal of improving the quality of life and if so, whether quality of life should be treated as a component of sustainable development or whether sustainability ought to be considered as a subdimension of quality of life (Noll, 1999: p. 15). Various authors have held the view that the overarching goal of sustainable development ultimately is to increase the quality of life for all people, not only of present but also of future generations (Wiman, 1999; Hart, 1998–99; OECD, 1998b). Hence, a major difference between the goals of quality of life and sustainable development can be seen in the emphasis put on intergenerational equity (IISD, 1998: pp. 1–2). Thus, sustainability considerations are essential for ensuring the quality of life of future generations. On the other hand, the concept of sustainability does not claim to develop a comprehensive formula of the ‘good life’ as the quality of life concept does. Human Development The concept of human development was originally developed by Miles (1985) in the framework of a development project of the United Nations University.5 It was further elaborated and became well-known in the context of the United Nations Development Programme (UNDP), where it was particularly influenced by the ideas of Amartya Sen, winner of the 1998 Nobel Prize for Economics, and Mahbub ul Haq, the former head of the Human Development Report Office.
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
61
In the first Human Development Report, human development has been defined as “a process of enlarging people’ s choices” (UNDP, 1990, p. 1). Three factors – basic capabilities – are considered as particularly important, because they strongly determine the range of available choices and opportunities: health, education/knowledge and access to resources needed for a decent standard of living.6 The concept of human development has been continuously refined, as reflected in subsequent editions of the Human Development Report. The various aspects and components of the concept have been explained in detail and new dimensions have been added: As early as 1990, human freedom was recognised as an essential precondition for exercising choices: “Human development is incomplete without human freedom” (UNDP, 1990: p. 16). In 1992, the Conference on Environment and Development in Rio de Janeiro and the agreement on Agenda 21 influenced the concept and sustainability was adopted as an important dimension of human development. Equality of opportunities for all people7 and intergenerational equity have been particularly emphasized. A further essential component of the concept is the empowerment of people. People should be qualified for participating in economic, social, and political activities and decisions that are relevant to them, since “greater participation enables people to gain for themselves access to a much broader range of opportunities and thus involves widening their choices” (HDR, 1993: p. 21). Besides human freedom the aspect of human security has been added. People should be able to exercise their choices freely and safely (HDR, 1994). Another important condition is economic growth which is regarded as a means to human development. However it has been stressed “that there is no automatic link between growth and human development” (HDR, 1996: p. 1). It has been explicitly emphasised that the human development paradigm puts people at the centre of its concerns and ultimately aims to improve human well-being. Thus, human development actually represents a welfare concept focussing on the individual who is perceived as an active agent and participant rather than as a passive
62
HEINZ-HERBERT NOLL
beneficiary in the development process (Doraid, 1997). However, the scope of the human development concept goes beyond individual welfare and also embraces supra-individual qualities such as equality of opportunity, equity, and solidarity. Although the concept of human development hardly incorporates any goal dimensions of welfare not yet covered by other welfare concepts previously discussed, it should be underlined that this approach directs the attention to concerns such as freedom, security and the empowerment and participation of people, which are sometimes neglected by other approaches. An essential merit of the concept of human development is certainly its broad and comprehensive perspective which successfully integrates individual and societal dimensions of welfare. Social Quality An equally comprehensive notion of welfare, which has been promoted only recently, is represented by the concept of social quality. The concept has been elaborated by the European Foundation on Social Quality which has been established under the Netherlands Presidency of the European Union in 1997. The Foundation has framed the “Amsterdam Declaration on the Social Quality of Europe” which has been signed by European social scientists in order to call attention to the attainment of the social objectives as part of the European Treaties (Beck, van der Maesen and Walker, 1998). Social Quality is defined “as the extent to which citizens are able to participate in the social and economic life of their communities under conditions which enhance their well-being and individual potential” (Beck, van der Maesen and Walker, 1998a: p. 3). The social quality experienced by citizens is considered to be based on four conditions:
the degree of socio-economic security; the extent of social inclusion; the strength of social cohesion and solidarity between and among generations; the level of autonomy and empowerment of citizens. The four conditions or components of social quality have been characterised with respect to two dimensions which constitute the
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
63
axes of the so-called social quality quadrant (Beck, van der Maesen and Walker, 1998b). The first dimension concerns the distinction between the micro-level (individual) and the macro-level (social structures); the second dimension concerns the distinction between institutions/organisations and communities/groups/citizens. The concept of social quality refers only to welfare concepts discussed previously and does not add further dimensions not yet covered. However, social quality has to be conceived not primarily as a new conceptualisation of welfare, but rather as an effort to integrate the ideas of social cohesion, social exclusion and human development under a common policy perspective. The primary purpose of the concept of social quality is to foster a discussion on issues of social quality in Europe and to direct attention of policy makers to the social dimension of the process of European Integration. Goals and Objectives of European Policies
Values and goals of societal development are not only dealt with on a conceptual level within the social sciences, but they are also part of political programmes and measures. The reference to agreed upon societal goals as well as political relevance are major characteristics of social indicators. Thus, social indicators are frequently considered as measures of goal attainment. In developing the conceptual framework of a European System of Social Indicators, this requirement can be fulfilled by considering the goals and objectives tackled by current policies of the European Union. These goals and objectives are agreed upon by the different Member States and – since they are ultimately the result of democratic decision processes – they may also be considered as common concerns of the majority of European citizens. By integrating these concerns into the welfare model, which is taken as starting point for elaborating a European System of Social Indicators, this indicator system is not only supposed to monitor the development of welfare in Europe, but will also serve the function to measure progress towards political goals and specific targets.8 The goals of European policy are first of all documented in the Treaty establishing the European Community (Rome, 1957), the Treaty on European Union agreed upon at Maastricht in 1992 and
64
HEINZ-HERBERT NOLL
in the amendments made by the Treaty of Amsterdam in 1997. Furthermore, there is a wide range of other official documents by the European Commission – White Papers, Action Programmes, Communications – which outline general and specific objectives of European policies. As the result of analysing these documents a large set of policy concerns has been identified. Some of them are formulated at a rather general level: The promotion of economic and social progress, the improvement of living and working conditions, the increase of the standard of living and the quality of life, the fight against social exclusion, the strengthening of economic and social cohesion, the promotion of equal opportunities, the commitment to the principle of sustainability are general goals which have been strongly emphasised. Other objectives are being articulated more precisely on a rather concrete level. By trying to classify the various objectives into broader categories, which are related to some of the welfare concepts previously considered, three main categories have been distinguished each covering several policy areas and specific issues (Figure 2). The first main category contains objectives aiming at the economic and social progress and the improvement of people’s living conditions and quality of life. The second category is concerned with issues of strengthening the economic and social cohesion. In a sense, the objectives of the second category are ultimately also directed towards enhancing people’s quality of life, but the difference to the first category is their focus on distributional and relational concerns. The third category covers goals related to the principle of sustainability. As to the first category, the promotion of employment and the combatment of unemployment are top priorities of European social policy (see Figure 2). There is a strong focus on these objectives in the European Treaties as well as in the “White Paper on Growth, Competitiveness, and Employment”, the “White Paper: European Social Policy” and other documents. The enhancement of education, initial vocational training and especially continuing training as well as improvement and adaptation of qualifications – that is life-long learning – are considered to represent important means to achieve the employment objectives and to increase the competi-
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
65
tiveness of the European Union in a global perspective. The same impact is ascribed to the objective of promoting the use of information and communication technologies and the acquisition of the
66
HEINZ-HERBERT NOLL
respective knowledge. The need for these investments in human capital is emphasised in particular in the “White Paper on Growth, Competitiveness, and Employment”, the Green Paper on “Living and Working in the Information Society: People First”, the “Social Action Programme 1998–2000”, and the “Agenda 2000”. The improvement of public health is another aspect of social progress which ranks high on the political agenda of the European Union. This is firstly reflected in the claim for an improvement of health and safety at work and a reduction of work accidents and occupational diseases, which can be found in the European Treaties and “The Social Action Programme 1998–2000”. Security and safety are further aspects treated in European policy documents. The social security of people has been mentioned in the European Treaties as social concern of the European Union. The urgent need to modernise and to improve the social protection systems has been explained in many other documents. An improvement of personal safety should be achieved in the area of crime prevention, the protection of data, and in the area of transport. Also included in the European policy objectives are the reduction of environmental pollution and the improvement of environmental protection. These goals are especially mentioned in the context of discussing the need for a sustainable development. The second category of policy objectives – strengthening the economic and social cohesion – concerns, generally speaking, the development of relations between people or groups of people. The previous discussion on the concept of social cohesion suggests to distinguish two main aspects: the reduction of inequalities and the strengthening of ties. Objectives which fit into the category “Reduction of Economic and Social Disparities between Regions and Social Groups” are improving economic and social conditions of backward regions, promoting equal opportunities and reducing social exclusion. The reduction of regional disparities is addressed by the EU’s Structural Funds. The main goal of the Structural Funds is twofold: first, the funds support the development of areas which are lagging behind or are being affected by industrial decline; second, the funds facilitate the structural adjustment of rural areas through special programmes and measures covering the improvement of human
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
67
resources, economic conditions and infrastructure. The “Treaty on European Union” requires from the Commission to prepare a “report on economic and social cohesion” every three years in order to monitor respective progress. A very prominent concern of European social policy, which has been strongly emphasised especially in recent years, is the objective of gender equality. Equal opportunities of women and men are a goal in the realm of the labour market – remuneration, sex segregation, qualification – and in regard to the compatibility of occupational and family responsibilities. Furthermore, the promotion of equality with respect to participation and decision-making in political and other public realms are part of the policy goals. The White Paper on European Social Policy has paid much attention to this issue and has called for an annual Equality Report from 1996 onwards. The idea of mainstreaming has also been stressed within the “New European Community Disability Strategy” (European Commission, 1996b), which considers issues of equal opportunities for disabled people not separately, but instead as integrated elements in mainstream policies. The goal of equal opportunities for disadvantaged groups in the population is strongly connected with the objective to combat social exclusion and discrimination and to promote the socioeconomic integration of excluded groups. Since the middle of the 1980s matters of social exclusion have gained growing attention in European social policy, as reflected in the shift of the focus of the Commission’s poverty programmes from poverty to social exclusion, the establishment of an Observatory on National Policies to Combat Social Exclusion in 1989, and in the growing research activities on the measurement of social exclusion launched by the Commission. In the European Treaties the fight against social exclusion and all forms of discrimination are explicitly mentioned as a policy concern, and the goal of an inclusive society belongs to the three main topics of the Social Action Programme 1998–2000. A second aspect of economic and social cohesion in European Policies concerns the strengthening of connections and relations between people and regions. This includes the strengthening of social ties, as for example the general objective of encouraging
68
HEINZ-HERBERT NOLL
solidarity between people which has been stressed in the White Paper on Growth, Competitiveness, and Employment. There are several specific European concerns which refer to the relationships between Member States. A strengthening of feelings of solidarity and of a common European Identity are among the major concerns of the European Union, and there is a growing debate on the importance of a European citizenship and the development of a European constitution for the promotion of this objective (Welsh, 1993; Schäfers, 1999). Within the Treaty of Maastricht the notion of a European citizenship and respective civil and political rights have been outlined for the first time. A strengthening of the cohesion between the Member States is also intended by the objective of developing a European dimension of education and training which has been formulated in the White Paper on Growth, Competitiveness, and Employment. The knowledge of European languages, the mutual recognition of qualifications and skills, and the exchange of pupils and students promoted through EU programmes like Leonardo da Vinci or Socrates are considered to form important aspects of this dimension. Not the least, European cohesion is supposed to be further promoted by building a European labour market which turns out to be another fundamental goal of European Policy as – for example – stated in the White Paper on Growth, Competitiveness, and Employment and in Agenda 2000. The third category of policy objectives is related to the commitment of sustainable development. The challenge of a sustainable Europe is to achieve economic growth based on higher employment rates, reduced environmental pollution and improved resource efficiency of energy and raw materials. The “White Paper on Growth, Competitiveness, and Employment”, the “Communication from the Commission on Environment and Employment – Building a Sustainable Europe” as well as the Agenda 2000 outlined these goals in greater detail. It can be easily recognised that many of the welfare goals addressed by the various concepts discussed in the previous chapter are also being emphasised by European policies. However, the scientific use of these concepts is much broader and hence they cover more issues of relevance than those considered by policy
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
69
objectives. Several important components of quality of life are not yet or insufficiently taken into account by European policies as for example the living situation of families, housing conditions of the population, income and living standard, and participation and social relations as major determinants of social capital and social cohesion. Nevertheless, there are clear similarities between both perspectives and their integration into a common conceptual framework is rather obvious. A more precise definition of goal dimensions and an increase in the relevance of the resulting indicators system for policy use are among the advantages of including the policy objectives into the conceptual framework. The Conceptual Framework: Notion of Welfare and Goal Dimensions Considered As it turned out from the previous review of theoretical concepts of welfare and the analysis of goals of societal development at the level of European politics, our theoretical framework will be mainly based on the ideas of quality of life, social cohesion and sustainability. As has been shown before, these are also concepts which play a major role at the level of European politics and many policy objectives are closely related to them. Our main point of departure is the quality of life concept which focuses at dimensions of welfare at the individual level. This concept is considered to be still appropriate to cover current and future issues of individual welfare. However, we are proposing to widen the perspective by taking into account not only dimensions of individual quality of life but also dimensions of the quality of societies, as they are addressed by the more recent concepts of social cohesion and sustainability. The concept of quality of life is being used in terms of the rather comprehensive definition of the German approach as the constellation of objective living conditions and subjective well-being. By objective living conditions we include all aspects of the living situation which are relevant for the welfare of the individual regardless whether they are considered to be outcomes, resources, capabilities, or external circumstances. No effort is made to distinguish between these categories, since it often simply depends on the point of view whether certain living conditions represent outcomes or resources. The notion of subjective well-being embraces affective
70
HEINZ-HERBERT NOLL
and cognitive, positive and negative components. The principal goal dimensions extracted from the quality of life concept are then the improvement of objective living conditions of individuals as well as their subjective well-being in various life domains. Referring to the general objective of promoting economic and social cohesion in Europe, we are distinguishing two main goal dimensions as suggested above: (1) the reduction of disparities and inequalities, including social exclusion, and (2) the strengthening of connections and social ties including the enhancement of social capital. For each of the two main dimensions of the social cohesion/social capital perspective various subdimensions are to be distinguished, such as the reduction of regional disparities and the promotion of equal opportunities, the promotion of social and political participation and voluntary activities in networks and associations; the formation and strengthening of social relations between population groups or the improvement of the quality of relations including issues such as shared values, a common identity, trust, and solidarity. The European System of Social Indicators will also take into account the goal of sustainable development which is conceptualised with reference to the World Bank’s four capital approach. Thus, there are four major goal dimensions: the enhancement or preservation of social, human, produced, and natural capital. For each type of capital two aspects ought to be distinguished9: (1) the preservation or enhancement of the societal capital of current generations and (2) the provision for future generations. The first refers to the goal of promoting living conditions of the present generations, while the latter focuses on the means to preserve the societal capital for future generations, that is on the processes and measures necessary to secure equivalent living conditions for the future. This latter aspect actually represents the primary idea of sustainability which has to be conceived as a general principle shaping societal developments.10 The European System of Social Indicators is not only supposed to be a tool for measuring welfare and goal achievement, but also to monitor more general trends of social change, to register progress in modernisation and the related problems and consequences. Thus, the European System of Social Indicators shall also provide information on trends of social change concerning major elements of the
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
71
socio-economic and socio-demographic structure as well as trends of social change concerning individual values and attitudes. As far as the latter are concerned, relevant changes to be monitored by a European System of Social Indicators are – for example – changes in value orientations from materialistic to postmaterialistic values as well as changes in gender roles or party preferences to mention just a few. Dimensions of the social structure to be monitored by our indicators system are – for example – changes of the age structure, changes in the sectoral structure of employment or changes in the class structure.
Thus, the conceptual framework of the European System of Social Indicators overall results in two perspectives and two levels of measurement (Figure 3). The two perspectives of measurement are the measurement of welfare on the one hand and monitoring general social change on the other. For both of them an individual level and a societal level is being distinguished. Welfare measurement at the individual level addresses objective living conditions and subjective well-being as the two principal goal dimensions of the individual quality of life. Welfare measurement at the societal level covers several dimensions of the quality of a society as they have been derived from concepts as sustainability and social cohesion. Monitoring social change at the individual level puts the emphasis
72
HEINZ-HERBERT NOLL
on measuring changes in individual values and attitudes whereas monitoring of social change at the societal level will focus on the observation of socio-structural trends.
ELEMENTS OF THE ARCHITECTURE OF THE EUROPEAN SYSTEM OF SOCIAL INDICATORS
Life Domains, Goal Dimensions and Measurement Dimensions
The European System of Social Indicators will first of all be structured by life domains and goal dimensions as well as more general dimensions of socio-structural change as outlined above. The life domains considered correspond to some extend to the European policy concerns, but other domains, which also constitute important aspects of quality of life, have been added as well. Besides various domains of life, the total living situation will be included as well and covered by comprehensive measures, such as welfare indices or global evaluations. Thus, the European System of Social Indicators covers the following 14 life domains or modules: Population Household and Family Housing Transport Leisure, Media and Culture Social and Political Participation and Integration Education and Vocational Training Labour Market and Working Conditions Income, Standard of Living, Consumption Patterns Health Environment Social Security Public Safety and Crime Total Life Situation The conceptual framework outlined before determines perspectives (welfare measurement; monitoring of social change) and levels of measurement (individual level, societal level) as well as the dimensional structure of the European System of Social Indica-
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
73
tors. For each life domain, the following goal dimensions are being distinguished:11 improvement of objective living conditions enhancement of subjective well-being reduction of disparities, inequalities and social exclusion, promotion of equal opportunities strengthening social connections and ties – social capital preservation of human capital preservation of natural capital Besides the attainment of societal goals the European System of Social Indicators will also cover the major elements of the social structure and related changes and thus include demographic and socio-economic developments as well as changes in values and attitudes. For each goal dimension within a life domain, appropriate measurement dimensions will be derived. Figure 4 only presents the main categories of measurement dimensions for the life domain “Labour Market and Working Conditions”. The measurement dimensions derived from the goals of improving objective living conditions and enhancing subjective well-being concern characteristics of the life situation which are neither related to the objective of strengthening connections and social ties nor to the principle of preserving societal capital for future generations. These dimensions are related to the state of living conditions and personal well-being, such as the state of health, the level of education, environmental conditions or satisfaction with life. The goals of preserving human capital and natural capital have been operationalised by dimensions which refer to the measures and processes fostering these goals. These dimensions point to factors that influence the goal attainment such as the efficiency of energy consumption, preventive measures in the area of health, investments in education. Usually measurement dimensions will be further broken down into subdimensions as shown in Figure 5 by example of the life domain “Labour Market and Working Conditions” and the goal dimension “Improvement of Objective Living Conditions”. The measurement dimension “working conditions” covers – for example – the subdimensions “working hours”, “earnings” and
74
HEINZ-HERBERT NOLL
“work environment and job content”. The measurement dimension “unemployment and underemployment” is broken down into the subdimensions “level of unemployment”, “duration of unemployment”, “subsistence of the unemployed” as well as “level of underemployment”.
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
75
To sum up, the European System of Social Indicators covers 14 life domains (Figure 6). Within each life domain up to six dimensions of welfare and two dimensions of general social change are being distinguished. At a third level there are dimensions of measurement and at a fourth level subdimensions, which are going to be operationalised by one or more indicators each. Indicators
The European System of Social Indicators includes indicators of goal attainment as well as more general indicators of social change. The former are supposed to be direct measures of individual and societal welfare and thus – according to Mancur Olson – “subject to the interpretation that if (they) change(s) in the ‘right’ direction, while other things remain equal, things have got better, or people are ‘better off’ ” (Department of Health, Education and Welfare, 1969: p. 97). The latter are descriptive indicators measuring struc-
76
HEINZ-HERBERT NOLL
tural, attitudinal and value changes in a society and thus providing information which supports politics rather in an indirect way. The European System of Social Indicators also includes objective as well as subjective indicators.12 While the objective indicators used are for the most part supposed to measure the outcomes of societal processes in terms of living conditions and individual resources, some inputs related indicators are included too. The latter are particularly needed when it comes to evaluate the efficiency of societal institutions and policy measures. Subjective indicators are by nature outcome measures. They include first of all indicators of subjective well-being, but also other perceptional and evaluational measures as for example preferences and concerns or hopes and fears. While the European System of Social Indicators for the most part will be based on cross sectional indicators, longitudinal indicators will be used as well as far as appropriate and applicable due to the availability of longitudinal data bases.
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
77
To give an example of how indicators within the European System of Social Indicators will look like, Figure 7 presents a preliminary list of indicators related to just one measurement dimension (unemployment and underemployment) as part of the goal dimension “Improvement of Objective Living Conditions” within the life domain “Labour Market and Working Conditions”. At a later stage of the project of developing the European System of Social Indicators also the possibilities of constructing composite indexes will be considered, which are supposed to synthesise and summarise the detailed information provided by the multitude of indicators.
78
HEINZ-HERBERT NOLL
Countries and Regional Disaggregation Concerning the coverage of the European System of Social Indicators in terms of countries included, a decision has been taken to rely primarily on the 15 current member states of the European Union. However, as far as possible additional European nations will be included too: Norway, Switzerland and three Central European countries, which are going to join the European Union in the near future, that is the Czech Republic, Hungary and Poland. In addition the indicators will be provided for the European Union as a whole. Beyond the mayor European countries and the EU-average also two important reference societies – the United States and Japan – will be included as far as appropriate and comparable data will be available. For each of the 20 European countries regional disaggregations of indicators will be provided as far as regional splits seem to be reasonable and as far as respective data are available. For the European Union countries, the regional disaggregation will mainly follow the NUTS classification (Nomenclature des Unités Territoriales Statistiques) elaborated by Eurostat.13 This classification subdivides each member state of the European Union into territorial units by using a hierarchical structuring at five levels of differentiation which correspond to the national administrative units. At the NUTS-1 level altogether 78 regions of the European Union are being distinguished. These are again split up into 211 territorial units at level NUTS-2 and 1.093 units at level NUTS-3. Indicators within our European System of Social Indicators will be mainly disaggregated at the NUTS-1 level. However, since NUTS-1 represents a rather rough breakdown of several countries or at times is not available at all, a disaggregation at the NUTS-2 level has been chosen in some cases. The European countries which do not belong to the European Union will be disaggregated at a similar level according to the standards of the national statistical offices. Starting Point and Periodicity of Observations The European System of Social Indicators will present yearly figures for the included indicators, given that the data are available. As a matter of fact, for many indicators this will not be feasible, especially for those indicators which are based on surveys not conducted on a yearly basis or with varying thematic coverage.
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
79
The early 1980s have been chosen as the starting point of the time series for pragmatic reasons, taking into account that for five of the present Member States of the European Union the date of accession was not before the middle of the 1980s. An earlier starting point of the time series would have raised the problem of availability of comparable data for these countries, since they are included in statistics of the European Union only since their entry into the Union.
Data In order to implement the European System of Social Indicators no efforts of primary data collection are planned for the moment. Instead, various available data sources will be used and exploited. As a general principle, all efforts will be made to make use of the best available databases and to ensure comparability across national societies and across time as far as any possible. As a consequence the exploitation of cross national – and if at all possible harmonised – databases will be given high priority. As far as aggregated data are concerned, the databases of Eurostat – as for example NEW CRONOS and REGIO – are certainly among the most important and will be used to a large extend. Beyond aggregated data also various microdatasets will be exploited for the purposes of the indicators system. These microdatasets include data collected by the European Commission such as the Eurobarometer Surveys, data collected within the system of official statistics – such as the European Community Household Panel (ECHP) or the Luxembourg Income Study (LIS) – as well as data collected within science based survey programmes, as for example the International Social Survey Programme and the World Value Surveys, to mention but a few. SUMMARY AND CONCLUSIONS
This paper develops the conceptual framework for a European System of Social Indicators and outlines the major structural elements of its architecture. As far as the conceptual framework is concerned, the quality of life concept turns out to be most central and constitutes the overarching perspective of observation
80
HEINZ-HERBERT NOLL
and measurement. From this perspective, the indicators system puts its emphasis first of all on the objective living conditions as well as the subjective well-being of the individual citizens. In addition to these dimensions of the individual quality of life, the conceptual framework incorporates also the notions of social cohesion and sustainability, both of which are considered to represent major aspects of the quality of societies. From the social cohesion discourse two dimensions have been identified to be covered by the European System of Social Indicators: the amount of disparities and social inequalities on the one side and the strength of social connections and ties on the other. From the sustainability concept the conceptual framework of the indicators system adopts in particular the perspective to preserve the capital of the society – natural, human, and social capital – for future generations. As far as the architecture of the European System of Social Indicators is concerned, a life domain approach is most characteristic. The indicator system covers altogether 13 life domains and includes in addition a module on the total life situation. Within each life domain, the dimensions of measurement and indicators address different aspects of the individual quality of life, social cohesion and sustainability. Moreover also basic dimensions of the social structure as well as attitudes and value orientations will be covered. The indicator system will include 20 European countries, but also the U.S. and Japan as two important reference societies. If reasonable and possible, indicator time series will be disaggregated for regions at the NUTS-1 level. Indicator time series are supposed to start at the beginning of the eighties and will – given that appropriate data are available – provide information on a yearly basis. Using the conceptual framework and following the structural setup outlined above, the European System of Social Indicators will be completed domain by domain. For the moment the indicator system has been developed in full only for the domain “Labour Market and Working Conditions”. For this life domain 162 indicators have been selected and time series data have already been collected completely. Indicators and time series data for this life domain are available at the internet as PDF-documents.14 Work on constructing indicators and collecting time series data for further life domains is under progress. The European System of Social Indicators will
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
81
finally be presented as an electronic information system using the potential of modern information technology. A prototype of such an electronic European Social Indicators Information System will be available in the near future. Having developed the conceptual framework and the main elements of the architecture of the European System of Social Indicators, the scientific community as well as policy makers are supposed to examine and discuss the suggestions made. This process of reviewing and critical perception and reaction will be of crucial importance to validate and improve this new tool of social monitoring and reporting.
NOTES 1
Considerable parts of this article are drawn from Berger-Schmitt/Noll 2000. For a more extensive review of the rise and the meaning of the concept of quality of life see Noll 1999. 3 For a review of the respective literature see Rossing Feldman/Assaf (1999). 4 An overview can be found in OECD 1998c, Moldan/Billharz/Matravers 1997; Hardi/Barg 1997. See also U.N. Commission on Sustainable Development 1996. 5 For further details see Noll 1999, pp. 16–17. 6 The concept has been operationalised and measured by means of the Human Development Index which is based on these three criteria. 7 Gender equality was especially stressed and became the theme of the HDR 1995 “Gender and Human Development” (UNDP 1995). 8 In view of the self-image of the European Union as a community of shared values the European goals considered here represent more than just the smallest common denominator of goals of European policies respecting the principle of subsidiarity. A more detailed and further reaching comparative analysis of similarities and differences of welfare goals of the EU member states would in any case go far beyond the research programme of this project. 9 The category of produced/physical capital will not be included since the main objective of the European System of Social Indicators is the measurement of social developments. 10 Although goal dimensions have been derived from welfare concepts and measurement dimensions will be delineated from goal dimensions, it will not be possible to inversely draw inferences from the measurement dimensions of the indicators system to the underlying welfare concepts. The reasons are the substantial overlaps between the welfare concepts at the level of goal dimensions as well as at the level of measurement dimensions. The overlaps between the concept of social cohesion and quality of life – for example – concern the goal dimension of reducing social exclusion which has been conceptualized as an 2
82
HEINZ-HERBERT NOLL
individual state of economic, social and/or political deprivation. This goal can be subsumed under the heading of the social cohesion aspect of reducing disparities and inequalities, but at the same time it could also be treated as an aspect of the individuals’ quality of life. 11 This does not mean that all goal dimensions are included within each domain, since some dimensions are not relevant for particular domains. 12 See for example Noll 1996. 13 The most recent version of the classification is presented at http://www. europa.eu.int/en/comm/dg07/tif/nomenclatures/nomenclatures_nuts_99.htm. 14 See: http://www.gesis.org/en/social_monitoring/social_indicators/EU_ Reporting/indicators.htm.
REFERENCES Allardt, E.: 1993, ‘Having, loving, being: An alternative to the Swedish model of welfare research’, in M. Nussbaum and A. Sen (eds.), The Quality of Life (Oxford University Press, Oxford, pp. 88–94). Argyle, M.: 1996, ‘Subjective well-being’, in A. Offer (eds.), In Pursuit of the Quality of Life (Oxford University Press, Oxford, pp. 18–45). Beck, W., L. van der Maesen and A. Walker: 1998a, ‘Introduction’, in W. Beck, L. van der Maesen and A. Walker (eds.), The Social Quality of Europe (The Policy Press, Bristol, pp. 1–15). Beck, W, L. van der Maesen and A. Walker: 1998b, ‘Social quality: From issue to concept’, in W. Beck, L. van der Maesen and A. Walker (eds.), The Social Quality of Europe (The Policy Press, Bristol, pp. 301–340). Beck, W., L. van der Maesen and A. Walker: 1998, The Social Quality of Europe (The Policy Press, Bristol). Becker, E., T. Jahn, I. Stiess and P. Wehling: 1997, ‘Sustainability: A crossdisciplinary concept for social transformations’ (MOST Policy Papers 6, UNESCO, Paris) (http://www.unesco.org/most/discuss.htm). Bell, D., P. Halucha and M. Hopkins: 1999, ‘Sustainable development concept paper. Sustainability project’ (Government of Canada, Policy Research Initiative (PRI)) (http://www.schoolnet.ca/pri-prp/sustainability/concept%20papere.htm). Berger-Schmitt, R. and B. Jankowitsch: 1999, ‘Systems of social indicators and social reporting: The state of the art. EuReporting’ (Working Paper No. 1, Centre for Survey Research and Methodology (ZUMA), Social Indicators Department, Mannheim) (http://www.gesis.org/en/social_monitoring/social_ indicators/EU_Reporting/index.htm). Berger-Schmitt, R. and H.-H. Noll: 2000, ‘Conceptual framework and structure of a European system of social indicators’ (EuReporting Working Paper No. 9, Subproject “European System of Social Indicators”, Centre for Survey Research and Methodology (ZUMA), Social Indicators Department, Mannheim) (http://
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
83
www.gesis.org/en/social_monitoring/social_indicators/EU_Reporting/index. htm). Berghman, J.: 1995, ‘Social exclusion in Europe: Policy context and analytical framework’, in G. Room (ed.), Beyond the Threshold. The Measurement and Analysis of Social Exclusion (The Policy Press, Bristol, pp. 10–28). Campbell, A., Ph. E. Converse and W. Rodgers: 1976, The Quality of American Life (Russel Sage Foundation, New York). Coleman, J.: 1988, ‘Social capital in the creation of human capital’, American Journal of Sociology 94, pp. 95–120. Dahrendorf, R. et al.: 1995, Report on Wealth Creation and Social Cohesion in a Free Society (Commission on Wealth Creation and Social Cohesion, London). Department of Health, Education, and Welfare: 1969, Toward a Social Report (US. Government Printing Office, Washington). Doraid, M.: 1997, ‘Analytical tools for human development’ (Human Development Report Office, United Nations Development Programme (UNDP)) (http://www.undp.org/hdro/anatools.htm). Erikson, R.: 1993, ‘Descriptions of inequality: The Swedish approach to welfare research’, in M. Nussbaum and A. Sen. (eds.), The Quality of Life (Oxford University Press, Oxford, pp. 67–83). European Commission: 1993a, ‘Growth, competitiveness and employment’. White Paper COM(93)700 (Office for Official Publication of the European Communities, Luxembourg). European Commission: 1994, ‘European social policy – A way forward for the union’. White Paper COM(94)333final (Office for Official Publication of the European Communities, Luxembourg). European Commission: 1996a, ‘Living and working in the information society: People first’ (Green Paper COM(96)389, Brussels) (http://europa.eu. int/ispoosoc/legreg/docs/peopl1st.html). European Commission: 1996b, Equality of Opportunity for People with Disabilities – A New European Community Disability Strategy COM(96)406 (Office for Official Publications of the European Communities, Luxembourg). European Commission: 1997b, For a Stronger and Wider Union (COM(97) 2000, Agenda 2000, Volume I – Communication, Strasbourg) (http://europa.eu.int/ comm/enlargement/agenda2000/strong/1.htm). European Commission: 1997e, Communication the Commission on Environment and Employment (Building a Sustainable Europe) (COM(97)592 final, Brussels). European Commission: 1998a, Social Action Programme 1998–2000 (COM(98)259, Brussels) (http://europa.eu.int/en/comm/dg05/news/sapen. htm). European Communities: 1997a, Treaty of Amsterdam Amending the Treaty on European Union, the Treaties Establishing the European Communities and Certain Related Acts (Office for Official Publications of the European Communities, Luxembourg).
84
HEINZ-HERBERT NOLL
European Communities: 1997b, Consolidated Version of the Treaty on European Union (Office for Official Publications of the European Communities, Luxembourg). European Communities: 1997c, Consolidated Version of the Treaty Establishing the European Community (Office for Official Publications of the European Communities, Luxembourg). Gore, C. and J.B. Figueiredo: 1997, ‘Social exclusion and anti-poverty policy: A debate’ (International Institute for Labour Studies Research Series, No. 110, Geneva). Grootaert, Ch.: 1998, ‘Social capital: The missing link?. World Bank, Social Capital Initiative (SCI)’ (Working Paper No. 3, Social Capital Initiative, Washington, D.C.) (http://www.worldbank.org/poverty/scapital/wkrppr/sciwp3.pdf). Haan, A. de: 1999, Social Exclusion: Towards a Holistic Understanding of Deprivation (Deutsche Stiftung für Internationale Entwicklung (DSE), Villa Borsig Workshop Series, Berlin) (http://www.dse.de/ef/poverty/dehaan.htm). Hardi, P., S. Barg, T. Hodge and L. Pinter: 1997, ‘Measuring sustainable. Development: Review of current practice’ (Occasional Paper Number 17, International Institute for Sustainable Development (IISD), Winnipeg). Hart, M.: 1998-99, What is Sustainability Anyway? (http://www. sustainablemeasures.com/Sustainability/index.html). Immerfall, S.: 1999, ‘Sozialkapital in der Bundesrepublik. Thesen zu Konzept und Größenordnung’, in E. Kistler, H.-H. Noll and E. Priller (eds.), Perspektiven gesellschaftlichen Zusammenhalts. Empirische Befunde, Praxiserfahrungen, Meßkonzepte (Edition Sigma, Berlin, pp. 121–128). International Institute for Labour Studies (IILS): 1998, Social Exclusion and Anti-Poverty Strategies. Research Project on the Patterns and Causes of Social Exclusion and the Design of Policies to Promote Integration: A Synthesis of Findings (IILS, Geneva). International Institute for Sustainable Development (IISD): 1998, City of Winnipeg Quality of Life Indicators (IISD, Winnipeg) (http://iisd1.iisd.ca/ pdf/wpg.qoli.pdf). Jenson, J. 1998b: Mapping Social Cohesion: The State of Canadian Research Canadian Policy Research Networks, Ottawa) (CPRN Study No. (http://www.cprn.com/cprn.html). McCracken, M.: 1998, Social Cohesion and Macroeconomic Performance. Centre for the Study of Living Standards (CSLS), Paper presented at the Conference: The State of Living Standards and the Quality of Life, October 30-31, Ottawa, Ontario/Canada. Mega, V. and J. Pedersen: 1998, Urban Sustainability Indicators. European Foundation for the Improvement of Living and Working Conditions (EUR-OP, Luxembourg). Miles, I.: 1985, Social Indicators for Human Development (Frances Pinter, London).
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
85
Moldan, B., S. Billharz and R. Matravers (eds.): 1997, Sustainability Indicators: A Report on the Project on Indicators of Sustainable Development SCOPE report 58 (Wiley, Chichester). Munasinghe, M. and J. McNeely: 1995, ‘Key concepts and terminology of sustainable development’, in M. Munasinghe and W. Shearer (eds.), Defining and Measuring Sustainability (The Biogeophysical Foundations, Washington, D.C., pp. 19–56). Narayan, D.: 1999, Bonds and Bridges: Social Capital and Poverty (World Bank, Poverty Group,Washington, D.C.) (http://www.worldbank.org/ poverty/scapital/library/narayan.htm). Noll, H.-H.: 1996, ‘Social indicators and social reporting – The international experience’, in Canadian Council on Social Development (ed.), Symposium on Measuring Well-Being and Social Indicators. Final report (http://www.ccsd.ca/noll1.html). Noll, H.-H.: 1999, ‘Konzepte der Wohlfahrtsentwicklung: Lebensqualität und “neue” Wohlfahrtskonzepte’ (EuReporting Working Paper No. 3, Centre for Survey Re-search and Methodology (ZUMA), Social Indicators Department, Mannheim) (http://www.gesis.org/en/social_monitoring/ social_indicators/EU_Reporting/index.htm). Noll, H.-H. and W. Zapf: 1994, ‘Social indicators research: Societal monitoring and social reporting’, in I. Borg and P. Ph. Mohler (eds.), Trends and Perspectives in Empirical Social Research (Walter de Gruyter, Berlin, New York, pp. 1–16). North, D.: 1990, Institutions, Institutional Change, and Economic Performance (University Press, New York/Cambridge). O’Connor, P.: 1998, ‘Mapping Social Cohesion. Canadian Policy Research Networks’ (CPRN Discussion Paper No. Ottawa) (ftp://ftp.cprn.org/ family/msc_e.pdf). OECD: 1998a, ‘Sustainable development: A renewed effort by the OECD’ (OECD Policy Brief No. 8, Paris) (http://www.oecd.org//publications/pol _brief/1998/9808-eng.htm). OECD: 1998b, ‘OECD work on sustainable development’ (A Discussion Paper on Work to be Undertaken over the Period 1998–2001, Paris) (http://www. oecd.org/subject/sustdev/oecdwork.pdf). OECD: 1998c, ‘Sustainable development indicators’ (OECD Expert Workshop, 8–9 October 1998, Paris). Pearce, D.: 1993, Measuring Sustainable Development (Earthscan, London). Pearce, D.W. and J.J. Warford: 1993, World Without End. Economics, Environment, and Sustainable Development (Oxford University Press, New York). Policy Research Initiative, Government of Canada: 1999, Sustaining Growth, Human Development, and Social Cohesion in a Global World. A Report Prepared for the Policy Research Initiative (http://www.schoolnet.ca/pri-rp/ keydocs/sustain99/index-e.htm). Putnam, R.: 1995, ‘Bowling alone: America’s declining social capital’, Journal of Democracy 6(1), pp. 65–78.
86
HEINZ-HERBERT NOLL
Rodgers, G. (ed.): 1995, The Poverty Agenda and the ILO. Issues for Research and Action. New approaches to poverty analysis and policy – I. A contribution to the World Summit for Social Development (International Labour Office, Geneva). Rodgers, G., C. Gore and J.B. Figueiredo: 1995, Social Exclusion: Rhetoric, Reality, Responses (International Labour Office, Geneva). Rossing Feldman, T. and S. Assaf: 1999, ‘Social capital: Conceptual frameworks and empirical evidence. An annotated bibliography’ (Social Capital Initiative Working Paper No. 4, World Bank, Social Capital Initiative (SCI),Washington, D.C.) (http://www.worldbank.org/poverty/scapital/wkrppr/sciwp5.pdf). Schäfers, B.: 1999, ‘Komparative und nicht-komparative Ansätze zur Analyse der Europäisierung der Sozialstrukturen’ (WZB Arbeitspapier Nr. FS III 99–407, Wissenschaftszentrum Berlin fur Sozialforschung, Berlin). Sen, A.: 1993, ‘Capability and well-being’, in M. Nussbaum and A. Sen (eds.), The Quality of Life (Oxford University Press, Oxford, pp. 30–53). Silver, H.: 1994, ‘Social exclusion and social solidarity: Three paradigms’ (Discussion Papers Series No. 69, International Institute for Labour Studies (IILS), Geneva). Social Capital Initiative (SCI), World Bank: 1998, The initiative on defining, monitoring and measuring social capital. overview and program description’ (Working Paper No. 1, Social Capital Initiative, Washington, D.C.) (http://www.worldbank.org/poverty/scapital/wkrppr/sciwp1.pdf) UN – Commission on Sustainable Development (CSD): 1996, Indicators of Sustainable Development: Framework and Methodologies (CSD, New York). United Nations Development Programme (UNDP): 1990, Human Development Report 1990: Concept and Measurement of Human Development (Oxford University Press, New York). United Nations Development Programme (UNDP): 1993, Human Development Report 1993: People’s Participation (Oxford University Press, New York). United Nations Development Programme (UNDP): 1994, Human Development Report 1994: New Dimensions of Human Development (Oxford University Press, New York). United Nations Development Programme (UNDP): 1995, Human Development Report 1995: Gender and Human Development (Oxford University Press, New York). United Nations Development Programme (UNDP): 1996, Human Development Report: Economic Growth and Human Development (Oxford University Press, New York). United Nations Development Programme (UNDP): 1998, Human Development Report 1998: Consumption for Human Development (Oxford University Press, New York). Uusitalo, H.: 1994, ‘Social statistics and social reporting in the nordic countries’, in P. Flora, F. Kraus, H.-H. Noll and F. Rothenbacher (eds.), Social Statistics and Social Reporting in and for Europe (Informationszentrum Sozialwissenschaften, Bonn, pp. 99–120).
TOWARDS A EUROPEAN SYSTEM OF SOCIAL INDICATORS
87
Veenhoven, R.: 1996, ‘Happy life-expectance: A comprehensive measure of quality-of-life in nations’, Social Indicator Research 39, pp. 1–58. Veenhoven, R.: 1997, ‘Lebenszufriedenheit der Bürger: Ein Indikator für die ‘Lebbarkeit’ von Gesellschaften?’ in H.-H. Noll (ed.), Sozialberichterstattung in Deutschland (Juventa Verlag, Weinheim, pp. 267–293). Welsh, J.M.: 1993, ‘A peoples’ Europe? European citizenship and European identity’ (EUI Working Paper No. 93/2, Florence). Wiman, R.: 1999, Putting People at the Center of Sustainable Development. Proceedings of the Expert Meeting on the Social Dimension in Sustainable Development. Volume 1: Policy Themes – A Synthesis. National Research and Development Centre for Welfare and Health (STAKES) (Ministry of Social Affairs and Health, Helsinki). Woolley, F.: 1998, Social Cohesion and Voluntary Activity: Making Connections. Centre for the Study of Living Standards (CSLS), Paper presented at the Conference: The State of Living Standards and the Quality of Life, October 30–31, Ottawa, Ontario/Canada. World Bank: 1997, ‘Expanding the measure of wealth: Indicators of environmentally sustainable development’ (Environmentally Sustainable Development Studies and Monograph Series, No. 17, Washington, D.C.). World Commission on Environment and Development: 1987, Our Common Future (Oxford University Press, Oxford). Zapf, W.: 1984, ‘Individuelle Wohlfahrt: Lebensbedingungen und wahrgenommene Lebensqualität’, in W. Glatzer and W. Zapf (eds.), Lebensqualität in der Bundesrepublik (Campus, Frankfurt a.M./New York, pp. 13–26).
ZUMA, Social Indicators Department Mannheim, Germany E-mail:
[email protected] This page intentionally left blank
JOACHIM VOGEL
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING: A 30-YEAR EXPERIENCE
INTRODUCTION
Most developed countries have developed regular information systems to monitor general living conditions along the lines of the 1960s, the founders of this tradition were greatly influenced by the development of the System of National Accounts (SNA). The unity, integration, coverage, routine data collection, and standardised compilation and publication of economic SNA statistics appeared as a prototype for the social dimension as well. Similar systems were discussed under headings such as social accounting and a system of social and demographic statistics (SSDS), linking elements of social statistics to produce detailed as well as aggregate statistics of individual and societal well-being. The successful GDP concept demanded a social counterpart to monitor the social consequences of modernisation and economic change. In the absence of a unifying concept (such as money), and already established statistical systems (which at that time were expanding in economic statistics), the answer was a system of social indicators, general social surveys (data collection systems), and routine social reporting (descriptive statistics; analysis of living conditions). These are still the three major pillars of the social indicator movement. During the last forty years, the original focus on individual as well as societal welfare was realised in most developed countries by researchers who not always defined themselves as belonging to this tradition, but were definitely inspired by the early writings. Today, we have two mainstreams. One tradition is official and semi-official social reporting located at public institutions (national statistical institutes, special research institutes, and someSocial Indicators Research 58: 89–113, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
90
JOACHIM VOGEL
times ministries) with a distinct public mandate and usually regular long-term public funding. The other stream is the independent research tradition, with temporary funding, ad hoc and temporary studies, sometimes less comprehensive in nature, and more focused on special problems. The Swedish Social Indicator Program conducted by Statistics Sweden belongs to the first mainstream (official social reporting), and is today one of the oldest and best funded programs. This report is based on the experience of goals and methods used in the Swedish program since 1969, including the strategies for indicator selection, data collection and social reporting. This report summarises the early start and the strategic adjustments of the Swedish system to support a regular system of social reporting, as well as general social research and social planning. The author has been engaged in the development of the Swedish system and official social reporting over a period of 30 years, as well as in international comparative work. The general design of the Swedish system has to be understood from the political background and the statistical tradition in Sweden. In particular, I would like to emphasise the considerable advantages of: standardisation: a standardised yet flexible system of social indicators the value of integration: annual standardised comprehensive social surveys with simultaneous measurement, and maximum variable content flexibility to serve general social research, social planning and public debate focus on objective measures the uses of a combined cross-sectional and longitudinal design institutionalisation, long-term mandate and long-term funding: the development and preservation of a standardised, diversified, and flexible system of social reporting,
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
91
BASICS OF THE SWEDISH SURVEY SYSTEM
The Swedish surveys of living conditions (ULF) have been conducted by Statistics Sweden annually since 1974. Today, ULF is one of the oldest and most extensive information systems in the world in this field, both in terms of the dimensions of the data material (180 000 interviews 1974–99), and in terms of the variable quantity (700 variables). For almost three decades ULF has had a permanent staff of sociologists, statisticians and technical staff, which is responsible for planning, data collection and dissemination of an extensive publication program. These activities are financed in part by state appropriations (about 1.75 million $ per year) and in part by commissioned work from government agencies, unions, other interest groups, the mass media and special research funds. The Swedish survey system comprises a large standardised set of 125 social indicators and 600 additional indicators and background variables. The sample design is mixed cross-sectional and longitudinal, the annual sample size is 7600 in a population of some 8.8 million, the response rate is 80 percent. The reporting system is standardised since 1975 and produces general social reports as well as domain reports, reports on disadvantaged groups and special policy studies (92 titles), as well as a periodic journal. In addition, all databases are available to research institutions for secondary analysis. The general design of the Swedish system deviates much from that of most other countries, which is explained by the special historical context (the expanding welfare state of the early 1970s demanding new statistical information for its administration), as well as the early ambitions of Statistics Sweden to respond to this transformation process, and to incorporate the major users (in social planning and social research) into the work of designing a large scale information system in the vein of the social indicator movement. Since 1971 a series of 10 advisory user boards (one for each domain), in total including some 100 members (dominated by researchers and officials in planning), were involved in the development of the Swedish system. This strategy paid off in the successful institutionalisation and survival of a large and regularly funded survey and reporting system over 25 years. During this period ULF developed from a fixed general social survey towards an integrated
92
JOACHIM VOGEL
survey system, including a fixed set of 125 social indicators (the core module) and a fixed series of satellite domain surveys (health survey, victimisation survey, housing survey etc), as well as ad hoc focused research projects linked to the survey system. Today, ULF is the key information system within general social statistics, social reporting, and data base of general social research. Statistics Sweden has moved towards an integrated official survey system, in contrast to the irregular and unrelated data collection and reporting still prevailing in most countries. Lately, the national statistical institutes in several countries are moving in the same direction (e.g. Norway, United Kingdom, the Netherlands). SPECIAL FEATURES OF THE SWEDISH SYSTEM
Integration and Flexibility
Given the various user needs, it became obvious from the very beginning that a standardised general social survey would be insufficient to respond to the divergent user needs from the administration and from social research, as well as from the perspective of the system of social statistics in general, and from social reporting in particular. Technically speaking, the simultaneous measurement of large volumes of indicators and rich background information was given priority over large sample size, in order to give maximum flexibility for regular social reporting as well as policy analysis and general research. The survey system was also perceived as a flexible platform for extensions incorporating issues covered earlier by special surveys, for follow-up studies, complementary samples, and as a sampling frame for further studies. The general design is rather complex, but still flexible. The Swedish welfare surveys encompass for each year a core module including a general measurement of some 125 social indicators within 13 welfare components1 plus a large number of classification variables for distributive analysis. To this is added one of four in-depth programs within a group of components (modules concerning labour market, health/care, physical environment, and social relations), plus one register complementation (see below). The in-depth program is changed every second year, and is then repeated every eight years. In addition to the annual collection of
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
93
some 125 social indicators by interview, a large set of variables are collected from various administrative registers (see below). The modular design provides data bases with close to 12 000 observations collected over a two-year-period. The main part of the reporting is a (cross-sectional) account of the welfare changes over an eight-year period, using 24 000 observations (2 × 12000), which allows rather detailed disaggregations for cross-sectional trend analysis. Part of these observations (appr. 5000 cases) form a panel repeatedly interviewed every 8 years. The integrated Swedish survey system thus covers much of the needs for special studies within the various domains. In most other developed countries, such in-depth studies take the form of independent intermittent special studies with no direct co-ordination with general social indicators included in the welfare surveys. This connection between variables (domains, indicators, classificatory variables) by simultaneous measurement is one of the great advantages of the Swedish system, creating flexibility in analysis and reporting. The third element in the Swedish surveys are flexible additions included for one or two years or shorter periods, where the detailed welfare measurement is requested as valuable background information (for policy analyses, to research specific issues). Some of these studies will follow up earlier samples or add new subsamples targeting special groups. The fourth element are larger extensions financed by research funds or special commission. Some recent studies incorporated in the survey system are modules on consensual poverty, public opinion on welfare state provisions, and the European Welfare Module (Delhey et al., this issue). Finally, the surveys are also used to host basic data of wider research interest. Some recent examples are a retrospective module on life history data concerning family formation and labour market participation, and perceived health effects of working conditions. Forthcoming is a large (permanent) panel survey of ageing and the elderly, associated with the general welfare survey. Surveys and Administrative Registers
The statistical systems in Sweden and the other Nordic countries deviate from those in most other countries in their extensive use
94
JOACHIM VOGEL
of administrative registers. The register approach has a relatively long tradition in this region. The Nordic surveys are, for example, complemented with detailed information on all household members’ incomes, paid taxes, received transfers, wealth, real estate values, standard census data, and so on. Computerised central population registers are used in the Nordic countries in which every inhabitant is recorded by a unique personal identification number. This identification number is used in essentially all public administrative registers. Using the personal identification number, samples can be extracted from central computerised population registers (providing current address and census data for possible stratification). Established panels to be interviewed at a later date can easily be located by means of the identification number in the population register, which delivers the current address and present family circumstances. The use of registers could also be extended to provide the surveys with new register material. More generally, registers can be used for longitudinal analyses of interview material (retrospectively or prospectively). The interviewees’ personal income development can thus be followed in the registers. Mortality and causes of death in older samples are regularly added to the database. The personal identification number thus opens a variety of options. In the space remaining, this paper will address some of the major cross-roads of social indicator work, and the corresponding strategies developed in the Swedish case: the conceptualisation of social indicators: the Level-of-Living versus Quality-of-Life approach trend analysis: cross-sectional versus longitudinal approach systematic social reporting and institutionalised distribution analysis
THE CONCEPTUALISATION OF SOCIAL INDICATORS
The Basic Idea of Social Indicators
Social indicators go beyond the merely descriptive or “bookkeeping” approach of traditional social statistics, and incorporate a potential for evaluation, prognosis and planning. They usually
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
95
focus on individual living conditions, and they are typically outputrelated (e.g., the state of health, not investments in health services). Social indicators look beyond traditional global monetary indicators. Non-monetary aspects of social relations, the natural environment, and the distribution of health are among the additional variables introduced. This trend is especially pronounced in the Scandinavian countries; elsewhere, income statistics are still the primary index of living conditions. Much of the early thinking of the social indicator movement was influenced by the unity of economic statistics, as well as its success as a basis for planning. However, neither a unifying general social theory nor a corresponding linkage (such as money in economic statistics) was available. Social indicators are policy-related, and therefore normative in character. They are usually regarded as an integral component of social programmes. Ideally, social indicators should form a parsimonious system of measurements, making it possible to collect and analyse data on the main issues without exhausting available resources. Social indicators should facilitate systematic comparisons between groups (degree of inequality) and over time (trends). The emphasis on individual well-being puts the distributional aspects of well-being in focus. Issues on the political agenda such as social exclusion, poverty, inequality and equal opportunity, can only be addressed with parameters based on individual and household statistics, normally micro level data from surveys. This approach is particularly evident in Scandinavian social reporting (cf. Vogel, 1990, 1993; Vogel and Hall, 1997), and is related to the egalitarian tradition of social policy in these countries. The distributional aspects of well-being are for the most part exceptionally wellresearched in the Scandinavian countries. General social surveys were introduced early in Sweden (starting 1968) and social reporting incorporates a long tradition of systematic reporting on the situation of marginalized groups. The Role of Social Indicators in Social Planning Originally, social indicators were meant to signify a restricted set of official policy-related indicators. For example: Social indicators in the United Nations work are intended to monitor aspects of welfare;
96
JOACHIM VOGEL
that is, indicators are considered as tools in the planning process, while the OECD has emphasised only the importance of indicators as results in terms of well-being, with much less regard to inputs and resources, their transformation into social services, and their distribution and use (United Nations, 1979; emphasis added.) But now the term “social indicators” has been used in so many different ways that its meaning has become obscure. There are not many examples of comprehensive official social indicator systems that are institutionalised at the national level. In the programmes of the Scandinavian national statistical institutes, for example, no official set of social indicators is defined, but larger sets of institutionalised measures are produced on a regular basis, under the heading of living conditions. These sets include both social indicators and others with less obvious and direct policy implications. It is a logical consequence of the Swedish political tradition that social indicators should serve as monitoring tools. However, the selection of social indicators should not be based solely on policy relevance. “Relevance” often has conservative implication: welfare problems are defined within a framework of established institutions and traditional policies. Horn (1980) argues that this bias has “haunted the official government-sponsored publications”. It is a challenge for the scientific and statistical imagination to identify new issues and strategic indicators that go beyond present conceptions of social policy. Hence, the Scandinavian programmes also explicitly identify additional areas of concern (e.g. social contacts), even when there is no clear institutional basis for monitoring, or when direct intervention is even regarded as controversial. This means that measurements of well-being take precedence, no matter what the implications for public policy might be. Consequently, a separate concern, political resources, has been added to the programme; it is a measure of individual competence and access to collective resources for influencing political decisions. Objective and Subjective Indicators
The Scandinavian level of living approach defines welfare in terms of command over resources (money, wealth, consumer goods, health, education, social contacts, etc.) in the Titmuss tradition. Resources are objective entities that are used to gain control over
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
97
living conditions. They can be used in “arenas” (markets, relations) to produce goods, services and more intangible rewards, or to produce other resources to be used in other arenas. “Manipulating” individual or collective resources, by increasing or limiting them, is an important element in social planning. In the Swedish tradition with its strong orientation towards planning, social indicators are primarily designed to influence political action. This places objective (or descriptive) indicators of living conditions rather than subjective (or evaluative) indicators at the centre of interest, since it is the explicit purpose of political action to influence real living conditions. Parallel with the “objective” Scandinavian level-of-living approach, a competing tradition was developed, usually referred to as quality of life approach, based on subjective indicators derived from surveys of individual perceptions, emotions, satisfactions and welfare demands. This tradition notes discrepancies between “objective” and “subjective” indicators, e.g. that those who have most are often least satisfied, and vice versa, and argues that objective indicators are insufficient. Instead, a combination of objective and subjective measurements is recommended. Unfortunately, some QOL research operates without this principle anchoring subjective indicators in objective reality. The problem with subjective indicators, from a planning perspective, is that they tend to measure the social definition and control of aspirations. Since planning and distribution should be based on objective conditions and not on welfare demands, objective indicators predominate in Scandinavian official statistics as well as in social research. In the case of the Swedish system the principle is to give priority to objective measurement, but also to allow for a limited set of additional subjective indicators. The volume is a matter of resource allocation and the perceived value of the information. There is strong agreement between researchers, planning staff, and Statistics Sweden on this principle. Recent inclusions of subjective indicators were adjustments in the direction of comparative studies in which ULF participates (ECHP, the European Welfare Module).
98
JOACHIM VOGEL
THE GENERAL DESIGN: CROSS-SECTIONAL VS LONGITUDINAL TREND ANALYSIS
Reference data are essential in social indicator work. The Swedish system has been designed to support evaluation by comparing groups (inequality), periods (progress/decline) and nations (comparative evaluation). We need such reference data from a political, as well as methodological point of view. Traditionally, trend analysis in social statistics is based on repeated cross-sectional surveys adding new samples or register data; performance is then measured by increases or decreases in the values of social indicators year by year. The longitudinal measurement of transitions is valuable but a second priority. In the Swedish case, the strategy has been to combine a cross-sectional and longitudinal sampling design, where sufficient technical quality is guaranteed, and without jeopardising cross-sectional trend analysis. This strategy rests on studies of experience of the pace of change in social indicators, a series of studies of measurement errors in surveys, and observed panel attrition in various types of panel design. For these reasons the design of the Swedish system deviates much from other panel surveys. Most social indicator work is based on two types of survey design, using either the traditional cross-sectional sample design or annually repeated panel sample design. In most countries (with the exception of the Nordic countries and the Netherlands) the longitudinal approach requires a contact each year to be able to trace the panel members. In Scandinavia this is not necessary, since the total population register (sampling frame) carries information on the personal identity code, as well as current address and demographic status, allowing location of panels members after decades. Without this option annual interviewing would be required. Having this option allows for the choice of a design based purely on theoretical and methodological considerations. Consequently, the longitudinal design of the Swedish survey system is very different from the dominant solution applied by PSID (Panel Survey of Income Dynamics), ECHP (European Community Household Panel), SOEP (German Sozio-ökonomisches Panel) and other national panels. In our opinion, the pace of change, the level of measurement errors even in the best surveys, and volume of panel attrition are strong arguments against annual repeated panel design.
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
99
The Pace of Change The first series of Swedish surveys in the 1970s showed that the pace of change in most indicators of living conditions was far too slow to justify reports of annual change, given the affordable sample size (n = 6000), the basic disaggregations required, and the corresponding variances for cross-sectional change. This understanding, and the strong user demands on detailed information which could not be accommodated in one single survey, led to the decision to abort the ambitions of reporting annual estimates, and move towards a survey system with changing content. The survey system was then reorganised in an 8 year cycle (as mentioned earlier), the sample base increased (estimates based on two years), and trends usually reported over an 8-year period. These general considerations also apply to longitudinal trend analysis. The Swedish solution was to define 50 percent of the samples as panels to be re-interviewed every 8 years, following the general cycle of measurement. Hence, trend analysis in a typical report in the Swedish reporting system will then be able to utilise a variety of databases for crosssectional as well as longitudinal analysis: cross-sectional trend analysis based on samples interviewed over the 8 year cycle (n = appr. 60 000) for the set of 125 social indicators collected every year (usually regression analysis), cross-sectional change based on two samples interviewed in two two-year periods with an interval of 8 years (each with n = 13 000) for the entire range of variables (the in-depth modules, appr. 700 variables), a panel sample interviewed twice with an interval of eight years (n = 6000), also for social indicators as well as the entire range of variables, and a panel sample interviewed three times with intervals of 8 years (n = 5000), also for the entire range of variables; Measurement Errors A second obstacle associated with yearly repeated panels is the relation between the pace of real change and measurement errors. Based on a series of reliability studies (test-retest) conducted by Statistics Sweden in the 1980s, we came to the conclusion that the magnitude of measurement errors for most indicators would not
100
JOACHIM VOGEL
allow short-term longitudinal analysis. The level of measurement errors is related to the stability of phenomena itself, the reliability of the questionnaire items, as well as the general quality of data collection and data management. Crucial issues are the subjectivity involved in questions, versus objective classifications using wellknown definitions. Repeated Swedish test-retest studies (with an interval of 3 weeks) covering the entire range of social indicators and further variables, indicate that between 10 and 50 percent of the transitions (off the diagonal) are already in place after three weeks, as compared to 6–8 years. Hence, the message from these studies is that the quality of data has to be improved radically, or measurement errors have to be estimated by test-retest studies and then included in model-based estimation. Neither of this is usually done in the large-scale surveys based on annual panels. The effect of measurement errors is usually neglected, which means huge overestimation of transitions. This is acceptable for cross-sectional analysis, but unacceptable for longitudinal analysis, since it will introduce a systematic bias. Decreasing the relative level of measurement errors by expanding the period of observation, and thus the relative level of (real) change (in the Swedish case to eight years) will improve the situation. In the Swedish case the experience from methodological research (access to reliability coefficients) helps to avoid the most obvious mistakes.
Panel Attrition Repeated interviewing each year of the same samples creates panel attrition that gradually invalidates the use of such panels for crosssectional trend analysis, as well as for longitudinal analysis. In the European case, the annual panels (e.g. Eurostats ECHP, SOEP, the Dutch CBS panel), have much lower response rates than similar studies in North America and Scandinavia, in some cases beginning with first-wave non-response rates of 40–50 percent. After several years this approach will produce non-response rates which can be roughly estimated to 60–70 percent, as compared to the first-wave sample universe. Usually the proper calculation of response rates is not even reported. In case of comparative analysis (ECHP) we will face a variation of response rates which can be estimated between 25 and 80 percent.
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
101
In these cases the cross-sectional representativity has been sacrificed after a few years, and even longitudinal analysis can no longer be defended from a methodological perspective. This sacrifice of cross-sectional statistics is unacceptable for a national statistical institute responsible for official statistics. Longitudinal analysis is an extra, building up on and qualifying cross-sectional analysis. In the Swedish experience, social reporting with its strong orientation towards social planning, response rates such as in the eroded European panels would be unacceptable for the users, including researchers, planning staff, the media, as well as the general public, and hence impossible to publish. It should be noted that the costs of the Swedish survey system with its integration of cross-sectional and longitudinal design is of similar size and costs as the national (annual) panels or national ECHP surveys. In addition, key information for longitudinal analysis is collected from administrative registers. FEATURES OF SOCIAL REPORTING
The Democratic Function of Social Reporting
Social reports are most often descriptive documents which tend to avoid normative evaluations and straightforward conclusions. However, like social indicators, and despite efforts to maintain neutrality, social reports are normative but the normative aspect is implicit and largely camouflaged by the selection, ordering and presentation of data, by the use of tables, graphs and disaggregations, by the blending of different kinds of data, and by the organisation of the report. The strategy is to “let the figures speak for themselves” by organising the data in a logical order and relying on the reader’s ability to draw the appropriate conclusions. This quasineutrality is necessary to attain general acceptance, and it is also an expression of the official status of the publishers – in most cases the national statistical institutes. In the Swedish case, this strategy has been successful, in the sense that the reports are usually accepted as fair and impartial by most political interest groups and the general public, although there is an ongoing debate within the profession. For the founders of the social indicator movement, an important goal was to put quality-of-life issues on the political agenda, and
102
JOACHIM VOGEL
thus stimulate public debate on individual quality of life. Their aim was to compel a response from political, economic and intellectual elites. Thirty years later, this provocative function of social indicators has been applied in only a few developed countries. In some countries, social monitoring programs have been terminated; and they have never been fully accepted at the international level. Influencing the political agenda is the most important challenge facing the social indicator movement today. There is a need for publicly accepted social indicators, in the parsimonious format of the original conception, and as most clearly defined by the OECD in the 1970s. Social reports are aimed at “interested citizens”, as well as decision-makers and the research community. Social reporting plays an important role in a representative democracy, by providing the electorate with comprehensive information on social conditions. If we accept the basic notion that the most important task of politics is to improve living conditions for everyone, and in particular for disadvantaged groups, then regular social indicators and regular social reporting must be regarded as an essential tool for assessing progress toward that goal. To the extent that social indicators become accepted in public discourse, they give a voice to the voter. They provide an objective measure of progress towards the defined goals: a decent existence, economic security and a fair distribution of wealth. This is the role that social indicators and social reporting should play in a representative democracy, as discussed by Sten Johansson in Toward a Theory of Social Reporting (1983). He regards social reporting as a joint activity linking the various actors in the political arena (citizens, elected representatives, political parties and other interests). Social reporting informs everyone concerned about citizens’ living conditions and demystifies politics. It provides a common ground from which everyone involved can learn something about “the way things are”. That leads to the question of “What is to be done?”, which then can be discussed by representative bodies (including political parties), further clarified by experts, and finally resolved by the appropriate political institutions. Social reporting thus performs an important function in a representative democracy. All actions bearing upon our living conditions the formation of political opinion, voting, decision-making, etc.
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
103
should be based on objective information gathered by means of scientific methods, and that information should be available to everyone. This is the democratic mission of the social indicator movement. Social indicators and social reporting, at both the national and international level, illuminate trends, comparative dissimilarities and patterns of inequality. This is an inexpensive investment in democracy, as was evident to the movement’s founders in the 1960s. The program in Sweden, one of the most extensive in the world, is run at a cost of 20 cents (US) per capita annually. Such an expenditure is negligible, compared with the costs of marketing goods and services, the consumption of tobacco and alcohol, or the information budgets of political parties. Even in comparison with general social research, the cost of a fully established social indicator program is negligible. Data Needs Access to regular, integrated and large scale data is a pre-requisite of social reporting. This is still not the case in all countries. Most national statistical institutes still produce general information on social indicators and general living conditions as a by-product of other programs. In many cases, living conditions are represented by several unrelated data sources, often collected intermittently and for entirely different purposes. Accordingly, the form and content of social reports are largely determined by the limited available data. Although the resulting publications are often voluminous, the original idea behind social indicators – that is, to provide a regular, comprehensive, simultaneous and parsimonious picture of reality, within both a temporal and an international framework – often gets lost. The need is for regular, integrated and simultaneous measurement of social indicators, as well as extended data on living conditions. We also need to build up time series at the national and international levels. Unfortunately, this has not been accomplished with the European Community Household Panel. However, several of the national statistical institutes, including those of Sweden, Norway, Italy and the Netherlands, are planning to, or already have redesigned their general national survey programs into integrated
104
JOACHIM VOGEL
systems, in order to expand their analytical capacity, as well as to economise their statistical system. Integrated surveys with simultaneous measurement, where large sets of information are collected from the same sample will largely enhance the analytical potential for social reporting and for policy analyses. A comprehensive survey can be designed to match the characteristics of a carefully chosen social indicator list specially designed to relate to policy requirements, without many of the technical and legal limitations of existing national administrative registers and surveys. Such a program could not only include officially-adopted social indicators, but also a broader set of indicators of living conditions that could be used to identify new issues requiring attention, and thus assist in an ongoing formulation of social policy. Conducting a comprehensive survey on a regular basis (preferably annually) would also provide an opportunity to flexibly incorporate additional modules relating to special social concerns, as well as in-depth enhancements of selected topics; or even opinion modules. Such a modular design of the survey should not only enhance the coverage of social issues, but also analytical capacity and flexible analysis for research and policy purposes.
STRATEGY AND STRUCTURE OF SWEDISH SOCIAL REPORTING
The scope of social reports is largely determined by data availability: one takes what is at hand, which usually means the data held by national statistical institutes. Some social reports are limited primarily to input and aggregate indicators, with limited access to social indicators in the original sense. Others make extensive use of micro-data from surveys, and can concentrate on direct measures of living conditions at the individual/household level, which also yield detailed insights into the distribution of well-being. These reports underscore the great significance of the comprehensive social survey for social reporting. The Swedish publications form an integrated system of social reports, where the different types of reports are published according to a long-term plan, following the 8-year cycle of data collection. Swedish social reporting is financed within the framework for general parliamentary appropriations for official statistics, without
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
105
specific terms of reference or detailed review by the government. The reports are published as a subseries of the official statistics of Sweden (title: Living Conditions), issued by Statistics Sweden; hence reporting is not a temporary activity, but an institution since 25 years. The list of social reports include 93 titles produced between 1975 and 1999. Frequently, the reports have contributions by external researchers, supplementing the reports with special research topics. The latest general social report covering living conditions and inequality in Sweden over two decades edited by Vogel and Häll (1997) had 28 contributors from Statistics Sweden and external research organisations. The following types of reports are published regularly: General social reports: The yearly Swedish welfare surveys have been used for general social reports containing trend analyses and systematic comparisons of many different population groups, using 125 social indicators covering the complete range of social domains (Vogel and Häll, 1997). The data underlying the latest of these reports is 156 000 interviews collected in 1975–95. This report gives trend data for 1975–95 as well as international data against which the Swedish living conditions of the 1990s can be evaluated. This report is presented in more detail later. Sectoral analyses: By utilising the modular design (in depth modules appearing for 2 years in an 8 year cycle) a standardised series of 10 sectoral social reports is published over an 8-year period. In these reports, time series analyses make further use of the entire yearly material since 1975 for the main social indicators, and of the in-depth programs for complementary indicators. Analyses of the situation of vulnerable groups: Separate systematic analyses of the level of living are carried out within all domains for specified, often relatively small target groups. For example, separate Swedish reports deal with disabled persons, immigrants, public allowance recipients, students, the unemployed, farmers, young people, manual workers, women, etc.. The Swedish reports often use co-ordinated data collected during a series of 5–8 years to build up large sample sizes (n = 20–40 000).
106
JOACHIM VOGEL
In addition, with special funding, a number of policy evaluations have been published in the series. Sweden like several other countries, including Denmark, Norway, Finland and Germany, publishes journals on social indicators and social reporting for a general audience. In addition to the official report series, there is a large volume of other publications (reports, articles) issued by general social research, reports issued by ministries and other public institutions, by unions and other organisations, based on databases and/or analyses delivered by Statistics Sweden. All databases are available to research institutions as well (in anonymous form).
SELECT EXAMPLES OF SWEDISH SOCIAL REPORTING
The large Swedish data sets collected during a period of 25 years (a total of 180 000 interviews), and the combined cross-sectional and longitudinal design, allows for extensive and advanced analyses of trends in living conditions and inequality. For the closing section of this paper three examples of routine presentations included in the regular social reporting have been selected to illustrate the value of a long-standing and integrated survey and reporting system: A. Standardised tabular presentation of social indicators: The main objective of social reporting is the systematic analysis of trends in the level and distribution of living conditions. The selection of reporting categories should enable the identification of vulnerable groups and marginalising tendencies. Accordingly, Statistics Sweden has developed a standardised format for tabular presentation focusing on a large set of social indicators (125), within 13 social domains, all of which have been computed for some 120 subgroups of the population for each year since 1975. The set of observed subgroups includes disaggregation by gender, age, type of household, family cycle, region, employment status, working environment, trade union affiliation, socio-economic status, citizenship, education, earned income, disposable income, number of dependants, financial problems, social network, health, and various types of marginal or elite status.
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
107
Each of the 13 social domains is outlined in a summary table of six pages, each displaying some 10 social indicators for the 120 groups. More detailed tables following the same design are available on CD-ROM. The standard tables present the following information: 1. Current status for 125 social indicators based on the mean values for latest four years of data collection (pooled data from ca. 24 000 interviews) 2. Trend analysis for these indicators during the entire period since 1975 (regression estimates comparing the first and last year), based on ca. 180 000 interviews (cross-sectional data) 3. Systematic distribution analyses, including detailed disaggregations of living conditions based on comparisons of current conditions and trends for 120 sub groups (see above) 4. The standard format of each domain table is then a 10 × 120 matrix (10 social indicators by 120 displayed subgroups of the population), where each cell gives an estimate of the current status as well as the trend estimate for the last 20 years. B. Analyses of the dimensions of inequality However, in addition to this detailed display of inequality and change, we are particularly interested in summarising the overall development of inequality along the major cleavages in society (social class, gender, generation and region). In order to identify the changing inequality pattern we need to decompose the inequality structure. In other words, we require an index which measures the level of inequality separately between cleavages, as well as allowing comparisons over time and between nations. Such an index was constructed and included in Swedish Social Reporting (Ribe, 1987; Vogel, 1988; Vogel and Hall, 1997). This technique has been used to compare the determinants of inequality in material living conditions2 between nations as well as over time (Vogel, 1997). The inequality index is an extension of logistic regression with the following characteristics: Pairs of contrasting subgroups are selected from each dimension to represent the variation by gender, generation, social class, family and region; By logit analysis we control the structural differences deriving from other dimensions;
108
JOACHIM VOGEL
We consider an imaginary population where the two contrasting groups have equal structure in all other included dimensions; The inequality index equals the difference in the percentage in each of the two groups having a certain good (e.g. a car); Indexes are calculated separately for a set of material indicators, as well as an average for these indicators; The index varies between +100 and –100, where +100 signifies total inequality (all in the first group have the good, and none in the other group); –100 means that the situation is reversed; and 0 indicates perfect equality (equal percentages of the two groups have the good). The numerical index scores for each pair of groups (and dimension) are thus directly comparable. Hence, with this method we can compare the relative impact of the dimensions in different countries, and in the case of Sweden we can measure the corresponding trends since 1975. The trend analysis for Sweden over the period 1975–95 is presented in graph 1. The class dimension is analysed by comparing manual workers compared to upper non-manual employees including larger managers and employers (at least 10 employees). The numeric size of the index shows the relative impact of social class, or in other words, the contribution of market effects (employment, earnings) on the distribution of material living conditions. Graph 1 indicates that up until the mid-1980s, there was a narrowing of class differences in Sweden; after that followed stagnation. Generational differences have been increasing for a long time, but there was an especially sharp increase during the 1990s. The material living conditions of young adults aged 20–29 years deteriorated in relation to two older age-groups in the study (45–64 and 65–74 years). The material living standard of the oldest improved in relation to both young and middle-aged adults. From the mid1980s, the generational difference between young adults and the middle-aged increased to become as large as the corresponding class differences. The situation of single parents and families with young children became worse in consequence with the impact of labour market hazards during the 1990s on young families, along with
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
109
recent reductions in the transfer systems. Regional differences in material living conditions remained largely unchanged during the entire period from 1975–95. C. A systematic trend analysis of inequality between generations From example B can be seen that the growing generation gap is a major concern in Sweden, as well as in most other developed coun-
110
JOACHIM VOGEL
tries. In a recent Statistics Sweden report, Vogel and Hall (2000) developed generalised techniques for analysing the cross-sectional and longitudinal change in general living conditions. The analysis is presented in graphical form, estimating each of the social indicators for five-year age groups and for three periods. Logit analysis was used, with gender, social class, type of family and region included in the model. The cross-sectional trends were monitored by comparing three periods (pooled data for the early 1980s, late 1980s and mid 1990s). Each social indicator is analysed by a series of four graphs, displaying the overall level of inequality between age groups (example: see graph 2 below), followed by further disaggregation between men and women, between working and middle class, and finally between single and married people.
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
111
Graph 2 displays the changing generation gap for an index of material living conditions. The structure of these findings underline the growing generation gap, which is (primarily) explained by prolonged youth education, delayed entry to the labour market, and delayed income (young adults), and increased public pension rights (the elderly). Hence, material conditions have developed very differently for different generations. They have improved for persons over 45 years of age, but have deteriorated for the younger age groups during the early 1990s. Note the regression during the 1990s: In the Swedish case this was a period of rapid growth of unemployment, a decline of employment rates by one tenth, and a subsequent reduction of welfare benefits. The value of these presentations lies in the reduction of information, and focus on key issues of recent Swedish social change and reconstruction of the distributional structure in the 1990s. This graphic presentation was used for the entire set of social indicators as the main technique of presentation. NOTES 1
Including education, employment, working conditions, income, material living standards, housing, transport, leisure, social networks, participation, victimisation, health and social mobility. 2 The index is based on the following items: absence of overcrowded housing, high standard of housing space; dishwasher, car, second home, caravan, boat, video, freezer, access to a daily newspaper.
REFERENCES
Carley, M.: 1981, Social Measurement and Social Indicators: Issues of Policy and Theory (London). Horn, R.: 1980, ‘Social indicators’, International Journal of Social Economics 7, p.8. Johansson, S.: 1983, Toward a Theory of Social Reporting (Institute for Social Research, Stockholm). Journal of Social Policy, special edition: ‘Whatever happened to social indicators?’ Jnl Publ. Pol. 9, p. 4. Mapeta, W., K. Wallberg and J. Vogel: 1992, A Social Indicator Program for Zimbabwe. Identifying Objectives, Methods and Indicators for an Indicator Monitoring Survey. ZIMSTAT 1992: 5.
112
JOACHIM VOGEL
OECD: 1982, The OECD List of Social Indicators (Paris). Titmuss, R.: 1958, Essays on the Welfare State (London). United Nations: 1979, Studies in the Integration of Social Statistics: Technical Report; Studies and Methods, Series F, no. 24; New York. Vogel, J.: 1991, Social Report for the Nordic Countries. Living Conditions and Inequality in the late 1980’s. Statistical Reports of the Nordic Countries, Copenhagen. Vogel, J.: 1993a, ‘Class and Inequality. The Swedish Experience’, in R. Hansen, and Uusitalo och Eriksson (eds.), Welfare Trends in The Scandinavian Countries (Sharpe, Chapter 6). Vogel, J., H. Sauli and A. Andersen: 1993, Welfare Surveys and Social Reporting in the Nordic Countries. Experiences and Potential. Eurostat: doc E3/IS/5/93. Vogel, J: 1993c, ‘Wohlfahrtssurveys und Sozialberichterstattung in Nordeuropa. Systematische Dauerbeobachtung der objektiven Lebensverhältnisse’, in W. Glatzer (ed.), Einstellungen und Lebensbedingungen in Europa. Campus. Vogel, J.: 1994a, ‘Social indicators and social reporting. Traditions and current options for the development of comparative social indicators for the European Union’, Statistical Journal of the United Nations Economic Commission for Europe 11(4). Vogel, J.: 1994b, Living conditions, values, and attitudes of young adults in Sweden. Report prepared for The Commission on Living Conditions and Future Prospects of Young People (Stockholm). Vogel, J.: 1995a, Living conditions, values, and attitudes of young adults in Sweden. The Siena Group Seminar, Oslo, June 8–9th 1995. Statistical Journal of the United Nations Economic Commission for Europe 12(3,4). Vogel, J.: 1995b, Living conditions, values, and attitudes of young adults in Sweden. The Siena Group Seminar, Oslo, June 8–9th 1995. Statistical Journal of the United Nations Economic Commission for Europe 12(3,4). Vogel, J.: 1995c, Options for Integrated and Comparative Statistics on Living Conditions. Doc E2/IS/7/95, Eurostat, Luxembourg. Vogel, J.: 1995d, The Social Indicator Tradition. Doc E2/IS/2/95, Eurostat, Luxembourg. Vogel, J.: 1995e, Thoughts on the Dissemination of Social Statistics for the European Union. I. The Future of European Social Statistics. Proceedings of the third Mondorff seminar. Eurostat, Luxembourg. Vogel, J.: 1995f, Towards an Integrated Household Survey Programme. The Swedish Experience. Doc E2/IS/8/95, Eurostat, Luxembourg. Vogel, J.: 1997a, ‘The future direction of social indicator research’, Social Indicators Research 42(2), pp. 103–116. Vogel, J.: 1999, ‘The European ‘welfare mix’: institutional configuration and distributive outcome in Sweden and the European Union. A Longitudinal and Comparative Perspective, Social Indicators Research 48(3), pp. 245–297. Vogel, J. and L. Häll: 1997, Välfärd och ojämlikhet i 20-årsperspektiv 1975– 95 (Living conditions and Inequality 1975–95). Report 91, series Living Conditions. Statistics Sweden.
STRATEGIES AND TRADITIONS IN SWEDISH SOCIAL REPORTING
113
Vogel, J. and L. Häll: 2000, Äldres levnadsförhållanden 1980–1998 (Living Conditions of the Elderly 1980–1998). Report 93, series Levnadsförhållanden, Statistics Sweden. Vogel, J.: 1997d, Living Conditions and Inequality in the European Union. Eurostat: Population and Social Conditions E/1997-3. Luxembourg. Vogel, J.: 1998, Three Types of European Society. Internet: Nordic News Network,
[email protected]. Vogel, J.: 1998a, Coping with the European Welfare Mix. Welfare Delivery Systems, Family Formation and Material Inequality between types of Families in the European Union. Paper presented at the 1998 Siena Group Meeting on Families at the end of the Twentieth Century; Sidney, Australia, 7–9.12.1998. Vogel, J.: 1998b, Ambitions, problems and cure in comparative monetary poverty indices. Some recent experience from comparative research. High level think tank on poverty statistics, 1998-01-20-21, Stockholm. Vogel, J. and U. Davidsson: 1998, Monitoring Living Conditions: The Design of an Information System for Social Indicators and Social Reporting. Paper presented at the Joint IASS/IAOS Conference on Statistics for Economic and Social Development, Aguascalientes, Mexiko, September 1998.
Statistics Sweden Welfare Analysis Program Stockholm, Sweden University of Umeå E-mail:
[email protected] This page intentionally left blank
JEROEN BOELHOUWER
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
ABSTRACT. The four planning offices in the Netherlands form a network which covers all areas of government. This network matches with the policy-structure in Dutch administration. One of the planning offices, the Social and Cultural Planning Office (SCP), studies social and cultural aspects. Most of the SCP publications deal with several different aspects of the living conditions. As an integrative tool the SCP developed an overall monitoring instrument: the living conditions index (LCI). The LCI measures aspects of individual living conditions and is composed of indicators from eight areas: housing, health, leisure activity, ownership of consumer durables, sport activity, vacation, social participation and mobility. Education, income and paid employment are considered as resources and are used to describe the backgrounds of living conditions. Not only developments in living conditions for the Netherlands as a whole are monitored, but for various groups in society as well. Is the LCI useful for policy(makers)?
THE NETWORK OF PLANNING OFFICES IN THE NETHERLANDS
In The Netherlands there is a close network of planning offices, which covers almost all areas of government. In total, there are four planning offices, which co-operate in answering each major demand for information arising during the process of preparing policies. This planning organisation matches with the policy-structure in Dutch administration, where the ministries are the main centres for policymaking.1 Every ministry has its core-business, a specific area of concern. Each of the planning offices covers a specific aspect of government policies: The Central Planning Office (CPB) deals with economic affairs. It mainly forecasts economic developments and evaluates economic and budgetary policies. The National Physical Planning Agency (RPD) deals with the physical aspects like spatial planning and housing. Social Indicators Research 58: 115–140,2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
116
JEROEN BOELHOUWER
The National Institute of Public Health and the Environment (RIVM) takes care of the environmental planning. The Social and Cultural Planning Office (SCP) studies social and cultural aspects.2 The Social and Cultural Planning Office was established by Royal Decree in 1973.3 It is an inter-ministerial institute that conducts independent research on social and cultural developments and advises the policy-makers. The SCP considers all areas of government policy, as far as social aspects are involved. It concentrates on the fields of health, welfare, social security, labor market and education and especially on the intersections between these different social fields. The SCP monitors social changes and social services, evaluates social policies, gives information about expected developments in the future and gives recommendations for further policy. Key issues are the assessment of economic, demographic and social changes and the influence of these factors on the social and living conditions of the population. The government is the principal user of the SCP products. Ministers, the parliament and civil servants are the primary target groups of the SCP. Second categories are the regional and local authorities. They too need information about social trends and governmental social policies as a frame of reference for their own social policies: SCP reports frequently on local and regional developments. The publications are also used often for scientific purposes by universities and researchers. Other categories of consumers are the management and professionals of the service-sector, subsidized by government, like schools, hospitals and theatres. Though reports printed on paper are still the main products, today our website on the Internet becomes more and more important in publishing our findings. Each report is not only published on paper, but in its entirety on the Internet as well. A main target for the near future in this respect is to make publications specially designed for web publishing. Not only reports are published on the Internet. Most of the metadata on the surveys can be found there as well. Though still in a developing phase, some of our data are too: people can make their own cross-tabulations on the time use survey. They also can compute their own score on our living conditions index as well as compare it with average scores of others.
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
117
THE SCP AND SOCIAL INDICATORS
For a long time now, statistics are being used to express characteristics of society. In addition to statistics, indicators are used; a special kind of statistics. In the words of Bauer: “social indicators [... ] are statistics, statistical series, and all other forms of evidence that enable us to assess where we stand and are going with respect to our values and goals, and to evaluate specific programs and determine their impact” (Bauer, 1966). Or as the United States Department of Health put it in 1969: “A social indicator may be defined as a statistic of direct normative interest which facilitates concise, comprehensive and balanced judgements about the conditions of major aspects of a society. It is in all cases a direct measure of welfare [. . . ] if it changes in the ‘right’ direction, while other things remain equal, things have gotten better or people are ‘better off” ”(cited in Horn, 1993). The interest in social indicators and social studies in the Netherlands became visible not only by the establishment of the SCP in 1973, but by the earliest publications of the SCP as well. The first ever publication was titled “The Social and Cultural situation in the North of the Netherlands”. Other early publications dealt with education, care for the elderly, social indicators on health, car-use, social deprivation, time use, women (emancipation) and a description of state income and expenditure linked to social welfare claims by households. This was the period when the social indicator movement was at its peak. A little later, in the 1980’s, there was a dip in the attention paid to social indicators, especially the focus on an integrated index for well-being or for living conditions faded away. Researchers concentrated on the various sub-domains of living conditions like health and housing. After this dip, simultaneously with an increasing economy, the interest in social issues gradually revived. Not only for specific areas, but also for a measuring instrument that integrates various areas. An international example is the “Human Development Index” (HDI) of the United Nations Development Programme (UNDP), which integrates indicators in three areas: health, education and income, and compares various countries on the HDI.4 In The Netherlands too, governmental policy in the past years shifted from attention for different areas to a more integrative approach. Examples are the so-called “major cities policy” (with special atten-
118
JEROEN BOELHOUWER
tion for the specific, and diverse, problems of the four largest cities in The Netherlands) and the growing attention for livability (in the major cities as well as in the rural areas).5 Policymaking on improving the quality of life does nowadays not only aim at the struggle against unemployment, but also at improving living conditions, safety and social and societal participation. The most important publication of the SCP is the “Social and Cultural Report”. It is published every two years and deals with various aspects of Dutch society, such as: demographic developments; economy, work and social security; values and norms in society; health and health care; housing; education; and crime and justice. In this way it provides policymakers with an integrated overview of social and cultural developments in the Netherlands. Every Social and Cultural Report is based on a specific theme. In the most recent Reports the focus was on 25 years of social change in The Netherlands, European comparisons and the special policy for big cities. It contains analyses based on empirical data, official statistics and policy papers.6 Another important, periodical, publication is the “Social and Cultural Outlook” which is published yearly. The purpose of the Social and Cultural Outlook is to describe briefly (several aspects of) the living conditions of the Dutch population and to analyse current developments in important areas of government policy. In this way it combines longitudinal research with short-term areas of interest. Both the Social and Cultural Report and the Social and Cultural Outlook are published at the request of the Dutch Parliament. Other SCP-publications deal with special categories of the population (like the youth, the handicapped people and women) or with a specific domain of living conditions (like health, education, income or housing). The SCP also studies current topics, such as social exclusion, the consequences of long-term unemployment and the social position of the elderly. Sometimes these reports are explicitly asked for by policymakers, sometimes they are a result of our research and advisory programme, which is developed every two years and needs the consent of the Council of Ministers. Some publications are published on a regular base (like our yearly report on ethnic minorities, on the elderly or on the youth), others are one-time publications (like a report about the connection between
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
119
education and work in the quartary sector and the market or a report on backgrounds of employment of minority women). Each report gets more or less extensive press coverage, by which the general public (as well as the target groups mentioned above) get to know about the findings and results of SCP-research. Scientific aspects of SCP-research are highlighted in national and international journals. As said, the publications are also used for scientific purposes by universities and researchers. But the most important is of course the attention paid to our publications by ministries and members of parliament. They use SCP-publications very often; the ministries when they are preparing bills and the parliament when they discuss them, or when they question ministers. In the last couple of months SCP-publications were used when policies were discussed about various topics, like minorities, youth, social services, emancipation, education and (health) care. DATA
The SCP uses data from various sources. Statistics Netherlands (CBS) is the main provider of data. Besides, we collect some of the data ourselves, while other data come from ministries, universities, research institutes, international organisations and social organisations or institutions, etcetera. Examples of data we collect ourselves are surveys on time use, on cultural changes in the Netherlands and on facility use. Only recently we are expanding the data sources with focus groups and panel data. Via the Scientific Statistical Agency, the scientific world is supplied with a number of surveys on persons and households. This agency was founded to improve the accessibility and availability of data resources. In this way it is an intermediary between researchers and data providers. They co-operate with Statistics Netherlands and other data providers in providing micro-data for (scientific) research. A different source of data is the Steinmetz Archive.7 The archive holds a large collection of datasets in the social sciences, to make that collection available for secondary analysis as well as to the public.
120
JEROEN BOELHOUWER
A major, and growing, problem in the Netherlands is the low response rate; a response rate of less than 50% is not exceptional. Today SCP pays a lot of attention to finding ways of coping with this problem and to figuring out if people participating in surveys are different from those who don’t. At this moment the nonresponse of the Facilities Use Survey is examined by questioning – with a shortened questionnaire – a group of nonrespondents of the initial survey. Another data problem are the hard-to-get groups in society, like the homeless people or people in homes for the elderly or mental institutions. In regular surveys they are not covered. Sometimes we carry out special surveys on them (like people in homes for the aged), to be able to report on them as well. Data services are part of the information services at SCP. The first task of data services at SCP is determining what kind of data is required in a particular study. Researchers are urged to use existing data. For this purpose SCP has a wide range of data files available on-line: longitudinal series of population surveys from Statistics Netherlands; longitudinal series of surveys (partially) funded by SCP; on facility use, time budgets, cultural changes and opinions; international surveys (Eurobarometer waves, International Social Survey Programme (ISSP modules); institutional data (hospitals, schools); one-time data (troubled youth, opinions on social security, environmental behaviour); demographic data. For our reports we use all available data sources, creating something like a “best resources” situation. With this method we can explore the main theme of a specific report to a large extent. In a report on housing, data on housing of the youth can come from one source and data on housing of the elderly from another. In a report on women, data about income can come from another source than data about unpaid work. Though most of the surveys are multidimensional to some extent, it is hard to explore a broad range of subjects together from one source only. To get insight in the connections between various
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
121
parts of, for example, living conditions, a special survey is needed. As said, the need for integrated information in the Netherlands is growing. New policies on livability, social exclusion, deprivation and big cities are examples of social policies that need integrated data on various domains. When seeking solutions for problems connected with these topics, information about income or housing alone is not enough. Information about income, education, housing, social exclusion, deprivation and safety has to be available in the same data source if one wants to link different topics. The StateSecretary of the department of Health, Wellbeing and Sport stated that it would be good to “develop a monitor which can produce relevant figures for social policy. That is, figures about areas of living, safety, education, income, social networks, health and leisure activities” (policy letter on local social policy). To be able to report on developments on living conditions, the SCP started, in close co-operation with Statistics Netherlands, a specific survey on living conditions, as soon as in 1974. THE LIVING CONDITIONS INDEX
In every SCP report you will find a mixture of attention drawn to domains of living conditions and attention for special categories of the population. For example: a report on housing conditions deals not only with general housing conditions, but also with housing of the youth, the elderly or poor people. In reports on the youth, besides describing the general state of the youth attention is paid to their housing conditions, education, income, et cetera. As said: most of the SCP publications have a ‘best resources’ base. Most of the time, because of the different data resources being used, the connection between these aspects cannot be made. As an integrative tool the SCP developed in 1974 an overall monitoring instrument: the living conditions index (LCI), based on the survey on living conditions. The aims assigned to the study on living conditions in 1974 were fivefold: 1. The first aim was to depict the living conditions as a single entity. In the seventies integrated information was needed for bridging the opposite posed notions of ‘well-being’ and ‘deprivation’. Today
122
JEROEN BOELHOUWER
again there is a sharp focus on deprived groups in the Netherlands. At the same time, the approach to such problems is no longer in terms of a one-dimensional solution. Improving the quality of life is not only carried out through offering jobs, but via improving livability and social participation as well. To have the possibility to tell if things are getting worse or better, we have to 2. Evaluate this index in terms of positive and negative. To do so, we don’t want to publicise absolute figures each year. An important aim is to identify trends, so we want 3. To create a time series for observing changes. To get some clues on the cause of changes in the living conditions it can be a great help to see changes in the separate indicators. So the fourth aim set for the study was not only to monitor living conditions as a single entity, but 4. To monitor developments in the separate indicators over time as well. However, in order to explain changes over time we not only take into account changes in the separate indicators, we also use other information. We examine for example to what extent age, income, education and labour influence the living conditions. The last aim was therefore 5. To situate the description of social and cultural conditions in a broader context of background information. The next step in developing the LCI was to determine which areas were to be included in the umbrella notion of living conditions. An important starting-point was that a main task of the SCP is to make policy recommendations to the government. For that reason, clusters were selected that were (more or less) capable of being influenced by government policy. The actual indicators, as well as the variables comprising these indicators, were then defined within each cluster. Until 1993 the term used was ‘the index on well-being’, a term often associated with subjective feelings or with happiness. Since the index was designed to create an objectified picture, the SCP now prefers the term ‘living conditions index’.
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
123
Because no all-encompassing theory was available for making a selection, the SCP’s decision in 1974 was to choose indicators and variables which can be presumed in their totality to say something about a person’s current living conditions, or state of well-being. In so doing, the SCP adopted a position based more on pragmatics than on principle, keeping in view its primary objective of depicting living conditions as a comprehensive whole. We did, however, draw from the experiences of others, in particular the 1973 OECD list of indicators. A number of requirements were set for the indicators: They were to be focussed on ‘output’. The number of dwellings actually built was less important for the LCI than peoples housing conditions, and the number of doctors was not as important as the state of health of individuals. The indicators were to be general in nature. The indicators have to be applicable to the entire population rather than to specific groups. So there are, for example, no indicators about working conditions, as the unemployed obviously don’t enjoy them. They were to measure objectified characteristics of living conditions. Rather than speak of ‘objective indicators’, we use the term ‘objectified’ – as the choice of indicators, and the implicit assumption of what is good and what is not, is at least normative. Anyhow, the chosen indicators are not subjective nor satisfactions. This in order to avoid personal preferences and individual interpretations of concepts and, more important, because of our purpose, which is to be relevant for policymakers. It was to be clear, at least implicitly, whether they had a positive or a negative effect on well-being. Otherwise we wouldn’t be able to tell whether the living conditions as a whole is changing, whether it is getting better or worse. They also had to be measured at an individual level. This makes it possible not only to monitor developments at the national level, but to break down living conditions for different groups in society as well. Taken together, we believe the indicators provide an objectified description of individual living conditions. The LCI is nowadays composed of indicators from eight areas, as can been seen in Table I. This choice of clusters and indicators is, however, by no means final, and has changed over the years. I’ll come to this later on.
124
JEROEN BOELHOUWER
RELEVANCE OF THE LCI FOR POLICY(MAKERS)
What could be the relevance of the Living Conditions Index for policy and policymakers? Let’s start by recalling that the main aim of the LCI is to monitor the social and cultural developments in the Netherlands: is it getting better, or is it getting worse? In this way
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
125
the LCI is a descriptive instrument to identify trends in the living conditions. Moreover, the LCI can help us to keep track of deprived groups in society. If it goes better in the Netherlands, are there any groups lagging behind, and if so: which groups? This is related to an important goal of policy in the Netherlands: as many people as possible should profit from economic growth. Ideally, there should be an equal distribution of growth between groups (and individuals). When groups fall behind or do not profit for a longer period, it is possible that new policies have to be considered. However, the LCI can provide us only with a superficial glance of groups lagging behind and the causes of it. To tell something more about the exact reason of their deprivation, other resources are needed. With the LCI we can point out and identify deprived groups, but thereafter we have to look at other data to come up with a more complete answer about the reasons of their deprivation. That is, if it is possible at all to identify the exact causes of their deprivation. In social sciences it has proved very hard to identify exact causal relations between outcomes and policies. So many things influence developments in society that providing policymakers with an exact model of causalities is very near to impossible. An additional problem could be that not all the indicators comprising the LCI can equally be influenced by government policy. Though this was one of the requirements for the indicators set at the very beginning of the study on living conditions, we succeeded for some indicators better than for other. For example the number of psychosomatic symptoms is on the verge of being an objective and a subjective indicator. The same holds for the scale of social isolation. This is not to say that subjective indicators are unable of being influenced by policy at all, but it is very hard to tell what exactly caused a change in such indicators. On the other hand, this only is a minor problem, for a main goal of the LCI is to provide policymakers with a global image of living conditions, with which trends and deprived groups can be identified. Of course the worst off groups have an accumulation of arrears. Policymakers can judge the importance of arrears in one area different from arrears in another area. The judgement of people who don’t go to theatres or to museums (things government subsidises and likes people to do) might depend partly on other factors. It shall
126
JEROEN BOELHOUWER
be worse if they don’t go out to other places of entertainment as well, or if they don’t meet people in other ways, or if they are poor, etcetera. On average, arrears in more than one domain of living conditions are worse than arrears in one domain only. So, the relevance of the LCI for policy and policymakers can be divided in the following steps: 1. the LCI monitors the development of living conditions via an integrated way as well as via various domains and indicators, and for the Netherlands as a whole as well as for subgroups; 2. in this way we can also identify deprived groups; 3. when they are identified, causes of their deprivation can be looked for (if this is not possible by the living conditions survey, then other data have to be used); 4. on the basis of the importance given to the arrears and to the causes, new policies need to be found. However, we have to bear in mind that the LCI can only be interpreted in a relative sense. We can only tell if things are getting better or worse if we have a time series to look at. This goes for the development of living conditions in the Netherlands as a whole, as well as for the development of living conditions for subgroups. For example: traditionally the elderly have had bad living conditions compared to other age groups. By itself this is a not very worrying insight, for – on the average – their houses are smaller, and they are less able to participate in sport and in non-domestic entertainment activities than younger people. The situation becomes more alarming if their living conditions decline, or even worse, keep declining over a longer period of time. THE LIVING CONDITIONS BETWEEN 1974 AND 1999
Between 1974 and 1999 the living conditions of the Dutch population as a whole improved, as can be seen in Figure 1.8 Due to the 1980 (oil) crisis, there was a dip in 1983. Nevertheless, in 1986 the living conditions had reached the same level as before the crisis. The overall improvement in living conditions can be attributed primarily to a growing number of home owners and the greater number of household appliances. In addition, car ownership increased and more people were joining organisations and engaging in sports.9
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
127
In 1994 you can see a vertical line. Moreover, in 1993 as well as in 1997 the index equals 100 for ‘all’. This is done because of the changes that were made between 1993 and 1997. The index used in 1993 was somewhat different than the one used in 1997.10 Some of these changes will be discussed later on. Although the living conditions of most groups have improved, the pace of improvement is not equal. As said: with the LCI we can trace deprived or disadvantaged groups: who are they and what is the reason for their deprivation. As can be seen in the figure: the youngest age-group (18–24 years old) was confronted with deteriorating living conditions between 1989 and 1992. The reasons for this drop were diverse: their scores on most indicators declined. This conclusion led to questions in parliament (did the minister know about this?). We took a look at other data resources and found that their declining living conditions could be due to changes in the scholarship system, which gave students less spare time to spend at social, cultural and sport-activities and at paid work. Analysing the influence of the different clusters on the living conditions reveals that leisure activities and sport activities are the most important, whereas health is the less important cluster. Of the
128
JEROEN BOELHOUWER
indicators, non-domestic entertainment activities and the number of sports contribute the most to the living conditions. The least important indicators are a scary spot in the neighbourhood, the type of housing and the possession of a season ticket for the railways.11
CONSTRUCTING A SINGLE INDEX
An important question is how to integrate the different indicators into one single index. While money can be used as a counting unit to aggregate or compare economic indicators, no such unit is available for social indicators. Just like there is no theory available for selecting clusters or indicators, there is no theory for combining the indicators into one index as well. There are a lot of different possibilities to cope with this problem, such as consulting experts or summing the figures unweighted. One other method is to look at the change in percentages of the various indicators, and to construct an index as the unweighted average.12 For example: in year × 25% own a car and 30% participate in an organisation. In the next year of measurement, 50% own a car and 35% participate in an organisation. A combined index will rise from 100 to 115 [that is: 100 + ((50 – 25) + (35 – 30)/2)]. Another method is to define goals and calculate to what extent these goals are met. This method is used by the UNDP for their Human Development Index. Because of the lack of external criteria for weighting the indicators, the SCP chose a rather pragmatic solution. The starting point was the common dimension of the chosen indicators, which all contribute positively or negatively to the living conditions. Moreover, a single indicator which correlates better with the others should have a greater weight. Another starting point was that it had to be an individually based index, since we not only wanted to follow developments for the Netherlands as a whole, but for different groups in society as well. We decided to statistically construct the living conditions index from the eight clusters mentioned earlier. In the beginning we used factor analysis to construct the index. We now use nonlinear canonical correlation analysis (the program is called OVERALS) to do so. This procedure enables us to cluster the indicators not only theoretically, but in the analysis as
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
129
well. Besides that, the procedure can be used to weight the indicators. In short, this analysis – which is a variation on principal component analysis – calculates the weights so as to maximise the sum of the item-total correlations. OVERALS also has the advantage that variables do not have to be measured at the interval level (which is an assumption of PCA). Another advantage is that the categories of the indicators receive weights besides the indicators themselves, thus enabling us to compare the categories too. OVERALS looks first at the influence of the separate indicators within one cluster. Then every cluster is equally weighted in the combined index; irrespective of the number of indicators included in a cluster. This also is an advantage to previously used techniques. Thus, the pragmatic, statistical option used for constructing the index gives us the possibility to look at the scores on different clusters as well. We can do so not only for the Netherlands as a whole, but for groups in society or regions as well. For example: the four biggest cities13 in the Netherlands traditionally have low scores on the living conditions index. As Figure 2 shows, this is mainly caused by low scores on the clusters of housing and mobility (there are more small flats and less car-owners in these cities compared to the rest of the country). On the other hand: the scores in the four largest cities on the clusters leisure activities and sport are better than in the rest of the Netherlands. This can, however, not compensate the low scores on the other clusters. Figure 2 tells us also that the scores on all domains of the 21 bigger cities are below average. Let us take a brief look at people living in the four largest cities and make a further distinction between Dutch people and minorities.14 As Figure 3 shows, the Dutch have better living conditions than the minorities, whether they live in the biggest cities or not. Also the figure shows that people living in the largest cities have worse living conditions than people living elsewhere do. In Figure 2 we noticed that people living in the four biggest cities have better than average scores on the clusters of housing and sport. This is the case for minorities in these cities, but the Dutch living there also have better than average scores on social participation and holiday.
130
JEROEN BOELHOUWER
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
131
These examples reveal that it is worthwhile to look not only at overall developments, but at developments in separate domains as well. CHANGES THROUGH THE YEARS AND CONTROVERSIAL INDICATORS IN THE INDEX
As said before, the choice of indicators and clusters was never meant to be final. The index we use today is not the same as the index used in 1974. What variables, indicators and clusters should be included or omitted has always been the subject of debate. Because of the lack of a theory for selecting indicators, there has been some controversy about indicators from the beginning. There have been some minor changes as well as some major changes in the construction on the LCI. Though the indicators have changed, the SCP tried to take great care to preserve the comparability of the data over time. Unfortunately the time series could not be continued after 1993 because of changes that were made in collecting the data by Statistics Netherlands. It changed its surveying design as well as the questions, the order in which they are asked, and the methods of selecting and contacting respondents. These alterations logically caused a break in the trend registered so far. On the other hand, this moment has provided us with a good opportunity to carry out a thorough revision of the LCI. Only a single break in the trend occurred, and a new time series has started. From 1997 on new data is collected every other year. That changes were needed had several reasons: 1. some parts of it had grown out of date. Items that were important 20 years ago are not necessarily so any longer. A consumer article then considered a luxury may now be commonplace, or even antiquated. Furthermore, due to technological progress there are now articles available that did not exist, or were not available to the general public, in 1974. So slide projectors have made way for personal computers and CD players. 2. policy nowadays deals with aspects of life that affect living conditions, not all of which could be covered adequately by the LCI. Examples are social participation, residential environment and public safety.
132
JEROEN BOELHOUWER
3. some indicators have been subject to discussion. For example: there has been some controversy about the role of car ownership in the index. People with cars score higher on living conditions than those without, and some observers asked whether that was rightly the case, in view of the growing awareness of the environment and of vehicle pollution. Although the pollutive nature of cars obviously cannot be denied, having a car does give many people a sense of freedom, of being able to come and go as they please. This is presumed to have a favourable effect on their living conditions. In view of the arguments on both sides, it has now been decided to include possession of a public transport season ticket alongside car ownership, since the former can engender a similar sense of freedom. We also used the opportunity to broaden and deepen several clusters, such as whether the vacation was in the Netherlands or abroad, and with greater numbers of sports and leisure activities specified in the survey. Furthermore, in 1997 the cluster ‘leisure activities’ was split into two new ones. One of them now contains indicators which not necessarily lead to social contacts: organisational membership; number of hobbies; and number of nondomestic entertainment activities. The other cluster (‘social participation’) contains indicators which can be expected to contribute to making social contacts. The splitting up has to do with the greater attention attached to social participation nowadays. We also included a separate cluster ‘sport activity’. Sport activities were formerly added to the cluster of ‘leisure activities’. The reason for creating a new cluster is that sport participation can have an effect on leisure activities as well as on health and social participation. Changes have also been made in combining the variables into one single index. Not only has the statistical method changed from factor analysis to OVERALS, some indicators have been removed from the index sober, like education and income. LIVING CONDITIONS AND RESOURCES
As you may have noticed, in the index itself there are no indicators for education, income, source of income or paid employment. They were left out in 1989. This change was motivated by theoretical
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
133
considerations. Until 1989 the living conditions index contained indicators which can be seen as ‘output’ or ‘realised opportunities’, like the type of housing and the activities someone participates in. On the other hand the index contained some elements which contribute to realising the opportunities, like level of education, income and paid employment. We thought that these elements could better be seen as resources which enhance peoples opportunities to achieve certain living conditions (see Figure 7 for a graphical presentation). So we now use these resources to describe the backgrounds of living conditions and have removed them from the index. When describing these backgrounds we posit a causal relation: the more resources a person has, the more likely he or she is to enjoy better living conditions. In this way the living conditions index reflects the degree to which people realise their opportunities. This makes it possible to consider if government policy is needed to improve the conditions by which more people can realise more of their opportunities. The removal of the resources from the LCI did, however, not affect the main structure of the index. Without these indicators, the remaining ones correlated in the same way as they did before. The splitting up of living conditions and resources, and the theoretical considerations that motivated it, now raise the question of what to do with health. From the standpoint that resources create the preconditions for achieving certain living conditions, there is something to be said for classifying the health cluster under the resources. One might argue that good health is a prerequisite for good living conditions as defined by the index. On the other hand, one could also regard peoples state of health as an outcome of government policies or other factors that make for optimal living conditions. For this reason we have decided thus far to retain health as part of the index proper. Analysis of the relationship between resources and the LCI reveals that the assumed relationship holds true: having a high level of education, a paid job and a good income causes better living conditions. Looking at some background characteristics tells that people older than 75 years and one parent families as well as single person households, have bad living conditions. Taken together, the resources explain (in a multivariate analysis with age and house-
134
JEROEN BOELHOUWER
hold composition) for about 50% of the variance in the living conditions.15 In 1999 we constructed an individually based socio-economic deprivation index, made up of (low) income, (low) education and (no) paid work.16 Between 1983 and 1993 the percentage of not deprived people, who are younger than 65 years, rose from 29% to 48%. In 1993 only 6% are very deprived. The same development was found for people older than 65 years, though on a different level: a higher percentage of the elderly is deprived. In 1993 28% of the 65+ are not deprived on any of the indicators, whereas almost half of the younger people are not deprived at all. So not only were the living conditions better in 1993/1997 than they were in 1983, also less people were deprived. In figure 4 the relationship is shown between the deprivation index and the living conditions index. The combination between the two shows us that the group of people that are very deprived as well as younger than 65 have worse living conditions in 1993/7 than they had in 1983. Besides: the gap between the younger non-deprived group and the older very deprived group has widened -though not that much- in the same period.
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
135
OBJECTIVE LIVING CONDITIONS AND SUBJECTIVE SATISFACTIONS
There are also no subjective opinions of individuals in the index. The most important argument for leaving them out, is that we analyse living conditions in terms of policy-relevant indicators. Our purpose is to furnish the government with concrete handles for achieving a desired result. Identifying the exact sources and determinants of happiness would be so complicated, if at all possible, that research on subjective feelings would yield little information relevant to policy-making. But, because of their obvious relationship, we regularly examine the relationships between the living conditions index, satisfactions, subjective health and general happiness (see Figure 7 for a graphical presentation). Analysis of this relation shows that the better someone’s living conditions are, the happier he or she is. Moreover, the correlation between happiness and the overall SCP index is stronger than that between happiness and the separate indicators in the index.17 In other words, combining the different aspects of living conditions into a single index is also worthwhile for studying how those conditions are related to happiness. We also analyse the relationship between the living conditions and subjective health. The same relationship applies as with happiness: the better the living conditions are, the healthier someone finds him/or herself.18 There have been some changes over time: in 1977 84% of the Dutch population were said to be (very) happy, compared to 88% in 1997. The percentage not so happy and unhappy people remained stable at 3%. For self-reported health there is a similar development. Where in 1974 21% were said to have a very good health and 31% to have a good health, in 1997 these percentages have risen to 23% and 55%. The percentages of people whose health is only so-so, or even bad, fell from 12% and 9% in 1974 to 5% and 3% in 1997. The living conditions of people saying they are (very) happy or considering themselves to be (very) healthy improved between the 70’s and the 90’s, along the same lines as the overall development of living conditions in the Netherlands (see Figure 5). This is not the case for the (somewhat) unhappy and unhealthy people, their living conditions remained stable at the best. Moreover: the differ-
136
JEROEN BOELHOUWER
ence between people with (very) good health and bad health grew from 11 points in 1974 to 16 points in 1997. The gap between people that are (very) happy and unhappy people widened from 15 points in 1977 to 23 points in 1997. We also regularly examine the relationship between the LCI and satisfaction with various domains of the living conditions. In 1997 these questions had to be answered by a 10 point scale; 10 being most satisfied and 0 being least satisfied. As can be seen in Figure 6 only 2% were unsatisfied with their housing, whereas 25% were (very) satisfied. In the same figure the relationship between satisfaction and living conditions shows that the more satisfied people are, the better their living conditions are.
THE FUTURE OF THE LIVING CONDITIONS INDEX
The findings on the living conditions index have been published since 1974. At the beginning bi-annually in the Social and Cultural Reports, later on yearly in the Social and Cultural Outlooks. In the past years findings about the LCI have been published in a chapter called living conditions. In this chapter successively the develop-
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
137
ment of income, education, position on the labour market and the LCI were discussed. The first three topics on basis of other data sources than the living conditions survey in order to get the optimum findings. From 2001 on, there will be new design of the Social and Cultural Outlooks. From then on it will be called “The Social Side of the Netherlands”, which is going to be published bi-annually. In this volume the LCI will be related not only to the resources (income, education, labour market) and to subjective indicators, but to other indices as well. The draft social model that we will be working with is shown in Figure 7. Where the living conditions index is an individually measured gauge, the other indices are measured on a neighbourhood level. Nowadays policymakers in the Netherlands are very interested in figures about socially deprived neighbourhoods and districts. Because of the growing interest in developments between cities and between areas in cities, the SCP is working on an index on livability. In the future we shall link the LCI to this new index and to other social indicators and indices that are measured on a neighbourhood level. For example, in the past year findings have been published about indices on socio-economic deprivation, crime and lack of
138
JEROEN BOELHOUWER
security and level of services. In the future these indices will be linked with each other as well as with the living conditions index. This gives us the opportunity to see in which areas people with bad living conditions live. Do these neighbourhoods differ from where people with good living conditions live, or do they live in more deprived and unsafe areas? In this way we will get insights not only in accumulation of arrears at the individual level, but in possible spatial concentrations as well. NOTES 1
Speaking of planning can, in this context, be misleading, as the planning offices do not make a blueprint for the future society. They analyze developments, evaluate policy and make recommendations for improving that policy. 2 For more information about the four planning offices, see their websites. CPB: http://www.cpb.nl/eng RIVM: http://www.rivm.nl/index_en.html RPD: http://www.minvrom.nlSCP: http://www.scp.nl/defaultuk.htm 3 In the Royal Decree the official definition of the task is threefold: 1. to carry out research designed to produce a coherent picture of the state of social and cultural welfare in the Netherlands and likely developments in this area;
QUALITY OF LIFE AND LIVING CONDITIONS IN THE NETHERLANDS
139
2. to contribute to the appropriate selection of policy objectives and to provide an assessment of the advantages and disadvantages of the various means of achieving those ends; 3. to seek information on the way in which interdepartmental policy on social and cultural welfare is implemented with a view to assessing its implementation. 4 See the Human Development Reports. 5 Policy on livability is about the absence of crime, violence and garbage on the streets, about housing and environment, green spots, trees, basic facilities and services – like a school or a shop – within walking distance. It is about the situation of the neighborhood people live in and how they feel about it. 6 English translations of the latest Social and Cultural Reports are available and can be ordered via our website at http://www.scp.nl. 7 The Steinmetz Archive, founded in 1962, is part of the Netherlands Institute of Scientific Information Services (NIWI, http://www.niwi.knaw.nl/us/ homepag.htm) of the Royal Netherlands Academy of Arts and Sciences (KNAW). 8 It should be noted, however, that the survey for the living conditions index is only held among persons living on their own (and their household). This implies that people (temporarily) living in institutes or without a permanent address are not questioned, by which they are no part of the sample. 9 See Boelhouwer and Stoop (1999). 10 Until 1993 we only used four clusters (housing, health, ownership of consumer durables and leisure activities). 11 For more information about the contribution of clusters and indicators to the living conditions, see my paper for the ISQOLS conference 1998. Available at: http://www.scp.nl/users/Boelhouwer/ISQOLS.htm 12 See for example the Pierce Quality of Life Benchmark Project and the Quality of Life Index as used in Ontario. 13 The four largest cities in the Netherlands are Amsterdam, Rotterdam, The Hague and Utrecht. 14 Minorities meaning in this context that (one of) the parents of the respondent was not born in The Netherlands. 15 See Boelhouwer and Stoop (1999, p 63) and several Social and Cultural Outlooks. 16 A relative simple index is made, with a minimum score of 0 if someone has no low income, no low education and paid work A maximum score of 3 is given to the opposite situation. As only few people older than 65 have a paid job, their maximum score is 2. See the Social and Cultural Outlook 1999 for more information. It should be noted that the questions on education have changed several times since 1993. 17 See Boelhouwer and Stoop (1999: p. 62). 18 See several Social and Cultural Outlooks.
140
JEROEN BOELHOUWER
REFERENCES
Bauer, R.A. (ed.): 1966, Social Indicators (MIT Press, Cambridge, MA). Boelhouwer, J. and I. Stoop: 1999, ‘Measuring well-being in the Netherlands. The SCP index from 1974 to 1997’, Social Indicators Research 48, pp. 51–75. Horn, R.V.: 1993, Statistical Indicators for the Economic Sciences (Cambridge University Press, Cambridge). Ministry of Health, Well Being and Sport: 1997, Beleidsbrief Sociaal Lokaal Beleid (policy letter on social local policies). Mootz, M. and M. Konings-van der Snoek: 1997, De mate van welzijn: Verdeling en concentraties van welzijnsaspecten 1974–1989 (Social and Cultural Planning Office/VUGA, Rijswijk/The Hague). Ontario Social Development Council: 1997, Quality of Life in Ontario (http:// www.lks.net/cdc/spno/report.html). Organisation for Economic Co-operation and Development: 1973, List of Social Concerns. Common to Most OECD Countries (OECD, Paris). Pierce County Department of Community Services: 1998, Quality of Life – Benchmarks (http://www.co.pierce.wa.us/services/family/benchmrk/qol.htm). Social and Cultural Planning Office: various editions, Social and Cultural Outlook (Social and Cultural Planning Office/VUGA, Rijswijk/The Hague). Social and Cultural Planning Office: various editions, Social and Cultural Report (Social and Cultural Planning Office/VUGA, Rijswijk/The Hague). United Nations Development Programme: various editions, Human Development Report (Oxford University Press, New York/Oxford).
Social and Cultural Planning Office P.O. Box 16164 2500 BD The Hague The Netherlands Website: http://www.scp.nl E-mail:
[email protected] CHARLES S. KAMEN
“QUALITY OF LIFE” RESEARCH AT THE ISRAEL CENTRAL BUREAU OF STATISTICS: SOCIAL INDICATORS AND SOCIAL SURVEYS
ABSTRACT. “Quality of Life” research at the Israel Central Bureau of Statistics is headed by an integrated pair of projects: the series on Social Indicators and the Social Survey. The development of these two projects raises issues of harmonization of measures with other departments in the Israel CBS as well as with international recommendations. Quality-of-life research necessarily requires collecting information about respondents’ evaluations of their situation, but proposals to include such questions in surveys carried out by National Statistical Organizations may meet with opposition which argues that the information provided is not “objective.” The quality-of-life research program at the Israel CBS has a dual goal: to provide baseline information and to be policy relevant.
INTRODUCTION
Two projects involving Quality of Life Studies are currently underway at the Israel Central Bureau of Statistics: a program of social indicator reports and a social survey. I will discuss these projects under the following headings: 1. Quality-of-life research at the Bureau 2. A description of each of the projects 3. Coordination between the projects and harmonization of measures 4. “Objective” vs. “subjective” questions 5. Planning for policy relevance
QUALITY-OF-LIFE RESEARCH AT THE BUREAU
Like Moliere’s “bourgeois gentilhomme,” who realized that he had been speaking prose without knowing it, the Bureau has for years Social Indicators Research 58: 141–162, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
142
CHARLES S. KAMEN
carried out research on “quality of life” without being fully aware of it. What has changed, of course, is not the work that we are doing, but the way in which it can be conceptualised. The growing recognition that the success of government policies must ultimately be measured not only in economic terms but also by their effects on the daily lives of people, means that increasing attention is paid to measuring these effects in different areas of social life. The concept “quality of life” provides a lens through which we may re-view our activities, reorganize the presentation of the information we collect so that it serves new needs, and develop projects which provide new information. The Bureau’s data gathering activities are divided between field surveys and utilization of administrative records. Israel has a national population registry, and each citizen has a personal identification number which is recorded as part of the standard procedure in most of his contacts with public and private agencies and organizations. Our hope is to make increasing use of administrative records and data linkage to take advantage of the availability of ID number, and reduce respondent burden. We are also evaluating the best way to obtain this number from respondents in those household surveys whose samples are drawn from lists of dwellings and not from the population registry. The two Bureau programs I will discuss today, social indicators and the social survey, are responses to these changing needs. The social indicators project pulls together published data on specific social areas from many Bureau surveys, as well as from other sources, and presents it in a series of publications, each devoted to a particular topic. The social survey will provide a platform which collects standard measures, includes new questions, and brings together in a single data file items currently found in separate surveys. Before describing these programs in detail, let me begin by briefly reviewing the prose we have been speaking for years, focusing on some of the social data we collect regarding individuals (Not included in the list are surveys collecting institutional data – on libraries, for example, or on schools): 1. The continuing Labour Force Survey (LFS), which collects information on household composition, labour force participa-
“QUALITY OF LIFE”
2. 3. 4. 5. 6. 7. 8. 9. 10.
143
tion, educational attainment and, beginning in 2000, questions about persons who have given up looking for work because they’ve despaired of finding a job. Other surveys on specific topics often “piggyback” on the LFS. Utilization of health services, a piggyback on the LFS. Victimization from criminal activities, a piggyback on the LFS. A survey of persons aged 60 and older, covering many areas of social life. Surveys of holders of academic degrees, focusing on education and employment. Transportation survey, focusing on commuting and other transportation-related behaviour. Survey of leisure activities. Survey of leisure activities of youth. A time-use survey. Administrative files of educational participation (all the following files contain individual records identified by Population Registry ID number):
10.1 Annual files of high school students. 10.2 Annual files of matriculation examinees and examination results. 10.3 Annual files of applicants to colleges and universities, students and degree recipients. 11. Administrative files dealing with health (also identified by ID number):
11.1 Births 11.2 Cause of death 11.3 Hospitalization and cause This information is scattered throughout many files and publications. Some of it is available annually; other data is collected at less regular intervals. Taken together, it comprises a rich source of longterm indicators on many aspects of society in Israel. The “quality of life” perspective provides a new way of looking at this data, and suggests new directions for organizing and presenting it. Neither the social indicators project nor the social survey is unique in their content. They are the result of reconsidering existing
144
CHARLES S. KAMEN
approaches, using the concept of “quality of life” to guide our thinking. A DESCRIPTION OF EACH OF THE PROJECTS
Both the social indicator project and the social survey are part of the Bureau’s Social and Welfare Statistics branch, which is also responsible for statistics on education, health, culture, social welfare, and criminal justice. Planning of each of the projects is carried out in coordination with internal and external steering committees, composed of representatives of users within and outside the Bureau. Funding for the social indicator project comes from the Bureau’s internal budget. The social survey is funded by a grant from the Budget Division of the Ministry of Finance. Social Indicators
The current social indicators program at the Israel Central Bureau of Statistics represents a re-birth and expansion of a project which began in the late 1970’s and culminated with the appearance, in 1976 and 1980, of two volumes in a series which was called Society in Israel. That initial program did not continue beyond those first two volumes. Society in Israel was the response of the Bureau to the social indicator movement which first flowered in the early 1970’s, and an attempt to present information about social conditions in Israel from the “social indicator” perspective. Society in Israel collected in a single volume tables which had appeared in many different Bureau publications reporting on various aspects of social life. The volumes were intended to “[provide] data on important questions which deserve investigation by social analysts.” The 1980 volume differed from the earlier publication in that it identified “major issues or concerns” which served as a basis for selecting the tables and a framework for organizing them. No new data were collected for either of the two publications, no reanalyses were undertaken, nor were the tables accompanied by a descriptive or an analytic text. The series stopped after the second volume, probably because the project was carried out more or less in isolation from the existing
“QUALITY OF LIFE”
145
departments of the Bureau rather than being integrated into one of them. It was, in fact, a one-man show, run by a person who was specially recruited for the purpose and who hoped the results would be sufficiently convincing to establish social indicators as an ongoing program at the Bureau. Despite these intentions, the organizational setting of the project reflected the prevailing view that economic statistics were what really mattered, and its demise probably reinforced that understanding. In 1995 the Bureau recommitted itself to publishing social indicators, in circumstances which differed greatly from those which had led to the earlier undertaking. During the fifteen years which elapsed since the end of the first project, significant changes occurred in Israel which heightened the salience of and the need for information on social conditions. These changes included a trend toward privatization in different areas of social and economic life whose consequences included an expanded variety of institutional arrangements as well as growing economic polarization. Among the effects of such changes was an increased willingness to admit that existing social problems were not transitory phenomena but rooted in fundamental social-structural arrangements. This recognition was accompanied by a greater willingness to see interrelations between economic and social issues. Moreover, a growing environmental movement was accompanied by increased concern for quality-oflife issues, as a result both of developments within Israel and the influence of similar developments in Europe and in the United States. The social indicator publication program presently has two components – a series of subject volumes, and an annual social report. Originally, the intention was to issue a series of subject volumes covering the major areas of social life, on a rotating basis, over a period of (say) five years. After the appearance of the first two publications, on health and on education, it became clear that the resources available were insufficient to keep to the original timetable of a five-year cycle. In addition, it was felt that there was need for an annual summary volume in addition to the detailed subject publications. We decided, therefore, to issue this year the first annual publication. We were fortunate to have access to the social reports
146
CHARLES S. KAMEN
issued by other countries, which served as examples of alternative approaches to our own, and we used them extensively. Our social indicator reports are divided between text, tables and graphs. In the text we attempt to detail the significance for the welfare of the population of the conditions shown by the indicators. We are not yet certain what the best format is for the annual report. This year’s volume presents a summary of conditions in the main areas of social life, but our plan for next year is to focus on differences among the country’s regions. The current social indicators program at the Bureau is based on a conceptual scheme which identifies “domains” – areas of substantive interest such as health status, labor and employment, public order or social mobility – and “concerns,” which comprise, on the one hand, institutional means and resources, and desired policy outcomes on the other. Thus, the project is explicitly policy-oriented in its focus. Although it does not attempt to evaluate policy outcomes of specific government programs, the social indicators project identifies what might be called “societal goals” – education for all; a healthy life; adequate leisure opportunities; etc. – and examines progress toward them in two ways. First, it presents longitudinal data on change over time in basic indicators in each of the substantive areas. Second, it compares different social groups on the degree to which their members have realized these goals. The groups of interest may vary depending on the domain, but always include comparisons between Arabs and Jews, and between socio-economic groups within each of the communities. Identification of “goals,” of course, is less simple than I have made it sound. Social cleavages in Israel overlap in part with ideological cleavages, and the differentiation in Israeli society to which I referred above is accompanied by increasing legitimation of “multiculturalism,” a term originating in the United States and applied now in Israel. For many years there was a hegemonic “public ideology” which, even if not accepted by some sectors of the population, was seldom publicly challenged. This has changed, especially with respect to the three fundamental social divisions which have characterized Israeli society since the beginning of modern Jewish settlement in Palestine: between Jews of European origin and those
“QUALITY OF LIFE”
147
from North Africa and the Middle East; between Jews and Arabs; and between religious and secular Jews. So far, the social indicators project has operated as if there was consensus on basic societal goals, and the first two subject reports on health and education might be seen to refer to areas about which there could be little disagreement about desirable outcomes. While that may be true in the area of health, it is not the case regarding education. There is a segment of the Orthodox Jewish community which rejects secular education, especially for males, and emphasizes instead lifelong religious studies. The only “standard” measure which could provide comparable information on educational attainment for both Orthodox and non-Orthodox sectors of the Jewish population is “number of years of schooling.” But the meaning of this number is completely different for someone who is moving from high school through college and professional training, accumulating (say) twenty years of schooling along the way, compared to one whose twenty years of schooling have been spent in longterm, full-time religious studies which were not undertaken for the purpose of gaining occupational qualifications. While the number of persons involved in such religious studies may be relatively small, having only a minor effect on national statistics, their existence demonstrates the difficulty of arriving at consensus on goals. It is possible to present examples in other areas as well – for example, if more families have cars, is that a good thing or a bad thing? It may be good for individual family mobility, while bad for traffic conditions, open space, accidents and pollution. One solution to the problem of presenting data on the attainment of social goals when no consensus exists about them is to develop indicators which show the consequences according to alternative views of what is desirable: the relation between family car ownership and access to vacation sites or leisure activities; effect on commuting time; on residential dispersion; on pollution from motor vehicles; etc. Such an approach would recognize that there are both benefits and costs which can not necessarily be evaluated according to a single set of criteria. You might call this a “multicultural” approach to social indicators. We have not yet considered taking this route.
148
CHARLES S. KAMEN
The Social Survey
The social survey is a new undertaking; the first wave is scheduled to go into the field in January, 2002. It is planned as a continuing survey, with fieldwork spread over the course of an entire year. Each year a new wave will be introduced. The survey is based on a sample of persons from the population registry, not a sample of households. Data collection will be carried out via computer-assisted personal interviewing (CAPI); the interviewer will have a laptop computer containing the questionaire in BLAISE format. The desired final sample size is 7,000 persons, aged 20 and older. The lower age limit was set because almost all persons younger than 20 are in school, living in their parents’ household, or in compulsory army service, and not yet part of the “adult” world. Since the sample for the social survey is drawn from the population registry, which contains basic demographic information for each person, it would be possible to extend the coverage of a future wave of the survey to respondents younger than 20 if required by a particular topic. We had to decide in the initial planning stages whether the social survey would have a panel structure, either in whole or in part. There are, of course, many advantages to a longitudinal panel design, in particular the possibility of measuring change over time in characteristics of individuals, not just in the distribution of characteristics in the sample as a whole or in subgroups. We decided, at least for the initial waves of the social survey, to base it on a series of annual cross-sectional samples. Our decision was made entirely on practical grounds. The social survey is a new undertaking. We were already introducing two major new elements in its design which have not previously been used at the Bureau: sampling persons from the Population Registry, rather than sampling households; and carrying out the field work using CAPI. We felt that two major design innovations were more than enough for the first waves of the social survey, and that a panel design would impose a planning and operational burden we were unwilling to incur. Since it would always be possible to transform one of the samples, or part of it, into a panel for subsequent waves, we didn’t feel that we were making an irrevocable decision. On the other hand, moving to a panel design would require making substantial changes in an
“QUALITY OF LIFE”
149
ongoing project, and might need a stronger justification than would have been necessary were a panel design planned from the outset. The questionnaire has two parts, a set of core items and a variable module. The interview is designed to last less than an hour, with half the time devoted to each part. The core items will be repeated each year. They cover the main subject areas and collect information both on the circumstances of the respondents and their evaluation of these circumstances (e.g., satisfaction, expectations for the future, etc.). The areas covered include household composition; principal activity of each household member; educational attainment; housing conditions; car ownership; household help; health status; religiosity; computer use; employment; journey to work; vacations; family and social relationships; victimization; volunteering; income; environmental quality. The topic of the variable module for the first wave of the social survey will be pension coverage, at the request of the Ministry of Finance Budget Division. There is a growing range of pension schemes in Israel, including a basic social security payment, two basic arrangements for government employees; different employerfunded plans; and private schemes. No systematic data exists on the extent of pension coverage of the Israeli population, or on what people know about their pension rights. Among other reasons for choosing the topic of pensions was a fear among Budget Division staff that many people had inadequate pension coverage, and that they expected the government to provide them with an income once they ceased working. The process of defining the content of the core questions differed from that by which the topic of the variable module was determined. The core items were chosen to provide basic information on a range of social topics. We wanted the core items to reflect the different areas of social statistics at the Bureau as well as important topics not yet covered by the current program of social statistics. We also looked at questionaires from social surveys in other countries to get ideas for additional items which could be included. The main dilemmas we faced in deciding on the core questions were of two kinds: Is a particular topic of sufficient interest to be investigated each year? What degree of detail is needed to cover that topic? A draft of the proposed questions was distributed within the Bureau
150
CHARLES S. KAMEN
for comments, and it was also presented to the external steering committee. Determining the content of the variable module raised an unusual problem. Unlike all the other surveys carried out by the Bureau, the social survey had no “client,” in the sense of a funding ministry, agency or organization with specific data needs which the survey would have to meet. The core items had no client either, but this was less of a problem because the core was in many respects “standard” in content, similar to that of other social surveys, and the main problems which arose regarding it were, as I already noted, those of space limitations. By including a variable module in a survey whose funding was already covered, the Bureau was in essence providing, at no charge, an extremely valuable platform for the collection of social data on any issue which could appropriately be investigated through a sample of individuals. It was important to us that the Budget Division be strongly involved in the selection of the topic for the variable module, for a number of reasons. We saw the social survey as an instrument which combined the collection of basic data on social conditions for long-term monitoring with policy-relevant research. Social policy decisions in Israel are often based on insufficient data, and we believed that data provided by the social survey could serve as an example of the importance and utility of policy-relevant data. We also believed that providing useful data to the Budget Division would increase the likelihood of continued funding for the survey.
COORDINATION BETWEEN THE PROJECTS AND HARMONIZATION OF MEASURES
“Social indicators” and “social surveys” are complementary tools for describing “quality of life” and understanding the relations among QOL measures. As such, it makes sense to develop them jointly, in order that each can provide information needed by the other. Data collected in a social survey can enrich the social indicator reports. On the other hand, the preparation of social indicator reports leads to the identification of additional data needs, some of which may be met using the social survey. In our case, both projects are located in the same branch, under a single authority, which
“QUALITY OF LIFE”
151
in principle should facilitate coordination between them. Since, however, each of the programs exists as an independent department, coordination is not automatic. Moreover, the social indicator project had already developed a conceptual scheme and gotten underway more than three years before funding was assured for the social survey. Therefore, the two projects could not be planned according to the same timetable and approach. Issues of coordination exist not only between the social indicators program and the social survey, but between each of them and other departments at the Bureau. These issues are also connected to problems of the harmonization of measures, both within the Bureau and with international standards. Since “measures” exist in an organizational context, and their creation is the result of the organization’s procedures, problems of harmonization and coordination arise within the broader context of relations between different departments in the organization and cannot be understood in isolation from them. Projects such as the social survey or the social indicators program create problems of coordination for organizations such as the Bureau whose work is organized according to a division into subject areas. Similar problems arise in the case of other projects which focus not on a specific topic, but on a specific population. Examples include a survey of holders of academic degrees; a survey of persons aged 60 or older; and, of course, a census. Such projects are, in their subject matter, “interdepartmental.” They include topics drawn from the different subject departments – education, labor, health, leisure, etc. – and combine them to provide an overall picture of the social and economic characteristics of the group which is the focus of the study. In the case of the Bureau, responsibility for the different subject areas which are included in such a survey does not rest with the subject departments, but with the department running the survey, which will “borrow” standard items from them. Problems of coordination and harmonization immediately arise within the organization as a result of the way work is divided. While it is desirable to involve subject departments in planning the “interdisciplinary” survey, in order to benefit from their experience, the establishment of the survey in a separate department means that staff with subject expertise have no formal responsibility for it and are
152
CHARLES S. KAMEN
likely to have little free time. Their formal involvement, therefore, tends to be limited, although staff members expressing more interest are often consulted informally as the need arises. The “borrowing” of standard items also raises problems. If the standard item is copied in full, there is usually no problem. Sometimes, however, it isn’t necessary to use the entire set of standard questions. For example, the planners of the social survey may wish to classify respondents according to their main activity (work, schooling, etc.). One way of obtaining the necessary information would be to include the standard set of questions relating to participation in the labour force and, on the basis of the answers to them, characterize the situation of the respondent. If one of the aims of the social survey is to provide information on the extent of labour force participation, it is reasonable to include the full set of items. Otherwise, estimates of participation based on the social survey will differ from those provided by the Labour Force Survey. If, however, the aim is simply to classify respondents according to their main activity, but not provide estimates of labor force participation, including the standard set of items may be both wasteful and unnecessary. It might be sufficient to ask respondents to classify themselves. Such an approach might be felt to constitute an “infringement” on the expertise and judgement of the subject matter specialists. It may also lead to different estimates of the same phenomenon which would then have to be explained. No one denies the desirability of harmonizing measures across surveys. The preceding discussion suggests the difficulty of doing so within a single organization, so it should not be surprising that it is even more difficult to harmonize measures developed in different organizations, much less in different countries. We confronted this difficulty in defining the content of the social survey, and in essence abandoned any attempt to harmonize wording and concepts. I’ll try to illustrate some of the difficulties we faced using examples from the European Community Household Panel (ECHP). “Harmonization” has at least two aspects: conceptualizing the topics; and wording the questions. There are topics that all statistical bureaus are interested in, such as “labor force,” and at that level harmonization presents no difficulty. Israel conducts an ongoing labor force survey, and computes the labor force participation rate,
“QUALITY OF LIFE”
153
based on a standard set of questions which adhere to ILO recommendations. The ECHP asks these questions of all persons aged 16 or older. In Israel, the age threshold is 15. Thus, even if we asked the identical questions to those asked in the ECHP, our results would not be comparable without additional data processing to omit 15-yearolds from the tabulations, as well as correcting the population base used in the computation. These changes are not difficult, and the information needed to make them is readily available. But the wording of one of the questions we ask to determine labor force participation is not identical to that in the ECHP. For example, the ECHP asks whether the respondent is “presently” working. In Israel we ask whether the person worked “last week.” The difference in the wording might not make any difference in the results, but the wording still differs. It wouldn’t be easy for us to harmonize the wording with that of the ECHP, even though the change appears negligible. Other differences between the ECHP and our LFS can be seen in the table of possible reasons for not seeking work (Table I). There may be good reasons for these differences, related to the specific social conditions in Israel. It would be very difficult to replace the list of reasons in Israel’s LFS with that used by the ECHP, but harmonization suffers if this is not done. Were the Israeli social survey to include questions on the reason for not seeking work, it would not be possible for us to use categories other than those in our LFS, because our primary responsibility would be to harmonize within the Bureau. Problems of harmonization also arise in the development of new measures, and they may also have organizational aspects. Bureau surveys don’t now categorize respondents according to their principal “activity status” (employed, student, retired, keeping house, etc.). We hope to do so in the social survey. Information regarding activity status could be used by other departments in the Bureau, and the question could become a standard item in many surveys. The most natural location for a standard question on activity status would probably be the LFS, but the department responsible for the LFS is not currently involved in developing such a measure. It wouldn’t be desirable for the social survey to determine on its own the operational definition of a measure of activity status which would have general relevance. Were a decision to be made to add activity status
154
CHARLES S. KAMEN
to the LFS, a committee would be formed, representing the different departments having a substantive interest in the measure. In our case, we will also form a committee, but it will have a different standing, and lead to problems of coordination because of the need to adhere to the planning timetable for the social survey.
“QUALITY OF LIFE”
155
“OBJECTIVE” VS. “SUBJECTIVE” QUESTIONS
In planning the social survey, we believed that it had to include questions asking people how they feel, and how they evaluate their condition. In my view, it is misleading to characterize questions as “objective” or “subjective.” Asking someone, for example, how much he makes, whether he is satisfied with his income, whether it suffices for his needs and whether he expects it to change in the future, are all factual questions about different aspects of income. By continuing to use the terms “objective” and “subjective” to describe questionaire items, we implicitly accept the legitimacy of this distinction I don’t think it is either necessary or appropriate here to justify the inclusion of items which provide information about how people evaluate their own condition; I assume that participants in this meeting recognize the necessity of asking such questions, and the value of the data they provide. Social policies are implemented in order to affect people’s lives. The judgements of these people regarding the effects on them of such policies are as valid as inferences about effects which are based on reports of behaviors which are presumed to reflect responses to these policies. Let me, instead, suggest how objections to asking such questions may be overcome. In Israel, at least, the strongest resistance over the years has come from the Bureau itself, rather than from our “clients.” Our willingness to include in our surveys questions about how people evaluate their situation has slowly increased over the years. Even in the past, when opposition to such items was much stronger, they were included nonetheless. In fact, not only were questions on satisfaction with housing included in a series of surveys of new immigrants which began in 1969; the results were also published in the Bureau’s annual Statistical Abstract. The Bureau’s survey on immigrants was its first to include a range of such questions. In addition to asking about satisfaction with housing, we similarly asked about work, salary, knowledge of Hebrew, whether respondents expected to be satisfied in Israel, whether they felt at home, and whether they intended to remain in the country. We also asked them how they defined “belonging.” We asked those immigrants who had contact with public agencies
156
CHARLES S. KAMEN
responsible for immigrant absorption whether they were satisfied with the treatment they received. The content of the survey of immigrant absorption was unusual for the Bureau. In large measure it was due to the fact that the survey was a cooperative endeavour between the Bureau, the Ministry of Immigrant Absorption, and the Israel Institute of Applied Social Research, a private, non-profit research organization established and directed by Louis Guttman. The Ministry of Absorption wanted to know how the new immigrants felt, and Guttman had long been interested in the structure of attitudes. Together they overcame the Bureau’s long-standing reluctance to ask people for their opinions and evaluations. While such items did not become standard in Bureau questionnaires, they continued to appear. Surveys of foreign students beginning in 1969 asked them, among other topics, to evaluate their academic program, their social life, and whether they intended to remain in Israel after completing their schooling. A 1979 survey of tourists asked for their evaluation of the quality of services to tourists, and whether they would recommend that others also visit Israel. The 1984 post-censual survey of academic degree holders inquired about respondents’ satisfaction with employment abroad, aspects of their job in Israel, and the relation between their job and their field of study. A 1991 time-use survey collected information about attitudes toward active participation in sport, toward volunteering and about satisfaction with work. The 1996 survey of persons aged 60 and older obtained data on: satisfaction with life in general, with housing, contacts with one’s children and neighbours and with medical treatment, overall level of spirits; health self-evaluation; feelings of loneliness; and attitudes toward work and retirement. A survey of “victimization” currently in the planning stage may ask respondents whether they were satisfied with the way the police handled their complaints, how serious was the damage they suffered, and whether they believe they would benefit from government-sponsored victim services programs. If not all these topics are included it will be for reasons of space, not because they are considered to be illegitimate. It seems, therefore, that we often ask people to evaluate their situation; we just don’t like to admit that we do it.
“QUALITY OF LIFE”
157
A number of factors contributed to what sociologists used to call “collective ignorance” with respect to Bureau practice, as opposed to principle, regarding asking respondents to evaluate their situation, and to greater willingness to include such items. Perhaps the most important was one I already mentioned earlier, in another context. Subject departments at the Bureau are relatively autonomous, and final decisions regarding questionnaire content are in their hands. While, as noted earlier, this independence may lead to problems of coordination, it also permits initiative on the part of staff willing to try new things. If the director of a survey wishes to ask respondents to evaluate their condition, he or she is free to do so. In earlier years, when such items were included less frequently, insuring their inclusion may have required considerable stubbornness; today, their inclusion is much easier. A second factor is generational. As the staff of the Bureau comes more and more to be composed of professionals who have had experience in their academic training and other work settings with surveys collecting information on attitudes and evaluations, their willingness to include such items in Bureau surveys increases, as does dissatisfaction with their omission. This process has a cumulative effect; the number of surveys which include such items increases, together with their acceptability. The demands of clients are a third reason for including respondent attitudes and evaluations in Bureau surveys. As awareness grows among policymakers and planners of the importance of obtaining information on public perceptions and evaluations, more demands will be made on us to provide such information. The Ministry of Internal Security, responsible for the police, wants to know whether people are satisfied with the way the police handle their complaints. The Budget Division of the Ministry of Finance, concerned about the adequacy of existing pension arrangements, wants to know who the public thinks should be responsible for insuring a decent living standard for retirees. The need and desire for such information will only increase. Joint activities with researchers in universities and research institutions provide an additional incentive for including items on attitudes and perceptions. Such information is routinely collected in social science research projects. Partnerships between academic
158
CHARLES S. KAMEN
researchers and staff members of national statistical offices will inevitably lead to pressure on the latter to include such items. Researchers won’t accept a refusal which is based primarily on past practice. Since cooperation between NSO’s and researchers is important for both sides, change will come. Part of the Bureau’s reluctance to ask questions about attitudes stems from a concern that in doing so we may be seen as taking sides politically, a perception which would be very damaging to our ability to provide data which is seen as reliable, valid and neutral. We are particularly reluctant to ask respondents to evaluate the functioning of public agencies (“How well do you think the Ministry of X is dealing with problem Y?”). On the other hand, if results of surveys are to be relevant for decision-makers they must include evaluations of agency functioning. At present, we handle this problem by asking whether respondents who were in contact with public agencies were satisfied with the treatment they received. Such questions are specific, in that they refer to particular experiences which have been identified in earlier questions, rather than requests for a general evaluation whose reference is indeterminate. A number of lessons regarding the inclusion in NSO surveys – and not only in social surveys – of questions on perceptions, attitudes and evaluations may be drawn from this brief review of Israel’s experience. Such items will probably be seen as innovations, and in what follows I refer not specifically to steps which should be taken to introduce questions on attitudes and perceptions, but to innovation in general. First, much depends on the structure of authority and control in the organization. The greater the departmental autonomy, the greater the possibility for innovation. On the other hand, departmental autonomy may hinder the spread of innovations into other branches of the organization. This leads to the second lesson: innovation should be encouraged as part of organizational policy, and support for innovation should come from the highest levels of the NSO. An atmosphere should be created in which innovation is rewarded. But “changing the rules”: can also be disruptive, as staff who have learned the accepted ways of working begin to wonder “What do they want from me?”
“QUALITY OF LIFE”
159
In short, a policy of laissez faire may lead to innovations, but they are more likely to arise in response to an active policy of encouraging them. Specific solutions and possibilities for new surveys and new procedures depend on the structure, traditions and accepted ways of behaving of a statistical bureau, which don’t change easily. The potential for change must be developed.
PLANNING FOR POLICY RELEVANCE
There are different ways in which data provided by the Bureau are “policy relevant.” Surveys such as the LFS are designed to provide clearly-defined indicators of economic conditions in the country (such as the proportion of the civilian labour force which is not currently employed). Other activities we carry out provide data (such as number of students by field of study; proportions of women) to the Planning and Budgeting Committee of the Council of Higher Education which allocates funding for colleges and universities using this information. This data is based on student records which were generated not for planning purposes but for the administrative needs of the institutions. A client may come with specific questions which a survey is designed to answer. The Public Libraries Unit of the Ministry of Education, for example, asked us to collect information about the distribution of public libraries, the services they provide and their utilization. What role Bureau staff can appropriately play in designing data gathering and analysis oriented to policy issues is not always clear. “Professionalism” requires political neutrality; it isn’t the job of NSO staff to advocate particular policies. Yet NSO staff are also citizens, and may hold views on public issues. Even if they adhere to organizational norms requiring neutrality, their normal work as professionals may sometimes blur the line between neutrality and advocacy, even without their realizing it. Our work inevitably affects decision-makers because of the decisions we make at different stages in the design and analysis of projects. Even when a client has clearly-defined data needs, methodological suggestions by professional staff may have unintended policy implications. We may ask persons for their attitude on a particular subject, and obtain answers. But it might also be legitimate to ask
160
CHARLES S. KAMEN
the respondent how important the subject is to her, in order to gain a fuller picture of the meaning of the attitude. We ordinarily assume in investigating respondents’ attitudes that the issues we ask about are important to them, but we may be wrong. Asking directly how much respondents care about the issue provides additional information which helps understand the meaning of the attitude. If, however, we discover that many are indifferent, this also has policy implications, regardless of whether the client is pleased at this discovery. Both the social indicators project and the social survey assume, explicitly or implicitly, that it is important to improve the welfare of people, that information on level of welfare is important for decision-making, that it is possible to identify the components of individual and social welfare and measure them and that one of the tasks of an NSO is to act in accordance with these assumptions. We also assume that these assumptions are widely shared. These assumptions, and our belief about them, inevitably carry value implications which we can’t avoid. We affect decision-makers well before we present to them the results of our surveys. The Bureau’s experience in planning the social survey demonstrates some of the factors which have to be considered in implementing our desire to provide information which is policy relevant. As an agency which depends on external funding, we have to demonstrate that the money we get is well-spent, and produces results. Proven “policy relevance” is an example of such results. We also want to believe that what we are doing is worthwhile, and that our work has social value. The two components of the social survey meet these needs in different ways. As I noted in an earlier section of this discussion, the core module is designed to provide comparative annual data on a set of social indicators. The information collected is “policy relevant” only in the most general sense, since the indicators were not designed to supply answers to specific questions. By their nature, social indicators track conditions and changes over time. We didn’t ask ourselves what practical uses could be served by the information in each area covered. Rather, we hoped the indicators would enable us to identify changes in social conditions and provide information on them which is unavailable from other sources.
“QUALITY OF LIFE”
161
The variable module, on the other hand, is specifically intended to provide practical information to policy makers. For it to fulfil that function, we believed it was essential that their representatives be involved in the earliest stages of planning, especially in the selection of the topic. Since the Budget Division of the Finance Ministry was funding the survey, we assumed that its wishes should carry great weight. In our initial meetings with Budget Division staff we invited them to provide a list of topics of interest to them, but were not successful in eliciting specific suggestions. To overcome this difficulty, a series of meetings was held with staff responsible for specific budget areas, and on the basis of these meetings a list of appropriate topics was prepared. As noted earlier, the Budget Division decided that the first wave would focus on pension coverage. One of the problems which arose at the beginning of our discussions with the Budget Division involved the timetable for the survey. Although planned as a continuing, annual survey, the fact that some of the survey’s most important methodological components were either being undertaken for the first time at the Bureau (CAPI and direct transmission of data files from the field staff to central headquarters) or infrequently employed (sampling from the population registry) meant that a long lead-time was required for planning. Our discussions with the Budget Division on the topic of the variable module began more than two years prior to the scheduled beginning of data collection, and their staff expressed some discomfort at what they felt to be the inordinate delay in receiving results. We could only point out that the period of preparation was reasonable considering that the results were to contribute to planning long-term policy, and not aimed at providing information regarding immediate, short-term problems. A second problem involved coordination of planning activities between Bureau staff and the Budget Division. Our preference is for continuing contact and discussions during the process of designing the survey instrument, as well as at subsequent stages of the project, in order to insure that client needs are met. In our experience, working closely with the client usually improves the final instrument, helps prevent misunderstandings, and results in a joint effort from which both sides benefit.
162
CHARLES S. KAMEN
As of this writing, we have not been successful in gaining the cooperation we hoped for, not because of opposition on the client’s part, but because of what appear to be serious limitations in the staff time which the Budget Division can devote to the project. The danger is that, despite our best efforts, we may miss some aspects of the topic which are important to the client, even though we have consulted many experts in the field of pension arrangements in other agencies and organizations. We think it unlikely that there will be any major omissions, but continue to be concerned about the relative weight given to each area of concern. We are, essentially, operating independently. There can be situations in which such independence is preferable. In the present case, given our strong desire that the results of the survey be relevant to policy concerns, it is not desirable. Israel Central Bureau of Statistics Department of Social and Welfare Statistics E-mail:
[email protected] JAN DELHEY, PETRA BÖHNKE, ROLAND HABICH and WOLFGANG ZAPF
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE: THE EUROMODULE AS A NEW INSTRUMENT FOR COMPARATIVE WELFARE RESEARCH
ABSTRACT. In this article, a new survey instrument for comparative welfare research and social reporting is described, the EUROMODULE. It has been set up in intensive discussions among experts from several nations engaged in quality of life research and social reporting. By combining indicators of objective living conditions, subjective well-being, and quality of society, with this new survey central aspects of the quality of life of European citizens can be investigated. The EUROMODULE initiative aims at strengthening efforts to monitor and systematically analyze the current state and the changes in living conditions and quality of life in Europe in a comparative perspective. So far, data for eight European countries are available.
As Europe grows together politically and economically, the international perspective becomes more and more important in social reporting and welfare research. Research teams from 19 nations have set up a research initiative. As a result of this cooperation, the EUROMODULE came into being, a survey instrument for a European welfare comparison. At present, data from Germany, Hungary, Italy, Slovenia, Spain, Sweden, Switzerland, and Turkey are available. In this paper the development and conception of the EUROMODULE are described.1
HOW THE EUROMODULE CAME INTO BEING
The EUROMODULE is a research initiative of European researchers engaged in the field of social reporting and quality of life. The aim of this initiative is to strengthen efforts to monitor and systematically analyze the current state of affairs and the changes in living conditions and quality of life in Europe in a comparative perspective. Due Social Indicators Research 58: 163–176, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
164
JAN DELHEY ET AL.
to several developments, these issues gained importance in recent years. First of all, the predicted crisis of the welfare state has set off a lot of controversy about the state of the nation and citizens’ welfare in many European countries. There is growing public interest in how well people are doing in a period of ongoing modernization and globalization, and how severe disparities and social exclusion can be avoided. This renewed public interest is also stimulated on the European level. As a result of European integration, comparable information about living conditions in the member states is of great interest. In the Maastricht treaty, several objectives related to individual welfare, quality of social relations, the combat against poverty and exclusion as well as the convergence of living conditions within Europe are given high priority by the European Union (EU). Another development is the transformation of the former socialist countries. For obvious political reasons, monitoring their progress on the road from state socialism to market economy, and from dictatorship to democracy, is an important topic for years to come, especially for those countries wanting to access the EU within the next few years. These developments highlight the increasing demand for comparative European welfare research. In 1996, the Research Unit “Social Structure and Social Reporting” at the Social Science Research Center Berlin (WZB) and the Social Indicators Department at the Survey Research Centre Mannheim (ZUMA) started an initiative to develop a European Welfare Survey. In summer 1996, the WZB and ZUMA groups invited a number of colleagues from the social indicators and quality-of-life communities, but also from statistical offices, to discuss the feasibility of such a project. The response was far better than expected. Research teams from 19 countries – West European as well as East European countries – participated in three meetings in Berlin. In 1998, the concept of a European Welfare Survey as one of several projects which were part of a TSER application (‘Targeting Socio-Economic Research Programme’) titled “Towards a European System of Social Reporting and Welfare Measurement” was submitted. The addressee of this application was the European Commission. The expert advice given by the European Commission about the TSER application was positive in large
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE
165
parts. During further negotiations, however, it became clear that Brussels would recommend to concentrate on those parts of the TSER project which aimed at taking stock of already existing statistics from government institutions or other sources – official and nonofficial. Thus, money was raised to carry out three subprojects under the title “EuReporting. Towards a European System of Social Reporting and Welfare Measurement”: (1) European System of Social Indicators (EUSI), (2) Access to Comparative Official Microdata, and (3) Stocktaking of Comparative Databases in Survey Research. The project is coordinated by the Social Indicators Department at ZUMA, Mannheim, and carried out in collaboration with researchers from several European countries.2 Under these circumstances, the initiative quickly agreed not to follow the most ambitious idea of establishing full-fledged welfare surveys in many countries, which would have demanded a huge amount of central funding. Instead, at another meeting in 1998 they agreed to follow a stepwise, bottom-up strategy by establishing a smaller version of the originally planned European Welfare Survey. The revised idea was to develop a set of basic questions which could be implemented in different types of ongoing surveys in the participating countries. This set of basic questions – called EUROMODULE – was composed in intensive discussions considering a variety of interests. In its prototype version it consists of core questions plus core standard demography consuming approximately 25 minutes of interviewing time; and of optional questions of approximately 20 minutes. The idea was to run the EUROMODULE in as many countries as possible. So far, it has been carried out in eight European countries. The decentralized way the initiative is organized is very similar to the way the International Social Survey Programme or other international projects are organized. The initiative is coordinated by the Research Unit “Social Structure and Social Reporting” at the WZB under the direction of Wolfgang Zapf and Roland Habich. Since there is no central funding, each country team which is interested in running the EUROMODULE has to raise funds by themselves.
166
JAN DELHEY ET AL.
GOALS AND OBJECTIVES
The common interest of the participants in the EUROMODULE network is to gain comparative data about welfare and quality of life. The initiative stands in the tradition of the social indicators movement, which enjoyed its takeoff in the late 1960s and during the 1970s. The most practical and visible output of this movement has been and still is social reporting. “Social reports are social policy analyses with the clear-cut question if objective living conditions and subjective well-being, and beyond individual dimensions if the quality of society has improved” (Zapf, 2000: p. 8). Examples for such comprehensive social reports in Western Europe are Social Trends in Great Britain (since 1970), the French Données socials (since 1973), the Social and Cultural Reports of the Netherlands (since 1974), and the German Datenreport (since 1983). In Eastern Europe, Hungary recently started its series of Social Reports on Hungary (for an overview of social reporting activities and the social indicator movement in Europe, see Habich and Noll, 1994; BergerSchmitt and Jankowitsch, 1999). Many of these social reporting activities have been and still are joint activities by national offices of statistics and social scientists. Another line of activities can be found at the supra-national level of international organizations (cf. Vogel, 1994; Zapf, 2000). The OECD, the United Nations, Eurostat and other organizations gave rise to a multitude of social reports and a lot of continued periodic publications. Moreover, these organizations themselves produced huge compendia of social indicators for world regions, mainly consisting of aggregate data on the level of nation states. During its takeoff, the social indicators movement had a strong inclination to compare nations. The Social Indicator Development Programme of the OECD, for example, was launched with the objective of generating a comprehensive body of data for social indicators common to all OECD countries (OECD, 1982, 1986). A cross-national perspective was also followed by the 1972 pioneering survey directed by Erik Allardt, the Comparative Scandinavian Welfare Survey. This survey described various dimensions of welfare in Finland, Sweden, Norway and Denmark (see Allardt et al., 1972; Allardt, 1981). The EUROMODULE ties on to this crossnational research tradition. The use of social surveys is seen as the
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE
167
preferred method for studying living conditions and subjective wellbeing. Aggregate figures often used in social reporting (most of all in reports published by supra-national organizations) can not be related to individuals. Thus, microdata stemming from surveys are the ideal way to explore the distribution of welfare within a society, the relationship between different life domains, and the way quality of life is connected to socio-demographic characteristics. Moreover, survey research offers the possibility to combine individual living conditions and subjective evaluations – and it has also proved to be a flexible tool for comparative welfare research across nations. The EUROMODULE can fill a gap in European comparative social reporting and social structure analysis. International surveys that already exist are either primarily dedicated to political opinions, or they cover only indicators for few selected life domains, or they are hardly accessible to scientific analysis. Even though concepts such as life satisfaction or happiness are included in surveys like the Eurobarometer and the World Value Survey, they only appear as single indicators. With regard to the European Community Household Panel (ECHP), Eurostat has initiated and harmonized national household surveys. The main focus of the ECHP, however, is on labour market and financial situation and therefore covers only some areas of life. Moreover, the data are rather expensive for secondary analysis, they are no longer sufficiently up to date for many research questions and limited to the member states of the EU. Within the EUROMODULE project non-EU-countries such as Switzerland, Turkey and a couple of Central and Eastern European countries are also participating. Thus a number of additional cross-national comparisons have become possible. The aims of the EUROMODULE research initiative can be described as follows: 1. strengthening efforts to monitor and systematically analyze the current state of and changes in living conditions and quality of life in – as many as possible – European countries. 2. providing comparative representative survey data dealing with central aspects of quality of life and individual welfare. 3. bringing together different national traditions of welfare research, which we regard as complementary rather than conflicting.
168
JAN DELHEY ET AL.
4. using the competence and knowledge of the national teams to provide thorough and meaningful interpretation of the data. 5. providing accurate assessments of the quality of life for policy makers. 6. improving public understanding of welfare development.
WELFARE CONCEPTS AND CONCEPTUALIZATIONS
The EUROMODULE initiative considers welfare as a concept which applies not only to the rich West European countries, but also to less modernized countries. Although there are different opinions of what the right notion and conceptualization of welfare is – even within Western Europe – quality of life is “the most widely recognised and the most frequently used framework for analysing the welfare development of a society” (Berger-Schmitt/Noll, 2000: p. 8). It is a multidimensional concept which encompasses both material and immaterial, objective and subjective, individual and collective aspects of welfare. In principle, the EUROMODULE combines three kinds of welfare concepts: objective living conditions, subjective well-being, and (perceived) quality of society. During the 1970s and 1980s, the understanding of welfare was an “individualistic” one. Quality of life was conceptualized mainly as individual welfare or welfare of households (cf. Noll, 2000). Components of individual welfare are not only good objective living conditions, but also subjective well-being. Objective living conditions have been and still are prominent in the Scandinavian approach as well as in the above-mentioned Social Indicator Development Programme of the OECD (under the term “social concerns”). In the tradition of level-of-living research, welfare is defined as “the individual’s command over resources through which the individual can control and consciously direct his living conditions” (Erikson, 1993: pp. 72/73). Living conditions are measured in a variety of life domains: income, housing, education, family, work, and so on, some of them representing resources or capabilities, others outcomes or ends, and some of them both (e.g. income). The theoretical assumption of this objectivist approach is that there are so-called basic needs and that satisfying these basic needs determines people’s
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE
169
well-being (see Zapf et al., 1987). This approach was very influential for comparative social reporting, especially the Social Indicator Programme of the OECD, which was launched in 1970 (cf. OECD, 1973, 1977, 1982). Subjective well-being emphasizes another perspective, closely related to the socio-psychological approach. It is often associated with the Anglo-Saxon – mainly American – research tradition of mental health. Although American researchers also use objective indicators when assessing quality of life, there is a long-standing tradition to analyze subjective well-being, which “is concerned with individual’s subjective experience of their lives. The underlying assumption is that well-being can be defined by people’s conscious experiences – in terms of hedonic feelings or cognitive satisfactions” (Diener and Suh, 1997: p. 1991). Or, as Campbell (1972: p. 442) has put it: “Ultimately, quality of life must be in the eye of the beholder...”. Life satisfaction, pleasant affect and unpleasant affect are interrelated, but separable components of subjective well-being. That is, it includes not only positive feelings and experiences, but also negative affective experiences like anxieties and worries. During the 1970s, there was an intensive discussion within the scientific community about which concept might be the more appropriate one. Nowadays, there is a mainstream consensus that objective living conditions and subjective evaluations are actually just two sides of the same coin. Subjective evaluations of personal life circumstances can relate to life as a whole as well as to different life domains, like work or income. This underlines the complementary nature of the two approaches, objective welfare measurement, and subjective well-being. In the EUROMODULE survey, both approaches have “equal rights”. The main idea is to collect objective as well as subjective indicators in order to focus on their constellation. This combined approach has been used in several survey projects, e.g. in the above-mentioned Scandinavian Welfare Survey, and the German welfare research. The German Welfare Survey, which was initiated in 1978 and has been replicated several times since then (recently in 1998), is one of the central surveys for continuous observation of the German society (Habich, 1996; Habich, Noll and Zapf, 1999). This branch of welfare research combines the Swedish approach with its socio-political focus and
170
JAN DELHEY ET AL.
the socio-psychological approach of the American tradition. Welfare and quality of life are thus influenced by the constellation of objective living conditions and subjective well-being. “Quality of life can be understood as ... good living conditions that go along with positive subjective well-being” (Zapf, 1984: p. 23). Another aspect of welfare which is included in the EUROMODULE is “quality of society”. As human beings, our personal development and opportunities to a large extent depend on the “liveability” (Veenhoven, 1996, 1997) of the society we live in. In recent years, new concepts of welfare emerged, highlighting specific aspects of the societal components of welfare, namely social cohesion, social exclusion, and social capital (cf. Noll, 2000; Berger-Schmitt and Noll, 2000). These concepts refer to the quality of a given society, i.e. the quality of relations among the members of society and the binding effects of these relations, the rupture of the relationship between individual and society due to new forms of poverty, and the feelings of mutual commitment and trust created by common values and norms. The EUROMODULE integrates some aspects of these concepts in its questionnaire. Characteristics of society and its central institutions which have a positive or negative influence on individual welfare are subsumed under the term “quality of society”. When these characteristics, social conflicts, social and political rights, public safety, are evaluated by the population, we speak of the perceived quality of society. The different aspects of welfare covered by the EUROMODULE are illustrated in Table I.
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE
171
THE EUROMODULE QUESTIONNAIRE
In June 1998 and January 1999 two meetings were arranged at the WZB, where the participants agreed on a common questionnaire (Master Questionnaire) and on methodological standards for carrying out the project. The result of this international cooperation is the “EUROMODULE”. Its conceptualization is closely related to the German Welfare Survey. Beyond the “classic” concept of welfare research, more recent concepts regarding the societal quality have influenced the choice of indicators. The questionnaire consists of a core part and an optional part. The core part, which is obligatory for all participating countries, focuses on central life domains and their subjective evaluation: housing, composition of the household, social relations, participation, standard of living, income, health, work, education, personal environment and safety. Thus, private social concerns are covered as well as public ones. Moreover, well-established global measures of subjective well-being (life satisfaction, happiness, anomia, anxiety) as well as some aspects of the quality of society are included. A set of socio-demographic background variables is obligatory for all countries and should be asked in a uniform fashion, as far as possible. In the optional part, more detailed questions are available, which can be additionally asked if sufficient financial resources are at hand. This optional part offers supplementary questions, in particular with regard to the quality of society, for instance the issue of social integration. In addition questions with regard to the individual level are included, e.g. the importance of various life domains for well-being, or the evaluation of personal living conditions. The main indicators are listed in Table II. Since the EUROMODULE is planned as a “slim” survey likely to be attached to omnibus surveys, each life domain could be covered only by a few indicators. The intention was to cover as many social concerns as possible, rather than ascertain in-depth data for a few concerns. With regard to the measurement of the standard of living, however, a more detailed and time-consuming unit was developed. Following earlier British and German studies (Townsend, 1979; Gordon and Pantazis, 1997; 1999), a list of 19 commodities and activities was drawn up, which serve as indicators for the achieved living standard of the respondents. In addition, information
172
JAN DELHEY ET AL.
is gathered about the respondents’ notion of a decent standard of living. This gives the researcher the opportunity to explore not only cross-national differences in material well-being, but also differences in the definitions of “acceptable” and “unacceptable” living conditions. The emphasis on material living conditions is justified by the wide range of economic power which the participating countries command, from “rich” Switzerland to “poor” Turkey, and by the vital political and public interest in processes of social exclusion and poverty. The EUROMODULE may be carried out as a stand-alone survey as well as part of a multi-purpose survey. Till now it has been run in eight European countries: in Germany, Hungary, Slovenia, and Sweden (all in 1999), Spain, and Switzerland (2000), Italy and Turkey (2001). Other countries participating in the research network are Austria, Belgium, Czech Republic, Denmark, Finland, France, Great Britain, Latvia, the Netherlands, Norway. Although the initiative is a European enterprise, the idea of comparative welfare research has also attracted interest from outside Europe: in 2000, South Korea has joined the network and has carried out the survey in 2001. The South Asian “tiger state” will be an interesting extra-European case of comparison. With South Africa another non-European country is preparing to adapt the EUROMODULE questionnaire. In addition, the EUROMODULE project cooperates with the NORBALT project, a “level-of-living” survey in the Baltic countries directed by the Norwegian FaFo Institute. Another interesting opportunity for comparative research could result from the project “Living conditions, lifestyles and health” in eight former Soviet countries, coordinated at the Institute for Advanced Studies, Austria. This survey deals with the changing (and often declining) quality of life in the successor states of the Soviet Union and has adapted some parts of the EUROMODULE questionnaire. Thus, the EUROMODULE data facilitate international comparisons of levels of welfare, the relationship between different dimensions of welfare and the social situation of certain groups of people in various European societies, which differ in their level of modernization, the type of welfare state, and political traditions.
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE
173
OUTLOOK
In October 2001, another conference took place where the first comparative results were presented. The participants agreed that for the time being the documentation of the data as well as their harmonization and management should be coordinated and carried out by the WZB research unit. The harmonization of the data and the
174
JAN DELHEY ET AL.
integration into a common database is an important step to enable comparative research. Part of this package is the EUROMODULE codebook. This technical documentation gives an overview on the wording of the questions and the coding of the answers and offers unweighted marginals and means for all variables, broken down by countries. Furthermore, the national studies are described by giving information on fieldwork data, the principal investigators, sample type, fieldwork methods and institute, the context of the EUROMODULE questionnaire, sample size, response rates, weighting and national population characteristics. The participants of the network have agreed to exchange the EUROMODULE data within the network for the next two years. Later on, the data base should be shared with the broader scientific community. With data from eight countries by mid 2001, the EUROMODULE got off to a good start. With this enterprise, the research initiative hopes to contribute to social reporting in Europe and to a deeper understanding of the state of affairs of the nations and the mood of their population. However, several larger European countries are still missing, e.g. France and Great Britain. We cordially invite our European colleagues to join the project and fill the white spots on the EUROMODULE map. Besides a broader geographical coverage, the repetition of the surveys is envisaged within the next years. This might add another perspective, the perspective of comparisons over time. And it might provide a good opportunity for newcomers to join. A repetition would be another milestone for establishing the EUROMODULE as a continuous enterprise in the long run. NOTES 1
We are grateful to Uschi Gerlach for doing the language check. A more detailed description of the project including a list of participants, the EUROMODULE master questionnaire, a list of variables and other documents can be found in Delhey/Boehnke/Habich/Zapf (2001). 2 The description of the projects and bibliographies are available on the following website: http://www.zuma-mannheim.de/data/social-indicators/eureporting.
QUALITY OF LIFE IN A EUROPEAN PERSPECTIVE
175
REFERENCES
Allardt, E.: 1981, ‘Experiences from the comparative Scandinavian welfare study, with a bibliography of the project’, European Journal of Political Research 9, pp. 101–111. Allardt, E. and the Research Group for Comparative Sociology: 1972, The Scandinavian Welfare Survey 1972. Data Material (available at the Finnish Social Science Data Archive, Tampere). Andreß, H.-J.: 1999, Leben in Armut (Analysen der Verhaltensweisen armer Haushalte mit Umfragedaten, Opladen). Berger-Schmitt, R. and B. Jankowitsch: 1999, Systems of Social Indicators and Social Reporting: The State of the Art. EuReporting Working Paper No. 1, Centre for Survey Research and Methodology (ZUMA), Mannheim. Berger-Schmitt, R. and H.-H. Noll: 2000, Conceptual Framework and Structure of a European System of Social Indicators. EuReporting Working Paper No. 9, Centre for Survey Research and Methodology (ZUMA), Mannheim. Boehnke, P., J. Delhey and R. Habich: 2000, ‘Das Euromodul – ein neues Instrument fur die europäische Wohlfahrtsforschung’, Informationsdienst Soziale Indikatoren, Ausgabe 24, Juli 2000, S. 12–15. Campbell, A.: 1972, ‘Aspiration, satisfaction, and fulfillment’, in A. Campbell and P.E. Converse (eds.), The Human Meaning of Social Change (New York, pp. 441–466). Delhey, J., P. Boehnke, R. Habich and W. Zapf: 2001, The EUROMODULE. A New Instrument for Comparative Welfare Research. EUROMODULE working paper series, Social Science Research Center Berlin (WZB). Diener, E. and E. Suh: 1997, ‘Measuring quality of life: Economic, social and subjective indicators’, Social Indicators Research 40, pp. 189–216. Erikson, R.: 1993, ‘Descriptions of inequality: The Swedish approach to welfare research’, in M. Nussbaum and A. Sen (eds.), The Quality of Life (Oxford). EUROMODULE: 2000, Codebook (Social Science Research Center Berlin (WZB)). Gordon, D. and C. Pantazis (eds.): 1997, Breadline Britain in the 1990s (Aldershot). Habich, R.: 1996, ‘Die Wohlfahrtssurveys – ein Instrument zur Messung der individuellen Wohlfahrt’, in Statistisches Bundesamt (Hg.), Wohlfahrtsmessung (Aufgaben der Statistik im gesellschaftlichen Wandel, Band 29 der Schriftenreihe Forum der Bundesstatistik, Stuttgart, S. 121–147). Habich, R. and H.-H. Noll: 1994, Soziale Indikatoren in der Sozialberichterstattung. Internationale Erfahrungen und gegenwärtiger Forschungsstand. Bundesamt für Statistik, Bern. Habich, R., H.-H. Noll and W. Zapf: 1999, ‘Subjektives Wohlbefinden in Ostdeutschland nähert sich westdeutschem Niveau. Ergebnisse des Wohlfahrtssurveys 1998’, in Informationsdienst Soziale Indikatoren, Ausgabe 22, Juli 1999, S. 1–6.
176
JAN DELHEY ET AL.
Noll, H.-H.: 2000, Konzepte der Wohlfahrtsentwicklung: Lebensqualität und “neue” Wohlfahrtskonzepte. WZB Discussion Paper P00-505, Social Science Research Center Berlin (WZB). OECD: 1973, List of Social Concerns Common to most OECD Countries. The OECD Social Indicator Development Programme, 1 (Paris). OECD: 1982, The OECD List of Social Indicators. The OECD Social Indicator Development Programme, 5 (Paris). OECD: 1986, Living Conditions in OECD Countries. A Compendium of Social Indicators. OECD Social Policy Studies No. 3, (Paris). OECD: 1977, 1976 Progress Report on Phase II. Plan for Future Activities. The OECD Social Indicator Development Programme, 4 (Paris). Statistisches Bundesamt (Hg.) in Zusammenarbeit mit WZB und ZUMA: 2000, Datenreport 1999. Zahlen und Fakten über die Bundesrepublik Deutschland. Schriftenreihe der Bundeszentrale für politische Bildung, Band 365, Bonn. Townsend, P.: 1979, Poverty in the United Kingdom (Harmonds-worth/ Middlesex/Berkeley). Veenhoven, R.: 1996, ‘Happy life-expectance: A comprehensive measure of quality-of-life in nations’, Social Indicators Research 39, pp. 1–58. Veenhoven, R.: 1997, ‘Lebenszufriedenheit der Bürger: Ein Indikator fur die Lebbarkeit von Gesellschaften?’ in H.-H. Noll (Hg.), Sozialberichterstattung in Deutschland. Konzepte, Methoden und Ergebnisse für Lebensbereiche und Bevölkerungsgruppen (Weinheim/München, S. 267–293). Vogel, J.: 1994, ‘Social indicators and social reporting’, in Statistical Journal of the United Nations ECE 11, pp. 241–260. Zapf, W: 1984, ‘Individuelle Wohlfahrt: Lebensbedingungen und wahrgenommene Lebensqualität’, in W. Glatzer and W. Zapf (Hg.), Lebensqualität in der Bundesrepublik (Objektive Lebensbedingungen und subjektives Wohlbefinden, Frankfurt a.M./New York, S. 13–26). Zapf, W. et al.: 1987, Individualisierung und Sicherheit (Untersuchungen zur Lebensqualität in der Bundesrepublik Deutschland, München). Zapf, W.: 2000, ‘Social reporting in the 1970s and the 1990s’, in Social Indicators Research 51, pp. 1–15.
Social Science Research Center Berlin (WZB) Research Unit “Social Structure and Social Reporting ” Reichpietschufer 50 10785 Berlin Germany Internet: http://www. wz-berlin. de/sb/
AADNE AASLAND and GURI TYLDUM
THE NORBALT PROJECT: COMPARATIVE STUDIES OF LIVING CONDITIONS IN THE THREE BALTIC COUNTRIES
ABSTRACT. The NORBALT living conditions project consists of comprehensive living condition surveys that were conducted simultaneously in Estonia, Latvia and Lithuania in 1994 and 1999. The project was coordinated by Fafo Institute for Applied Social Science in Oslo, Norway, and implemented in cooperation with local partners. In this article we describe the NORBALT project, and discuss some of the choices that were made in developing a survey tool to be used for both in-depth descriptions of living conditions within individual countries, and analyses of cross-national and regional variation. The surveys were designed based on the Scandinavian model of living conditions analysis, but particular aspects important for transition economies and specific regional aspects were also taken into account. In view of a possible future integration into the European Union, it has also been a goal to enable comparisons with established European surveys. We argue that in cross-national surveys, due to language and cultural differences, objective indicators of living conditions are more robust than subjective ones, and we recommend greater care in interpretation of questions that ask for subjective evaluation.
The NORBALT project is a continuation of the pioneering effort made in 1990–91, when Fafo Institute for Applied Social Science carried out a living conditions survey in Lithuania. This was the first major survey of this kind ever conducted by a western research institution in the former Eastern Bloc. In the autumn of 1994 Fafo, in cooperation with the ministries of Social Affairs and the statistical authorities in the Baltic countries, conducted nation-wide living conditions surveys, not only in Lithuania, but this time also in Latvia and Estonia, as well as in two regions of Russia (NORBALT I). Five years later similar surveys were conducted in Estonia, Latvia and Lithuania with the same institutions involved (NORBALT II).1 In this article we discuss some of the major challenges associated with the implementation and analysis of such large scale surveys, involving interviews with more than 20,000 Baltic Social Indicators Research 58: 177–189, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
178
AADNE AASLAND AND GURI TYLDUM
households and more than 60,000 individuals, covering countries with different languages and cultures, which are also undergoing rapid economic and social change. The project is financed by the Norwegian Ministry of Foreign Affairs, the Norwegian Research Council/Council of Universities, the Nordic Council of Ministers, and the Baltic governments. The main aim of the NORBALT project has been to provide decision-makers responsible for policy planning and policy implementation in Lithuania, Latvia and Estonia with updated, reliable and relevant statistical information on living conditions. However, although the main target group for the research has been Baltic policy-makers, it has also been a goal to provide international agencies, NGOs, business organisations and enterprises, the research community as well as the general public in the Baltic, the Nordic and other European countries, with relevant and policy-oriented descriptions and analyses of living conditions in the three countries. The NORBALT teams have been concerned with finding an optimal balance between academic and political aims of the research. This means that the study was designed with the intention of increasing the scientific understanding of the process of social change in transition countries, as well as providing operational tools for improved policy planning. A main concern for Fafo has been to infuse the survey with the maximum amount of relevance for policy-makers in the Baltic area and beyond. To achieve this goal, the survey has been firmly anchored, in both thematic and organisational terms, with local authorities. This has been done to make sure that the survey design is realistic and politically germane, and to increase the likelihood that survey results are put to good practical use by decision-makers. Also, by cooperating with state institutions and researchers in the three countries Fafo believe we have contributed not only to developing competence and skills in the monitoring of living conditions in relevant institutions, but also to building Baltic-Nordic networks concerned with the advancement of applied social science, and generally to encourage links between academics and politicians in the region. Following the breakdown of the system of planned economy, the Baltic institutions and networks for production of statistical data
THE NORBALT PROJECT
179
were to some extent discredited and in need of reform. The Baltic governments soon realised that they would increasingly come to rely on modern survey research and monitoring to cover their most acute information needs in terms of good governance. However, the capabilities and methods for data collection were unevenly developed in the countries concerned. Fafo was asked to contribute to the development of a permanent survey capability in the Baltic countries employing state-of-the-art survey techniques. In line with the wishes of local authorities, Fafo concentrated its efforts on institutions having the potential to become reliable future producers of data for public use. THE SCANDINAVIAN LIVING CONDITIONS MODEL
The NORBALT surveys have been designed using established Scandinavian methods of living conditions analysis, and the living conditions components in the NORBALT questionnaire are developed taking as a starting point the components of the living conditions surveys which are routinely carried out in the Nordic countries. The living conditions studies were first introduced by Finnish social scientists in the 1960s and later adopted by the Swedes. In Sweden and Norway the living conditions surveys became institutionalised through giving the responsibility for design and analysis to the central bureaux of statistics. In Norway the publication Sosialt Utsyn, with analyses along the main living conditions dimensions, has been published regularly since 1974, the most recent in 2000 (Statistisk sentralbyrå, 2000). Sweden has also had regular reporting of findings from the living conditions studies, and the most recent general analyses of developments in living conditions were published in 1997 (Vogel and Häll, 1997). In Denmark the Central Bureau of Statistics has not been so much involved in the living conditions surveys, and the surveys in 1976 and 1986 were executed by researchers at the Socialforskningsinstitutet (Hansen, 1986). Some work has also been done comparing living conditions in all the five Nordic countries (e.g. Vogel, 1990).
180
AADNE AASLAND AND GURI TYLDUM
The Scandinavian living conditions surveys are conceptually influenced by the British social scientist Richard Titmuss’s definition of standard of living as: The individual’s command over resources in money possessions, knowledge, physic and physical energy, social relations, security, etc., by the help of which the individual can control and consciously direct her/his conditions of life (Titmuss, 1958).
Advocates of the Scandinavian model of living condition surveys question the use of only economic indicators to describe allocation of resources in a society, and to determine vulnerable groups. Planners need other types of information to evaluate whether progress has been made in important aspects of life. The living conditions surveys imply a multidimensional approach, where indicators of different domains of life are included. More weight is placed on objective resources than subjective feelings and opinions, although the latter are also included. The main components in the NORBALT questionnaires2 are: Housing conditions Income and economic resources (including subjective evaluation) Education Labour force participation Informal economic activity Working conditions Health and health care services Exposure to crime Migration Social contact and networks Political opinions and value orientation One-dimensional analyses are seldom able to provide a true picture of more general welfare trends. Living conditions are produced as interactions of several components. One of the main objectives of the NORBALT project is therefore to investigate the relationships between the various components in the survey. In a number of surveys attempts have been made to create a compound
THE NORBALT PROJECT
181
measurement of living conditions, where an individual (or a household, community, country, etc.) is given a total score based on the aggregate sum of the score on each component in the survey. However, one important problem with such an approach is to estimate the relative importance of each component and thus to know how much each component should count in the final analysis (Andersen and Barstad, 1990). In the analysis of the data from the NORBALT surveys different ways of looking at interactions between components have been made in order to shed light on the relationships between different aspects of welfare and social life. These authors do not believe, however, that it would be useful or meaningful to construct a general measure of living conditions. Instead we seek to identify vulnerable groups, i.e. groups that experience deficits along one or several living conditions dimensions – and clustering of such deficits is identified. The way these deficits tend to cluster, may vary greatly between societies and social groups. Thus, when used in different societies they do not construct a compound index, but rather different indicators that may, or may not appear together.
ADAPTING AND DEVELOPING THE DESIGN
Although based on the Scandinavian surveys, the indicators used in NORBALT have been adapted to the special conditions of the Baltic countries. Four elements were crucial in the design-phase of the NORBALT II: 1. To ensure comparability with the 1994 (and the Lithuanian 1990–91) surveys. 2. To develop social indicators which take into account the particularities of countries undergoing rapid social and economic transformations (in areas such as the unofficial economy, social networks, and compensation strategies of households when there is a loss of job or government transfers, etc.). 3. To provide opportunities for each of the Baltic countries to include indicators in the questionnaire which they see as particularly important.
182
AADNE AASLAND AND GURI TYLDUM
4. To include living conditions indicators used in surveys in the Scandinavian and other European countries. This has, among others, been considered highly relevant for a possible future integration into the European Union. Comparable data from the Baltic countries can be useful for European countries for the same reasons. The use of standard European and Scandinavian indicators facilitates comparative studies and contributes to putting the findings in each country into a wider regional and European perspective. UNDERSTANDING DEVELOPMENTS AND TRENDS – TIME SERIES
Although the survey approach for studying living conditions has proven a valuable analytical tool in Scandinavia and now also in the Baltic countries, the method involves some limitations of which one should be aware. The approach is best suited for giving a bird’s eye view of the overall situation and variations among main groups at a given point in time (Hernes and Knudsen, 1991). With a general survey like NORBALT little can be said about very small groups of the population. Even if one, as was the case in Estonia in both 1994 and 1999, increases the sample size considerably, the method is not suitable for comparative analyses of developments in smaller regions. If there is a need for analyses on specific groups, such as for example the physically or mentally handicapped, it is necessary to increase the number of these respondents in the sample, or to conduct a specialised study. Increasing the sample size to 5000, however, enables the Estonians to carry out more in-depth regional analyses than what is possible in the other two countries. The living conditions surveys in 1994 gave a snapshot of living conditions in the Baltic region at a special point in time. To get the full effect of the surveys it was recognised that they should be repeated on a regular basis, as has been done in the Scandinavian countries. The need to monitor the changes in living conditions is even more acute in the Baltic countries where social and economic change takes place very quickly. One may criticise this type of time-series data for being static and unable to explain how and why changes take place in a society, as one cannot follow the same individuals over time, and thus lack the
THE NORBALT PROJECT
183
possibility of dynamic analyses that panel data gives. However, the costs and validity problems associated with panel surveys has made this an option not considered for the NORBALT project. In order to bring some dynamic aspects into the study, we have included some prospective (about the future) and retrospective (about the past) questions.
METHODS
The data were collected in personal interviews, using a 30-page questionnaire, containing approximately 250 questions. Data were collected both for one individual (randomly selected individual) and on the household level, including information about all the members of the households. On average, an interview took 60 minutes. The data were transformed into computer readable form on a customized data entry template (Blaize), and are processed using advanced statistical software (SPSS, SUDAAN and Stata). The sample was a combination of a single stage stratified sample in the larger cities, and a two stage stratified cluster sample in smaller cities and rural areas. Sample sizes varied from 2500 to 5500 (see Table I), depending on budgetary constraints and need for comparisons between smaller regions. The public population registers were used as sample frame. These are updated regularly, but not systematically, and largely on private initiative. Since households that have recently moved may not be correctly registered, we decided to treat the population registers as a list of addresses, interviewing the household living in the dwelling, irrespective of whether this was the selected household or not. This may still have created a slight bias, as households that have recently moved (without registering) to newly constructed dwellings, or dwellings that were formerly used as shops or offices, could not be included in the sample frame. Overall there was a non-response due to frame imperfections (non-existing or vacant buildings) of between 2 and 6 per cent. Non-response due to refusals and no-contact was below 10% in all three countries, in all waves, but higher in 1999 than in both 1994 and 1999 (see Table I). The local teams had the overall responsibility for carrying out the field work in their own country, using a detailed quality control
184
AADNE AASLAND AND GURI TYLDUM
system. In all countries there were networks of interviewers locally reporting to a central office in the capital cities. During NORBALT I, Fafo played a very active role in the training of supervisors, interviewers and data-entry operators. However, as the Baltic countries have gained considerable experience in carrying out large-scale and complex surveys, comprehensive participation from Fafo in these training activities was not required during NORBALT II. Much attention was nevertheless paid to the Field Work Manual which had to be more or less identical in all three countries in order to ensure that the questions were asked and interpreted in the same way in each country. Methodological Problems in Comparisons and Measurements
The objective of the NORBALT surveys is both to asses the living condition situation in each country at present, to compare the situation with other Baltic countries, as well as with other countries in Europe, and to look at developments over time. This raises a number of methodological challenges. A large amount of time and resources was put into the survey design phase in order to ensure validity of the measurement. Pilot
THE NORBALT PROJECT
185
surveys were carried out in the spring of 1999 where the survey instruments were tested. Relatively few problems were encountered during the pilot surveys, but some important improvements were made. Much work was put into assuring that questions match the social institutions (health care, education, the unofficial economy) in all three countries, and at the same time remain comparable within the three countries, and with other European countries. The inclusion of project teams from each country and thorough (and time-consuming) discussions about each question and indicator in the questionnaire has reduced the scope for cultural differences in interpretation of the questions. Still – much care must still be taken in interpreting the findings, – the educational level is, in the NORBALT surveys, measured by different classes of education (primary, basic, secondary, etc.), but it is much more difficult to assess the quality of the education, which is concealed in the way the data are collected. There is for instance little room for a survey team to make sure universal rules are applied in deciding which types of post secondary studies should be registered as higher education. The survey research in itself is developed in order to provide accurate knowledge of general tendencies and differences, but gives little room for nuances in measuring the value or importance of a living condition component. Such nuances are particularly important to take into account when comparing results between different political systems, or cultures. A survey question will always give a response, but lacks the flexibility to allow the respondent to elaborate on why he/she answers the way they do. Questions are asked in the same order and in an identical way to all the respondents, whether or not the questions fit the situation of the respondent or not. Naturally, it is possible for a respondent to answer ‘do not know’ or to say that the question is not relevant, but in most cases respondents will seek to fit into a category. Also, the meaning of a question may differ from one person to another – for instance it may vary greatly between individuals and social groups what is understood by poverty, to be disturbed by depression or to experience threats. Using standardised measures does not guarantee a high validity of the instrument used. This becomes particularly difficult as interviews are carried out in four
186
AADNE AASLAND AND GURI TYLDUM
different languages, and much of the analysis takes place in a fifth – in English. The questionnaires and manuals were first developed in English, and then translated into the three national languages locally. Thorough back-translations were conducted to discover, and when possible correct the nuances that had disappeared, or perhaps been added in the translation. In addition the three national teams developed a Russian version from their own questionnaire (for use by the large Russian-speaking populations), which again were compared between the three countries, and final adjustments made. In spite of thorough back-translation one should not believe that the questionnaires will be totally identical. Words and nuances in use in one language may not exist at all in other, or may only exist with an additional meaning/nuance. Through the back-translation we thus get a better awareness of such language nuances, and realize which cross-national comparisons may be done, and which ones can only be done with great care. The most problematic questions are the ones asking the respondent to evaluate an item or situation, and in particular when answer alternatives make use of an ordinal scale (for instance very often, often, sometimes, seldom and never). Our experience shows that these answer alternatives tend to change with each translation, and with substantial change of meaning. In retrospective we regret not making more use of questions of the type: on a scale from 1 to 5, where 1 is very often and 5 is never, how often would you say that . . . There are some problems associated with using this type of scales in comparisons between countries as well. However, as the Baltic countries have close to a one hundred percent literacy rate, one can expect everybody to know the numbers, and understand the concept of a scale. As the three countries are not too culturally differentiated, one may expect that to the extent that there are cultural associations with certain numbers (for instance number 13 as an “unlucky” number), these will be similar across the region. Questions asking for evaluations of items or situations, both had a tendency of changing during translation, and were reported back by interviewers as being understood differently by different types of respondents. We have therefore chosen not to put much emphasis on
THE NORBALT PROJECT
187
questions asking for subjective evaluations – when the objective is to compare between nations, questions on objective living condition are more reliable. To get the nuances on what people think is important or not, one would need to supplement the data with qualitative assessments from the countries involved. Because of these problems we recommend not comparing frequencies from some of these variables directly between countries, but rather to compare distributions between social groups or developments over time, in order to control for the potential language specific or culturally created nuances in the way people respond.
MAIN FINDINGS AND REPORTING
The survey gives clear evidence of the welfare developments in the three Baltic countries. A large proportion of the households are still on the verge of poverty – a poverty which restricts their daily lives. In the 1994–99 period differences between the three countries have increased. Estonia has moved ahead, Lithuania is lagging behind. However, differences between social groups within each of the three countries are still larger than differences between them. Particularly vulnerable are families with many children, and especially single parent households. In 1999 households where one or more members are unemployed stand out as among the worst off groups. Retired persons were particularly bad off in 1994, but their situation has improved, at least relative to other groups. A positive trend is that there has been a strong reduction in stress and psychological problems in all three countries, perhaps due to a more stable and predictable economic and political situation. The results are presented in three national tabulation reports – compiled in an identical manner in order to enable direct comparison between the three countries. The same set of background variables are used throughout the reports, in order to identify to what extent social groups experience clustering of living condition deficits. The most important variables are sex, age, place of residence (urban/rural), ethnicity, citizenship, education and labour force status. For analysis on the household level, information is presented for various levels of household income and household types as well. Some additional background variables are only used
188
AADNE AASLAND AND GURI TYLDUM
for selected modules – where they are necessary to shed light on this particular aspect. In addition to the national tabulation reports we have prepared a 34 page brochure with presentations of trends and comparisons between the three countries, aimed at persons with a need for brief, surface knowledge of the living condition situation in the three countries. More in-depth analyses of the situation in each country will be conducted by the three national teams, while Fafo is responsible for compiling a collection of articles, with various comparative perspectives, based on the NORBALT material. For more information on the NORBALT project – consult our web site http://www.Fafo.no/norbalt. NOTES 1
The project has resulted in a large number of publications (e.g., Hernes and Knudsen, 1991; Aasland, 1996; Hansen, 1996; Grøgaard, 1996; Hernes and Knudsen, 1991; Knudsen, 1996; Aasland and Cesnuityte, 1997; Aasland et al., 1997; Aasland and Tyldum, 2000). 2 The English version of the NORBALT questionnaire and more information about the project can be found at the NORBALT web-site: www.Fafo.no/ norbalt.
REFERENCES
Aasland, A. (ed.): 1996, Latvia: The Impact of the Transformation (Fafo, Oslo). Aasland, A. and V. Cesnuityte: 1997, Living Conditions in the Baltic Countries Compared (Fafo, Oslo). Aasland, A., K. Knudsen, D. Kutsar and I. Trapenciere: 1997, The Baltic Countries Revisited: Living Conditions and Comparative Challenges (Fafo, Oslo). Aasland, A. and G. Tyldum: 2000, Better or Worse? Living Conditions Developments in Estonia, Latvia and Lithuania 1994–99 (Fafo, Oslo). Andersen, A.S. and A. Barstad: 1990, ‘Om sammenfattende analyser av levekår’, in Nordisk statistisk sekretariat (ed.), Nordiska seminaret om levnadsvillkor och välfärdsstatistik (Nordisk statistisk sekretariat, København). Grøgaard, J. (ed.): 1996, Estonia in the Grip of Change (Fafo, Oslo). Hansen, E.: 1996, Coping with it: St. Petersburg and Kaliningrad Facing Reform (Fafo, Oslo). Hansen, E.J.: 1986, Danskernes levekår – 1986 sammenholdt med 1976 (Hans Reitzel, København). Hernes, G. and K. Knudsen: 1991, Lithuania Living Conditions 1991 (Fafo, Oslo).
THE NORBALT PROJECT
189
Knudsen, K.: 1996, Lithuania in a Period of Transition (Fafo, Oslo). Statistisk sentralbyrå: 2000, Sosialt utsyn 2000 (Statistisk sentralbyrå, Oslo). Titmuss, R.A.: 1958, Essays on the Welfare State (Allen and Unwin, London). Vogel, J.: 1990, Leva i Norden (Nordisk Statistisk skriftserie, København). Vogel, J. and L. Häll: 1997, Välfärd och ojämnlikhet i 20-årsperspektiv 1975– 1995 (Statistics Sweden, Stockholm).
Fafo Institute for Applied Social Science P.O. Box 2947 Tøyen N-0608 OSLO Norway E-mail:
[email protected],
[email protected] This page intentionally left blank
THOMAS ANDERSEN and BIRGER POPPEL
LIVING CONDITIONS IN THE ARCTIC
ABSTRACT. This paper introduces a model for conducting research on living conditions among peoples that have experienced rapid social, cultural and economic change in countries where a non-parallel development has occurred. This model was developed by the researchers of SLICA, A Survey of Living Conditions in the Artic; Inuit, Saami and the Indigenous Peoples of Chukotka, which was initiated by Statistics Greenland in 1997. The point of departure for this model is a critique of contemporary living conditions surveys carried out by national statistical bureaus in economically, technologically and culturally segmented areas. The point of view is that these studies erroneously assume that the populations they investigate are homogeneous, and that consensus concerning individual social and economic objectives exists. This usually leads to research designs and indicators of individual well-being that reflect the dominant culture, or the prevalent way of living and thinking in these countries. The focus of this paper is on the research design of SLICA. The implementation of two important methodological challenges is discussed. Namely, (1) how to secure a contextspecific concept of well-being which also mirrors the life forms and the priorities of the respondents and (2) how to measure impacts of structural change on individual well-being.
1. INTRODUCTION
Towards the end of the 1960s and in the 1970s the first living conditions studies were carried out in the Nordic countries. The Swedish Institute of Social Studies published the “Levnadsnivåundersøkning”, which was the first study of living conditions, in 1968 (Johanson, 1970) and it was followed by Allardt’s pan-Nordic study in 1972 (Allardt, 1975), Norway in 1973/74 (NOU, 1976), Denmark in 1976 (Hansen, 1978) and Finland in 1978. Such studies now form part of the plan of work in the national statistical bureaus in Sweden, Norway, Finland and Denmark. In Greenland studies of living conditions have never been institutionalised in the way they are in the Nordic countries. The Danish Social Indicators Research 58: 191–216, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
192
THOMAS ANDERSEN AND BIRGER POPPEL
National Institute of Social Research did carry out the first real study of the Greenlanders’ living conditions in 1970–73, but more than twenty years passed before Statistics Greenland carried out the next living conditions study in 1994. The need for such a study arose partly from a desire to measure the distributive effects of 15 years of Home Rule politics on different population groups and partly because of a lack of register statistics in different areas. Due to the small number of studies in Greenland, research on living conditions was thus at a beginner’s level in 1994. The 1994 study temporarily closed a number of “black holes” in register statistics, but several invalid results prompted Statistics Greenland to evaluate the research design for future living conditions studies in Greenland. In this paper we introduce the result of these considerations; Namely SLICA: “A Survey of Living Conditions in the Arctic; Inuit, Saami and the Indigenous Peoples of Chukotka”. In section 2 we briefly describe the parameters of the Greenland survey from 1994. In section 3 we introduce the specific Arctic context, while in section 4 we present a general list of ground rules for living conditions studies among indigenous peoples in the Arctic based on our experiences from the 1994 survey. In section 5 we introduce SLICA which is a circumpolar study of living conditions among Inuit and Saami peoples, and in section 6 we discuss how to secure (a) context-specific and valid concepts of well-being and (b) focus on the relation between structural change and individual well-being within SLICA. Furthermore, we define our use of concepts like well-being and living conditions in this section. Finally, in section 7, the relationship between living conditions research and political planning in the Arctic in general and sustainable development more specifically is discussed. 2. THE GREENLAND LIVING CONDITIONS STUDY 1994
The research design of the 1994 study was to a large extent a copy of the Scandinavian model which had been developed in connection with the earlier Nordic living conditions studies in the 1960s and 1970s (Eriksen and Uusitalo, 1987; Hansen, 1990). The selection of living condition dimensions and indicators was characterised by a “garbage-can” process in which those director-
LIVING CONDITIONS IN THE ARCTIC
193
ates, which were interested and had the time had most questions included in the final study. When operationaling the questions, the living conditions study carried out in 1992 by the Danish National Institute of Social Research and Statistics Denmark (Danmarks Statistik and Socialforskningsinstituttet, 1992) had a considerable influence. This meant, although it was not so intended, that large parts of the Greenland questionnaire were characterised by the social indicator approach on which the Scandinavian statistical bureau’s base their living conditions studies (ibid; Vogel, 1996; Statistiska Centralbyrån, 1995) as well as by the factor which Eriksen and Uusitala in 1987 described as The Scandinavian Model. During the analytic work on the data of the Greenland survey a number of problems arose which should lead to a discussion of the application of the Scandinavian model to future Greenland living conditions studies. In part it was a question of specific problems connected with transferring a concept of well-being between different types of society, and in part general problems connected with the application of the social indicator approach and The Scandinavian model on the Greenlandic reality. Thus the conclusion was that it is problematic to transfer a research design which presupposes a basic value consensus in society, and hence also implies the presence of universally necessary needs, from a relatively homogeneous late industrial society (such as Denmark) to a country which first of all is characterised by a non-parallel internal development, and secondly can hardly be described as industrialised, and which in addition has a completely different cultural and historic profile from the Nordic countries. (For a detailed critique of these approaches, see Andersen, 1998, 2001.) The main conclusion was that to obtain valid measures of wellbeing in the Greenlandic population it was necessary to develop a new research design including identification of new living conditions dimensions and new operationalisations of traditional dimensions adapted to the Greenland context. In section three, we describe the main features of the Greenland empirical context. As will be shown, the history of Greenland post-war development is very similar to that of other indigenous peoples in the Arctic.
194
THOMAS ANDERSEN AND BIRGER POPPEL
3. THE ARCTIC CONTEXT
The political and economic changes in the period after the Second World War led to crucial changes in the traditional Greenland fishing/hunting culture, and hence in the traditional social structures. The changes in the industrial culture, the transition to a monetary economy, the educational mobility as well as the increased urbanisation have meant that the norm, value and prestige systems in the Greenland society today are linked less to a subsistence production in extended families in small closed communities and primarily to wage earning in a more globalised and open society. In 1945 it was estimated that 66 per cent of the labour force in a population of 21.412 individuals was involved in hunting and fishing. In 1996 this proportion had decreased to approximately 25 per cent. A figure which also includes persons working in the modern fishing industry. Without maintaining that the mode of production is the only decisive factor for the culture of a given people, it should be stressed that in Greenland it has been possible to identify a direct connection between changes in the mode of production and cultural and social changes (Kruse, 1999; Andersen, 2001). It has been a question of a process of modernization which in some areas is reminiscent of industrialization of the modern Western societies – but also different as far as pace and point of departure and results are concerned. The changed industrial structures and the urbanisation in the 1950s and 1960s resulted in changes in the social structures which then led to a shift in the traditional norm and value systems, especially for people in the towns. New arrivals typically became unskilled workers in the fishing industry or fishermen in the new modern fishing fleet, and they lived in modern blocks of flats. This process of change, and the population’s attempt to adjust to it, was not without problems. Not least because it was extremely rapid. The changes in the economic, political and social structures led to a break with the traditional norm and value systems. Due to the rapidity of the process these systems were not naturally replaced by new norms and values. A lack of norms, or anomie, appeared in large segments of the population. This was accompanied by a general feeling of powerlessness or alienation in large parts of the population who felt they could not influence the process of modernisation (Hansen, 1999), and the result was a number of serious
LIVING CONDITIONS IN THE ARCTIC
195
life style problems. The increase in number of suicides was thus pronounced (approx. 12 per cent of all deaths in 1997 was due to suicide), crime, social problems (including the removal of children from their home and child abuse), and the consumption of alcohol and marijuana. This has not been characteristic of Greenland only but has also been occurring in other indigenous populations in the Arctic where a similar modernisation process has taken place. Jack Kruse, who is the director of US branch of SLICA, described the transformation of living conditions among indigenous peoples in the Arctic like this: The last quarter of a century has brought a radical transformation of living conditions among Native people across the circumpolar world. There are few remaining doubts about the social-historical forces behind this revolution. Government policies in the late fifties initiated trends of modernisation. The growth of the market economy, social welfare policies, consumerism, and mass communication accelerated these trends to their full momentum in the early seventies. The basic processes of changing living conditions are also well understood: settlement in permanent communities; monetization of work, trade and mutual aid; technological emancipation from toil and uncertainty (housing, clothing, feeding, transportation, hygiene and heath maintenance); bureaucratization of social organization; personal mobility through schooling, vocational training, or professional experience; growing class stratification and status differentiation; increasing separation between public and private life; institutionalization of politics and the rule of law; concomitant development of civil society (voluntary associations, independent press, and private enterprises); and conversion of public discourse to conflicting ideologies of progress, mastership of destiny, formal rights and a self-conscious quest for identity. Almost in the space of one generation, most of the practical or mental references of what it had always meant to be an Inuit, Iñupiat, or a Saami became indeterminate, fluid, less reliable. At the same time, widening opportunities opened the door to rising expectations, both personal and collective. Inasmuch as the past no longer provided a handbook for the future, people sought – if only partially – to write their own futures. But the acceleration of change also frayed social bonds and weakened the bearings of social regulation. Today Native Arctic peoples are experiencing some of the highest rates of social problems (see, for example, Kirmayer, 1994; Berman and Leask, 1994). These global changes had deep implications for the living conditions of the Arctic’s inhabitants. As Scardigli (1983) put it, there was a general shift from a cultural order to an economic order; in other words, from a closed society to a market society, and from a society governed by customs to a society governed by laws and regulations (Kruse, 1999).
196
THOMAS ANDERSEN AND BIRGER POPPEL
Thus, the political, economic and social developments in the indigenous societies in the period after the Second World War have been characterised by rapid and regionally different structural changes which have had crucial effects on the living conditions of these peoples. 4. A NEW SET OF GROUND RULES FOR LIVING CONDITIONS RESEARCH IN THE ARCTIC
On the background of the criticism levelled at the Greenland living conditions study 1994 (Andersen, 2001), and hence at the Scandinavian model and the social indicator approach, the Scandinavian model was revised and a general list of ground rules for living conditions studies among indigenous peoples in the Arctic produced (Ibid.): 1. The identification of the dimensions and indicators for living conditions must be context-specific so that the well-being concept reflects the way of life and the priorities of the population. 2. The research design must be able to capture the effects of social, cultural, political and economic changes. Thus a correlation between structural events and changes in individual living conditions must be established. 3. The living condition dimensions and indicators must be identified, if not on the background of theories at least on that of hypotheses about and models of causal relationships between the indicators, and between the indicators and the structural variables. 4. The research design must be based on explicit definitions of living conditions and well-being so that the researchers, the respondents and the politicians will be able to evaluate the validity of the data. 5. Both evaluative and descriptive measures for well-being must be applied. 6. Both material and non-material indicators must be applied. 7. Data must be representative, reliable and valid. 8. The focus should be on resources rather than on the satisfaction of demands and preferences.
LIVING CONDITIONS IN THE ARCTIC
197
9. The selected dimensions and indicators of living conditions must be coherent and exhaustive. 10. It is assumed that it is feasible to talk of universal or generally necessary resources within that period of time to which a given living conditions study refers. 11. The study must be quantitative. 12. The analytical focus must be the allocation of resources between various segments of the population. 13. Data must be statistically comparable. This means that the same theoretical variables are operationalised, as far as it is possible, in the same manner. The above critique may be extended to include living conditions studies among indigenous peoples in the Arctic. As mentioned in the introduction our preparatory work for the new Greenland living conditions study showed that especially the problems connected with the use of non-adequate concepts of well-being did not apply to the Greenland living conditions research only – but was a general problem connected with the study of living conditions among indigenous peoples in the Arctic. As these populations – in contrast to the Greenlanders – are relatively small, ethnic minorities, the national statistical bureaux have only used few resources for developing specific assessment instruments in connection with living conditions studies among these peoples. SLICA – A SURVEY OF LIVING CONDITIONS IN THE ARCTIC; INUIT, SAAMI AND THE INDIGENOUS PEOPLES OF CHUKOTKA
The Arctic Living Conditions Study among Inuit and Saami (SLICA) came into being as a reaction to the problems we had experienced in connection with the analysis of data from the Greenland living conditions study 1994. Thus at the beginning of 1997, Statistics Greenland initiated preparations for carrying out a new Greenland living conditions study. However, as described above we very soon discovered that our own problems with the living conditions study were similar to those encountered by other living conditions studies among indigenous peoples in the Arctic. We also encountered an increasing need among these populations – including Greenland – for opportunities for comparing ones own
198
THOMAS ANDERSEN AND BIRGER POPPEL
living conditions with those of other populations with the same developmental characteristics – instead of comparisons with the majority populations in the various countries, or, as in the case of Greenland, with Denmark. With this in mind Statistics Greenland initiated the comparative Arctic living conditions project instead of carrying out an exclusive Greenland study at the beginning of 1997. At present approx. 40 international senior researchers and 15 research institutions in and outside the Arctic participate in the project. Furthermore a number of organisations representing the indigenous peoples are also acting as active participants in SLICA. The main purposes of SLICA are: 1. To develop a new research design for comparative studies of the living conditions among the Inuit and Saami population in the Arctic region. This includes drawing up a selection of nominal and operational living conditions indicators based on earlier theoretical literature, the experiences from the Greenlandic living conditions study and consultations with the indigenous peoples’ organisations. 2. To carry out a comparative, dynamic social analysis of the cause and effect relationship between the various political, economic, cultural and technological structures and individual living conditions, and between individual resources and individual conduct. 3. To map the living conditions among Inuit and Saami in Greenland, Inuvialuit, Nunavut, Nunavik, Labrador, Alaska, Sweden, Norway, Finland and Chukotka and the Kola Peninsula. 4. To create a better basis for local as well as national decisionmakers for political planning and implementation 5. To establish an interdisciplinary network of researchers and research institutions interested in living conditions research in the Arctic region. 6. To increase knowledge among the indigenous peoples concerning their own and other indigenous peoples’ history of development and living conditions. 7. To train and include local students and researchers in the project (Andersen, 1999b).
The project is a quantitative cross section study which will carry out 23,000 personal interviews (14,000 of these will be carried out in
LIVING CONDITIONS IN THE ARCTIC
199
Canada by Statistics Canada) and which is expected to be completed at the end of 2002. The group of respondents is made up of Inuit and Saami in the Arctic who are at least 15 years old at the time when the collection of data takes place. The participating regions are shown in Figure 1. In September 2001 SLICA’s phase 1, which include the development of a new research design and carrying out pilot studies, will be concluded. The actual data collection will take place during October 2001 to February 2002 while analyses and publishing are scheduled to 2002–3.
5. FOCUS ON CONTEXT, SOCIAL CHANGE AND THE CONCEPTS OF WELL-BEING WITHIN SLICA
In this section we focus on how we implement three crucial items of the list of ground rules presented in section 4 within SLICA. Firstly, we specify SLICA’s concepts of well-being and living conditions. Secondly, we demonstrate how SLICA through the process of identifying living conditions indicators ensure that these are context-specific and thereby valid. In this connection we introduce
200
THOMAS ANDERSEN AND BIRGER POPPEL
the concept of household production as an example of contextspecific and model based identification of living conditions indicators. Finally, we briefly discus how to include and measure the effects of structural changes on individual living conditions. 5.1. The Concepts of Well-being and Living Conditions in SLICA
We define living conditions by means of two components. First of all we define, as in the Scandinavian model, individual living conditions based on the individual’s control of material as well as non-material resources which may be applied in order to influence one’s own living conditions (Titmuss, 1958). Focusing on resources rather than on demand satisfaction is a result of the recognition of the fact that the priorities of any individual between the satisfaction of social, cultural and physical needs are not universally given, but the level and the demand for satisfaction of the same types of need will vary with the historical, cultural and material context. This is true both inter- and intranationally. It is not the same as saying that it is impossible to establish a universal hierarchy of demands in a given society at a given time. It simply emphasises that apart from a few basic needs, no universal hierarchy of demands separated from time and space exists. By focusing on resources it is emphasised that the individual is not simply a victim of a mechanical need for satisfaction but is equipped with a free will which will lead to a situation where the application and appropriation of resources at a conscious as well as at an unconscious level will change with time and place, and that different individuals may choose to apply the same resources in different ways. Secondly, the definition of living conditions should include an assumption of the fact that the application of resources to a certain extent is influenced by structural conditions. In this connection it is relevant to include the arena approach which has developed as an extension of Coleman’s social theory (Coleman, 1971) and which was used in the first Norwegian study of living conditions at the beginning of the 1970s (Ringen, 1975, 1995). The core of this point of view is the fact that well-being cannot be defined only as the individual possession of a number of resources – but as the individual possession of resources which may be applied in various arenas (= social structures as for example the labour
LIVING CONDITIONS IN THE ARCTIC
201
market, the private sphere, the political/public life etc.). It is especially relevant for this project as the process of development in the regions may be of a very heterogeneous nature so that the arenas, and hence the opportunity for applying specific resources may differ considerably from the one region to the next. Thus the recognition of structural limits is important in the process of identifying the composite parts of individual well-being and for the analysis following. Thus living conditions are defined as: individual possession of resources in the form of money, goods, services, mental and physical energy, social relations, physical security etc. by means of which the individual person may control and consciously direct his/her living conditions in so far as the necessary arenas are available (Andersen, 1999a).
Well-being The purpose of measuring living conditions is ultimately to obtain a picture of individual well-being (Allardt, 1975). The concept of well-being is a broader concept than that of living conditions because it also includes the person’s subjective evaluation of his/her objective resources. Only using objective measuring had some unfortunate consequences for the Greenland study of living conditions 1994 (Andersen, 2001). We found, like Erik Allardt (Allardt, 1975), that to measure individual well-being adequately, it is necessary to combine the objective resources of the individual with the person’s evaluation of these resources. Below is the outline of Allardt’s conceptual apparatus for measuring well-being in the Nordic countries. Allardt’s argument was that well-being should be measured subjectively (well-being) and objectively (happiness), and that the measuring should include material (level of living conditions) as well as non-material (quality of life) variables. Although well-being in this connection only signifies the objective measure, he actually includes all four dimensions in his concept of well-being (ibid.). As the concept of well-being in SLICA, as argued above, should be based on resources and not on needs, we have adapted Allardt’s model to fit this. In summing up, the following may be said about measuring wellbeing and living conditions in SLICA:
202
THOMAS ANDERSEN AND BIRGER POPPEL
(1) We are interested in all aspects of living as experienced by individuals in the context of social groups; (2) individual experiences with living conditions have objective and subjective components, the latter including satisfaction, expectations, and aspirations; (3) we are interested in the resources individuals can apply, the ways in which they choose to apply these resources, and the barriers and incentives to the application of resources; (4) we want to organise our dimensions of living conditions in terms that are both meaningful to Arctic indigenous peoples and that can be associated with hypotheses about regional, community, household, and individual differences in living conditions.
LIVING CONDITIONS IN THE ARCTIC
203
5.2. Context
Thus the concept of living conditions is constituted by individual resources as well as by access to the relevant arenas. Not all resources are equally important, however, and not all arenas are equally relevant. The obvious question, therefore, is how to select the adequate dimensions and indicators of living conditions. As argued in Section 4, the first requirement is that the underlying concept of well-being in a reasonable measure corresponds with the idea of well-being in the group of respondents – in other words that the concept of well-being is context-specific. Our argument is that this can only be achieved by an interactive selection process between the researchers and representatives from the group of respondents. In Figure 2 we have described this process and the elements of the process, which are used within SLICA. Broad social goals The starting point is to identify broad social goals that are generally accepted either by the researchers or by majorities of the respondents. These can be divided into two sub-categories (Kruse and Hannah, 1998):
204
THOMAS ANDERSEN AND BIRGER POPPEL
1. Social concerns/goals that are dominant in all areas of the project. Physical security, health, and sufficient and healthy nutrition are types of universally accepted social goals that are included in this category. These are goals that are identified by researchers in the project and/or on the basis of previous literature in the field of living conditions research. While these goals are universal by nature, the ways of reaching or satisfying them may vary over time and place. In addition we identify social goals that are not universal – but dominant among Inuit and Saami peoples in all areas of the project. These are defined by researchers and by Inuit and Saami representatives. 2. Region-specific social goals. While the basic social goals mentioned above are legitimate in all geographical areas of this project – there exist of course region-specific social goals in each area due to variations in material and technological conditions and cultural and political differences. These social goals are identified in an interactive process between researchers and Inuit and Saami representatives. The implication for SLICA is that we develop two questionnaires: One that contains a common core of questions which is employed in all areas of the project and 11 questionnaires containing region-specific questions. Dimensions of living conditions Dimensions of living conditions are operationalised social goals in terms of variable clusters that comprise important areas of the wellbeing of the individual. The proper “translation” of social goals into dimensions of living conditions is very context-dependent. Food security is defined differently among Inuit in Chukotka than e.g. in Western Europe. The identification of dimensions of living conditions therefore takes place as an interactive process between researchers and Inuit and Saami representatives. Individual/collective resources Relevant individual and collective resources must be derived and operationalised from the dimensions of living conditions. The question is: Which individual resources are needed to obtain individual well-being within the different dimensions of living conditions.
LIVING CONDITIONS IN THE ARCTIC
205
Again, this translation is very context-dependent which calls for a joint effort from researchers and the respondents. We want to underline again that the actual measurement of resources will be at the individual level while we recognise the influence of collective resources on the well-being of the individual. Collective resources are examined in socio-economic and cultural analyses of the different regions and will be related to the individually based survey data in the final analyses. Living conditions indicators Living conditions indicators are the actual questions to be posed as operationalisations or representations of the individual resources mentioned above. These will constitute the questionnaires. The questionnaires will be constructed by the researchers in the project and then pilot tested in all areas. The above model for inclusion of the group of respondents in the process of selecting dimensions and indicators of living conditions ensures that the final research design will be context-specific and thus will reflect also the perceptions of well-being in the groups of respondents so that all relevant dimensions have been included, and that the theoretical variables are correctly and validly operationalised. The household economy – an example. To illustrate the importance of this process and of the application of context-specific models for the identifications of living conditions indicators in general SLICA has introduced an alternative way of measuring among others income among indigenous peoples in the Arctic. This model is described below. In living conditions studies the unit of analysis is most often the individual as the main focus of the research is the individual wellbeing. This is also the case for SLICA. Studies (e.g. Hertz, 1995; Nuttall, 1998; Dahl, 2000) of ways of living and living conditions in small communities in the Arctic, however, document that more traditional family and household structures still play an important role in the Arctic region. Hence the SLICA research team has developed a household production model from which it is also possible to derive information at the individual level.
206
THOMAS ANDERSEN AND BIRGER POPPEL
The household is in this model chosen as a central unit of analyses primarily because of three circumstances: 1. Even though family patterns and ways of living together through several generations have been – and still is changing – it is still common, that a household consists of family members besides the nuclear family. It is also assumed that the share of three generation households is significantly bigger in the Arctic than in e.g. Scandinavia, Canada and the USA. 2. Despite the rapid economic, social and cultural change of the Arctic societies, a large part of the families are still dependent on subsistence hunting and fishing1 or at least a combination of subsistence hunting/fishing, wage income and transfer payments.
In the 1994 Greenland Living Conditions Survey, one of the topics analysed was subsistence activities – measured as the respondents participation in subsistence hunting and fishing activities as well as the share of the food supply stemming from such activities. Whereas less than 9 per cent of the Greenland part of the labour force has subsistence hunting or fishing as their main occupation (Danielsen et al., 1998), the degree to which people participate in those activities at all, is much more extensive. This is especially true for the settlement inhabitants. According to the 1994 Greenland Living Conditions Survey, 67 per cent of the population in the settlements indicating wage income as main income source were engaged in small-scale fishing and/or hunting activities (the corresponding figure for the town inhabitants was 28 per cent). 80 per cent of the households in the settlements had members participating in hunting sea or land mammals and/or fish for the consumption of the household (and/or the sledge dogs) as a necessary supplement to their wage incomes (Andersen and Statistics Greenland, 1995). These figures show both the importance to the Greenland Inuit of having access to hunting and fishing activities and the significant difference between the more traditional living in the small settlements and the more modern ways of living in the towns. (There is no official distinction between the concepts of “towns” and ‘settlements”. In practice, however, the “capitals” of the 18 municipalities are classified as towns – while other inhabited places are classified as settlements. The number of persons living in different towns
LIVING CONDITIONS IN THE ARCTIC
207
range from 550 to 14.000, whereas the population figures in the settlements vary from 25 persons to 500.) Most of all, however, the figures underline that contributions from subsistence activities are crucial to the survival of many households. In traditional statistical studies the outcome from these activities are not part of the income measurement due to the monetary concept of income most often applied by researchers and statisticians. The ultimate consequence is that from a GDP-point of view subsistence hunting and fishing does not contribute to the wealth of a nation (Poppel et al., 2000). To grasp the importance of subsistence and how the household in the mixed subsistence-based economy works as a microenterprise in organising productive activity and allocating the factors of production (land, labour, capital) to optimise income the household production model was developed (Usher, 2000a) to and through the SLICA-process. The model (Figure 3) illustrates how an essential socio-economic unit below the regional economy level – the household – works as a micro-enterprise in the Arctic and makes it possible for living conditions researchers to capture both monetary, non-monetary production and consumption within the household.
208
THOMAS ANDERSEN AND BIRGER POPPEL
There are three types of questions relevant to the household economy model: characteristics of the individual members (their western and traditional education for example); activities of the individual household members (wage labour and hunting for example); and flows of different types of cash and harvests into and out of the household (Usher, 2000b). Households have access to land – not in the conventional sense of ownership – but to the traditional land base used for harvesting. Households own items used in production: boats, outboard motors, snow mobiles, trucks. Household members work to produce money and harvests: they have paid jobs, they hunt and fish, and they process foods. We try to understand how everyone in the household fits together to do these things – this is the challenge theoretically as well as methodologically (Ibid.). As can be seen from the short description of and comments to the model, the main purpose is to understand how the household as a production unit works: the individuals’ resources, the arenas they are being used in and the output of the process. How the process is being perceived and evaluated by the household members will be addressed by questions derived from “outside the model”. 5.3. Focus on Structural and Individual Change As mentioned in Section 5 it is one of SLICA’s main goals to carry out a dynamic social analysis of the effects of structural change on living conditions. As we are not able to measure the relationship between structural changes and individual well-being over time, we shall have to do it over place. Thus the value of a circumpolar comparison of living conditions is found in, among other things, a better understanding of how different structural conditions and changes affect individual resources in culturally comparable groups. For example, how has the exploitation of mineral resources affected the living conditions of the Inuit compared to areas where no such exploitation takes place? Or how have the G50 and G60 plans of the Danish government affected the living conditions of the Greenland population in contrast to areas where no concentration policy, for example, was initiated? The first task is thus to identify the most important structural changes or “forces of change” which have influenced the freedom of
LIVING CONDITIONS IN THE ARCTIC
209
choice, either by (1) changing the resources people possess, or (2) affecting them so that new arenas are chosen where their resources are more or less sufficient, or (3) creating completely new arenas. Kruse and Hannah (Kruse and Hannah, 1998) have identified the following important causes of structural changes among Saami and Inuit in the Arctic: Climate change Government policies Native self-government Resource development Technology Animal rights groups Contaminants Increased access to the north Religion Education The question is whether these structural and institutional factors affect individual living conditions and well-being. Mohatt, McDiarmid and Montoya (1988) presented an overview of the impact of social change among Alaska Natives on human health. They pointed out that forces for change are cumulative, a combination of “big” history (e.g. disasters, social change, economics); and “little” history (e.g. family and individual trauma, developmental trauma, biological/hereditary events within family history, emotional events). They emphasised the importance of understanding how these forces of change affect the coherence of individuals, families, and communities. Quoting Antonovsky (1980), they defined coherence as, . . . a global orientation that expresses the intent to which one has a pervasive, enduring though dynamic feeling of confidence that one’s external environments are predictable and that there is a high probability that things will work out as well as can reasonably be expected (p. 123).
We raise Mohatt’s point here because it would be a delusion to think that we can partition the effects of concurrent forces for change on arena outcomes and, even if we could, we would lose sight of the cumulative effects. These cumulative effects may be the most important in that they become a resource limitation in all living
210
THOMAS ANDERSEN AND BIRGER POPPEL
conditions dimensions at the same time. This means that the loss of a sense of confidence in a degree of social and cultural continuity, for example, can undermine the ability and the will to act, and hence the freedom of choice of the individual in general. However, we assume, as mentioned above, that we are able to identify some direct effects of well-being based on the structural changes which have taken place in the Inuit and Saami societies during the period after the Second World War. Figure 4 describes our view of the overall co-variation between structural change and a changed individual well-being. The model is an illustrative one and not an attempt at reproducing reality precisely. Thus the list of the structural and institutional forces of change is not exhaustive. The model may look rather deterministic at first glance. To underline the individual choice or the free will of man we have added a box to indicate that acquiring and employing individual resources is not only conditioned by forces of change or access to arenas, but also by the individual choice. As a consequence of Mohatt’s argument one might draw an arrow which indicates a correlation between the forces of change and the general will and ability of the individual person to manage his/her resources.
LIVING CONDITIONS IN THE ARCTIC
211
Two consequences arise from the above model. First of all, the identification of specific causes of change should contribute to steering the selection and operationalisation of the living conditions indicators. Secondly, SLICA will produce region-specific socioeconomic reports which will make it possible to correlate and compare these structural events with the living conditions of the individual so that in the analysis phase it will be possible to carry out dynamic social analyses and thus identify how the forces of change have affected the living conditions and the well-being in the various regions. In this way focus remains on change and not on static descriptions of the Inuit and Saami living conditions in the Arctic. 5.4. An Outline of SLICA’s Research Design
The above sections of the paper have mainly been oriented towards (1) a discussion of the experiences from the Greenland Living Conditions Survey 1994, (2) the basic model for living conditions research among Inuit and Saami within SLICA, and (3) a selective discussion of how this model may be implemented in a comparative study of these populations. The drawing up of the actual questionnaire is not discussed here. However, below we have listed the living conditions dimensions included in SLICA’s research design Part A: Family (1) Family relationships and (2) Household economy; Part B: Background (1) Mobility, (2) Language, and (3) Education; Part C: Lifestyle (1) Employment, (2) Harvest, (3) Leisure, (4) Spirituality, (5) Identity and (7) Health; Part D: Environment (1) Housing, (2) Income and expenses, (3) Technology, (4) Safety and justice, (5) Resource management, (6) Environmental health, (7) Political resources and (8) Community viability
212
THOMAS ANDERSEN AND BIRGER POPPEL
This list contains not only living conditions dimensions specific to indigenous peoples in the Arctic but of course also a number of more traditional topics. The point is that the important theoretical variables or resources do not always differ between different kinds of societies – but that they have to be operationalised differently to measure well-being among indigenous peoples in the Arctic and e.g. well-being of the inhabitants of western, industrialised countries as was shown in the example regarding income measurement in section 6.2.
6. THE PERSPECTIVE: LIVING CONDITIONS RESEARCH AND POLITICAL PLANNING IN THE ARCTIC
If living conditions research is going to play a role in connection with political planning and implementation in the Arctic, it will have to become relevant again, and in order to be relevant research must again become theory-propelled and must to a greater extent focus on the interplay between events at system level and individual well-being. Furthermore, living conditions research must be based on context-specific research designs in order to obtain meaningful and realistic measurement of well-being. If these conditions are included in the living conditions studies, living conditions research may contribute with new and essential knowledge about the indigenous peoples in the Arctic. Not least in connection with planning, implementation and evaluation of the sustainable process of development which plays such an important role in setting the political agenda in the Arctic. In 1996, the Arctic Council, whose members represent all the 8 countries geographically linked to the Arctic region, defined sustainable development as a concept not only related to the environment but also to economic and social development and as a question of securing the cultural welfare of the Arctic inhabitants. (Arctic Council, 1996). In order to acquire a correct understanding of how living as well as non-living resources should be regulated in a sustainable way, it is necessary to focus in a greater measure on the social aspects of sustainable development. In order to understand how resources were regulated in the past, how they are
LIVING CONDITIONS IN THE ARCTIC
213
regulated today, and how they should be regulated in the future, it is important to include human activity, politico-economic systems, cultural conditions and technological aspects in the concept of sustainability. Only by including in the mental picture the social, cultural, political, economic and technological differences in the various Arctic societies it is possible to plan and implement a sustainable development in the Arctic. A sustainable regulation of the resources is not a question only of adapting the political and economic control systems. In order to avoid “the tragedy of the commons” (Hardin, 1968) and “the tragic necessity of Leviathan” (Ophuls, 1973), it is necessary to learn more about the living conditions, informal institutions, political resources, attitudes, values and welfare priorities among the Arctic peoples who actually manage the resources. The majority of the indigenous peoples in the Arctic have, as shown above, experienced rapid changes in a number of exogenic variables which to a great extent have affected their lives. Today many of these people mix traditional activities with wage labour, for example. They adapt the traditional lifestyle. Seen from the perspective of sustainable development it is important to study how these events at the system level affect habits of diet, consumption patterns, occupational conditions, incentive structures, and, in continuation of these factors, the ways in which the individual person relates to the living and the non-living resources. When the concept of sustainability does not exclusively relate to animals, plants and minerals but to human beings also, it becomes relevant to study how the above events have affected the living conditions of the indigenous peoples, that is their health, opportunities for work, housing, level and sources of income, educational level, family structures, social networks as well as their attitudes, values and priorities in general. By carrying out a representative, comparative study of individual living conditions, it becomes possible to expose how various indigenous peoples with different cultural, technological, political and economic structures manage their resources. In this way it becomes possible to learn more of the causal relations between economic, political and cultural systems and the way in which the resources are managed. A comparative study of these conditions makes it
214
THOMAS ANDERSEN AND BIRGER POPPEL
possible to identify some of the key variables that decide whether the management of resources is a success or a failure. Thus living conditions research may contribute considerably to political planning in the Arctic - not least in connection with the implementation of the sustainable development process if it is based on adequate and meaningful research designs. We are convinced that the comparative study of living conditions among Inuit and Saami in the Arctic is a major step in the right direction. NOTE 1
The concept of subsistence (hunting and/or fishing) as “hunting and/or fishing for the household economy, with a distribution system which secures that the community shares the products” is based on the definition by Kapel and Petersen (1982).
REFERENCES Allardt, E.: 1975, Att Ha, Att Älska, Att Vara. Om Välfärd i Norden (Lund). Andersen, T.: 1995, Work and Unemployment in Greenland (Statistics Greenland, Nuuk). Andersen, T. and J. Jensen: 1998, ‘Ethnical and demographic aspects of the circumpolar living conditions project’, in T. Greiffenberg (ed.), Development in the Arctic (Danish Polar Center Publications no. 5, København). Andersen, T.: 1999a, ‘Living conditions research and sustainable development’, in Inussuk, Arctic Journal of Research 2, Nuuk. Andersen, T.: 1999b, ‘A circumpolar study of living conditions among Saami and Inuit peoples’, in H. Petersen and B. Poppel (eds.), Autonomy, Dependency and Sustainability in the Arctic (Ashgate, London). Andersen, T.: 2001, ‘Living conditions in the arctic – How to measure living conditions among indigenous peoples living in the arctic’ (will be published fall 2001). Antonovsky, A.: 1980, Health, Stress and Coping (Jossey-Bass Publishers, San Francisco). Berman, M. and L. Leask: 1994, ‘Violent death in Alaska: Who is most likely to die?’ Institute of Social and Economic Research, University of Alaska Anchorage, Alaska Review of Social & Economic Conditions 29(1). Coleman, J.S.: 1971, Resources for Social Change (New York). Dahl, J.: 2000, Saqqaq: An Inuit Hunting Community in the Modern World (Toronto). Danielsen M., T. Andersen, T. Knudsen and O. Nielsen: 1998, Mål og strategier i den grønlandske erhvervsudvikling (Sulisa A/S, Nuuk. Danmarks Statistik og).
LIVING CONDITIONS IN THE ARCTIC
215
Socialforskningsinstituttet: 1992, Levevilkår i Danmark 1992 (København). Erikson, R. and H. Uusitalo: 1987, ‘The Scandinavian approach to welfare research’, in Eriksen, Hansen, Ringen and Uusitalo (eds.), The Scandinavian Model. Welfare States and Welfare Research (M.E. Sharpe, New York, London). Hansen, E.J.: 1978, ‘Fordelingen af levevilkårene’, SFI publikation 82 (København). Hansen, E.J.: 1990, Levekårsforskningen i Norden: Et historisk overblik, DDANYT, Nr. 53 (Odense). Hansen, E.E.: 1999, Udvikling og Forsorg i Gyldendal, Grønland, Gyldendals Boghandel, Nordisk Forlag A/S Copenhagen. Hardin, G.: 1968, ‘The tragedy of the commons’, Science 162, pp. 1243–1248. Hertz, O.: 1995, Økologi og levevilkår i Arktis: Uummannamiut (Copenhagen). Johansson, S.: 1970, Om levnadsnivåundersökningen. Utkast til kapitel 1 och 2 i betänkande att avgives af låginkomstutredningen (Allmänna Förlaget, Stockholm). Kirmayer, L.: 1994, ‘Suicide among Canadian aboriginal peoples’, Transcultural Psychiatric Research Review 31(1), pp. 3–58. Kruse, J. and V. Hanna: 1998, ‘Forces for change in the arctic North America and the sustainability of northern communities’, paper presented at the International Arctic Social Sciences Association, Third International Congress of Arctic Social Sciences, Copenhagen. Kruse, J.: 1999, Proposal to the National Science Foundation (Washington D.C.). Mohatt, G. et al.: 1988, ‘Societies, families and change’, American Indian and Alaska native Mental health Research, Monograph 1, pp. 325–365. NOU: 1976, ‘Slutrapport’, Nou: 28 (Oslo). Nuttall, M.: 1998, Protecting the Arctic: Indigenous Peoples and Cultural Survival (Amsterdam). Kapel, F.O. and R. Petersen: 1982, Subsistence Hunting – The Greenland Case, in Rep. Int. Whaling Commission (Special Issue 4). Ophuls, W.: 1973, ‘Leviathan or oblivion’, in H.E. Daly (ed.), Toward a Steady State Economy (Freeman, San Francisco, pp. 215–230). Poppel, B., T. Andersen and P. Lyster: 2000, ‘Living conditions of indigenous peoples of the arctic – The Greenland Inuit as an example’, paper presented at the IAOS conference: Statistics, Development and Human Rights. www.statistik.admin.ch/about/international/03iaos_det07.htm Ringen, S.: 1975, ‘Noen hovedtrekk ved den velferdsteoretiske diskusjon i Norden’, i A. Ringen (red.), Velferdsforskning og sosialpolitikk (INASrapport). Ringen, S.: 1995, ‘Well-being, measurement and preferences’, Acta Sociologica 38, pp. 3–15. Statistiska Centralbyrån: 1995, Levnadsförhållanden – appendix 15 (Stockholm). Titmuss, R.: 1958, Essays on the Welfare State (London).
216
THOMAS ANDERSEN AND BIRGER POPPEL
Usher, P.J.: 2000a, Understanding the household in the mixed, subsistencebased economy: Towards a survey questionnaire module (draft), www. arcticlivingconditions.com. Usher, P.J.: 2000b, ‘The household economy’, in Iqaluit Summary: meeting of the SliCA-Canada National Steering Committee Iqaluit Nunavut, 18–19 April 2000, www.arcticlivingconditions.com. Vogel, J.: 1996, ‘Social indicators and social reporting’, Statistical Reports of the Nordic Countries no. 66, Copenhagen.
Statistics Greenland Postbox 1025 3900 Nuuk, Greenland E-mail:
[email protected] MICHAELA HUDLER and RUDOLF RICHTER
CROSS-NATIONAL COMPARISON OF THE QUALITY OF LIFE IN EUROPE: INVENTORY OF SURVEYS AND METHODS
ABSTRACT. The intention of this paper is to provide the scientific community with information on research carried out in the course of the comparative research programme EuReporting – Towards a European System of Social Reporting and Welfare Measurement. This research is being funded by the European Commission and is being conducted by research teams throughout Europe. In this paper, we will report on the findings of the research carried out by the team in Vienna. The Viennese research team is to assess the existing stock of comparative and cross-sectional surveys in the field of social reporting and welfare measurement. Of concern here is not only documentary work, but also the quality and comparability of existing survey data. The focus of the Vienna project is upon such cross-national and cross-sectional surveys as the World Values Survey, the International Social Survey Programme, the Eurobarometer and the New Democracies Barometer, as well as national longitudinal studies.
INTRODUCTION
The goal of the EU-funded comparative research programme EuReporting – Towards a European System of Social Reporting and Welfare Measurement is to contribute to the development of a science-based European system for social reporting and welfare measurement. The actual research in this comparative programme is carried out within the framework of three individual projects: Developing a European System of Social Indicators, carried out by the Zentrum für Umfragen, Methoden und Analysen (ZUMA) in Mannheim; Stocktaking of Comparative Databases in Survey Research by the Paul Lazarsfeld Gesellschaft für Sozialforschung in Vienna; and Access to Comparative Official Microdata by the Mannheimer Zentrum für Empirische Sozialforschung in Mannheim. The Social Indicators Research 58: 217–228, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
218
MICHAELA HUDLER AND RUDOLF RICHTER
common aim is to improve knowledge in the field of social reporting and welfare measurement at the European level.
AIMS OF THE PROJECT ON INVENTORY OF COMPARATIVE DATABASES IN SURVEYRESEARCH1
Within this project, the goal is to take stock of the above-mentioned databases, focusing on the social-welfare data currently available in Europe. Our perspective is that quick and easy access to information on social welfare should be provided for interested users. Much of the data on welfare and social reporting are already available in European data archives; but to create a science-based system of social reporting, it will be important to establish an appropriate database with thematic information of cross-sectional surveys, and especially on survey questions covering welfare issues. The objective of the Vienna project is to develop and establish a Dynamic Information Centre (which will be available through the website at: http://members.aon.at/plg/eureporting/) to provide the European Commission and the scientific community with a database of retrievable information on surveys and survey questions in the area of social welfare. The creation of the Dynamic Information Centre on welfare and social reporting can only be achieved in close co-operation with data archives and research institutes that have diligently compiled meta information on surveys for users. Beside the documents work, the Dynamic Information Centre will also be able to provide information on the comparability of existing data. Comparative survey programmes covering welfare issues have been carried out throughout Europe. However, despite the fact that these surveys seem to cover comparable questions, they often do not measure the same thing despite equivalent wording in questions and equivalent response scales within a country over the same period of time. Though at first glance the survey data may appear comparable, equivalent questions can turn out to be problematic in cross-cultural comparisons. Hambleton and Patsula (1998) report on the problems that arise when comparing data gathered in different societies and cultures. They point out that, despite equivalent question wording in the original and translated versions of a given questionnaire, ambigu-
CROSS-NATIONAL COMPARISON OF THE QUALITY OF LIFE IN EUROPE
219
ities still tend to arise in terms of construct equivalence, administration, format, response style, the tendency to guess, and the social desirability factor. All these can exert a significant influence on the comparability of survey data in cross-national surveys. According to Hambleton and Patsula, a further cause of invalidity in survey results can be traced to differing levels of familiarity with item formats, technical designs and methods. One’s cultural background influences how people are accustomed to answering attitudinal or self-rating questions, for example. The Vienna project deals with such multinational cross-sectional surveys as the Eurobarometer surveys of the European Commission, and with additional special-interest surveys from Directorate General X. Special emphasis will be placed upon such scientific comparative studies as the International Social Survey Programme (ISSP), the World Values Survey (WVS) and the New Democracies Barometer. The latter was organised and carried out by the Paul Lazarsfeld Gesellschaft für Sozialforschung. The Vienna project will also focus on longitudinal national studies with repeated crosssectional samples, insofar as they are available in archives, and will attempt to locate large-scale studies. National studies will only be considered if the questionnaire is offered in an English or German version. The contribution of the Czech Academy of Sciences, as an associate partner, will be to document the survey information on social reporting in the Czech Republic. Five other institutions are also supporting the research of this project: the Observatoire Français des Conjunctures Économiques, Strathclyde University, the Institute for Advanced Studies, and the Erasmus University Rotterdam.
SUBJECTIVE AND OBJECTIVE INFORMATION ON SOCIAL WELFARE COVERED BY THE VIENNA PROJECT
As a rule, welfare surveys nowadays cover both subjective and objective indicators (Zapf, 1984; Berger-Schmitt and Jankowitsch, 1999; Noll, 1999). Such different theoretical approaches towards welfare measurement as Allardt’s approach to welfare (1973) – as well as various new welfare concepts like “Livability” (Veenhoven, 1997), “Social Cohesion and Social Exclusion” (Noll, 1999), and
220
MICHAELA HUDLER AND RUDOLF RICHTER
the “Human Development” approach (UNDP, 1990) – all consider subjective and objective indicators as being relevant for welfare measurement. However, the attempts made to contribute to welfare survey research at the beginning of the social reporting movement varied greatly. Those that preferred objective indicators for welfare measurement comprise the Scandinavian approach (Erikson, 1973) to survey research, in part Sen’s approach towards welfare (Sen, 1993; also see Berger-Schmitt and Jankowitsch, 1999), and the new concept of sustainable development (see Noll, 1999). A third approach is the quality-of-life concept, which bases welfare on subjective indicators (Campbell, 1972). Detailed information on the theoretical discussion of welfare concepts and on recent approaches to social reporting can be found in Working Paper 1 (see Berger-Schmitt and Jankowitsch, 1999) and Working Paper 3 (Noll, 1999) of the EU-Reporting Working Paper Series. FIRST STEPS TOWARDS AN INVENTORY OF COMPARATIVE DATABASES IN SURVEY RESEARCH
Many cross-national and cross-sectional surveys (such as the International Social Survey Programme or the World Values Survey) assess welfare topics by both subjective and objective indicators. Surveys in the Scandinavian tradition, as was mentioned above, tend to focus exclusively on gathering objective information on living conditions. Most of the surveys focusing on social welfare make enquiries as to objective living conditions as well as subjective well being. In this project, a preliminary inspection of comparative survey programmes – both cross-national and national – focuses on questions concerning the subjective and objective aspects of daily life as well as the quality of life. Documented are the following: objective living conditions (i.e., the exclusively objective component); questions on social status and social behaviour (mainly objective with a minor subjective component);
CROSS-NATIONAL COMPARISON OF THE QUALITY OF LIFE IN EUROPE
221
questions on one’s perception of his/her individual situation in different areas of life (again, mainly objective with a minor subjective component); and questions on how satisfied one is with the different areas of his/her life: happiness, contentedness, positive (or negative) attitudes towards other people, institutions, issues, and services in various areas (all of which indicate the subjective quality of life) (see Figure 1).
DESCRIPTION OF CROSS-SECTIONAL SURVEYS ON SOCIAL WELFARE
Socio-economic and political changes are the main causes for the development of an infrastructure to monitor welfare and standard of living. Since 1970, such supranational government organisations as the European Commission – as well as international research groups like the World Values Study Group or the International Social Survey Programme Research Group – have all carried out numerous science-based transnational, cross-sectional surveys to monitor societal progress in various domains. In addition to international activities, there are also many national approaches to monitor
222
MICHAELA HUDLER AND RUDOLF RICHTER
social change. In order to ensure a certain standard, only surveys stored in data archives have been considered for our inventory. We obtained all the surveys currently covered by our project through online catalogues from the various data archives. Our Czech partners provided us with the surveys from the Czech Republic, Hungary and Poland. The documented surveys are the result of searches in archives using the keywords “welfare”, “standard of living” and “quality of life”. The time period covered spans from 1980 to the present. Information on standard of living surveys from Scandinavian countries is taken from the web pages of the national statistical offices. The Polish surveys are based on descriptions provided by Adamski et al. (1999). The information from the Czech surveys is based on publications from the Sociological Data Archive of the Czech Academy of Sciences (http://archiv.soc.cas.cz, 2000-06-14). There are certainly a number of surveys that are not available in archives. They are stored in a variety of different institutions and cannot be found with an archive enquiry. To obtain access to these surveys, we requested assistance from our international partners in the EuReporting project, who all are deeply into welfare research. The regional coverage of the surveys comprises the 15 member states of the European Union, Switzerland, Norway, the Czech Republic, Hungary and Poland. Up to now, the surveys dealing with welfare issues have not been exhaustive. National surveys are frequently not available to interested users and, in most cases, neither is an authorised English or German translation of the questionnaire. The next section presents comparative cross-sectional crossnational surveys since 1980. This is followed by information on cross-national surveys that have been carried out once, but allow a comparison between different countries. The last section covers ongoing national studies that are stored in archives; these receive only scant mention in this paper but do provide data on many welfare issues at the national level (for details, see Hudler and Richter, 2000). To qualify for consideration, the national surveys need to be conducted over a minimum of three periods since 1980 and must cover social issues related to welfare measurement. At least three periods are necessary for comparison to be able to observe any real
CROSS-NATIONAL COMPARISON OF THE QUALITY OF LIFE IN EUROPE
223
changes in society. Data comparison from only two waves might result in interpreting changes as societal changes though they are simply due to different data-collection methods over a given period. To reduce this bias, researchers should compare data from surveys carried out more than three times. The preliminary results of our inventory are the following surveys, which have been categorised as follows: (1) repeated cross-national and cross-sectional surveys, (2) cross-national and cross-sectional one-time surveys, and (3) repeated national cross-sectional surveys. 1. Repeated Cross-national and Cross-sectional Surveys
The European Commission’s Eurobarometer Programme (http:// europa.eu.int/comm/dg10/epo/eb.html, 2000-06-14) focuses on long-term attitudes regarding health, education, leisure, communications, citizen participation, employment, income, consumption, social security, public safety, housing, environment and values. Part of nearly every survey are so-called “trend questions” investigating Europeans’ level of satisfaction with life, democracy and how values are prioritised. The Level of Living Surveys carried out in Scandinavian countries provide information on almost every area of life, though mainly on an objective level. However, a few questions on subjective issues are added, too. We also analysed the Central and Eastern Eurobarometer (http:// europa.eu.int/comm/dg10/epo/ceeb.html, 2000-06-14) conducted on behalf of the European Commission in Central and Eastern European societies. Some of the questions are equivalent to those in the Eurobarometer surveys, so that a few topics are indeed comparable between East and West. Concerning welfare, the surveys assess political participation, satisfaction with the democratic system, trust in the media, use of the media, and educational topics. The New Democracies Barometer (http://www.cspp.strath.ac.uk/ catalog4_0.html, 2000-06-14) monitored the social and political changes during transformation in Eastern Europe on five different occasions. It, too, deals with issues of social welfare that have also been investigated by surveys in Western Europe. Topics of concern
224
MICHAELA HUDLER AND RUDOLF RICHTER
are standard of living, satisfaction with the household’s economic situation, citizen participation, and attitudes towards migration. In the same vein, the International Social Survey Programme (http://www.issp.org, 2000-06-14) gathers data for comparison across time and space every five years. Time series reveal the development of certain social issues to pinpoint differences among societies. Thus, the International Social Survey Programme offers opportunities for cross-national comparisons between Western and Central and Eastern European countries. Such surveys as the International Social Survey Programme and the World Values Surveys provide valuable information for both politicians and researchers. These surveys enable detailed comparisons between member states of the European Union and the so-called “Applicant Countries”. The topics investigated in the International Social Survey Programme over time have been the role of government, social networks and support systems, social equality, family and changing gender roles, work orientations, religion, environment, and national identity. The World Values Surveys (http://www.icpsr.umich.edu, 200006-14) inquires across a vast spectrum of values and attitudes throughout various areas of life in various societies of the world. The International Crime (Victimization) Surveys, carried out only three times thus far, cover a single topic, i.e. public safety and crime. 2. Cross-national and Cross-sectional One-time Surveys
Several cross-national survey programmes have been identified that were carried out only once but provide important information for welfare measurement. Research programmes on this topic carried out in both Eastern and Western Europe are the comparative project on Class Structures and Class Consciousness, the Reader’s Digest Eurodata 1990, the Pulse of Europe, TARKI-INTER, the International Social Justice Project, Social Stratification in Eastern Europe after 1989 and the Social Consequences of Economic Transformation. All of these provide comparable data for social reporting in different countries. 3. Repeated National Cross-sectional Surveys
An impressive variety of surveys on social reporting issues have been carried out at the national level. We have selected a number of
CROSS-NATIONAL COMPARISON OF THE QUALITY OF LIFE IN EUROPE
225
national cross-sectional surveys both relevant and repeated. These describe the individual’s well being based on objective indicators, and sometimes also on subjective indicators. We analysed the repeated national surveys on topics relevant to social reporting. A preliminary overview of repeated national, cross-sectional surveys from the countries that are covered in our project can be seen in the following list: Great Britain and Northern British and the Northern Ireland Ireland: Social Attitudes Survey Czech Republic: Economic Expectations and Attitudes (EEA) Finland: Finnish Level of Living Survey Finnish Social Thinking France: Living Conditions and Aspirations of the French Germany: German Welfare Survey German General Social Survey (ALLBUS) Netherlands: Living Conditions in The Netherlands Cultural Changes in The Netherlands Norway: Norwegian Level of Living Survey Polish General Social Survey Poland: The Poles Life Conditions and Aspirations Social Structure and Consciousness of Polish Society Swedish Level of Living Survey Sweden: The Swedish National Survey of Living Conditions (ULF) Latinobarometer Spain: Centre for Research on Social Reality Survey
226
MICHAELA HUDLER AND RUDOLF RICHTER
The questions asked in these comparative survey programmes can be categorised as falling between providing information on an exclusively subjective or purely objective level. To sum up our inventory activities, we investigated essential topics relevant to social reporting and welfare measurement both over time and across countries, in a number of cross-national and cross-sectional as well as in repeated national cross-sectional surveys. The topics were often assessed using identical questions. Although the comparability of data seems obvious, it will require analysis if a question measures the same concept in different countries. Many factors can influence the comparability of survey results: translation and cultural effects (see Hambleton and Patsula, 1998), different positions of the questions in the questionnaire, datacollection methods, and so on (Scherpenzeel and Saris, 1997; Saris, van Wijk and Scherpenzeel, 1998). Besides documenting, another task of this project is to provide quality estimates for data in comparative analyses. Although these multi-national or national surveys measure the same topics over the same period in different countries, there is still the question of the comparability of the data obtained. Do the data really measure the same concept in different societies? Do they really measure the same concept over time? Are differences and changes real differences and changes, or are they due to different languages, for example, to a different wording of the questions, to different positions in the questionnaire, to different response scales, sampling methods or interviewer effects? Therefore, the intention of this project is to provide information on the quality and comparability of survey questions and data within the database of the Dynamic Information Centre, insofar as they have been published. NOTE 1
Hudler, M. and R. Richter: 2000, State of the Art of Surveys on Social Reporting in Western and Eastern Europe. EuReporting Working Paper No. 7, Vienna.
CROSS-NATIONAL COMPARISON OF THE QUALITY OF LIFE IN EUROPE
227
REFERENCES Adamski, W. and W. Pelczynska-Nalecz, Zabowrowski: 1999, System of Social Indicators, Social Reporting and Polish Society Transformation. State of the Art Report. EuReporting Working Paper No. 2, Warsaw. Allardt, E.: 1973, About Dimensions of Welfare. Research Group of Comparative Sociology. Research Report No.1. University of Helsinki. Berger-Schmitt, R. and B. Jankowitsch: 1999, System of Social Indicators and Social Reporting: State of the Art. EuReporting Working Paper No. 1, Mannheim. Campbell, A.: 1972, ‘Aspiration, satisfaction and fulfilment’, in A. Campbell and P. Converse (eds.), The Human Meaning of Social Change (Russel Sage Foundation, New York). Erikson, R.: 1993, ‘Descriptions of inequality: The Swedish approach to welfare research’, in M. Nussbaum and A. Sen (eds.), The Quality of Life (Clarendon Press, Oxford). Hambleton, R.K. and L. Patsula: 1998, ‘Adapting tests for use in multiple languages and cultures’, in Social Indicators Research (Kluwers Academic Publishers). Hudler, M. and R. Richter: 2000, State of the Art of Surveys on Social Reporting in Western and Eastern Europe. EuReporting Working Paper No. 7, Vienna. Noll, H.-H.: 1999a, Konzepte der Wohlfahrtsenwicklung: Lebensqualität und “neue” Wohlfahrtskonzepte. EuReporting Working Paper No. 3, Mannheim. Saris, W.E., T. van Wijk and A. Scherpenzeel: 1998, ‘Validity and reliability of subjective social indicators’, Social Indicators Research 45, pp. 173–199. Scherpenzeel, A.C. and W.E. Saris: 1997, The validity of survey questions’, Social Methods and Research 25, pp. 341–383. Sen, A.: 1993, ‘Capability and well-being’, in M. Nussbaum and A. Sen (eds.), The Quality of Life (Clarendon Press, Oxford). United Nations Development Programme (UNDP): 1990, Human Development Report 1990 (Concept and Measurement of Human Development, New York). Veenhoven, R.: 1997, ‘Livability of nations’, in H.-H. Noll (ed.), Sozialberichterstattung in Deutschland (Juventa, Weinheim). World Values Study Group: 1994, World Values Survey, 1981–1984 and 1990– 1993 (Inter-university Consortium for Political and Social Research, Ann Arbor). Zapf W.: 1984, ‘Individuelle Wohlfahrt: Lebensbedingungen und wahrgenommene Lebensqualität’, in W. Glatzer and W. Zapf (eds.), Lebensqualität in der Bundesrepublik (Campus, Frankfurt / New York, pp. 13–26).
References on the World Wide Web Central and Eastern Eurobarometer at the European Commission – http:// europa. eu. int/comm/dg10/epo/ceeb. html
228
MICHAELA HUDLER AND RUDOLF RICHTER
Eurobarometer at the European Commission, Directorate-General X, Public Opinion Analyses – http://europa.eu.int/comm/dg10/epo/eb.html ICPSR / Inter-university Consortium for Political and Social Research – http://www.icpsr.umich.edu ISSP / International Social Survey Programme – http://www.issp.org New Democracies Barometer at the Centre for the Study of Public Policy – http://www.cspp.strath.ac.uk/catalog4_0.html Sociological Data Archive of the Czech Academy of Science in Prague – http://archiv.soc.cas.cz World Values Survey in “World Values Survey, 1981–1984, 1990–1993 and 19951997”, published by the World Values Study Group, Inter-university Consortium for Political and Social Research, 1999 – http://www.icpsr.umich.edu
Paul Lazarsfeld Gesellschaft für Sozialforschung Maria Theresienstrasse 19/9 A–1090 Vienna, Austria E-mail:
[email protected] ABBOTT L. FERRISS
TELESIS: THE USES OF INDICATORS TO SET GOALS AND DEVELOP PROGRAMS TO CHANGE CONDITIONS
ABSTRACT. TELESIS – the application of intelligence to attain desired goals – employs social indicators to set goals and monitor progress toward achieving goals. The model links culture values to desirable goals which are identified by a change in social indicators statistic. Understanding appropriate steps to take, based upon experience and experimental knowledge, is vital in developing programs. Human and fiscal resources also are critical. Healthy People, a U.S. program to extend life, is presented as an example of the process, with the program to stop cigarette smoking reviewed in detail. Another example is the U.S. Goals for Education program with its 50 objectives to improve learning. The Kids Count, a program supported by a private foundation, is designed to improve the well-being of children and youth. It is an example of the telic process that engages local initiatives. From these examples, a summary of lessons learned provides suggestions for the application of the telic process.
INTRODUCTION
Friedrich August von Hayek, an Austrian economist who died in 1992, advocated free-market capitalism as the most efficient means of distributing goods and services. Many in the world now agree with him (Cassidy, 2000). The free market system must be supported by norms and values that uphold private property, respect for contracts and honesty in dealings. Through such a system, human activity in distributing information, producing goods and services, buying and selling, needs are satisfied efficiently. The market is the mechanism. When the market fails to satisfy needs, corrective action is appropriate (Cassidy, 2000, p. 51). To gauge the satisfaction of needs, the condition of the U.S. population now is monitored by social indicators. Indicators provide a system of information on the population by age, sex, location, and a host of other defining characteristics. We evaluate informaSocial Indicators Research 58: 229–265, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
230
ABBOTT L. FERRISS
tion on present conditions and needs in terms of our values. Values, based largely on Judeo-Christian traditions, define a condition as desirable or undesirable. The undesirable becomes the object for remedial or corrective action. Remedial steps include services that the market system has not provided. Thus, the capable young who can not pay the price of an education under normal market prices are provided scholarships so that talents are not wasted. Indigent persons in need of medical care are provided emergency treatment. When the business cycle dips, demand falls, and employees are laid off; the unemployed are supported to allay suffering. These and myriad other adjustments are required when the market fails to provide the services that the norms and values of society define as desirable. Social indicators supply trends in these conditions and needs.
AN INFORMED ELECTORATE: AN ABSOLUTE REQUIREMENT OF DEMOCRACY
The free flow of information is critical for efficient functioning of the free market and of democratic systems. Dr. Theodore D. Woolsey stated this position most eloquently when he stepped down as chairman of the (U.S.) Council of Professional Associations on Federal Statistics: “The citizen reads or hears about the rate of inflation; the rate of unemployment; changes in the cost of living; the growth and movement of the population; estimates, however poor, of illegal immigration; crime rates; the degree of success we are having in education of our children; the rate of divorce; the increasing numbers of single-parent families; the spread of AIDS; the number of teenagers killed in auto accidents or committing suicide; and hundreds of other descriptors of today’s life. These social indicators, if you like, absolutely must be made available to any interested citizen with a minimum of expense and difficulty because an informed electorate is an absolute requirement of a working democracy” (Woolsey, 1987). While these indicators inform the electorate to animate a working democracy, they also may be used to set goals for future change. Identifying such goals and setting about altering their direction or rate of change is a process called “telesis.” Telesis means: “Progress
TELESIS
231
that is intelligently planned and directed; the attainment of the desired ends by the application of intelligent human effort to the means.” (Webster’s New Collegiate Dictionary, 1977; Ward, 1903, used the term in a broader sense; Commager, 1967).
VALUES AND GOALS
In the 1960’s, scholars of social indicators began linking social indicators with national goals for guided change in the service of public policy. President Eisenhower’s Commission on National Goals (1960) gave concerted attention to goals as ingredients in setting policy and priorities. Later, in the seminal volume edited by Raymond A. Bauer (1966) appears “Social Indicators and Goals” by Albert D. Biderman, analyzing the practice and setting the stage for more deliberate consideration of goals in social planning. A note of economic cold water had been struck by Leonard A. Lecht (1966) in his analysis of the cost of achieving goals in education, health, welfare, urban development, private plant and equipment expansion, etc. He anticipated the 1975 cost to exceed 15 percent of GNP. Robin M. Williams, Jr. (1967) clearly set forth the valuesgoals-policy linkage and defined values as “those conceptions of desirable states of affairs that are utilized in selective conduct as criteria for preference or choice or as justifications for proposed or actual behavior” (p. 23). Values are “high-level guidance signals” that lead to setting goals, he said (p. 27). Williams’ scheme aligns values, goals, policy formation, resources, etc. in much the same arrangement as Map 1. Of values, he added: “. . . the primary location of values at the macroscopic level of social structure is as one of the sources of goals and criteria for policy formation.” In Social Goals and Indicators of American Society, a host of authors set forth goals and indicators for policies affecting civil liberties, voting, the mass media, discrimination, the natural environment, urbanism, poverty, education, science and technology, health and well-being, and others areas of public concern (Gross, 1967). Thus, scholars and public policy analysts have an extensive history of engaging values to identify goals in the interest of developing policies and programs to bring about social change, the Telic process to which we now turn (For additional background, see Ferriss, 2000).
232
ABBOTT L. FERRISS
A TELIC MODEL
The Values of society, cultural values, are the starting point in the initiation of social change, Map 1. For example, the following values were ranked among the top five in recent surveys of the American public: “A world at peace (free of war and conflict); family security (taking care of loved ones); freedom (free choice, independence); happiness (contentedness); and self-respect (self-esteem)” (Inglehart, 1990, p. 119). Such cultural values as these define the desires (wants) of people in society. Thus, “family security” translates into
TELESIS
233
the need, among other things, to preserve life, to live free from harm, and identifies the goal of “longer healthy life,” shown in Map 1. When people realize that their values are not being realized, they may desire a change in such conditions. Social indicators help to establish the discrepancy between the actual condition and the desired. Trends in the indicator reveals the direction of change, whether improving or declining. As pointed out by Woolsey (1987), social indicators are “absolutely necessary for a working democracy.” That a lag exists in social conditions relative to the desired is a call to action. The pattern of optimism that change is possible, also, must be present. If pessimism prevails, believing that things always will be the same, that change is impossible, the Telic model would not apply until the more optimistic value is accepted. Social indicators are measures of the human condition with respect to particular traits, say, the unemployment rate, the mortality rate from breast cancer of 45- to 49-year-old females, or the percent of 18-year-old youth not in school and not in the labor force, as suggested by Woolsey in the quotation above. Social indicators, then, are used to establish the reasonable extent of change in a rate by x-amount that is to be achieved by y-years time. In thus setting the goal, social indictors serve the development of policies to bring about the desired change. Policies are implemented through specific programs. Map 1 illustrates programs currently underway in the U.S.: Healthy People 2000, Educational Goals for America, and KIDS COUNT, to be described later.
PROGRAM IMPLEMENTATION
Program implementation of the teleological process is presented in greater detail in Map 2. Map 2 shows that it is necessary to know what steps will alter the direction or rate of change in an indicator. The “cause-effect sequence” refers to the understanding that if a certain step is taken it will have a predictable effect. Thus, the early identification of breast cancer in females leads to eliminating the condition through surgical intervention; or, training the unemployed in new and needed skills will lead to employment in new technological industry.
234
ABBOTT L. FERRISS
Knowing the steps of the cause-effect sequence for any indicator requires knowledge, knowledge gained through practical experience, experiments, demonstrations and tests. For example, in the case of the Healthy People program to reduce cigarette smoking, many studies had proved the adverse consequences of cigarette smoking and others had shown that stopping the habit led to improved health (CDCP, March 24, 1989, Feb. 26, 1999). Such information about causes and effects helped establish the goal. The next problem was to determine what programs held promise to
TELESIS
235
effect change in the indicator. The implementation of the program is designed to alter the social condition, as measured by indicators. This cycle is illustrated in Map 3. The Program Implementation Cycle begins with the setting of a goal to change a rate by x-amount by time t. The process of determining appropriate interventions will rest upon knowledge of the cause-effect sequences, past experience, and wisdom to select reasonable steps. For example, the use of mass media to influence behavior has been effectively employed in some cases. In others, community organizational efforts involving face-to-face interaction has been more effective. In the Healthy People example, to be presented next, the clinical approach employing the doctor-patient relationship was attempted and found inadequate to the problem. More effective methods were then introduced.
236
ABBOTT L. FERRISS
Lester Ward, who first advanced the concept of telesis in social affairs, believed that change could most effectively be accomplished through legislative initiatives (Ward, 1906; Commager, 1967), Whether through legislation or through private initiatives, resources – money, workers, cooperation of actors, etc. – are required. As with any social movement, the generation of public interest and support is necessary. Dependence upon governmental support is one approach, as Healthy People illustrates. But private interests can also be successful, as shown by the Kids Count program, below. Without resources of funds and manpower, the telic process will falter. As time passes, social indicators will reveal change in social conditions. This process is illustrated below in graphs. Monitoring progress involves identifying these changes, not only in the aggregate, but also with respect to sectors of the target population. This is illustrated below, especially, in the case of cigarette smoking. Segments of the population differ in prevalence rates. Attention must be directed toward the most critically affected segments. If progress is not being realized, the interventions should then be evaluated for their effectiveness, and, if found lacking, new steps initiated. Thus, the sequence of the implementation cycle would begin again (Map 3).
HEALTHY PEOPLE
In 1979, the U.S. Surgeon General published Healthy People. It set forth a strategy for improving the health and quality of life of the American people, a Telic strategy. Indicators in time series of mortality, morbidity and other survey indicators gave evidence of persistent health concerns. To establish objectives, regional “hearings” were held to assemble evidence from experts as to the best experience and understanding of the underlying causes of the several medical conditions. These working groups of professionals identified objectives for the program to reduce prevalence. The best available expertise identified steps to take toward achieving the objectives (U.S., Department of Health and Human Services, 1990). These steps are identified in our model (Maps 1 and 2) as the application of experiments and experience to setting goals and
TELESIS
237
developing programs to alter the direction or rate of change of the social indicator. The Surgeon General set forth the following major goals of the program: (1) Increase the span of healthy life of Americans. (2) Reduce health disparities among Americans. (3) Achieve access to preventive services for all Americans. These goals are implemented through 319 specific objectives (U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, 1998).
SPECIFIC GOALS
Goals are specific in terms of the health indicator and, frequently, in terms of the age, sex, ethnic group or other identifying traits of the population segment most deficient. For example, for Goal 1, the life expectancy at birth of the population is supplemented by an indicator of the death rate of people aged 74 and younger, and the infant deaths per 1,000 live births. An average years of healthy life is calculated by combining death rates with rates of acute and chronic illness, impediments and handicaps. For example, with a life expectancy at birth of 73.7 years in 1980, the years of dysfunctional life was estimated, using life table methods, at 11.7 years, reducing expected years of healthy life to 62 years. To extend life expectancy, one of the goals, shown in Figure 5, was to reduce infant mortality from about 10 per 1,000 live births to 9 by 1990. Figure 5 shows that the goal was achieved. As will be illustrated later, not all goals were achieved. For Goal 2, life expectancy is separated into color (white, black), the life expectancy being lower for blacks than whites (in 1987: whites 75.6 years, blacks 69.4 years). By 1987 infant mortality rates since 1970 declined 19 percent for both white and blacks. Death rates for people 74 years and younger, also, had declined during this period, but at a slower rate. This illustrates the necessity of employing indicators, not only of the gross measure, but also of sub-categories, in this case, color, sex, and age. Indicators for Goal 3 included the percentage of pregnant women receiving first trimester prenatal care, by color; percentage of children immunized by time of school entry; and the percentage of
238
ABBOTT L. FERRISS
people who lack a source of primary health care (U.S. Department of Health and Human Services, 1990, pp. 43–51). Thus, the program set forth the three general goals of the program became 319 specific objectives to improve health, each with steps designed to alter the course of given health conditions of segments of the population. The morbidity and mortality statistical systems and surveys of practices of the population were in place to monitor progress toward the goals. Let us now examine a specific goal, cessation of cigarette smoking.
CIGARETTE CESSATION PROGRAM
The Surgeon General has considered tobacco a health hazard for some time and in 1964 he issued a landmark report linking smoking to disease. His position advocating smoking cessation was supported by some 7,000 studies showing the effects of tobacco use on health (CDCP, March 24, 1989). Tobacco use causes one in six deaths in the U.S. Cigarette smoking causes 434,000 deaths yearly. It causes 21 percent of coronary heart disease deaths, 87 percent of all lung cancer deaths, and 82 percent of deaths from chronic obstructive pulmonary disease. These causes of mortality affect different segments of the population. Consider Figure 1, showing the percentage smoking cigarettes in 1991 for different population segments for which such data were available from the National Health Interview Survey. The total smokers among persons 20 years of age and over were 26 percent in 1991. The goal for 2000 was 15 percent. Rates were higher among blue-collar workers, persons with less than a high school education, blacks, men more than women, Indians more than Hispanics, and women in general more than pregnant women, etc. In fashioning programs to reduce prevalence, these differentials provided clues to target population segments.
TELESIS
239
MODELS OF CHANGE
Initially a universal medical model sought to reduce smoking. Health professionals were to change smokers’ attitudes one at a time. When the monitors of the program observed that the clinical model did not work, public health officials took an environmental approach, advocating increases in taxes on cigarettes and restricting of smoking in public areas, and other programs. The U.S. Congress passed the Federal Cigarette Labeling and Advertising Act of 1965, and the Public Health Cigarette Smoking Act of 1969. They required that cigarette packages carry a warning that they injure health. They banned TV and radio broadcasting of cigarette advertising. By June 30, 1995, state legislatures had joined the fight and had passed 1,239 laws addressing tobacco use (Shelton, et al., Nov. 3, 1995). [As noted above, Lester F. Ward, who originally used the term, “telesis,” in his studies, advocated legislative initiatives as the most direct means of bringing about change (Ward, 1906; Commager, 1967).] Laws restricted minors access to cigarettes after 1988. Requiring a smoke free indoor air began as early as 1984. Legislatures introduced marketing restrictions more slowly but they were introduced. The cumulative total of these provisions has been increasing since 1982 but has increased quite rapidly since 1992 (CDCP. June 8, 1999). More recently the private sector began offering devices to aid abstinence from smoking. In recent years States have sued tobacco
240
ABBOTT L. FERRISS
companies to recoup costs of treatment of illnesses attributed to tobacco use. With these funds some States have initiated multimedia campaigns designed to reduce smoking, especially among youth. These efforts have reduced prevalence of cigarette smoking, Figure 3. The percent of smokers, by sex, shows the long-term success of the program. By 1997, only 25.5 percent of males and 21.3 percent of females smoked cigarettes. Rates of smokers among males varied by state from 16.1 percent in Vermont to 31.7 percent in Missouri and 33.1 percent in Kentucky (U.S. Centers for Disease Control and Prevention, 1998), reflecting differences in effectiveness among States. SPECIFIC REDUCTIONS IN INDICATORS
The preceding merely illustrates one effort of the HPDP program. Other efforts were stated as specific reduction of indicators, as illustrated as follows: Objective 1.1 Reduce coronary heart disease deaths to no more than 100 per 100,000 people. Figure 4 illustrates the progress to 1993. Objective 1.1 a Reduce coronary heart disease deaths among blacks to no more than 115 per 100,000 people (U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, 1998).
TELESIS
241
The goal for reducing infant mortality is being achieved, Figure 5. The rate stood at 7.2 in 1997. The goal for death rates of children 1–14 years of age, likewise, is being achieved. Success has not shone upon the death rates for ages 15 to 24. Reducing young adult mortalities from accidents and violence has been more intractable, but some progress lately has been observed. However, reduction in the death rates among those 25 to 64 years is being achieved. In addition to attention to use of tobacco, health promotion goals include attention to physical activity and fitness, nutrition, alcohol and other drugs, family planning, mental health and mental disorders, violent and abusive behavior, and a community-based program of health education (U.S. Department of Health and Human Services, 1990). These and other efforts have brought about an increase in life expectancy, Figure 9. Life expectancy in 1999 at birth was 72.5 years for males and 78.8 years for females. CAUSE–EFFECT SEQUENCES
The use of the telic process to stimulate change in the direction and rate of change in indicators rests heavily upon knowledge
242
ABBOTT L. FERRISS
of cause-effect sequences. Some 7,000 experimental and observational studies established that cigarette smoking causes health hazards. Cancers, in particular, became the focus of studies and the objective of control efforts (Greenwald, 1986) Other research identified approaches to changing behavior. Now, research on prevention of cancer is establishing a linkage with nutrition, and efforts of the private sector are afoot to educate the public in the salutary effects of fruits and vegetables in the diet (American Institute for
TELESIS
243
Cancer Research, 2000). As experimental research and epidemiological observations uncover new knowledge of causes, preventive efforts will enter the telic process. in the program implementation cycle (Map 3). Map 2 indicates that Resources are applied to Programs. This implies both Federal, State and Local fiscal and personal resources, and private as well as public efforts. While resources applied to the improvement in life expectancy, itself, cannot be separated from other health efforts, in 1997 the magnitude of health expenditures, to illustrate, was $1,082.4 billion, of this public expenditures were $507.1 billion. Out of pocket private expenditures for health care in 1997 was $1,841 per consumer unit. The total health expenditures in 1997 in the U.S. was 13.5 percent of the Gross Domestic Product. This compares with 10.4 percent for Germany’s GDP and, 4.0 percent in Turkey (National Center for Health Statistics, 1999). Thus, massive public and private expenditures were devoted to improving health. THE GOALS FOR EDUCATION PROGRAM
The National Education Goals program is another illustration of the telic process in action. Initiated in 1989, it has not operated long enough to show dramatic results. In September 1989, President George Bush met with governors of the states at Charlottesville, Virginia, in a National Education Summit. They devised a plan to improve the quality of American education that relied upon states and local communities to determine how the goals would be achieved. They set six goals to which Congress later added two more. The eight national goals are, as follows (National Education Goals Panel, 1999; Creech, 2000): 1. All children will start school ready to learn. 2. The high school graduation rate will increase to at least 90%. 3. All students will become competent in challenging subject matter. 4. Teachers will have the knowledge and skills that they need. 5. U.S. students will be first in the world in mathematics and science achievement. 6. Every adult American will be literate.
244
ABBOTT L. FERRISS
7. Schools will be safe, disciplined, and free of guns, drugs, and alcohol. 8. Schools will promote parental involvement and participation. These goals are more general and less subject to measurement than goals of the Healthy People program. However, some 50 subgoals were identified that are subject to measurement by the educational reporting system, even though some of the indicators are not currently routinely collected. Table I presents these goals. It shows the 50 subgoals, the “baseline” value of the indicator, the most recent value and the long-term goal. Over the varying short periods of observation, improvement in the U.S. educational goals has been spotty, as evidenced by the following summation of results: Improvement in the indicator Regression in the indicator No change in the indicator Data not available Total
10 8 10 22 50
However, some State educational systems have improved while others have not. Twenty-three States significantly improved in 10 or more goals, eight States improved in 12 or more goals. The State reports list the States improving, those with highest levels of improvement. State reports cover each of the 50 subgoals. In addition, practices of States that have proven successful are described, so that other States may consider adopting the successful practices (Wurtz, 1999). Thus, successful practical experience of a State becomes the model to be emulated by other States. This follows the traditional U.S. pattern of States being experimental laboratories to demonstrate feasibility of some instrumentality of change that then is emulated by other States. Two education subgoals also are goals of the Healthy People program, numbers 1.2 and 1.2 (Table I). Seven goals are not measures of student learning but reflect the school environment, numbers 7.1 through 7.7. Three goals (8.1–8.3) concern parental involvement rather than student achievement, apparently assuming that stimulating parental interest will impact student learning. Other observations: (1) The Goals for Education program is relatively new. Expecting immediate goal-attainment is not realistic. (2)
TELESIS
245
Data to assess 22 subgoals are unavailable. Standardized, periodic collection of indicators is essential for the success of a program of telesis (Creech, 1998). (3) Some 11% of enrollment, grades K – 12, are private school students. Indicators in Table I, sections 1, 2, 6, and 8, appear to include both public and private school students; the remainder reflect only public schools. (4) Reports on progress toward attaining goals do not include information on the strategies employed. While some State reports give details of a State’s efforts (Cooney, 1999, September 1999), it is not possible to assess overall the dollar increase in funding nor other interventions toward achieving the goals. The Wartz report (1999) is the best source of successful interventions nationally. (As this is written the U. S. Congress is deliberating on providing additional resources to public schools, and also, perhaps, to private schools.) (5) Tests of the National Assessment of Educational Progress are used by many States to provide some uniformity in assessing student learning, but some States use other instruments (National Education Goals Panel, 1999; Bottoms, no date). (6) Successful strategies employed by a school system are publicized through informing State education departments of the strategy. Other means, such as the Wartz report, are used. However, there is no standard program to improve performance common to the entire system. For example, one intervention is teacher training (Cooney, 1998; Cornett, 1999); the indicator being the percent of teachers qualified in the subjects taught. Owing in part to the decentralized organization of the educational system, there is no uniform system of interventions. This decentralization gives local school boards control over the curriculum, subject to influence by the State department of education. While this arrangement contributes to local involvement in school affairs, it depletes the power of the federal establishment to influence school policies, except through the federal distribution of funds with “strings attached.” Some school districts resent this intrusion. Many European school systems follow a more uniform scheme that imposes national standards. In summary, the National Education Goals program has lacked knowledge of appropriate interventions that would rapidly bring about change in the schooling of children and youth (the knowledge of cause-effect sequence of Map 2). Programs have not had the
246
ABBOTT L. FERRISS
TELESIS
247
248
ABBOTT L. FERRISS
TELESIS
249
250
ABBOTT L. FERRISS
uniform approach of the Healthy People program, partly because of the decentralized character of the educational system, control of the public system resting with the 50 States and the District of Columbia school superintendents and the individual school districts. Private schools are less subject to intervention from without than are public schools. In addition, resources to effect change were not forthcoming in sufficient strength during the period for which data are now available. However, it is an on-going program. The Congress is considering infusing the program with additional resources, and future progress will be monitored by appropriate indicators. KIDS COUNT
Kids Count is a private program initiated with foundation support, not government moneys. The Annie E. Casey Foundation in 1990 initiated a national program to improve the well-being of children and youth in the United States. The Casey family created the Foundation with funds primarily from the United Parcel Service. It is thus a private initiative to employ social indicators to identify and monitor trends in measures of child well-being, to stimulate public
TELESIS
251
interest, to set goals for change nationally. By funding state-bystate initiatives, the Foundation extends the program to the local county and community levels. The aim is to excite policy makers, legislators and local leaders to provide services that improve child well-being, and monitor them with social indicators. The indicators currently exhibited are shown in Figure 10. The chart shows decline in three indicators but improvement in seven over the 1985–1996 period (the latest for uniform data (Annie E. Casey Foundation, 1999). Trends in six of the indicators show remarkable progress. The infant death rate, child death rate, deaths rates from accidents, homicides and suicides, and the teen birth rate are objectives also of the Healthy People program. Indicators of these show satisfactory progress. In addition, the percent of teens not attending school and not working and the percent of children living with parents not fully employed have improved. These trends are exhibited in charts, only two being presented here for lack of space, Figures 7 and 8. Other influences on child well-being have not been as successful. The percent low birth-weight babies continues to climb. The percent of families with children headed by a single parent has increased. The percent of children in poverty has declined, then increased and now has declined again but progress is erratic. Similarly, the percent
252
ABBOTT L. FERRISS
of teens that are high-school dropouts has not followed a steady downward course and is now at about the same level as 1986. Over the years the Kids Count program has developed criteria for selecting indicators for a telesis program (Annie E. Casey Foundation, 1999, p. 175) They are, as follows: 1. Data must be from a reliable source. 2. The statistical indicator must be available and consistent over time. 3. The statistical indicator must be available and consistent over time across all states.
TELESIS
253
4. The statistical indicator should reflect a salient outcome or measure of well-being. 5. The statistical indicator must be easily understandable to the public. 6. The statistical indicator must have a relatively unambiguous interpretation. 7. There should be a high probability that the measure will continue to be produced in the near future. These selection criteria restrict the items admissible in the national comparison. Other indicators are included by states to widen the scope. These include: demographic change by ethnic group, relevant social and economic characteristics of children, child health and family risk factors. The latter are especially revealing for they enable the classification of families according to the number of risk factors. Risk factors are: child is not living with two parents; household head is high school dropout; family income is below the poverty line; child is living with parent(s) who do not have steady, full-time employment; family is receiving welfare benefits; child does not have health insurance. Any one of these risk factors may not in itself be detrimental, but, so the theory goes, the family facing four or more of them are “at risk.” One of the virtues of the Kids Count program lies in the decentralization of the program to the state and county levels. The Annie E. Casey Foundation has provided funds to state organizations that are concerned with the quality of life of the child. These typically are private, not governmental, agencies. Each state develops its own data book showing the same indicators by county as shown in Figure 10. It distributes the indicators to legislators and other policy-makers, to county officials and public and private county and community groups. The data book presents data at the county level, where actions are initiated to improve the condition of the child. A number of the state programs focus upon legislative committees in an effort to increase funding for programs deemed beneficial to child well-being. The policies and programs of concern, of course, vary from state to state, depending upon the most pressing issues in child well-being of the state. Thus, decentralization by state has the advantage of flexibility depending upon the local needs. Alto-
254
ABBOTT L. FERRISS
gether, the Kids Count program is a model in the use of indicators to stimulate local action to solve local problems, following the telic model. LESSONS
We may draw several conclusions from our study of the three cases of teleological process, above. 1. The telic process, while originating at the broad, national level, must be enacted at the grass-roots, that is, from national to regional, state, county, community, neighborhood, family (school. class room, teacher-pupil, etc.). Any use of mass media, printed matter, or e-mail distribution must be supported by organization at the primary level of social interaction where intervention must take place. 2. Goals should be specific in terms of a statistical indicator. They should be set at attainable levels within the time frame. Long range goals should be distinguished from more limited goals. By achieving more limited goals, incremental progress will provide stimulus for achieving the next level of the goal. For motivational reasons, goals should be set at reasonable, attainable levels, for upon attaining a limited goal, the social group will be enlivened to set a higher goal in the next phase of the implementation cycle (Map 3). 3. Subgoals should be set that identify the population subgroup most deficient in the indicator. Such deficient subgroups should be the target of special programs. (See Figure 1 and related discussion.) 4. Selection of indictors should follow the rules of the Kids Count program: Indicators should be reliable, consistent over time, available at local levels, should represent a salient outcome, should be understood by the common man, of unambiguous interpretation, and be routinely produced by the statistical system into the future. 5. The planners of the program should take great care to establish from experimentation or from practical experience the strategy of intervention. Intervention at several levels – interpersonal, legal, organizational, etc. – may be appropriate, but each step
TELESIS
255
should be securely based upon a demonstrated cause-effect sequence. The success of the Healthy People program attests to the importance of this admonition. The faltering start of the Goals for Education program attests to the need for sound pedagogical cause-effect understanding to bring about learning in schools. The continuing success of the Kids Count program apparently rests upon the stimulation and freedom of local initiatives. 6. Through analysis, the subgroups of the population that adversely affect the indicator should be identified. These subgroups, then, become the targets of special programs for change, as specified in 3, above. As an aside, values are not uniform universally; and some deficiency in progress of a telic intervention may be attributed to the weakness of the value in a subgroup of the target population. As an example a program to reduce alcohol consumption among the Irish might face obstacles of norms that approve alcohol consumption as a male prerogative. The Program Implementation Cycle, Map 3, for subgroups may require special programs. 7. A program of telesis requires adequate human and financial resources. As was discussed in connection above with the Healthy People program, the Surgeon General’s initiative in bringing together technical experts to advise the targets for change and the aspect of the problem requiring change, did much to insure the success of the program. For example, low birth-weight newborn was identified as a factor in infant mortality. Steps then were initiated to improve pre-natal care of pregnant women. As another example, the Goals for Education program has faltered because of the lack of certainty in knowledge of how to improve student learning, even though much research in the learning process was available. 8. Trends in some social phenomena may be stronger than other trends. This may require extensive analysis of the socio-cultural system to devise approaches for change. For example, in the Kids Count program, over the period 1985 to 1996, the percent of families with children headed by a single parent was observed (Chart 8) to have become 20 percent worse. The trend has been affected by both the rise in out of wedlock births and the
256
ABBOTT L. FERRISS
prevailing high divorce rate, both nearly intractable changes in the family institution. Both these problems would require attention in order to reduce the percent of single parent families. Another example, is the lack of improvement in the Education Goals number 2.1 “Increase 18 to 24 year olds with high school credentials”. The goal was set at a modest 90%, but in stood at 86% both in 1990 and 1997, showing no progress. Research has shown that high school dropouts are greater during periods of economic prosperity, as typified the period following the brief dip in the economy in the early 1990s. Against such an incentive to drop out of school as employment, the percent of 18–24 year old without high school credentials persisted at 86%. Thus, general influences of the socio-cultural environment may create difficult hurdles for a telesis process, hurdles that may require broad changes in the system. 9. The programs here reviewed employed specific indicators rather than composite indicators. Only the Kids Count program employs a composite indicator summarizing the effects of the ten specific indicators (Figure 10) for interstate comparison. The same is true for state-level Kids Count indicators. In the case of the goals of the Healthy People program, a single indicator, length of life at birth, provides an overall indicator of progress
TELESIS
257
of many individual efforts, but it is not a composite indicator in the sense that several individual indicators have been statistically combined. The “span of healthy life” is a combination of two rates (length of life and years of disability), but such a combination is not a combined index in the sense employed by quality of life studies. Thus, for purpose of goal setting and monitoring in the telic process, individual indicators rather than composite indicators are more useful. 10. The above suggests that composite indices are not useful in the telic process. However, in the case of interstate comparisons in showing state-to-state variability in the well-being of children, the Kids Count combined index enables the ranking of states
258
ABBOTT L. FERRISS
and the demonstration of change from one period to another. In this sense, the composite index is useful. Undoubtedly a composite index of the condition of health of the people of a state would be possible. It would bring together indicators of the prevalence of acute and chronic conditions and include subjective judgments of the quality of life. Subjective data from the CDC Behavioral Risk Factors Surveillance System has been suggested for a combined index, by a paper, entitled, “Health Related Quality of Life” of the Hagerty report (forthcoming). Composite evaluations of schools, in terms of meeting specified standards, have been devised, including performance of students on standardized tests and other aspects of school organization. Consequently, while individual indicators of individual characteristics are more useful in the telic process, there is a place for combined indicators in long-range planning and evaluation. Finally, in examples here reviewed objective measures were employed: in Healthy People, in Goals for Education, and Kids Count. While no subjective measures are evident, some determinations, for example item 7.4 of Table I – student reports of being threatened – may have involved subjective judgment on the part of the student. But, as a whole objective measures prevail.
TELESIS
259
260
ABBOTT L. FERRISS
TELESIS
261
262
ABBOTT L. FERRISS
TELESIS
263
REFERENCES
American Institute for Cancer Research: 2000, Diet and Health Recommendations for Cancer Prevention (American Institute for Cancer Research, Washington, DC) (The pamphlet presents results from the Cancer Research – World Cancer Research Fund report, Food Nutrition and the Prevention of Cancer: a Global Perspective, by an expert panel, chaired by John D. Potter.) Annie E. Casey Foundation: 1999, Kids Count Data Book (The Annie E. Casey Foundation, Baltimore, MD). Biderman, Albert D.: 1966, ‘Social indicators and goals’, in Raymond A. Bauer (ed.), Social Indicators (The M. I. T. Press, Cambridge, MA and London). Bottoms, Gene: no date, State Leadership in Improving High Schools for More Students (Southern Regional Education Board, Atlanta, GA). Cassidy, John: 2000, ‘The price prophet’, The New Yorker, Feb. 7, 2000, pp. 44– 51. CDCP: March 24, 1989, The surgeon general’s 1989 report on reducing the health consequences of smoking: 25 years of progress (executive summary)’, MMWR: Morbidity and Mortality Weekly Report, 38/S-2 (Centers for Disease Control and Prevention, Atlanta, GA). CDCP: Feb. 26, 1999, ‘Decline in cigarette consumption following implementation of a comprehensive tobacco prevention and education program – Oregon, 1996–1998’, MMWR: Morbidity and Mortality Weekly Report, 48/7, (CDC, Atlanta, pp. 140–143).
264
ABBOTT L. FERRISS
CDCP: June 8, 1999, ‘Preemptive state tobacco-control laws – United States, 1982–1998’, MMWR: Morbidity and Mortality Weekly Report, 47/51 & 52 (Centers for Disease Control and Prevention, Atlanta, GA). Commager, Henry Steele: 1967, Lester Ward and the Welfare State (The BobbsMerrill Co., Inc., Indianapolis and New York). Cooney, Sandra: 1998, Improving Teaching in the middle Grades: Higher Standards for Students Aren’t Enough (Southern Regional Education Board, Atlanta, GA). Cooney, Sandra: 1999, Leading the Way: State Actions to Improve Student Achievement in the Middle Grades (Southern Regional Education Board, Atlanta, GA). Cornett, Lynn M.: 1999, Getting Beyond Talk: State Leadership Needed to Improve Teacher Quality (Southern Regional Education Board, Atlanta, GA). Creech, Joseph D.: 1998, Educational Benchmarks, 1998 (Southern Regional Education Board, Atlanta, GA). Creech, Joseph D.: September 1999, A Challenge for SREB States: Increasing the Percentage of Adults with a High School Diploma (Southern Regional Education Board, Atlanta, GA). Creech, Joseph D.: 2000, Linking Higher Education Performance Indicators to Goals (Southern Regional Education Board, Atlanta, GA). Ferriss, Abbott L.: 2000, ‘A brief history of social indicators developments’, SINET 61 and 62, pp. 8–9, 14. Greenwald, P. and E.J. Sondlk (eds.): 1986, Cancer Control Objectives for the Nation: 1985–2000 (National Cancer Institute, NCI Monographs, Bethesda, MD). Gross, Bertram M. (ed.): 1967, ‘Sopcial goals and indicators for American Society’, vol. I, The Annals, vol. 371, May; and vol. II, September. Hagerty, Michael R. and others: forthcoming, ‘Quality of life indexes for a national policy and review and agenda for research’, in Social Indicators Research. Inglehart, Ronald: 1990, Culture Shift in Advanced Industrial Society (Princeton University Press, Princeton, NJ). Lecht, Leonard A.: 1966, Goals , Priorities, and Dollars: The Next Decade (The Free Press, New York). National Center for Health Statistics: 1999, Health, United States, 1999 (U.S. Government Printing Office, Washington, DC). National Education Goals Panel: 1997, National Education Goals Report: Building a Nation of Learners 1997 (U.S. Government Printing Office, Washington, DC). National Education Goals Panel: 1998, National Education Goals Report: Building a Nation of Learners 1998 (U.S. Government Printing Office, Washington, DC). National Education Goals Panel: 1999, National Education Goals Report: Building a Nation of Learners, 1999 (U.S. Government Printing Office, Washington, DC).
TELESIS
265
President’s Commission on National Goals: 1960, Goals for Americans (Prentice Hall, Englewood Cliffs, NJ). Sheldon, Dana M. et. al.: Nov. 3, 1995, ‘State laws on tobacco control – United States, 1995’, MMWR: Mobility and Mortality Weekly Report, 44/SS-6 (CDCP, PHS, USDHHS, Atlanta, GA). Ward, Lester F.: 1903, Pure Sociology (Macmillan, New York). Ward, Lester F.: 1906, Applied Sociology (Ginn & Co., New York). Williams, Robin M., Jr.: 1967, ‘Individual and group values’, The Annals Social Goals and Indicators for American Society, vol. 371, May. Woolsey, Theodore D.: 1987, quoted in ‘Social Indicators: An Absolute Requirement of a Working Democracy . . .’, SINET 9 (February 1987), p. 5. Wurtz, Emily: 1999, Promising Practices: Progress Toward the Goals 1999 (National Education Goals Panel, Washington, DC). U.S. Department of Health and Human Services: 1990, Healthy People: National Health Promotion and Disease Prevention Objectives, Public Health Service (U.S. Government Printing Office, Washington, DC). U.S. Department of Health and Human Services: 1992, Health United States and Healthy People 2000 (National Center for Health Statistics, Hyattsville, MD). U.S. Centers for Disease Control and Prevention, National Center for Health Statistics: 1998, Healthy People 2000, Statistical Notes, ‘Operational definitions for year 2000 objectives: priority area 1, physical activity and fitness’, No. 18, December, U.S. Department of Health and Human Services (National Center for Health Statistics, Hyattsville, MD). U.S. Centers for Disease Control and Prevention, National Center for Health Statistics: 1998, Mortality and Mortality Weekly Report 47(43), November 6.
Department of Sociology Emory University Atlanta GA 30322 USA E-mail:aferris @emory.edu
This page intentionally left blank
VALERIE MØLLER and HELGA DICKOW
THE ROLE OF QUALITY OF LIFE SURVEYS IN MANAGING CHANGE IN DEMOCRATIC TRANSITIONS: THE SOUTH AFRICAN CASE* ABSTRACT. The South African Quality of Life Project has tracked subjective well-being, life satisfaction and happiness, since the early eighties at the national level. In most democratic countries around the globe, the average citizen says he or she is satisfied with life in general. In South Africa this is not the case. Since the early 1980s, the trend study shows up disparities between one sector of the South African population that is satisfied with life in general and various aspects of life and another sector that is very dissatisfied. Generally, the better-off report higher levels of satisfaction and happiness than the worse-off. The most plausible explanation for the South African quality-of-life constellation is the huge gap in living standards between rich and poor, a legacy of the apartheid era, which discriminated against blacks and to a lesser extent against Indian and coloured people. Euphoria following on the first democratic elections in April 1994, which registered equally high aggregate levels of happiness and life satisfaction among all sectors of the population, was short-lived. Under democracy, expectations “for a better life for all”, the election slogans for the 1994 and 1999 general elections, has risen. South Africa has one of the most enlightened constitutions, which guarantees basic human rights and supports advancement of the previously disadvantaged. As long as South Africans perceive barriers to accessing the material rewards of democracy, they do not see justice has been done. South Africa is currently grappling with problems common to other societies in transition to democracy. Since 1994, government programmes and policies have been devised to address the critical twin problems of poverty and inequality in society. The latest round of research for the South African Quality of Life Trends Project probes popular assessments of the policies and programmes aimed at improving the quality of life of ordinary South Africans. Interviews with a panel of 25 opinion leaders in the run-up to the June 1999 general elections were followed by a nationally representative opinion survey in October 1999. The paper outlines the role of social indicators in monitoring quality of life in South Africa and reports findings from the elite and rank-and-file surveys. Generally, the winners and losers in the new political dispensation see changes from a different perspective. The disadvantaged are more likely to have seen material gains and recommend increased delivery of services and opportunities for social mobility. The advantaged, who have mainly experienced non-material or no gains since 1994, are more likely to be pessimistic about the future. It is concluded that the groundswell of optimism will sustain the majority of South Africans who are still dissatisfied with life until their dreams of the good life are fulfilled. * Revised version of a paper presented at the Third Conference of the International Society for Quality of Life Studies, 20–22 July 2000, Girona, Spain. Social Indicators Research 58: 267–292, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
268
VALERIE MØLLER AND HELGA DICKOW
INTRODUCTION
Developing Social Indicators for Democratic South Africa
Social indicators serve many different purposes. They provide descriptive information on society, are used for setting goals for society, for monitoring or tracking outcomes to inform policy, to hold agencies accountable for improving outcomes, and to evaluate whether programmes are efficient and effective (Brown and Corbett, 1998 cited in Sinet 61/62, 2000: p. 6). Descriptive statistics were a contentious issue in South Africa under apartheid. National accounts excluded the bantustans, the so-called independent and self-governing homelands of the apartheid era, and thus distorted the overall picture. Alternative accounting systems to make more accurate population estimates were developed in myriad fashion by independent agents, such as the development bank, marketing and media, and academic institutions. There was great demand for the poverty datum lines developed by academics that were updated on a regular basis. Drawing on all these sources of social indicators, the South African Institute of Race Relations published a regular handbook to inform the general public on political and economic developments and living conditions throughout the country, the equivalent of a social report for general public consumption (see Schlemmer and Møller, 1997). Efforts to centralise South Africa’s social accounts began on the eve of the first non-racial general elections held in April 1994. The so-called poverty study, officially called the ‘Living Standards and Development Survey’, funded by the World Bank but co-ordinated and published by the University of Cape Town (Saldru, 1994), brought together a wide range of interested parties in academia, government, politics and non-governmental organisations throughout the country to resolve the problem of fragmented statistics. The survey instrument went through over 13 drafts in consultation with bodies countrywide until South Africans were prepared to call the study their own. The results from the survey of 9000 households provided the first baseline statistics on living conditions throughout South Africa for the first democratically elected government-in-waiting (RDP, 1995). The University of Cape Town led poverty study provided the model for central data collec-
THE ROLE OF QUALITY OF LIFE
269
tion combined with wide consultation of users in the public and private sector regarding methods and contents. Thereafter the Central Statistical Service, later renamed Statistics South Africa, took over the task of updating statistics on living conditions through the annual Household Survey and the 1996 census. The latter included many questions on access to infrastructure and household services. The Household Survey was to become the major source of ‘performance indicators’ for government departments to monitor service delivery in terms of the Reconstruction and Development Programme (Møller, 1995). A second study of poverty commissioned by the new government, again funded by the World Bank but executed by South Africans, followed on the first 1993 poverty study. This study, designed to complement the earlier study, was qualitative in nature and used case studies and participatory methods to elicit popular experience and views on poverty. A synthesis report on poverty and inequality drawing on the two studies was produced in 1998 (May, 1998a,b). Since 1994 Statistics South Africa1 has gained credibility for producing sound statistics. A change of headship and transformation of Statistics South Africa signified the move towards production of unified statistics for the new democracy. Credibility was initally established through consultation with stakeholders when developing new statistics and through the involvement of reference groups of users from all sectors of society and all nine provinces. The taking of the first post-apartheid census, which drew over 100 000 enumerators from the ranks of the unemployed, was introduced as a nationbuilding exercise with the slogan: ‘count us in’. The preliminary report on the census recalled that “In October 1996 all the people of South Africa were counted as members of a unified democratic nation for the first time” (CSS, 1997: p. 7). The advisory committee for the 1996 census included academics, representatives of civil society, other government departments, parastatal organisations, non-governmental organisations and the business community. Statistics South Africa has transformed its operations to develop capacity within its own ranks and among users of statistics. It has received generous expert and financial assistance from statistics offices in the Scandinavian countries, Canada, Australia and New
270
VALERIE MØLLER AND HELGA DICKOW
Zealand, among others, to fine-tune its methods and add to its repertoire. For example, Statistics South Africa has conducted its first national time-use study aimed at providing more information on women’s work and informal sector activities. However, constant upgrading of methods, staff turnover, and changing reference groups among users combined with financial constraints has also created problems of continuity. Thus some statistics are not available in time series or are not strictly comparable. The sophisticated and expensive software used to produce instant access to South African census data comes at a cost. Users in academia and government departments, many of whom have been involved in data collection and related tasks, are not prepared to pay for what they consider should be a free public service. It is probable that many local authorities have never made use of the census data available for their jurisdiction. Currently, too few South African users make full use of the statistical data available for further scientific and policy analysis. It was estimated that perhaps less than one dozen academics had conducted further analysis on the data sets from the quantitative poverty study of 1993. Similarly, datasets produced by Statistics South Africa since 1994 tend to be under-utilised. As mentioned users are reluctant to pay for the privilege of analysing data, which was formerly issued free or at cost. Another obstacle is the lack of expertise among potential users in handling large datasets, a problem which the National Research Foundation is seeking to address. Statistics South Africa has adopted a new image based on the 4 R’s of relevance (to policy making for growth and development), reliability (in producing timely and accurate statistics), responsiveness (to the statistical needs of users), and representativeness (in staffing and advisory committees). It has sought to package social indicators in ‘user-friendly’ descriptive reports based on the Household Survey to inform the general public. The descriptive report on ‘Living in South Africa’ (CSS, 1996) was followed by a similar report on Gauteng Province (Orkin, 1997), which is the country’s centre of commerce and industry, but the series lapsed thereafter. Since 1994, public debate on statistics, such as the unemployment rate and the population count, has become the order of the day. Public reliance on statistics to monitor performance of govern-
THE ROLE OF QUALITY OF LIFE
271
ment in achieving the goals of the new democracy has increased. An indication of the value attached to statistics in the democratic era is the public outcry following a government announcement of an indefinite moratorium on crime statistics while data collection methods were improved (International Herald Tribune, August 4, 2000). Although successive governments have projected a good quality of life for the population as an overall goal, there have been few systematic attempts to monitor quality of life and these have tended to rely on objective indicators. The new government has set great store by ‘key performance indicators’ of the objective variety as tools for goal-setting and monitoring activities of government departments. The main goals are given by the election promises. The typical performance indicators are ones measuring delivery of housing and infrastructure in terms of these promises. Government departments involved in public works programmes and land redistribution have commissioned evaluation research under the ‘quality of life’ heading to report on the achievements of their human resource development activities in terms of efficiency and effectiveness (see May, Stevens and Stolls, this issue). A more controversial application of social indicators arises from pressure put on private sector employers to keep staff records with prescribed racial tags. Racial quotas are used to indicate social transformation and the adoption of affirmative action2 and equal opportunity policies in the workplace. Since 1994 commercial and not-for-profit organisations which attract overseas donor funding have been in the forefront of opinion polling and the collection of subjective indicators to document attitudes and opinions during South Africa’s transition to democracy. Examples include the Opinion ’99 project by the Institute of Democratic Alternatives in Cape Town which tracked voter choice and knowledge in the run-up to the second national elections in mid1999 (Mattes et al., 1999). The Reality Check survey (MacGregor, 1999) was initiated by a consortium of media agencies to monitor the mood and pace of attitude change in the new democracy. For the most part, South Africa follows the tradition of data collection by means of cross-sectional studies. High geographic mobility makes panel studies impractical, although GIS technology
272
VALERIE MØLLER AND HELGA DICKOW
may assist in overcoming this problem to a certain degree. A notable exception is a university-driven follow-up study of the 1993 poverty survey in KwaZulu-Natal, which contacted the original households some years later (May et al., 2000). Local planners and policy makers generally set great store by hard indicators; subjective indicators are perceived to be unreliable. With few exceptions Statistics South Africa only collects hard facts. Statistics South Africa’s Household Survey and its forerunner, the 1993 poverty study, monitor improvements in living conditions and access to services and employment in line with the government’s election promises of 1994 and 1999 using only objective indicators. A notable exception are the two or three items on perceptions of improvements and deterioration of community safety and household living conditions, which tellingly fall under the heading of ‘quality of life’ indicators. Metropolitan Durban’s local government has taken the lead in monitoring quality of city life from the perspective of residents and users of public services with a mix of soft and hard indicators. Feedback to city residents is by means of a delightful, richly illustrated, jargon-free ‘easy reader’ booklet, available free or at cost to local organisations (Nicholson, 2000). THE SOUTH AFRICAN QUALITY OF LIFE TRENDS STUDY
The Quality of Life Trends project which commenced in the early eighties can be seen as an effort to complement South Africa’s hard indicator tradition with systematic studies of trends in life satisfaction, happiness and, more recently, expectations of future life satisfaction or optimism. The study was launched by a consortium of Human Sciences Research Council (HSRC), a state-sponsored research organisation, and university scholars with HSRC funding (Møller and Schlemmer, 1983). The project continues to operate under the auspices of the Institute of Social and Economic Research at Rhodes University (see www.rhodes.ac.za/iser). The study uses cross-sectional national surveys to generate its database. Since 1994 attempts have been made to add depth to the study and thus gain a better understanding of the dynamics underlying perceptions of changing quality of life. This has been achieved by adding items on topical issues that may influence overall life satisfaction and domain
THE ROLE OF QUALITY OF LIFE
273
satisfactions during the transition to democracy (see Møller, 1998, 1999; Møller et al., 1999). This paper reports on the descriptive results of the September 1999 update for the Quality of Life Trends project. The 1999 survey inquired into life satisfaction, happiness and projected future life satisfaction in five year’s time. These items give the most parsimonious yet balanced picture of quality of life and cover both the ‘personal’ and ‘social’ dimension. As Eckersley (2001) points out, how people feel about their own lives (subjective personal quality of life) may be different from how they feel about the society they live in (subjective social/societal quality of life). It is therefore important to capture both. In addition, respondents were asked to review the past five years in terms of positive and negative outcomes and to make policy recommendations to ensure a better quality of life for South Africans in future. The latter item meets the perceived need for public participation in social indicators research in South Africa. Method The October 1999 update for the South African Quality of Life Trends Project was commissioned to MarkData, Johannesburg. A nation-wide panel of 25 opinion leaders was interviewed personally in February and June 1999 to inform the study. The panel included persons from all race groups and representatives from across the political and institutional spectrum including education, health, the media, the church, national/provincial and local government, politics, labour, commercial agriculture, mining, business, small business development and training, safety and security, research, and non-governmental organisations. Some panellists drew on their experience of serving in different institutions. The interviews with elites were unstructured. Panellists were initially asked to describe the current situation in the country. The discussion then moved on to achievements and shortcomings of the new democracy, personal experience of change over the past five years and the manner in which government policies and programmes had impacted on the lives of ordinary South Africans. The results from the round of interviews with elites were used to shape survey items which were placed with a national syndicated study. Questions put to sample survey respondents included 11 items on subjective well-being,
274
VALERIE MØLLER AND HELGA DICKOW
national pride and related issues, and perceptions of progress in uplifting quality of life. This paper reports on select findings on subjective well-being, personal gratifications and disappointments, and views on important policies to uplift quality of life in the general population. MarkData’s October 1999 syndicated survey used a multiple stage cluster probability sample design to select 2219 South African residents of 16 years and older in metropolitan, urban and rural areas countrywide. The fieldwork was carried out between 23 September and 8 October 1999. Personal interviews were conducted by trained interviewers in the respondents’ homes in their language of choice. The items on subjective well-being, satisfaction with lifeas-a-whole and happiness, were identical to ones used in earlier studies. Responses were recorded on five-point scales with response categories ranging from ‘very satisfied/happy’ to ‘very dissatisfied/happy’ with a neutral mid-point. The optimism item read: “How do you think things will be for people like yourself in five years’ time? Taking all things together will things be better, worse or about the same as today?” Responses were recorded on a five-point scale, similar to that used for life satisfaction and happiness, which ranged from ‘much better’ to ‘much worse’ over a neutral mid-point ‘much the same’. All other items were closed. Response categories were read out to respondents who chose one option. Results presented are rounded percentages weighted to population size. In 1999, the official population count was 43.1 million; the racial breakdown was some 77% black, 9% coloured (mixed race), 3% Indian/Asian, and 11% white. Columns do not always add to 100% due to rounding. RESULTS
In the first section we report on select results of the preliminary discussion with elites. Results are grouped under the headings: Current situation in the country, perceptions of personal change and disappointments, policies for improved quality of life and challenges for the future. In the second section we shall report results on broadly corresponding items in the sample survey.
THE ROLE OF QUALITY OF LIFE
275
I. Results from the Elite Survey3
Elites on the current situation Five years after the first non-racial general elections many of the elites still marvelled at the South African ‘miracle’. The transformation had gone far more smoothly than anticipated. South Africa had a new constitution and a “liberal democratic framework” which created opportunities for creative solutions. The government had respected constitutional court decisions. The changes were dramatic; from an international perspective South Africa had become a different country. The power-sharing arrangements and the sunset clauses had provided a bridgehead to democracy. The work of the Truth and Reconciliation Commission (1998) had assisted South Africans to confront the past and build trust for the future. A number of panellists observed a major transformation of attitudes, which in their view signalled a process of “normalisation”. Slowly South Africans were looking at how issues affected the country as a whole “rather than whites and blacks”. President Mandela had succeeded in forging a broad national identity. South African society had opened up and many black people had moved into positions of power and responsibility. There was visible “deracialisation” of the elite. Many panellists saw the emergence of a black ‘middle class’. The very fact that Mandela had taken a principled decision not to extend his presidency beyond the first term set a useful precedence for the rest of Africa. Although the new African National Congress (ANC) government was inexperienced, it had learnt market-related lessons fast and proved itself capable of governing. The new government was praised for its good housekeeping. The economy had been managed responsibly if not imaginatively. On the eve of the elections the ANC government in waiting had announced the highly idealistic and ambitious Reconstruction and Development Programme (RDP) (ANC, 1994). Two years later it introduced Gear (Growth, Employment and Reconstruction) as its new macro-economic policy. A business leader aptly described Gear as South Africa’s “selfimposed structural adjustment programme”. Although there was less foreign investment than expected, a macro-economic policy was in place which made South Africa an acceptable partner to the western world. However, most prescriptions of Gear remained
276
VALERIE MØLLER AND HELGA DICKOW
incomplete in terms of implementation. Gear’s growth targets of 4–6% had never been reached and jobs had been shed rather than created since 1994. While some elites cited economic stability and modest economic growth as a major achievement for South Africa’s new democracy and praised the government for its restraint in not succumbing to populist demands, others chided the new government for not addressing the urgent needs of the poor. On balance, however, elites agreed that South Africa had achieved a lot in its first five years of democracy. In the words of one panellist: “We’ve come a long way. We have a framework established. We have the structures and plans. The new government has to focus on delivery now. Now is the time to build.” Elites on perceptions of change at the personal level Without prompting, a number of panellists spoke of different perceptions among rank-and-file and better-off South Africans and the different racial groups, whites and blacks in particular. Depending if you had been advantaged or disadvantaged under the former political dispensation your viewpoint would be worlds apart. For example, service delivery had made a difference to some sectors of society that had had no services previously. When asked how South Africa’s transition had affected them personally, elites responded that they had experienced few dramatic changes. Gains had been mainly immaterial ones. Generally, panellists noted a more relaxed atmosphere. There was now more “openness”, “transparency”, and a “tendency towards friendliness”. Both black and white panellists remarked that they felt a newfound freedom of movement. Elites on policies to promote better quality of life Elites were asked which government policies and programmes had had the most impact on the quality of life of ordinary South Africans since 1994. Three major themes emerged: the rapid succession of policies and programmes, large differences in service delivery, lack of conceptual clarity and vision, and problems of sustainability of services. It was generally agreed that government had been too eager to follow through the RDP manifesto. The election promises of a million houses, and jobs and services for all, had been overambi-
THE ROLE OF QUALITY OF LIFE
277
tious. Programmes had been launched with too little thought given to implementation. The new government lacked the ability to prioritise. The Masakhane campaign was singled out as the perfect example of a failed campaign launched too quickly. The intention was to empower people at local government to take responsibility for community services. Practically, this involved reversing the practice of non-payment for services, which had been adopted as a very successful strategy to make the black townships ungovernable under apartheid. Panellists noted that old habits die hard. Report cards. Many respondents gave ratings on service delivery and to the various ministers in charge. The approximate rank order started with water, then electricity, housing, health, transport and roads, and lastly education. Water delivery was singled out as a success story. While some services had improved, others had been downgraded in an attempt to bring services to everyone. This was most evident in the area of health where the emphasis on primary health care had cut the budgets of major hospitals including teaching hospitals. Education had suffered from overzealous transformation, which had allowed experienced teachers to take retirement packages. The unmotivated and inexperienced teachers left behind were not coping with the curriculum reform. Panellists across the political spectrum raised the issue of sustainability. Not enough attention had been given to outcomes when embarking on ambitious development plans: “They are putting in four phones and taking out three.” Indicator speak. It was noted that social indicators were widely used to monitor service delivery in terms of the RDP manifesto. Ministers of departments spiked their speeches with statistics on numbers of houses and clinics built, and numbers of water and electricity connections. One panellist declared that he had tired of performance indicators and ‘achievement talk’. Affirmative action and equal opportunities. According to mainly panellists on the political left, affirmative action and equal opportunities policies had had the more significant impact than the delivery of infrastructure and housing. Their argument was that households would only be able to sustain a higher level of living if their incomes improved. Labour market reforms, which opened up the way for
278
VALERIE MØLLER AND HELGA DICKOW
redistribution of income, were very important. A spokesperson for white miners had a different viewpoint: affirmative action policy was a major problem for members of his constituency. It had sowed the seeds of pessimism and despondency: “People feel side-lined, lost out, alienated.” A political scientist disagreed that affirmative action had acted as a catalyst for speeding up the movement of blacks into managerial positions which was happening in any case. In contrast to an earlier survey of elites (Kotzé, 1993), the members of the 1999 panel generally accepted the need for affirmative action as self-evident. The discussion now focussed more on the appropriate manner in which to implement affirmative action. Elites on future challenges for South African society In the course of the interview many of the elites enumerated the major problems which faced South Africa’s fledgling democracy and the challenges for the president-elect, Thabo Mbeki. Crime and corruption featured on most lists of negative facets of post-apartheid society. The background causes were seen to be quite similar. Other problems included many of the issues named as disappointments, namely economic issues such as growth, job creation, and skills development and education to enable South Africans to compete in the global market. Corruption had always been a part of South African society, according to some panellists, but the transparency of the new system had made corruption more visible. Transformation of South African society had contributed to corruption. What amazed former antiapartheid activists among panellists was how quickly many of their former comrades had made the switch from upholding idealism to becoming slaves to materialism. Elites on mindsets, morality and optimism Most panellists were over forty years of age and some very much older. A black panellist who had spent his youth in opposition to the then government spoke of “post-struggle exhaustion” among his contemporaries in the movement. Among whites, a former opposition politician noted that many of his outspoken colleagues had withdrawn from politics. The coloured community was thought to be in “a bit of a vacuum at the moment. It’s not quite sure where they fit in.” Democracy and its responsibilities were new for coloured
THE ROLE OF QUALITY OF LIFE
279
people. Among whites, Afrikaners were described as “deeply traumatised” and alienated. The majority felt disempowered and had not come to grips with their loss of influence. The historian/politician on the panel attributed increasing political intolerance and opportunism to the dashing of high expectations that things were going to change dramatically when the new government came into power. Many panellists referred to the legacy of the anti-apartheid struggle in shaping mindsets in contemporary society. The lack of respect for authority spawned in the antiapartheid movement continued in disrespect for the courts of law, jail breaking, and police killings. The lack of experience of new civil servants coupled with a poor work ethic (“people do very little but expect a lot”), and the ‘culture of entitlement’ was holding back productivity in South African society and the distribution of services to the poor. A black development worker despaired about the loss of values in the new South Africa: “I was hoping reconstruction would go as far as bringing back those values...” The archbishop on the panel wistfully questioned how you developed policy for a revitalised South African morality: “The greatest vision is hope ... that goodness can actually triumph. How do you encapsulate that in a policy?” On the other hand, panellists saw optimism for the future as an important factor determining the outcome of South Africa’s transition to democracy. In one panellist’s opinion, it was a “religious” optimism fired by the notion that things will continue to be better. Under apartheid, when there was no hope of getting a house, people felt despair. The success of the Ministry of Water Affairs was to make people to believe they were part of “a moving queue”. To be part of such a queue, no matter how slowly it moved, was a psychological advantage. II. Results from the Sample Survey
In this section we first give a graphic overview of trends in South African subjective well-being over time by way of background. We then present findings from the October 1999 national survey on life satisfaction, happiness, and expectations for the future. In correspondence to the broad themes explored in the elite study we review perceptions of positive changes and disappointments, the rationale
280
VALERIE MØLLER AND HELGA DICKOW
underlying these evaluations, and lastly, recommendations for future policy options. Results are broken down by race, levels of income, life satisfaction and optimism. Trends in South African well-being Figure 1 shows trends in happiness, satisfaction and expectations for the future over time. During the strife-torn eighties dissatisfaction rose in 1988. Subjective well-being peaked in the aftermath of the first non-racial elections of 1994 only to drop the following year. Levels are rising again slowly. Noteworthy is that optimism is rising faster than current satisfaction. Table I shows satisfaction and happiness for the current survey for the four race groups and two income groups. Satisfaction with life-as-a-whole and happiness conform to the pattern observed ever since the first measures were taken in the early eighties. On objective measures of living conditions and subjective measures of well-being and domain satisfaction whites score highest on positive indicators and blacks score lowest. Coloureds and Indians score somewhere in between. In contrast, current satisfaction and happiness are negatively correlated with optimism. Blacks are most optimistic for the
THE ROLE OF QUALITY OF LIFE
281
future and whites the most pessimistic. Better-off South Africans are more satisfied than poorer South Africans. Winners and losers What factors have increased the quality of life of ordinary citizens since 1994? As discussed earlier, the new government’s service delivery record is by no means negligible (Nedlac, 2000). Between 1994 and 1999 more than 3 million South Africans gained access to clean water, some 2 million people were provided with electricity, some 750 000 houses were built or under construction, and nearly 600 primary health clinics erected. Among the seven options presented to respondents in the October 1999 survey, four dominated in the responses (see Table II): ‘RDP’ factors, that is, infrastructure and service delivery in the areas of housing and electricity, greater freedom of movement, and a new mindset. Access to basic and improved services such as housing, electricity, health care, education, and water, taken together, accounted for some 58% of responses. The survey confirms that access to
282
VALERIE MØLLER AND HELGA DICKOW
THE ROLE OF QUALITY OF LIFE
283
electricity had a major impact on South Africans’ lives in the first five years of democracy. Surprisingly, access to water received few votes. The provision of clean water to rural households was regarded as one of the most successful and popular campaigns to meet 1994 election promises by the opinion leaders in the elite survey. As respondents in the October 1999 survey could only cast a single vote, it is possible that basic needs were crowded out by the non-material and less tangible gains of the past five years. For a substantial group of South Africans the greatest change in their lives since 1994 was greater freedom of movement and an improved self image and confidence when interacting with fellow citizens. No doubt increased freedom is also related to greater choice in all aspects of life. Only 5% endorsed better job prospects in the new South Africa, which might be taken as a reflection of the objective economic situation in the country. Just under one fifth stated they had not experienced a change for the better in their lives since 1994. Pessimists, who have a negative view of the future, were more likely than others not to have seen any positive changes in their lives since 1994. Men tended to stress the advantages of freedom of movement. Women and satisfied South Africans were more likely to report better access to health care. Perceptions of a gain in freedom of movement cut across the racial and income divides. Electricity made a difference mainly in the lives of poorer South Africans. White and Indian South Africans were most likely to state that the major change in their lives had been non-material, such as a change in their projection of self which had positively affected interpersonal relations in society. Although numbers are small, twice as many optimists than pessimists (6% versus 3%) had experienced better job prospects. Perhaps the most striking finding is that between eight and nine out of ten black and poor South Africans named a positive change in their lives over the past five years while two out of three white South Africans did not. Objective and subjective indicators of quality of life are still racially polarised five years into democracy. Race is still a defining factor influencing attitudes as revealed in the 1998 Reality Check survey (MacGregor, 1999: p. 26).
284
VALERIE MØLLER AND HELGA DICKOW
THE ROLE OF QUALITY OF LIFE
285
Disappointed South Africans What are South Africans’ greatest disappointments in the democratic era? Two issues stand out: Crime, corruption and nepotism, on the one hand, and unemployment, on the other (see Table III). Poorer South Africans are disappointed that their hopes for jobs and infrastructure have not materialised. Mainly richer South Africans felt disappointed that the new elite had been quick to trade their idealism for greed. Unemployment is the primary disappointment of black South Africans followed closely by crime. Among white, Indian and coloured South Africans crime disappointed more than job scarcity. Dissatisfied South Africans voiced their feelings of social exclusion and thwarted expectations as indicated in above-average endorsement of the responses of ‘unemployment’, ‘slow delivery of services and infrastructure’ and ‘lack of concern by the newly rich for the situation of the poor’. Satisfieds and optimists were among the few who volunteered that they had experienced no disappointments. The best way forward Some 56% of black and 52% of poorer South Africans were optimistic about their personal future. What are their recommendations for realising their own and South Africa’s potential? South Africans have divergent visions of the best route to improve the quality of life of South Africans in the longer term (see Table IV). Optimists set their hopes on better material standards of living through RDP-type measures coupled with an acceleration of black advancement and labour-friendly policies to achieve sustained quality of life. A substantial proportion of optimists also advised fiscal discipline as the best medicine to achieve the growth needed to provide a better standard of living for South Africans. A higher than average proportion of whites stressed the importance of sound economic policies to improve long-term quality of life while blacks backed improved service delivery. Obviously, deciding on policy options is a daunting task for many ordinary citizens. Between 10% and 15% were not capable of projecting policy outcomes for the future or unwilling to try.
286
VALERIE MØLLER AND HELGA DICKOW
THE ROLE OF QUALITY OF LIFE
287
DISCUSSION AND CONCLUSIONS
Consumerism and Quality of Life Gains
Water and electricity were the election promises that captured the imagination of the nation in 1994. However, their fulfilment appears not to have affected happiness gains among the poor in the first five years into democracy. In fact, dissatisfaction and unhappiness remain above-average among respondents who named access to water as the major change in their lives. Rapid adjustments to changing life styles along with rising expectations of what constitutes a decent life style may have intervened. Class differences appear to be the main explanation for perceptions of change. The South African living standards measure divides the population in eight groups according to urbanisation and consumption patterns using some twenty indicators. Bigger consumers are consistently more satisfied and happier than smaller consumers. Among ‘rural have-nots’ who have the lowest living standards, some 41% indicated life satisfaction and 45% happiness compared with 80% satisfied and 79% happy among the ‘progressive affluents’ with the highest living standards. As the elites in the panel predicted, perceptions of substantive gains are linked to occupational rank and living standards. The bottom occupational rank in society cites mainly access to water as the major change in its life since 1994. The ranks of the semiskilled and skilled and the intermediate living standards categories (4–6) mention predominantly health care, education, housing and jobs, in approximately that order. The most affluent consumers and the highest occupational ranks refer mainly to non-material gains or no positive changes in their lives. Worth noting is that housing is the most frequently mentioned change in the ‘emerging market’ consumer category rather than among poorer categories. While electricity appears to be the major gain since 1994 for rural dwellers, better access to housing is the major gain for urban blacks who may already have access to modern conveniences such as electricity, flush toilets, insurance, banking services and credit cards. There are signs that consumption has become an important reference standard against which poorer South Africans assess their
288
VALERIE MØLLER AND HELGA DICKOW
living circumstances. Among black South Africans, dissatisfaction and unhappiness are definitely associated with consumerism. While it is the most affluent among all South Africans who tended to see no changes in their lives since 1994, it is the socially excluded among blacks. The most popular consumer goods in South African society include a television set, a refrigerator, and a music centre along with electricity to operate these amenities. In the total sample, 79% have access to electricity, 68% own a television set, 60% a fridge and 58% a music centre. Black consumption lags somewhat behind; the corresponding percentages are access to electricity (73%), television (58%), refrigerator (47%), and a music centre (50%). It appears that perceptions of social exclusion hinge on access to popular consumer goods, which in turn influence overall evaluations of life. Mainly black South Africans who do not have access to electricity, television, fridge and a music centre see no positive change in their lives. Thus, the main contrast in quality of life gains appears to be between perceptions of improved material conditions and ego integrity. Poorer South Africans who gained their freedom in 1994 emphasise RDP factors and service delivery while richer South Africans, mainly whites, stress relief in shedding pariah status and being citizens of a country which is once again integrated into world society. Perceptions of no change in society since 1994 are concentrated among better-off South Africans and might be interpreted as a sign of disillusionment or disappointment with the new dispensation. More worrying is that a substantial proportion of poorer South Africans, some 13%, said they had seen no changes. Among blacks who said they had seen no change, dissatisfaction and unhappiness was significantly above average: 58% dissatisfied versus 50% average and 52% unhappy versus 38% average. Outlook
What are the prerequisites that all South Africans can experience happiness and life satisfaction in the new millennium? One factor, already observed by opinion leaders on the panel, is the creation of a stronger black middle class, neglected in the past. The 1999 survey category assigned 10% of blacks to the top two living standard
THE ROLE OF QUALITY OF LIFE
289
categories along with 47% coloured, 89% Indian, and 98% white. For the first time since the launch of the Quality of Life Trends Project, there were sufficient numbers of better-off blacks in the 1999 survey to make meaningful comparisons between black and white elites. The comparison is striking. Near identical percentages among black and white higher income earners, some 80%, were happy and satisfied with life. In addition, wealthy blacks have the extra advantage of optimism. According to opinion leaders, affirmative action measures will benefit only a small proportion of job seekers. So far, the changing face of wealth in South Africa has created only a small elite. According to a recent study, the proportion of black households in the richest 10% of South Africans households increased from 9% in 1991 to 22% in 1996. However, the gap between rich and poor widened, especially in the black community (Whiteford and van Seventer, 2000). Our 1999 study found that rank-and-file South Africans set their hopes on RDP-type interventions and affirmative action to uplift quality of life. Many of the elites saw RDP merely as a cushion but not a solution for poverty. Be that as it may, optimism is the dominant mood of South Africa’s transition period. Graph 1 shows that optimism is rising in the post-apartheid era. Other research confirms the mood. Commenting on findings from a survey conducted immediately after the 1999 elections, political analyst R W Johnson describes “the persistence of a deep well of optimism” among ANC voters, particularly black voters, as “perhaps the most fundamental feature of the current political situation”. Despite any material disappointments that ANC voters had experienced since 1994, they had not relinquished “the vision of a better future around the corner” (1999: p. 34). Many of the elites interviewed in 1999 identified optimism as the driving force for the transition which might sustain poorer South Africans until their expectations of the good life are fulfilled. The 1999 sample survey confirmed widespread optimism, especially among black South Africans who are still far less contented than white South Africans in spite of having achieved political freedom. There may be some years before poorer South Africans catch up
290
VALERIE MØLLER AND HELGA DICKOW
on happiness. The survey highlights the perceived importance of living standards and consumption as a quality-of-life booster. Social spending on the poor has been increasing since 1993 and economic growth, sluggish as it has been, has raised black incomes. These factors alone may be sufficient grounds for optimism among all South Africans.
ACKNOWLEDGEMENTS
The study was undertaken as part of the Quality of Life Trends Study that has monitored the subjective well-being of South Africans since 1983. The financial support of the French Institute in South Africa (IFAS) and the Rhodes University Council to carry out the study of elites and place items with the MarkData October 1999 syndicated survey is gratefully acknowledged. Views are those expressed by the authors and should not be attributed to sponsors or any other persons.
NOTES 1
More information on policy, research instruments, and publications are available on Statistics South Africa’s website: http://www.statssa.gov.za 2 See Nzimande and Sikhosana (1996) for an introduction to affirmative action as a tool to redress discrimination in the workplace in the South African context. 3 A more detailed account of results from the elite survey is given in Møller and Dickow(2001).
REFERENCES
African National Congress: 1994, The Reconstruction and Development Programme: A Policy Framework (Umanyano Publications, Johannesburg). Brown, B.V. and T. Corbett: 1998, Social Indicators and Public Policy in the Age of Devolution (IRP Special report series. Madison, WI: Institute for Research on Poverty). Central Statistical Service (CSS):. 1996, Living in South Africa: Selected findings of the 1995 October Household Survey (Central Statistical Service, Pretoria). Central Statistical Service (CSS): 1997, Census ’96: Preliminary Estimates of the Size of the Population of South Africa (Central Statistical Service, Pretoria).
THE ROLE OF QUALITY OF LIFE
291
Eckersley, R.: 2001, ‘The mixed blessings of material progress: Diminishing returns in the pursuit of happiness’, Journal of Happiness Studies 1, pp. 267– 292. International Herald Tribune: 2000, ‘South Africa to withold data on crime’. August 4, p. 12. Johnson, R.W.: 1999, ‘How to use that huge majority’, Focus 16, pp. 33–36 (Helen Suzman Foundation, Johannesburg). Kotzé, H.: 1993, ‘Attitudes in transition II: Elites on economy’, Indicator South Africa 10(3), pp. 23–28. MacGregor, K.: 1999, ‘ “Sunny with some stormclouds”: The Reality Check survey’, Indicator South Africa 16(2), pp. 20–28. Mattes, R., H. Taylor and C. Africa: 1999, ‘Judgement and choice in the 1999 South African election’, Politikon 26(2), pp. 235–247. May, J. (ed.): 1998a, Poverty and Inequality in South Africa (Praxis Publishing, Durban, South Africa). May, J. (ed.): 1998b, Experience and Perceptions of Poverty in South Africa (Praxis Publishing, Durban, South Africa). May, J., M.R. Carter, L. Haddad and J.A. Maluccio: 2000, ‘KwaZulu Natal Income Dynamics Study (KIDS) 1993–98: a longitudinal household database for South African policy analysis’, Development Southern Africa 17(4), pp. 567–581. Møller, V.: 1995, ‘Indicators for Africa: The October household survey’, Indicator South Africa 12(3), pp. 86–90. Møller, V.: 1998, ‘Quality of life in South Africa: Post-apartheid trends’, Social Indicators Research 43, pp. 27–68. Møller, V.: 1999, ‘South African quality of life trends in the late 1990s: major divides in perceptions’, Society in Transition 30(2), pp. 93–105. Møller, V. and H. Dickow.: 2001, Five Years into Democracy (Les cahiers de 1’IFAS No 1, French Institute of South Africa, Pretoria). Møller, V., H. Dickow and M. Harris: 1999. ‘South Africa’s “rainbow people”, national pride and happiness’, Social Indicators Research 47, pp. 245–280. Møller, V. and L. Schlemmer: 1983, ‘Quality of life in South Africa: Towards an instrument for the assessment of quality of life and basic needs’, Social Indicators Research 12, pp. 225–279. National Economic Development and Labour Council Forum (Nedlac): 2000, ‘Infrastructure delivery’, Indicator South Africa 17(2). Nicholson, J. (ed.): 2000, Quality of Life of Durban’s People (Urban Strategy Department, Durban Metropolitan Council, Durban, South Africa). Nzimande, B. and M. Shikosana (eds.): 1998, Affirmative Action and Transformation (Indicator Press, University of Natal, Durban). Schlemmer, L. and V. Møller: 1997, ‘The shape of South African society and its challenges’, Social Indicators Research 41, pp. 15–50. Southern Africa Labour and Development Research Unit (Saldru): 1994. South Africans Rich and Poor: Baseline Household Statistics (School of Economics, University of Cape Town, Rondebosch).
292
VALERIE MØLLER AND HELGA DICKOW
Truth and Reconciliation Commission: 1998, Truth and Reconciliation Commission of South Africa Report. Orkin, M.: 1997, Living in Gauteng: Selected Findings of the 1995 October Household Survey (Central Statistical Service, Pretoria). Reconstruction and Development Programme (RDP), Ministry in the Office of the President: 1995, Key Indicators of Poverty in South Africa (Government Printer, Pretoria). Whiteford, A. and D.E. van Seventer: 2000, Winners and Losers: South Africa’s Changing Income Distribution in the 1990s (Wharton Econometric Forecasting Associates, Johannesburg).
Institute of Social and Economic Research Rhodes University Grahamstown 6140 South Africa E-mail:
[email protected] Valerie Møller
Arnold Bergstraesser Institut Windausstrasse 16 D-79110 Freiburg Germany
Helga Dickow
JULIAN MAY, THILDÉ STEVENS and ANNARETH STOLS
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE: A SOUTH AFRICAN CASE STUDY *
ABSTRACT. This paper outlines the approach that is utilized by the Monitoring and Evaluation directorate of the Department of Land Affairs (DLA) in South Africa in assessing the quality of life for the land Reform beneficiaries. The paper begins with an overview of the three Land Reform programs in South Africa. The paper then moves on to outline the original design for monitoring and evaluating the quality of life for land reform beneficiaries. It then proceeds to detail the current Monitoring and Evaluation design being utilized, highlighting its strengths and weaknesses. The last section discusses some of the findings of the quality of life study undertaken in 1999.
1. INTRODUCTION
In 1994 the first democratically elected government of South Africa committed itself to the Reconstruction and Development Programme (RDP), the policy framework through which it was hoped, a broad transformation of South African society could be achieved. The overall goal of the RDP was the promotion of a fundamental transformation of the social, economic and moral foundations of South African society (ANC, 1994; GNU, 1994). Moreover, the RDP was seen to be a statement of intent, not only for government but also for other sectors of South African society, including the private sector, Non Government Organisation’s (NGOs) as well as local communities. Land reform is the third element of South Africa’s RDP policies that focuses on targeted transfers, the others being a housing grant and a suite of welfare transfers including state pensions and child support. Compared to land reform programmes in other countries, many of which are focused more on productive development, the South * A revised version of a paper presented to the International Society for Quality of Life Surveys conference, Girona, Spain, 20-22 July 2000. Social Indicators Research 58: 293–312, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
294
JULIAN MAY ET AL.
African programme has places a strong emphasis on equality and the redress of historical inequities. In the formulation of policy, attention has been paid to the interests of the rural poor generally, and the interests of rural women in particular. The appropriateness of this approach has been the subject of considerable debate (van Zyl et al., 1996; Levin and Weiner, 1997; Lipton, Ellis and Lipton, 1996). To the extent that access to land improves the well-being of poor households, the poverty profile of South Africa shown in Table I lends support to this emphasis. While approximately half of South Africa’s total population of 40 million people can be categorised as being poor, most of the poor live in rural areas of South Africa, with the poverty share of rural areas (i.e. the percentage of poor individuals that live in rural areas) being equal to 72%.1 The poverty rate in rural areas (i.e. the percentage of individuals classified as poor) is about 71%, compared with 29% in urban areas. This high poverty rate is combined with deep levels of poverty among the poor in rural areas. Consequently, rural households account for 76% of the total poverty gap, which measures the severity of poverty, although they only make up 50% of the population. Despite the potential offered by land reform for improving the quality of life in rural communities, and an improvement in the delivery performance of government in recent years, the land
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
295
reform initiative is currently under review. Information provided by the monitoring and evaluation of the programme can potentially assist the review process. Monitoring and evaluation (M&E) has become an important aspect of all policy analysis in South Africa and provides both insight into management and implementation processes, as well as the effectiveness of targeting and the provision of support. This has been recognised by the Land Reform Programme since the first planning exercises were initiated and M&E has been implicit in the programme since its inception. This paper uses the third land reform monitoring study undertaken since the introduction of the policy in 1994 as a case study to show how the quality of life of land reform beneficiaries is evaluated. Although many of the premises for monitoring remain consistent, the study represents a departure from the previous M&E system through its use of a single integrated instrument for the collection of data, a more complex sampling procedure, and the manner in which the data was collected. 2. THE LAND REFORM PROGRAMME IN SOUTH AFRICA
In being operationalised, the land reform programme has been broken into three elements: redistribution, restitution and tenure reform. Restitution: The goal of the restitution policy is to restore land and provide remedial options to people dispossessed by past racially discriminatory legislation and practises. Cases are dealt with through the Land Claims Court and Commission, established under the Restitution of Land Rights Act of 1994. Eligible cases are largely the victims of forced removals since 1913. The vast majority of restitution claims are still pending, either with the Commission on the Restitution of Land Rights, or with the Land Claims Court. Tenure reform seeks to improve tenure security for previously disadvantaged people of South Africa. This programme includes a review of current land policy, administration and legislation with a view to accommodating more diverse forms of land tenure. Tenure reform was still proceeding by way of pilots or test cases in 1999.
296
JULIAN MAY ET AL.
Redistribution was established with the aim of providing opportunities for the large number of black households who wanted to access land but did not have specific documentation to prove that their ancestors were forcibly dislodged or who were not immediate beneficiaries of programmes of tenure reform. The programme provides a grant of R 16 000 (originally R 15 000), equal to the national housing subsidy that can be used by rural households to acquire land in the regular market. This is expected to provide these households with the means to establish their own productive enterprises. Of the three branches of the Department of Land Affairs’ land reform programme, tenure reform is likely to have the most farreaching consequence due to the large numbers of people involved. Although no reliable figures exist as to how many people have insecure land rights, it can be surmised that most of those living in the former ‘homeland’ areas can be characterised as having insecure tenure. This amounts to some 3,9 million black rural households. In addition, there are presently around 1,3 million households living in informal and squatter housing in and around urban areas, and roughly 800,000 permanent farmworkers and their on-farm households whose lodging is only as secure as their jobs. This yields a rough sum of around 6 million households. However, the legislative reforms required for tenure reform has yet to be finalised, and there has been little progress in the implementation of this policy. By contrast, land redistribution is to affect about 1,5 million households over the next ten years. The initial land reform target for the redistribution programme was massive – to transfer 30% of South Africa’s 99 million hectares of farmland, or about 30 million hectares, between 1994 and 1999. After the first three years of operation, about 200 000 hectares of land have been transferred to about 20 000 households. This represented 0.6% of the target, and 0.2% of the households demanding land. However, as Table II shows, even with a moratorium imposed on land reform projects during 1999, there has been a rapid increase in both the number of projects and the number of beneficiaries.2 By the end of 1999 there were some 26 000 households involved in 245 land reform projects. This represents a ten-fold growth from the 245 households
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
297
that received land during 1995 and had risen to 36 394 households using 684 914 hectares by October, 2000. Finally, land restitution, which is mandated by the Constitution, is unlikely to affect more than 500 000 households, both urban and rural, over the same amount of time. Due to the complex legal processes involved, although 67 500 cases had been lodged by October, 2000, just on 8288 cases affecting 20 473 households had been resolved.
3. MONITORING AND EVALUATION OF LAND REFORM IN SOUTH AFRICA
The original design for monitoring and evaluation (M&E) was based on a series of questionnaire formats, each of which covered a different aspect of the Land Reform Programme. These were developed during 1994 through a series of workshops convened by the Land and Agricultural Policy Centre (LAPC).3 The system that was developed focused on two main elements: Measurement of the quality of life enjoyed by land reform beneficiaries. A household questionnaire was developed for this purpose; Assessment of the targeting and equity components of the land reform programme as a whole. A community level questionnaire was developed for this purpose supported by an on-line management information system.
298
JULIAN MAY ET AL.
A number of other Formats were also developed as supporting instruments, including environmental impact studies, land demand and supply, and land invasions. While this system was first implemented in 1996, it was only in October/November, 1997 that the first comprehensive and systematic study was undertaken. A total of 62 land reform projects was surveyed, but information was collected for only 217 households. In May 1998, the Monitoring and Evaluation Directorate of the Department of Land Affairs (DLA) completed the Quarterly Land Reform Monitoring Report. An independent assessment of the report concluded that, in addition to a number of implementation problems, the information collected was not sufficiently detailed to permit the type of evaluation analysis required by DLA. In July 1998, the Monitoring and Evaluation Directorate decided that the Quality of Life instruments previously utilised to monitor and evaluate the progress and impact of the land reform programme would be rationalised. In line with this decision, the previous system of multiple formats was replaced by a single integrated system with two data gathering instruments: A revised and expanded household survey to collect quality of life data based on the Living Standards Measurement Surveys implemented in many countries by the World Bank (Grosh and Munoz, 1996). The purpose of this questionnaire was to investigate the individual and household characteristics of land reform beneficiaries; A new community questionnaire to collect information concerning the diverse projects that have been established by the land reform programme as well as data concerning community level attributes. This questionnaire also collected price data to facilitate the calculation of a community specific price index. The redesign of the instruments was undertaken in collaboration with the Monitoring and Evaluation (M&E) directorate and a Technical Advisory Group (TAG) of land reform specialists. Fieldwork was undertaken between August and October, 1999, and after extensive cleaning of the data, the first revised Quality of Life report has been completed and submitted to the DLA.
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
299
4. FRAMEWORK FOR ANALYSIS
The conceptual framework that underpins the redesigned M&E system for land reform in South Africa is grounded on four elements. Together these reflect not just a concern for measuring improvements in the ‘standard of living’ of land reform beneficiaries, but following Dreze and Sen (1989: p. 13), also of assessing the effect of the land reform on the capabilities of beneficiaries in their productive and social lives. Food security is the first element of the framework and is an important determinant of well-being directly affected by land reform. This can be through a direct relationship such as the growing of food or cash crops that are either eaten by the household or traded. Well-being may also be affected through an indirect relationship when, because of access to secure tenure, households are able to reallocate their income towards greater food security, receive services and invest in improved shelter. This in turn, improves their health, enhances their quality of life and frees up time for productive activities (Chambers, 1988). Secondly, although the mandate of the DLA does not extend to the provision of services, such as water and electricity, and facilities, such as schools and clinics, these are central determinants of the physical quality of life of land reform beneficiaries. As a result, these are thought to be sufficiently important for inclusion into the DLA M&E system as indirect outcomes from land reform. Another aspect of the land reform policy relates to the formation of different types of land management committees. The functioning of these associations is critical not only to the effective use of the land that is transferred, but also to the ability of the community to mobilise and organise for the delivery of services. Further, although difficult to measure, empowerment is an important outcome that is sought by the land reform programme. Institutional capacity is thus key both as on outcome of development, and as a mechanism for the facilitation of development. Finally, while agriculture is not the sole, nor even the most important, activity that is addressed by land reform, the particular form taken by the South African land reform programme demands that agricultural activities be examined in detail. At a general level, land reform is concerned with the regeneration of an
300
JULIAN MAY ET AL.
agrarian economy, of which agriculture, whether for subsistence, for exchange or as a source of employment, is an important component. At a specific level, the anticipated comparative competitiveness of small-scale farmers over larger scale farming activities is central to the logic of a market-assisted land reform such as that adopted in South Africa (van Zyl and Kirsten, 1997: pp. 180–182). The South African land reform programme has adopted a targeted approach in its implementation, and thus an additional concern for M&E relates to the effectiveness of the targeting mechanisms. Generally targeting may carry two error types: 1. people are excluded as beneficiaries who should be included, in
other words, failure to reach those from whom the intervention is primarily intended; 2. people are included as beneficiaries who should be excluded, in other words, excessive coverage whereby groups not intended as beneficiaries received assistance from the intervention (Cornia and Stewart, 1995: p. 351). For this reason, a beneficiary profile is a necessary component of the M&E strategy, and, wherever possible, a comparative analysis with non-beneficiaries should be included. 5. QUESTIONNAIRE DESIGN
In translating this analytical framework into a survey instrument, both the household and community questionnaires have been structured to examine three themes. Different modules are used to collect information within these themes which combine objective and subjective measurements: Project, community and household composition; Project and household income, livelihoods and well-being; Project, community and household institutional involvement, satisfaction and expectations. Composition of the household is tracked as a means of checking the representivity of the sample against large sample official statistics, as well as to monitor changes in project, community or household composition that might arise due to the land reform. Well-being measures monitor longer-term trends in the economic and social
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
301
status of the beneficiaries. Within this theme, livelihood indicators show the various activities that the project or household engage in, and stresses activities, such as agricultural production, in which land is a productive input. Finally, satisfaction indices track the institutional structures and involvement of beneficiaries, the beneficiaries’ short-term views concerning the processes that are followed and the beneficiaries’ views of the changes in well-being that they anticipated when entering the programme. The size, structure and demographic composition of the household was collected by means of a standard household roster. The information collected includes age, gender, occupational status, residence and education. In addition, the relationship of each household member to the main decision-maker or household head, as well as to the land titleholder has been identified. In the case of the community questionnaire, the number and composition of households in the various community projects has been collected, along with movements onto and from the project. The well-being theme of the household questionnaire tracks different aspects of the quality of life of the household. Several modules have been used, with household income being tracked by means of a simple expenditure module and a livelihoods module which includes shares in joint production schemes. The expenditure component permits the calculation of a money-metric measure of well-being, that is measurements expressed in a unit of currency. The argument behind such measures is that these best reflect the economic well-being of the household by measuring the ability of the household to purchase the commodities and services required for some minimum standard of living. The arguments against such measurement include the fact that not all requirements for an acceptable quality of life can be purchased (for example justice, freedom from violence), that some requirements may be met through recourse to common property (water, wood fuel), and that this measure looks at the inputs to a satisfactory quality of life rather than the outcomes (such as health, education and so forth). Despite these problems, money-metric measures are widely used and are readily understood (Lipton and Ravallion, 1997). Agricultural production has received detailed attention to investigate the productive use to which land has been put. Household
302
JULIAN MAY ET AL.
wealth has been tracked by collecting information on assets, savings and debt, and by using the education data collected in the household roster. Access to services has been tracked looking at a check list of important services and assessing access, quality and reliability of each of the following: School Health facilities Energy Extension Telecommunications
Transportation and roads Water and sanitation Credit Markets services
Household empowerment was examined by looking at the following indicators, especially for women: Involvement in institutions Knowledge of institutions The functioning of land administration authority The security of household land rights has been traced by listing the various land plots to which the household might have rights, and then examining whether these rights are exercised, how they were acquired, whether they can be traded or leased. The following is examined: Land use rights Grazing Residential
Commercial Arable
This section also examines why some households do not exercise the land rights that they do have. The community questionnaire considers a similar set of indicators, although in the case of income and livelihoods, only the income and activities generated by community projects is gathered. It should be noted that the projects established under the land reform programme have assumed many different organisational structures. Sometimes these have involved group land allocation (and production), while in other cases individual land allocations have been adopted, while a mix of both forms of allocation is also possible.
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
303
Similarly, projects vary in terms of their purpose, with some being only for settlement, while others are equity share schemes with the clear objective of profit maximisation. The satisfaction indices consider two dimensions. Firstly, satisfaction with the process followed by DLA and the land administration body during the transfer, as well as in the post-transfer stage (after-care). Secondly, respondents were asked about their expectations of the land reform programme in terms of livelihoods, access to services, and empowerment.
6. RESEARCH DESIGN
The research design consists of a baseline survey in 1999 in which both household and community questionnaire have been administered. The sampling frame included all beneficiaries of the land reform programme who joined the programme up until Decemberl998. Further to this, the sample that was drawn was stratified in such a way that the beneficiaries between January 1998 and December 1998 can be analysed separately from the previous years. In 2000 and 2001, it is envisaged that the household questionnaire will be administered to a sample drawn from the beneficiaries of projects that commenced in 1999 and 2000 respectively. However, the community questionnaire will be undertaken in all projects that have been sampled for the current and previous years. In 2002, and each year thereafter, both questionnaires will be administered to a sample of new projects as well as to the sample from three years prior, thus forming a panel or longitudinal study. Community questionnaires will still be administered to the remaining sample. Conceptually, this approach may be represented as in Figure 1. This research design has a number of advantages. Firstly, information collected at the community level, such as the operation and income generated by communal project, can be integrated into the household data. Furthermore, targeting of land reform can be analysed by comparing the beneficiary profile drawn from different years. In each case, the new group is drawn randomly, thereby enabling a probability sample to develop in the face of an expanding universe. Analysis of the quality of life of land reform beneficiaries can be tracked by comparing the results of the community question-
304
JULIAN MAY ET AL.
naire that are completed each year. Since this questionnaire will be administered to all existing and new projects each year, analysis of the changes in quality of life that are experienced by communities in the land reform programme can take place in 2001, and every year thereafter. Analysis of quality of life changes at the household level can be assessed from 2002 as the households from each wave are resurveyed. There are a number of technical advantages as well. Due to the existence of panel data, future analysis will be able to take into account the effect of time-invariant household fixed effects. These include unobserved characteristics of households, such as willingness to innovate, which might lead to changes in their quality of life that are not attributable to land reform. Panel data also improves the precision of the results since data are collected from more than one time period allowing enumerator and respondent errors to be identified. Finally, the attrition rate of the household sample can be reduced, since information on migration can be collected during the implementation of the annual community questionnaire.4 The design has at least one important shortcoming. In each of the first three years of data collection, an impact assessment of the land reform can be only tentatively undertaken at the household level. This is because households will only just have established themselves after the transfer of land, and it is unlikely that the benefits
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
305
from increased agricultural or non-agricultural production will have been realised. Likewise, it will not be possible to identify failure to improve quality of life. However, at the project level, some analysis of the impact of the land reform will be possible after the first wave of data collection.
7. SUMMARY OF FINDINGS
In terms of the capabilities based framework discussed earlier, Table III summarises the key findings of the study. Food security is one the most important determinants of well-being that could be directly affected by land reform. The data show that the basic headcount ratio or incidence of poverty for the sample is 78%, while the headcount using a lower poverty line, indicating extreme or ultra poverty, stands at 47%. These levels are exceedingly high relative to other poverty estimates for South Africa. For example, Klasen (1997) estimates from the 1993 Saldru study that South African poverty rates range between 44 and 57%. This would seem to indicate that households involved in land reform projects are relatively poorer than the national average. In addition, the data also reveal that the depth of poverty (measured by the poverty gap ratio) is higher for land reform beneficiaries that for the South African population as a whole. These high levels of poverty among land reform beneficiaries confirms that this is a group in need of more secure access to assets and the livelihoods that these might bring (Carter and May, 1999). The results of the study do suggest that the land reform programme is making a contribution in this regard. When the value of agricultural production is subtracted from household expenditure, the incidence of poverty increases to 79% and more importantly, there is an increase in the percentage of the sample who are ultrapoor to 49% as well as a sharp increase in the severity of poverty as measured by the poverty gap ratio.5 Furthermore, twenty five different kinds of crops were listed as being grown by the land reform beneficiaries, with most households growing multiple crop types. Over 67% of those who were growing crops cultivated two or more items, and 20% growing four or more items. The implication is that agricultural production is an important source of income to
306
JULIAN MAY ET AL.
the very poor, which boosts the income of this group but in 1999 was still not sufficient to raise them out of poverty. It is noteworthy that 75% of land redistribution beneficiaries have access to more than one plot of land, usually a residential plot and a plot for production while just over 70% of sampled households currently have access to a plot of land for the cultivation of crops.6 Overall, grazing is the most commonly found productive form of land use. Using these data to explore possibilities for fostering sustainable growth by simultaneously improving equity and effi-
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
307
ciency, Deininger and May (2000) conclude that although the South African land reform program has not lived up to the quantitative goals set, the programme has led to a significant number of economically successful projects that have already generated sustainable revenues. These projects have involved significantly larger shares of poor people than less viable projects, suggesting that increased access to productive assets could be an important path to poverty reduction. However, much land remains under-utilised with neither grazing nor cultivation occurring. It is also disturbing that the data show that women-headed households are less likely to use land for production than male-headed households. The specific constraints faced by women on land reform projects may require further investigation. The data does suggest that there are opportunities for income generation that are being missed. For example, more households are holding land that is fallow or vacant than renting out their land for others to use. Services are an indirect outcome from land reform. Despite the high levels of unfulfilled expectations, the land reform beneficiaries enjoy comparatively high levels of services when compared to African rural households as a whole.7 Land reform beneficiaries are located farther from reliable transport networks but have better access to means of communication and services. A greater proportion of households have electricity connections, access to piped water and access to telecommunications than is the case reported by the October Household Survey 1997. However, in terms of the quality of housing, as measured by the type of building materials and number of rooms, land reform beneficiaries perform less well. In many ways this is not surprising since a time lag is inevitable before new houses can be fully constructed, and it is to be expected that this situation will improve over time. Finally, there is provincial variation, suggesting that land reform projects are better integrated into service delivery programmes in some areas than in others, and that improvements are possible in certain provinces. The functioning of land management committees and other local associations is critical not only to the effective use of the land that is transferred, but also to the ability of the community to mobilise and organise for the delivery of services. The community trust is
308
JULIAN MAY ET AL.
the favoured form of legal entity within most provinces with the exception of the Northern Cape and Gauteng provinces which have a higher level of common property association (CPA). While the study has found that community and committee meetings do occur on a regular basis, only half of the respondents felt that report backs from the management committee were regularly carried out. Further, 68% of the land reform beneficiaries were not aware whether the total land grant funds had been utilised and almost 30% of the respondents did not know what type of land management structure had been adopted. This is an important issue since clear awareness of a project’s governance structure, as reflected by beneficiaries’ knowledge of the rules of the “trust” governing the project, was found to be an important determinant of project success. The land reform programme is concerned with the regeneration of an agrarian economy, of which agriculture, whether as for subsistence, exchange or as a source of employment, is an important component. Land reform beneficiaries seem to be comparatively well endowed with agricultural resources. Provincial differences are quite marked in some instances – for example, while the average plot size owned or used by households in the Provinces of Mpumalanga, the Free State and the Northern Province exceeds 120 ha, in KwaZulu-Natal, the Northern Cape and the Western Cape it is less than 15 ha. With respect to the ownership of livestock, some 39% of surveyed households own large stock (excluding poultry). It is interesting to note however that only 10% of the sample had either bought or sold cattle over the past year, and only 4% had bought or sold poultry. This indicates subsistence rather than commercial use of these assets. Just less than two-thirds of households in the survey sample own agricultural equipment, which appears high. Interestingly, female-headed households tend to have a greater tendency towards agricultural equipment ownership than do male-headed households. A total of 118 projects were identified in the 86 communities that were enumerated, with average of 1.4 projects per community. Just over 35% of the households that were surveyed had at least one household member participating in these projects, with an average of 1.2 people per households amongst that that were participating. Roughly half of these communal projects are generating an income,
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
309
although few are making any profit. This is not unusual for enterprises that are only 2–4 years old, but does suggest the need for support from institutions specialising in micro-enterprise support, as well as from the Department of Agriculture. Nonetheless, although many projects do not yet show any signs of economic potential, fourteen projects (or about 15% of the total) were characterised by very high profits, generating a median income for the typical beneficiary of R 10 000 per year. This would not only provide a very favourable return on the land acquisition subsidy of R 16 000 but also be more than sufficient to lift beneficiaries out of poverty. Although only 13% of respondents indicated that there had been moderate or severe conflict in the community over the past year, only half of the sample said that people in the community trust one another over most things. It is also noteworthy that over 80% of land reform beneficiaries had expected to plant crops and to generate an income from agriculture, although only 22% actually realised this expectation. More than 90% of land reform beneficiaries expected better services and homes after the land reform process, and only 15% felt that these expectations had been realised. Lastly, 29% of the respondents described themselves as being unhappy or very unhappy with the land reform process, suggesting that despite the high level of unrealised expectations, in 1999, beneficiaries were broadly satisfied with the process. Finally, the data show that the land reform programme has largely succeeded in reaching its target of the rural poor and has not been hijacked by the rich and powerful. Regression analysis undertaken by Deininger and May (2000) supports the evidence that land reform targets labour-abundant households who are poorer than the average but have higher level of productive agricultural assets (mainly animals). Household income and expenditure emerge as being highly significant and negative determinants of participation in the land reform programme. It was also found that observable beneficiary characteristics had not been an important determinant for the economic success of specific land reform projects. Finally, women-headed households are at least proportionally represented in the land reform programme, although it does seem that maleheaded households have access to larger plot sizes on average. The educational level of heads of land reform households is lower than
310
JULIAN MAY ET AL.
the national average. It can be concluded from this that, if land reform can be made to be effective, it has considerable potential to contribute to overcoming the legacy of apartheid and to a sustainable reduction of rural poverty. 8. CONCLUSION
The South African land reform programme has developed as a process of learning by doing. The first five years of the postapartheid period have witnessed the establishment of an administrative infrastructure for land transfer, and the gradual adaptation of the initial programme design to the requirements of the real world. All of these, together with the change of government in 1999, make this an ideal moment to take stock and try to assess the lessons from the past for future implementation of land reform in this country. The collection of a comprehensive monitoring and evaluation data set has made such analysis possible. Analysis of the M&E data from the Quality of Life survey undertaken in 1999 shows that there has been an improvement in both the performance and impact of land reform since the previous monitoring activities undertaken in 1998. This has taken place in a number of ways, with an increased rate of delivery, targeting of the most poor has taken place and there is less evidence of institutional problems than was found in the 1998 study. In addition, both agricultural and non-agricultural production is occurring and services delivery to land reform beneficiaries seems better than to the rural population as a whole. However, poverty levels among land reform beneficiaries remain high, as do the levels of dissatisfaction that they express. Many projects do not yet show any signs of economic potential and many participants in the land reform projects have little knowledge of the management of the project and how funds have been utilised. This opens opportunities for corruption and the misuse of community funds. It is recommended that the land reform programme continue to be supported, and perhaps even expanded, but that consideration be given to the redesign of some elements of the programme. Areas for attention include simplifying the administrative procedures that are followed, increasing the flexibility of the programme to allow for larger grants and linking to other programmes of livelihood
MONITORING THE IMPACT OF LAND REFORM ON QUALITY OF LIFE
311
support and service delivery and finally. Finally, ways of targeting sub-groups of the rural poor whose current participation in the land reform is limited should receive specific attention. NOTES 1
The conventional set of money-metric poverty measures developed by Foster, Greer and Thorbecke (1984) have been calculated using data derived from Stats SA (1997), Income and Expenditure Survey, 1995. 2 Derived from DLA records and the M&E sample frame. 3 The system is explained in the following memoranda: Coleman, G. (25/10/1995), A reporting structure for the M&E System in the Directorate of Land Reform Information and Evaluation: First Thoughts; Coleman, G. (October, 1995), Implementing the M&E System in the Directorate of Land Reform Information and Evaluation: Third Thoughts; Coleman, G. (27/10/95), Report on Work at the Directorate of Land Reform Information and Evaluation, Department of Land Affairs, 1–28 October, 1995; Coleman, G. (18/10/95), Sampling, confidence levels and precision in the M&E system of the Directorate of Land Reform Information and Evaluation, First Thoughts; Coleman, G. (25/10/95), Monitoring and Evaluation Information as a Public Resource: First Thoughts. 4 Baulch and Hoddinott (2000) provide a useful review of examples of panel studies from developing countries. 5 The poverty gap ratio for the ultra-poor increases from 0.186 to 0.476. 6 Land access implies that the household either owned the land, or had been given the right to use the land by a land regulation authority. 7 Although it is recognized that not all land reform beneficiaries are African, and that not all beneficiaries are rural, for the purposes of comparison with the Statistics South Africa’s October Household Survey, only this group have been selected.
REFERENCES
ANC: 1994, The Reconstruction and Development Programme: A Policy Framework (Umanyano Publications, Johannesburg). Baulch, R. and J. Hoddinott: 2000, ‘Economic mobility and poverty dynamics in developing countries’, Journal of Development Studies, p. 366. Carter, M.R. and J. May: 1999, ‘Poverty, livelihood and class in rural South Africa’, World Development 271, pp. 1–20. Chambers, R.: 1988, ‘Sustainable rural livelihoods: a key strategy for people, environment and development’, in C. Conroy and M. Litvinoff (eds.), The Greening of Aid (Earthscan, London). Cornia, G.A. and F. Stewart: 1995, ‘Two errors of targeting’, in D. Van de Walle and K. Nead (eds.), Public Spending and the Poor: Theory and Evidence (Johns Hopkins Press, Baltimore).
312
JULIAN MAY ET AL.
Deininger, K. and J. May: 2000, Is There Scope for Growth with Equity: An Initial Assessment of Land Reform in South Africa. Policy Research Working Paper, No. 2451, World Bank, Washington, DC. Dreze, J. and A. Sen: 1989, Hunger and Public Action (Clarendon Press, Oxford). Foster, J., J. Greer and E. Thorbecke: 1984, ‘A class of decomposable poverty measures’, Econometrica 52(3), pp. 761–766. Government of National Unity: 1994, RDP White Paper Discussion Document (Government Printers, Pretoria). Grosh, M. and J. Munoz: 1996, A Manual for Planning and Implementing the Living Standards Measurement Study Survey. Living Standards Measurement Study Working Paper No. 126, World Bank, Washington DC. Klasen, S.: 1997, ‘Poverty and inequality in South Africa: An analysis of the 1993 SALDRU Survey’, Social Indicators Research 41, pp. 51–94. Levin, R. and D. Weiner: 1997, No More Tears, Struggles for Land in Mpumalanga, South Africa (World Press, Trenton NJ). Lipton, M., F. Ellis and M. Lipton (eds.): 1996, Land, Labor and Livelihoods in Rural South Africa, 2 vols (Indicator Press, Durban). Lipton, M. and M. Ravallion: 1997, ‘Poverty and policy’, in J. Behrman and T.N. Srinivasan (eds.), Handbook of Development Economics, Vol 3 (North Holland, Amsterdam, pp. 2553–2658). Stats SA: 1997, The October Household Survey (Statistics South Africa, Pretoria). Stats SA: 1995, Income and Expenditure Survey (IES). Van Zyl, J. and J. Kirsten: 1997, ‘Economic empowerment in South Africa’, in L. Haddad (ed.), Achieving Food Security in Southern Africa: New Challenges, New Opportunities (International Food Policy Research Institute, Washington, D.C). Van Zyl, J., J. Kirsten and H.P. Binswanger: 1996, Agricultural Land Reform in South Africa: Policies, Markets, and Mechanisms (Oxford University Press, Cape Town).
School of Development Studies Memorial Tower Building University of Natal King George V Ave Durban, 4001 South Africa Department of Land Affairs Private Bag X833 Pretoria, 0001 South Africa
Julian May
Thildé Stevens Annareth Stols
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG: THE PROCESS OF COMMUNITY ENGAGEMENT
ABSTRACT. Stakeholder involvement is an essential component in the development of effective measures of national and local development. The significance of this aspect of social indicator model-building in the construction of national/local assessment tools is illustrated through the creation of a Social Development Index (SDI-2000) for Hong Kong SAR (Special Administrative Unit of the People’s Republic of China). In all, the process leading to the SDI-2000 required two years to complete and the involvement of more than 100 political, administrative, academic and community leaders representing all aspects of collective social life in the SAR. Consisting of 47 indicators distributed across 14 sectors of social development, the SDI-2000 revealed patterns of development that both affirmed the validity of certain aspects of Hong Kong’s current development priorities – mostly those associated with its economic, educational and scientific priorities – and, via another set of related indicators, identified areas of considerable social instability for which new development initiatives are needed – especially in reducing growing social inequalities within and between population groups that largely have been left out of the SAR’s recent economic prosperity – i.e., lowincome households, children and youth, and the elderly. The article also identifies ways in which future versions of the SDI can be strengthened – including through the addition of subindexes that measure the SAR’s changes over time in relation to Rule of Law, Subjective Quality of Life, and the impact of the Social Service in reducing social inequality. The SDI-2000’s current Environmental Quality Subindex also needs to be revised. KEY WORDS: community engagement, Hong Kong, model-building, national social reporting, quality-of-life, social indicators, social progress
The research summarized in this paper would not have been possible without the contributions of my colleagues at the Hong Kong Council of Social Service: Chua Hoi Wai, John Y.C. Fung and Anthony K.W. Wong. Each is thanked for their enthusiasm for and the countless number of hours invested in the effort. Dr. Frans Lammertyn, Vice Rector for Educational Policy of the Katholieke Universiteit Leuven, is also acknowledged for providing me with the opportunity to complete work on this paper while serving as a Visiting Scholar with that university’s Faculty of Social Science. Social Indicators Research 58: 313–347, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
314
RICHARD J. ESTES
INTRODUCTION
Hong Kong SAR1 is an international metropolis located at the mouth of the Pearl River delta on the southeast coast of China with a population of some seven million people. Since at least the mid-1960s, Hong Kong has been regarded as one of East Asia’s major economic centers – a so-called “small dragon.” Since the 1980s, Hong Kong also has developed into a major trading and financial center for the world-as-a-whole. Today, with a per capita GDP exceeding USD $25 000 Hong Kong enjoys a level of affluence that places the SAR among the most prosperous societies worldwide (UNDP, 2000; World Bank, 2000/2001). The social benefits associated with Hong Kong’s economic transformation are obvious – a magnificent skyline, a first class transportation system, world class shopping districts, a major financial center, modern apartment blocks, beautiful entertainment centers and a network of highly regarded institutions of higher education. Hong Kong’s economic successes also are reflected in improvements in the day-to-day living conditions of her people – longer life expectation, increased access to preventive and curative health services, better schools, a rich network of sports and recreational services, a sophisticated system of social services, more locally available library services, significantly improved income security programs, and more public parks and other “green spaces.” These and other important advances have added measurably to improvements in the quality of life experienced by the people of Hong Kong (Lau et al., 1999). Following decades of rapid economic expansion, however, Hong Kong has arrived at an important crossroads. Economic expansion is no longer as rapid as during earlier decades. Jobs are more difficult to secure, pay less well, and are associated with less security than in the past. The cost of housing, food, health care and other basic necessities consume a larger share of total household earnings for which people must work longer hours. Meaningful savings are more difficult for all but a privileged few. And the SAR’s problems with crime, homelessness and the environment have proven less solvable than anticipated. In addition, “new” problems with family violence, drug abuse, youth suicide, and sex tourism have surfaced. Rapid
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
315
increases in population aging also are posing special challenges for Hong Kong’s already heavily pressed income security system. The Need for a Social Report
By early 1999 recognition existed in Hong Kong of the need to conduct a comprehensive assessment of its changing social conditions. The broad purpose of such an assessment was to arrive at an understanding of the SAR’s social environment more or less comparable to those which exist of its changing economic environment (Hong Kong Commission on Strategic Development, 2000). With that purpose in mind, the Hong Kong Council of Social Service (hereafter HKCSS) commissioned this author to undertake work on Hong Kong’s first comprehensive Social Report. The goals associated with the effort included:
1. the creation of an analytical tool unique to Hong Kong for use in assessing the SAR’s changing social, political and economic needs over time; 2. an analysis of the SAR’s social development trends at various points in time; 3. comparisons of the SAR’s social development patterns vis-à-vis those of other well-off societies; 4. a systematic analysis of the social development “performances” of Hong Kong’s historically vulnerable population groups – women, the aged, children, youth, and low-income households – vis-à-vis those of the society-as-a-whole; 5. with others in Hong Kong society, to initiate a process of working toward more balanced social and economic development; and, 6. to the fullest extent practical, to foster the participation of social leaders in Hong Kong throughout all stages of the data collection and analysis effort. This paper illustrates the contribution made by social reporting in helping Hong Kong’s social, political and economic leaders promote a more balanced approach to development, i.e., one that seeks to extend the benefits of its rapid economic development to all of the SAR’s residents.
316
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
317
RESEARCH METHODS
Table I identifies the project’s various timelines and tasks: (1) the conceptualization of social development in the context of a small, but well-off, East Asian society; (2) the creation of an analytical tool for use in measuring changes over time in that development; (3) data collection; (4) preliminary testing of the initial model; and (5) refinements to and re-testing of the final analytical model. Once fully tested, the Social Development Index (hereafter SDI-2000) was used to conduct an analysis of Hong Kong’s development trends for
318
RICHARD J. ESTES
the 17-year period 1981–1998 with particular attention given to the 12-year period 1986–1998. A. Time Frame The time periods covered by this analysis coincide with Hong Kong’s most recent five-year censuses, i.e., 1981, 1986, 1991, and 1996. Comparable data also were available for 1998 and, therefore, these data also were collected and used in the analysis. B. Levels of Analysis Development trends were examined at four levels of analysis: (1) for Hong Kong society-as-a-whole; (2) for each of 14 sectors of development included in the SDI-2000 (Table II); (3) for each of five “historically vulnerable” population groups of special concern to the HKCSS (Table III); and (4) for each of the five time periods indicated in (Figures 1–3). C. Data Sources The vast majority of the data used in this analysis were obtained either from the Hong Kong Census and Statistics Department or from other governmental bodies in Hong Kong (including the SAR’s Departments of Health, Education, Labor, Transport, Inland Revenues, Immigration, the Coroner’s Office, the Commission on Strategic Planning, and so on). Data also were obtained from non-governmental organizations that monitor specialized aspects of Hong Kong’s development, including from environmental groups, human services agencies, human rights groups and other civil society organizations. In all, time-series data were collected for approximately 400 social, political and economic indicators.
SOCIAL DEVELOPMENT: ITS MEANING AND MEASUREMENT IN HONG KONG
The major tool used to assess Hong Kong’s social development changes over time was the SDI-2000. The index is unique to Hong Kong and was developed through a collaborative process between
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
319
320
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
321
322
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
323
324
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
325
326
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
327
328
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
329
the author, senior staff members of the HKCSS, and approximately 50 members of the project’s local Panel of Experts. Figure 4 is a graphical representation of the conceptual framework used in the development of the SDI-2000. The framework emphasizes three distinct aspects – i.e. “domains” – of social life in Hong Kong: (a) Hong Kong’s collective memory and experience; (b) Hong Kong’s contemporary social, political, economic and environmental realities; and (c) the basic needs and higher level aspirations of individual residents of Hong Kong and the SAR’s complex network of social collectivities. Each of the SDI-2000’s 14 subindexes is associated with one of the model’s domains of collective social life. THE SOCIAL DEVELOPMENT INDEX-2000 (SDI-2000)
Table II reports the raw score values for each of the 47 indicators used to operationalize the SDI-2000. The following criteria were used in selecting these indicators: face and construct validity, reliability, representativeness, accuracy, timeliness, availability and, when possible, comparability with social measures used in other international and comparative analyses.2 An additional 31 indicators were used to create supplemental indexes that measure development trends for five of Hong Kong’s
330
RICHARD J. ESTES
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
331
most socially vulnerable population groups, i.e., children, youth, the elderly, women, and low-income households (Table III). Though complementary to the SDI-2000, social performances on these indexes were analyzed separately from the SDI-2000. The project’s remaining 320 indicators were placed in a social indicator “data bank” for future use.
332
RICHARD J. ESTES
Measuring Hong Kong’s Social Progress
Changes over time in Hong Kong’s social development were assessed using two methods: (1) cross-sectional analyses of the state of Hong Kong’s development at each of five time intervals – 1981, 1986, 1991, 1996 and 1998 (SDI-2000); and (2) the pace of Hong Kong’s social development between 1986 and 1998 (T-98). The statistical procedures used to compute both types of scores were comparable to those developed by the author for analyses of global (Estes, 1988, 1990, 1998) and regional development trends (Estes, 1996, 1997, 1999). 1. Standardizing indicator values For ease of comparison, all indicators were “standardized” such that indicator values for 1991 were set to equal 100, i.e., 1991 = 100. As a result of this process, indicator values for 1996 and 1998 could more easily be contrasted with one another and, in turn, with Hong Kong’s performances on the indicators in 1981 and 1986. The identical process was used in standardizing raw score values for each of the SDI-2000’s 47 indicators and for the 31 indicators used to create the population-specific indexes. 2. SDI-2000 item weights Statistical weights for SDI-2000’s 47 indicators were derived through consultations with the project’s 50-member Panel of Experts. The process was carried out in four steps: (1) experts were identified for each of the SDI-2000’s 14 subindexes, e.g., health care, education, economy, etc.; (2) working independent of each other, each subindex specialist was asked to assign statistical weights ranging from “0” (i.e., of no importance) to “100” (i.e., the most important) for each subindex indicator relative to its importance in explaining patterns of social development in Hong Kong, e.g., on the Health Subindex to distinguish between the relative importance of Infant Mortality vs. Childhood Immunizations, and so on; (3) subindex specialists were further instructed to treat each indicator as though it were independent of all other indicators contained in the subindex; and (4) the statistical weights assigned by each specialist to each indicator were summed and, then, averaged.
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
333
Table II reports the average statistical weights that resulted from this process. 3. Population-specific indicator statistical weights The process of deriving statistical weights for the study’s five population-specific indexes was identical to that used in obtaining item weights for SDI-2000 indicators: (1) local experts were identified who were most closely associated with the content area of each population-specific index, e.g., children, youth, the aged, women or low-income households; (2) working independent of one another, population specialists were requested to assign statistical weights ranging from “0” (i.e., of no importance) to “100” (i.e., the most important) for each index indicator relative to its importance in explaining patterns of social development in Hong Kong; (3) population specialists were further instructed to treat each indicator as though it were independent of all other indicators contained in the index; and (4) the statistical weights assigned by each specialist for each indicator were summed and, then, averaged. Table III identifies the average population-specific indicator weights that resulted from this process. 4. SDI-2000 subindex statistical weights The process of deriving statistical weights for the SDI-2000’s fourteen subindexes was similar to that used in obtaining item and subindex weights. However, and unlike the previous procedures, the judgments of all 50 members of the Panel of Experts were taken into account. The process involved the completion of three steps: (1) all members of the Panel of Experts were asked to assign a statistical weight ranging from “0” (i.e., of no importance) to “100” (i.e., the most important) indicating the relative importance of each subindex in explaining patterns of social development in Hong Kong; (2) in arriving at their judgments, panelists were instructed to treat each subindex as though it were independent of all other SDI-2000 subindexes; and (3) the separate statistical weights assigned by the panelists were summed and, then, averaged across all participants. Table II also identifies the resulting average statistical weights assigned by the project’s local experts to the SDI-2000’s 14 subindexes.
334
RICHARD J. ESTES
5. Standardized and statistically weighted subindex scores Prior to computing composite SDI-2000 scores, the statistical weights derived from Steps 2 and 4 were applied to the standardized indicator values created in Step 1. The following formula then was used to generate Weighted Standardized Subindex Scores (and, in turn, Weighted Standardized Population Index Scores):
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
335
6. Composite SDI-2000 scores Composite SDI-2000 scores for all five time intervals were computed using the formula reported in Table IV. Thus, composite SDI-2000 scores reflect: (a) standardized raw score values; (b) the indicator/item weights identified by specialist members of the Panel of Experts; and (c) the subindex statistical weights assigned by the Panel-of-Experts as-a-whole. Composite SDI-2000 scores for all five time periods studied are reported in Figure 1. 7. T-98 scores Social development trend analysis scores are identified in this study as “T-98.” The methodology used to compute these scores was developed jointly by Estes and Guilfoyle and, in essence, contrasts Hong Kong’s social development performance in 1998 with its performance during the preceding 10-year period, i.e., 1986–1996. The following formula was used to compute T-98 scores:
T(Z)98 = [(Value 1998 – (mean of values for 1986, 1991, 1996))/ (standard deviation 1986, 1991, 1996)] The z-scores (i.e., T-98 or “time” scores) that result from application of this formula, provide an estimate of the pace at which changes in development occurred over time for each of the SDI’s indicators and subindexes. Independently derived T-98 scores also were used to analyze changes over time in the social development profiles of each of the study’s five historically vulnerable population groups. T-98 scores are reported by SDI-2000 subindex in Figure 2 and for Hong Kong’s historically vulnerable population groups in Figure 3. FINDINGS
A. SDI-2000: General findings
The study’s major findings are summarized in Figure 1. Data are reported in the chart in two forms: (a) as Weighted Social Development Index (WSDI) scores for each of the study’s five time periods; and (b) as standardized versions of the WSDI in which 1991 is used
336
RICHARD J. ESTES
as the baseline year against which earlier and later development changes were compared (1991 = 100). The data reported in Figure 1 confirm that substantial social progress occurred in Hong Kong between 1981 and 1998, i.e., a net gain of more than 75% in SDI scores over the 17-year period. These gains were steady and significant, albeit the magnitude of the gains varied for each time period. Overall, though, Hong Kong’s most rapid improvements in social development occurred between 1991 & 1996 (+25%) and between 1986 & 1991 (+15%). Net improvements also occurred between 1981 and 1986 but these changes were slower and smaller (+10%). Additional unexpected gains on the SDI-2000 occurred between 1996 and 1998 (+8%), i.e., during a period of considerable political uncertainty when sovereignty over Hong Kong was returned to the People’s Republic of China (hereafter PRC). When viewed macroscopically, then, social progress in Hong Kong since 1981 has been steady and impressive. When viewed more microscopically (Figure 2), however, development trends in Hong Kong between 1986 and 1998 tend to be rather more asynchronous and asymmetric than the data reported in Figure 1 would suggest. B. SDI-2000: Sectoral performances – Areas of substantial net improvement
As reported in Figure 2, Hong Kong’s most impressive 12-year development achievements occurred on two critically important subindexes: Science and Technology (+107) and Education (+102). The gains on these subindexes were substantial and reflect dramatic improvements in Hong Kong’s scientific and educational infrastructure. Between 1986 and 1998, for example, the number of patents granted by foreign patent offices to persons residing in or doing business in the SAR more than doubled from 1,055 to 2,680. Similarly, the number of scientific articles published in international journals by Hong Kong affiliated researchers increased from about 1,000 in 1986 more than 5,400 in 1998 – a rise of more than 400% in just 12 years! Cellular phone, computer, and internet access in Hong Kong is among the highest in region; indeed, Hong Kong ranks 12th out of all High Human Development (hereafter HHD) nations3 reporting in
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
337
1999 with respect to the number of sponsored internet hosts, i.e., 21 per 1,000 population. The magnitude of these gains reflect considerable success on the part of Hong Kong to reposition itself from being a net consumer of innovations developed elsewhere to being a major contributor to scientific innovation. The results of Hong Kong’s consistent investments in Education also are evident in Table II. Adult literacy, for example, rose from 87% in 1986 to 91% in 1998 – an increase that is all the more remarkable given the high rate of migration into the SAR from rural China where rates of adult literacy tend to be lower. Similarly, the percentage of Hong Kong adults aged 20 years and older with upper secondary school educations increased from 35% in 1986 to 50% in 1998. And the percentage of Hong Kong’s residents with tertiary level educations increased by 80% between 1986 and 1998, i.e., from 8% to 14%. Clearly, Hong Kong is emerging as one of the East Asia’s most human resource rich societies. C. SDI-2000: Sectoral performances – Areas of significant net improvement
The SDI-2000 also confirmed that significant social development changes took place in five additional development sectors between 1986 and 1998: Arts and Entertainment (+80); Internationalization (+75); Health (+59); Personal Safety (+55); and Strength of Civil Society (+49). In the Arts and Entertainment sector, Hong Kong’s extensive film and book publishing industries continued to gain strength with net productivity improvements of 217% and 467% between 1986 and 1998, respectively. Similarly, the number of historic buildings and other sites preserved for future generations by the Hong Kong government increased from just 30 in 1986 to 67 in 1998 – a mark of increasing maturity for a society that previously choose to destroy its older buildings – often at the expense of losing part of its own cultural heritage. Appreciable increases also occurred in the number of persons that visited the SAR’s diverse museums and cultural venues – averaging, today, at least one visit per Hong Kong resident per year. With respect to Hong Kong’s role as an international city, the SAR remains one of East Asia’s major international centers. Hong
338
RICHARD J. ESTES
Kong continues to serve as an important crossroad for the conduct of international business and is a regular venue for a large number of international conferences, conventions and scholarly meetings. The number of international businesses with offices in Hong Kong also increased appreciably between 1986 and 1998, from fewer than 2,500 to more than 5,300. Additionally, the number of countries to which Hong Kong residents can travel without visas increased over the 12-year period from 61 to 101 – yet another form of recognition by the international community of the continuing importance of the SAR in international exchange. At the more personal level, the health status of Hong Kong’s growing population also improved markedly between 1986 and 1998. Gains in this sector were most evident in: sharp reductions in infant (–58%), child (–58%) and maternal mortality rates (–32%) – today, Hong Kong ranks 1st among HHD countries with respect to its low rate of infant deaths (along with Norway, Japan, Sweden, Finland and Singapore), 5th with respect to its rate of child deaths, and 4th most favorable with respect to the number of women who die in child birth; unprecedented reductions in the rates of infectious and communicable diseases, including tuberculosis (–15%); appreciable reductions in the prevalence of adult cigarette smokers (–17%); important increases in childhood vaccination rates; and at least modest reductions in the adult suicide rate, i.e., from 16 per 100,000 adults in 1986 to just over 12 in 1998 (–24%). These advances in the health sector contributed, in turn, to increases in years of overall life expectation for Hong Kong’s population-as-a-whole, i.e., from an average of 77 years in 1986 to an average of 79 years in 1998. Gains in average life expectation were even more impressive for Hong Kong’s elderly population for whom years of life expectation on reaching age 65 increased from 17 in 1986 to 19 in 1998. Today, average life expectation in Hong Kong compares favorably with that found in Japan (Average = 80 years) and exceeds that which exists in Singapore (Average = 77 years) and many economically advanced Western societies.
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
339
The SDI-2000 also revealed that significant 12-year improvements occurred in the strength of Hong Kong’s extensive civil society sector – the oldest, largest, and most deeply developed in East Asia. Substantial gains in this sector are reflected both in the increasing numbers of charitable organizations and trusts that are recognized by the Hong Kong government as tax exempt entities and in the level of financial support received by these organizations from private sources (including voluntary support from individuals). The number of charitable, educational, cultural and political organizations granted tax exempt status in Hong Kong now stands at more than 3,000 compared with only 1,400 in 1986 – a net gain over the 12-year period of some 109%. At the same time “countable” private charitable donations as a percentage of gross domestic product (GDP) more than doubled between 1986 and 1998 while the ratio of private donations to government subsidies for these organizations remained stable. Clearly, civil society institutions serve an important function within Hong Kong society and, on the basis of their record of achievement, receive support from both the governmental and private sectors. D. SDI-2000: Sectoral performances –Areas of moderate net improvement
Moderate levels of 12-year social progress occurred for Hong Kong on three additional SDI-2000 sectors: Political Participation Subindex (+39); Housing (+39); and Crime and Public Safety (+30). Owing to its unique political situation, issues related to politics and political participation have been among the most contentious issues confronting Hong Kong society – both prior to and following the return of political sovereignty over Hong Kong to the PRC. Prior to 1997, political activism within Hong Kong tended to be vested more in the personalities of individual office holders than in any identifiable system of political parties. Turnout on the part of the eligible electorate in district-wide votes never exceeded 39% and, in the main, the ratio of candidates to available district-wide positions tended to be low. Data reported on the SDI-2000 show that comparatively little changed in the structure of political life in Hong Kong between 1986 and 1998 with the sole, but important, exception that by 1998 a larger percentage of candidates for district-wide offices
340
RICHARD J. ESTES
identified themselves as members of a political party than in the past – 58% identified such an affiliation in 1998 compared with fewer than 2% in 1986. In 1998, however, voter turnout for district-wide elections dropped to 36%. Hong Kong’s housing sector represents another highly troublesome sector for the SAR. The sources of the trouble are several: (1) limited geographic space on which new housing stock can be developed; (2) a steady flow of new arrivals into the community – each household placing its own demands on the existing supply of low-income housing; (3) the cost of housing which, for many residents can consume 50% or more of their total household expenditures (Average = 31%); (4) long, and politically sensitive, waiting lists for publicly subsidized housing (averaging five or more years for the more than 130,000 households on the waiting list in 1999); and (5) the absence of flexible financing schemes for purchasing housing on credit over the long-term. As is the case for many other rapidly developing Asian societies, issues of housing supply are likely to remain high on Hong Kong’s social agenda well into the future. In terms of crime and safety, Hong Kong historically has been a comparatively safe place in which to live and work, albeit a certain level of crime does occur. Crime trends are particularly problematic among Hong Kong’s youth population and, increasingly, many of these crimes are associated with youth unemployment and drug use. Even so, Hong Kong succeeded in reducing the prevalence of both violent (–14%) and non-violent crimes (–30%) between 1991 and 1998. Today, Hong Kong ranks 4th worldwide among HHD societies with respect to intentional homicides, i.e., only 1.5 per 100,000 population. Between 1991 and 1998 the number of convictions for crimes involving public corruption increased from 3.8 per 100,000 population to 4.6 (+30%) – a pattern that reflects Hong Kong’s reduced tolerance for public corruption.
E. SDI-2000: Sectoral performances – Areas of modest net improvement
Less significant, but still important, gains in social development were reflected on the Economic (+23), Sports (+23), and Environmental (+14) subindexes.
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
341
In the economic sector, Hong Kong’s most important gain occurred in its increases in gross international reserves (as measured by months of import coverage). In 1986 Hong Kong held reserves sufficient to cover the cost of six months of imports; in 1998, these reserves had increased to 18 months (+187%). Obviously, Hong Kong has acted judiciously in both reducing public spending and increasing public savings – especially during a period of considerable economic instability both within and outside East Asian. In doing so, though, Hong Kong has incurred a variety of social costs associated with these policies including: (a) slower rates of growth in per capita income levels; (b) further concentrations of total wealth in the hands of a relative minority of Hong Kong individuals and institutions; and (c) nearly frozen personal wages for those at the lowest end of the earnings spectrum. Between 1986 and 1998, for example, the percentage of total household income earned by the bottom 50% of income earners in Hong Kong dropped from an already modest 23% to 19% of total household income earned. All three patterns make it extremely difficult for low income earners to hold onto their previous economic positions let alone move out of poverty. Hong Kong’s performances on the Sport and Recreation subindex confirm that important 12-year improvements occurred with respect to the development of additional public sports and recreation facilities. The number of such public facilities increased from a little more than 1,000 to 2,300 (+120%); however, the utilization of these facilities increased by only 12% during the same period. No appreciable 12-year changes occurred in the number of delegates sent by Hong Kong to important regional or international sporting events. With respect to social progress in the environmental sector, Hong Kong’s most important 12-year gains are reflected in the percentage of municipal waste that is recycled (+100%) and in the decision to set aside more “green” and “open” spaces for public use (+58%). Hong Kong lost ground with respect to the number of units of fresh water consumed on a per capita basis (–33%), however, as well as in the number of public beaches where water quality was rated as either “poor” or “very poor” (–34%). Improving the quality of Hong
342
RICHARD J. ESTES
Kong’s environment remains one of toughest challenges confronting the SAR. F. SDI-2000: Sectoral performances – Areas of substantial net loss
Hong Kong’s most dramatic, and most troubling, social losses occurred in a sector in which Hong Kong traditionally has been a world leader, i.e., family solidarity (–166). The very severe decline in Hong Kong’s performance on this subindex is associated with: (1) sharp decreases in the marriage rate (–36%); (2) even sharper increases in Hong Kong’s divorce rate (defined to be number of divorces as percentage of that of marriages) (+270%); and (3) rapidly increasing rates of reported family violence – including between spouses, between adults and children and, in the case of the elderly, between adult children and their aging parents. Clearly, “the family” in Hong Kong is undergoing tremendous change. For many families traditional family norms and roles appear to be no longer functional. G. Hong Kong’s vulnerable population groups
Considerable variation also exists in the 12-year social development performances of Hong Kong’s historically vulnerable population groups, i.e., populations that for one reason or another have not been able to participate adequately in either the social, political or economic life of the community. Chart 4 summarizes 12-year trend data for the five historically vulnerable population groups of special concern to the HKCSS: children, youth, the elderly, women, and persons living in low-income households. These data confirm that net social gains between 1986 and 1998 occurred for only two groups, i.e., women (+71) and the elderly (+13). But the pace of social change observed for the elderly was so marginal as to be considered insignificant given their complex financial and social care needs! Especially troubling in these data are the highly negative time series findings for children (–17), youth (–52), and persons living in low income households (–77). The size of all three populations is quite large and confirms the need for dramatic new initiatives that can respond more effectively to the difficult circumstances confronting the many people that make up these groups.
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
343
The situation is especially urgent for Hong Kong’s population of low-income persons whose numbers doubled between 1986 (N = 524,000) and 1998 (N = 1,093,000). These persons include many children, elderly and others with serious physical and emotional disabilities. The data reported in Table III also confirm that between 1986 and 1998: The percentage of children aged 1–14 years living in low income households increased by 75% from 13% to 23%; The number of child victims of physical, sexual, or emotional abuse doubled from fewer than 13 children per 100,000 to nearly 26 children per 100,000; The percentage of youth aged 15–19 years living in low income households increased 110% from 9% to 20%; Rates of unemployment for youth aged 10–19 years doubled from 10.3% to 20.5%; The prevalence of drug use among youth aged 15–19 years also doubled from 239 per 100,000 to 492 per 100,000; The number of youth arrested for having committed a violent crime increased by 8% from 371 per 100,000 to 402 per 100,000; The number of youths aged 10–19 years committing suicide increased by 68%, i.e., from 2.5 per 100,000 to 4.2 per 100,000; The percentage of women living in low income households increased 61% from 10% to 17%; and, The percentage of elderly living in low income households increased by 50% from 22% to 34%. All of the preceding trends should be regarded as being extremely serious; considered together, they affect a disproportionately large number of all the people living in Hong Kong.
DISCUSSION
The social development patterns uncovered for Hong Kong in this investigation are quite profound. Some offer reassurance that Hong Kong is on the “right track” with respect to achieving at least a portion of its social agenda – especially its goal of rapid
344
RICHARD J. ESTES
economic development. Other findings, though, are more disturbing – including those that find entire segments of the Hong Kong population being left out of the economic prosperity being enjoyed by others. Disparities in development are particularly pronounced among the SAR’s growing numbers of low-income households, its sizable population of youth in trouble, and those tens of thousands elderly residents who have not been able to achieve either financial or emotional security. Large numbers of Hong Kong’s children also are considerably under-developed relative to their peers in other well-off societies. Despite some important gains in the employment sector, social progress for many of Hong Kong’s women also is far from secure. And one of Hong Kong’s most important and traditional sources of stability and strength, its extended family system, is under serious assault. Based on the results obtained from this investigation, Hong Kong appears already to have developed into four distinct and unequal societies: (1) the “well off” whose social positions are largely immune from rapid shifts in the local economy; (2) the “socially secure” who occupy important positions of influence and are able to insulate themselves against all but the most extreme societal downturns; (3) the “socially insecure” who are employed . . .but are only “just making it”; and (4) a growing “underclass” of impoverished persons who possess neither the resources nor the possibility of extricating themselves from poverty. Many in the SAR’s social underclass are drawn from its new arrival community; others have lived in Hong Kong for decades . . . even longer. All are poor. Most are under-educated. A good many are ill. Some are drug dependent. Comparatively few of the SAR’s poor possess the educational or employment skills required to compete for other than the lowest paying jobs. A disproportionate number occupy the SAR’s infamous “caged homes;” some simply sleep on the streets. This study suggests that the harsh conditions that exist for many Hong Kong’s residents could swell the size of its underclass over the near-term.
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
345
FURTHER REFINEMENTS TO THE SDI-2000
As demonstrated by the preceding analysis, the SDI-2000 succeeded in achieving all of the goals associated with its development: (1) the index served as an effective tool for measuring changes in Hong Kong’s social development over time; (2) the index provided a basis for comparing Hong Kong’s social development accomplishments (and frustrations) both against its own past performances and in comparison with those of other societies located in and outside Asia; (3) the index succeeded in identifying the continuing social development needs of highly vulnerable population groups within Hong Kong; (4) use of the index sparked considerable public debate concerning the SAR’s current development priorities; and (5) the process of developing the index served as a vehicle for bringing together a broad spectrum of social, political and economic leaders in Hong Kong who shared a common goal of working toward further improvements in the quality of life for all the territory’s people. Even with these successes firmly in place, as a planning tool that seeks to impact strategic public policy decisions, the SDI-2000 can be further improved. More specifically, the SDI-2000 could be strengthened in the following ways: A “Rule of Law” subindex needs to be added to the SDI-2000. The following dimensions need to be included as part of the new subindex: (1) degree of independence of the judiciary; (2) changes in the nature and extent of constraints on attorneys in either the selection and/or prosecution of cases; (3) the extent to which legal precedents are used to refine existing civil and criminal codes; and (4) the extent to which Hong Kong’s local courts retain authority over all legal matters affecting the SAR. A “Subjective Quality of Life” subindex needs to be added to the SDI-2000. At a minimum, the subindex should seek to assess changes in: (1) people’s perceived sense of satisfaction with their lives; (2) the amount or level of “happiness” they experience; and (3) the extent to which they assess the social structure of Hong Kong to either contribute to or inhibit the pursuit of individual happiness. A “Social Service” subindex needs to be added to the SDI2000. This subindex should seek to: (1) identify new and
346
RICHARD J. ESTES
emerging social needs that exist in the SAR; (2) determine the nature and extent of these new needs; (3) assess the effectiveness of the existing service network in responding to these needs; and (4) give direction for how either the public or private sectors, or both, should respond in seeking to solve the development dilemmas posed by Hong Kong’s emerging social needs. The “Environmental Subindex” needs to be strengthened. In particular, better measures of air and water quality need to be added to the index. More sensitive measures of the extent to which solid waste and water is recycled also need to be incorporated into a revised subindex. Finally, one of the most difficult problems encountered in completing this project was the unavailability of politically sensitive data focused on Hong Kong’s changing political and legal structures. The absence of such data was compounded by the reality that that virtually all macroscopic data collection in Hong Kong is undertaken by one or another office of the Hong Kong government. The SAR’s several universities, think tanks and independent researchers do collect such data but rarely disseminate their findings either on a timely basis or via forms that are useful for shaping public policy debates. Future applications of the SDI-2000, in any case, would benefit from having more data sources from which to draw data in assessing changes in Hong Kong’s future development.
NOTES 1
Since its return to Chinese sovereignty on June 30, 1997, Hong Kong is referred to as “Hong Kong SAR” (Special Administrative Region). 2 Plus (+) and minus (–) signs are used to indicate the directional relationship of each indicator to social development, i.e., a higher number of “charitable institutions and trusts qualifying for tax exemption” is conceptualized as a indication of increased strength of the Civil Society sector which, in turn, is conceptualized as an indicator of a higher level of social development. 3 HHD refers to “high human development” societies – a designation assigned by the United Nations to countries characterized by high levels of social, political and economic development (UNDP, 1999: p. 134). In the UNDP model, Hong Kong ranks 24th out of 174 countries in overall level of human development.
TOWARD A SOCIAL DEVELOPMENT INDEX FOR HONG KONG
347
REFERENCE S
Community Chest of Hong Kong: 1999, Annual Report 1998/99 (Community Chest, Hong Kong). Estes, RJ.: 1988, Trends in World Social Development (Praeger, New York). Estes, R.J. : 1990, ‘Social development under different political and economic systems’, Social Development Issues 13(1), pp. 5–19. Estes, R.J.: 1996, ‘Social development trends in Asia, 1970–1994: The challenges of a new century’, Social Indicators Research 37(2), pp. 119–148. Estes, R.J. : 1997, ‘Social development trends in Europe, 1970–1994: Development prospects for the new Europe’, Social Indicators Research 42, pp. 1–19. Estes, R.J.: 1998, Trends in world social development, 1970–1995: Development prospects for a new century’, Journal of Developing Societies 14(1), pp. 11–39. Estes, R.J. : 1999, ‘Social development trends in the Middle East, 1970–1997: The search for modernity’, Social Indicators Research, pp. 1–31. Estes, R.J. : 2002, Social Development in Hong Kong: The Unfinished Agenda (Oxford University Press, London), in press. Hong Kong Commission on Strategic Development: 2000, Bringing the Vision to Life: Hong Kong’s Long-Term Development Needs and Goals (the Government of the Hong Kong Special Administrative Region, Hong Kong). Lau, S.K, M.K. Lee, P.S. Wan and S.L. Wong: 1999, Indicators of Social Development: Hong Kong, 1997 (Chinese University of Hong Kong, Institute of Asian-Pacific Studies, Hong Kong). Samaritan Befrienders: 1998, Annual Report by the Samaritan Befrienders H.K., 1997 (Samaritan Befrienders Hong Kong, Hong Kong). United Nations Development Programme (UNDP): 2000, Human Development Report, 2000 (Oxford University Press, New York). World Bank: 2000, World Development Report, 2000/2001 (Oxford University Press, New York).
University of Pennsylvania 3701 Locust Walk Philadelphia, Pennsylvania 19104-6214 E-mail:
[email protected] This page intentionally left blank
LARS OSBERG and ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
ABSTRACT. This paper develops a new measure of economic well-being for selected OECD countries for the period 1980 to 1996 and compares trends in this new Index to GDP per capita. We argue that the economic well-being of a society depends on the level of average consumption flows, aggregate accumulation of productive stocks, inequality in the distribution of individual incomes and insecurity in the anticipation of future income. However, the weights attached to each component will vary, depending on the values of different observers. We argue that public debate would be improved if there is explicit consideration of the aspects of economic well-being obscured by average income trends and if the weights attached to these aspects were made visible and were open for discussion.
Has economic well being increased or decreased in recent years? How would one know and why might it be useful to know? In 1980 Ronald Reagan asked the American people a seemingly simple question: “Are you better off today than you were four years ago?” Although U.S. per capita disposable real income was, in 1980, some 7.6 per cent higher than in 1976, his audiences answered “No!” Similarly, Bill Clinton in 1992 ran on the slogan “It’s the economy, stupid.” Both politicians were implying that their government would “do better”. Whether or not it is possible for a country to do better is an issue that is often resolved by pointing to how well other countries have done. International comparisons of trends in * An earlier version of this paper was presented at the annual meeting of the American Economic Association, January 5–7, 2000, Boston, Mass. In order that this paper be self-contained and provide a full explanation of the methodology used to estimate the index of economic well being, it draws on material from earlier papers which develop the index (Osberg, 1985; Osberg and Sharpe, 1998, 1999). These latter two papers are available on the CSLS website at www.csls.ca under indexes on economic well-being. We would like to thank Dmitry Kabrelyan for the extensive data development he did for this paper. Social Indicators Research 58: 349–382, 2002. © 2002 Kluwer Academic Publishers, Printed in the Netherlands.
350
LARS OSBERG AND ANDREW SHARPE
well being are thus crucial to knowing whether a country could “do better”. In modern democracies, national systems of social and economic statistics have become a crucial part of the informational feedback loop of public policy. By providing measures of social and economic outcomes, statistical agencies provide decision-makers and voters with the information that often defines the success or failure of public policies. Evidence on such successes or failures can be used to reallocate resources, or to replace governments; hence the calculation of measures of economic well being is an important issue. However, as the Ronald Reagan quotation above indicates, current measures – such as trends in per capita disposable income – may not necessarily be a good guide to popular perceptions of trends in economic well being. The core problem of statistical agencies is that of deciding what information to record and how to present it. Knowing that all statistics summarize a complex reality, and that there are wide variations among the public in which aspects of social reality are considered to be of greatest importance, statistical agencies still have to decide what to count, and what not to count, as part of a measure of economic well being. For many years, the System of National Accounts (SNA) has been the accounting framework within which most discussions of trends in economic well- being have been conducted, and Gross Domestic Product (GDP) per capita has been an often used summary measure of economic trends. The compilers of the national accounts have often protested that their attempt to measure the aggregate value of marketed economic output was never intended as a full measure of economic well being. Nevertheless, it has often been used as such, and the GDP accounting exercise has attracted a great deal of criticism as being a misleading indicator of economic well-being (e.g. Waring, 1988). Dissatisfaction with the GDP as a measure has led to a number of proposals for substitute measures (e.g. the Genuine Progress Indicator). Summarizing the economic well being of a complex society inevitably requires a series of ethical and statistical judgements. There are many different dimensions to well being, which are valued to different degrees by different observers. With a single index
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
351
number it may be difficult to disentangle the relative importance of value judgements in the construction of the index. Furthermore, in thinking about the appropriate public policy response, it is not particularly useful to know only that well being has gone “up” or “down”, without also knowing which aspect of well being has improved or deteriorated. This paper sees construction of measures of economic well being as a problem in the optimal aggregation of information. If the objective is to improve the quality of public decision making and political debate, excess aggregation is not helpful, because it does not enable value judgements and statistical judgements to be separated. Furthermore, excess aggregation offers no guide to policy priorities. Osberg (1985) therefore proposed that an index of economic well-being should be based on indices of consumption, accumulation, inequality and insecurity, with the explicit recognition that the weights attached to each component will vary, depending on the values of different observers. The underlying hypothesis is that public debate is likely to be improved if issues of fact, analysis and values are as clearly separated as possible. Measurement of the current level, or trend, of economic well being can be seen as the first stage of a three stage discussion in which a society asks: (1) Where are we? (2) Do we want to go somewhere else? (3) How do we get there? Issues of measurement, of values and of analysis may be conceptually distinct, but in a single index of economic well being, they often become hopelessly entangled. If the democratic debate on economic policy is to be fruitful, it would seem desirable to separate issues of measurement (question 1) from the debate on ends (issue 2) or the discussion of means (item 3). If the discussion is organized in this way, those people who fundamentally care most about a particular aspect of well being can discuss the facts about that aspect of well being, without confusing the discussion with other issues. Such discussions of measurement issues are of a fundamentally different nature from discussions of values – which aspect of economic well being should receive greatest weight.
352
LARS OSBERG AND ANDREW SHARPE
This basic framework – that a society’s well-being depends on societal consumption and accumulation and on the individual inequality and insecurity that surround the distribution of macro economic aggregates – is consistent with a variety of theoretical perspectives. We therefore avoid a specific, formal model. As part of a larger project on the state of living of standards and the quality of life in Canada, the Centre for the Study of Living Standards (CSLS) has constructed the index of economic well-being proposed by Osberg (1985) for Canada (Osberg and Sharpe, 1998) and for all provinces and for the United States (Osberg and Sharpe, 1999). This paper now extends the index to other countries, with a base year of 1980. Data availability and comparability problems mean that the construction of an international index of economic well being is a much more arduous undertaking than construction of an index at the national level, or even for two similar countries such as Canada and the United States. A crucial data requirement for the index is comparable estimates of poverty rates and poverty gaps for all members of society, the elderly, and single-parent families. Comparable estimates require comparable micro-data files and the only international source of such files is the Luxembourg Income Study (LIS). Consequently, the 14 countries covered by the LIS (Australia, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, and the United States) represent the maximum number of countries for which estimates of the index of economic well being can be constructed. Because the number of years of micro-data tapes available for many of these countries is limited, the construction of a reliable time series for all 14 countries is not possible. Only some countries (Canada, the United States, the United Kingdom, Australia, Germany, Norway and Sweden) have a large enough number of public-use micro-data files accessible from the LIS for construction of reliable long-run time series. Consequently most of the discussion in the paper focuses on trends in these seven countries. Appendix B found in the website version of this paper presents tables on seven other countries (Belgium, Denmark, Finland, Italy, Netherlands, France and Spain) in which the data is largely based on backward
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
353
and forward extrapolation and interpolation techniques (extrapolated and interpolated estimates are given in Italics). The data in both the main body of the text and Appendix A of the unabridged version of the paper are derived from underlying raw data which are presented, for all 14 countries in tables available on the Web at www.csls.ca. We hope that over time, it will be possible to “fill in the blanks” and that more reliable estimates for more countries will become available.1
AN INDEX OF ECONOMIC WELL-BEING
GDP is a measure of the aggregate marketed income of a society and most of its proposed substitutes (such as the GPI) are also primarily measures of adjusted average annual “income” flows [where the adjustments are meant to capture issues (such as environmental degradation) that GDP now ignores]. However, “income” is a flow variable that does not directly consider the aggregate value of the bequest which this generation will leave to its descendants. Although those now alive clearly care about the level of their own consumption, they also care (in varying degrees) about the well being of future generations. Furthermore, although trends in average income are important, individuals are justifiably concerned about the degree to which they personally will share in the prosperity of the average, and the degree to which their personal economic future is secure. The four components or dimensions of economic well being in the ideal index of economic well being are, therefore: effective per capita consumption flows net societal accumulation of stocks of productive resources income distribution – poverty and inequality, economic insecurity, We recognize that the System of National Accounts has, thanks to many years of development effort by international agencies, produced an accounting system for GDP which is rigorously standardized across countries. Internationally comparable statistics on other dimensions of economic well being are far less complete. However, using GDP per capita as a measure of well being such would implicitly set the weight of income distribution or economic
354
LARS OSBERG AND ANDREW SHARPE
insecurity to zero, by ignoring entirely their influence, which is surely a bad approximation of their actual importance to economic well-being. Average Consumption Flows Marketed personal consumption The starting point for this component of the index is aggregate real personal consumption per capita. Estimates of personal consumption per capita, expressed in national currency units, and in constant prices (base years differ among countries) have been taken from the OECD National Accounts publication. All countries experienced increases in real per capita marketed personal consumption over the 1971–96 period, but there was a large variations in the increase, ranging from a high of 100.2 per cent in Germany to a low of 26.4 per cent in Sweden. The System of National Accounts provides a strong basis for estimating the consumption of marketed goods and the cost of providing government services. Estimates are more imprecise when one considers the value of a number of other factors that also influence consumption flows, such as life expectancy, economies of scale in household consumption, and working time.
The value of increased longevity. Life expectancy has increased significantly in recent years, and we have every reason to believe that having a long life is an important component of the well being. The economic value of these extra years of life should be included in the total consumption flows of individuals, since presumably people care both about how much they consume per year, and how many years they get to consume it. Between 1971 and 1996, all 14 countries enjoyed increased life expectancy. For all years after 1971, personal consumption per capita is adjusted upward by the increase in life expectancy relative to 1971. Reduced economies of scale in household consumption. When individuals cohabit in households, they benefit from economies of scale in household consumption. When comparing the average effective consumption of individuals over time, the implication is that as households have shrunk in average size, economies of scale have
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
355
been lost. Trends in average per capita consumption should therefore be adjusted for the average loss over time of economies of scale in household consumption. Since economies of scale diminish in family size, the extent of change in economies of scale depends on where change occurs in the distribution of family sizes. The “LIS” equivalence scale (i.e. the square root of family size) has been applied to average family income to construct an index of equivalent family income (1981 = 100), which is used to adjust personal consumption per capita. The size of the downward adjustments to real per capita personal consumption are small, with the United Kingdom having the largest downward adjustment in 1996 relative to 1971, (10.8 per cent). Differences in working hours and economic well-being. In developing the Index of Economic Well-being for Canada, Osberg and Sharpe (1998) recognized the importance of leisure for economic well-being and hence the potential impact of changes in working time on economic well-being. However, in Canadian data, the absence of any significant trend in average annual hours of work meant that inclusion of the aggregate value of leisure in the index would have had no major effect on the overall trend of the index. This paper extends the Index of Economic Well-being to OECD countries among which there are major differences in both the initial level and trends over time in the average annual number of hours worked. The methodology for integrating variation in working hours into an international index of economic well being is developed in the unabridged version of the paper. Government services The provision of non-marketed or heavily subsidized services by the government is part of the consumption flow. Current expenditure by all levels of government including defense and capital consumption allowances, but excluding debt service charges and transfer payments (which influence marketed consumption) are used. These data were taken from the OECD national accounts, expressed in constant prices in national currency units. The importance of government final consumption expenditures relative to personal adjusted consumption expenditures differs markedly among OECD countries. In addition, over the 1971–96 period there were major
356
LARS OSBERG AND ANDREW SHARPE
differences in the rate of growth of real per capita government final consumption expenditures, ranging from a high of 172.3 per cent in Spain (which admittedly had a low base) to a low of 16.4 per cent for the United States. Total consumption flows Total per capita consumption is defined as the sum of personal consumption (adjusted for changes in average household size and life expectancy), government services, and the adjusted relative value of leisure.2 Between 1980 and 1996 (since leisure adjustments are relative to 1980, this year will be used as the base), the increase in the real per capita total consumption flows has ranged from a high of 65.6 per cent in Germany to a low of 11.7 per cent in Sweden (Table I). Wealth Stocks, Sustainability and the Intergenerational Bequest
In our view, measurement of trends in well being should include consideration of changes in the well being of generations yet unborn. This consideration of future generations can be justified either on the grounds that those now living care about the well being of future generations or on the grounds that a concept of “society” should include both present and future generations. Either way, wealth accumulation by this generation will increase the bequest left to future generations, and is an important component of well being. We would emphasize that this component of economic well being consists of those stocks of real productive assets that can generate real income for future generations – not the financial instruments that will determine the allocation of the return from those assets. The stocks of “wealth” left to the next generation, broadly conceived to include environmental and human resources as well as physical capital stock, will determine whether a society is on a long-run sustainable trajectory. Physical capital stock The physical capital stock includes both residential structures and machinery and equipment and non-residential, and both the business sector and the government sector. Between 1980 and 1996, the
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
357
increase in the fixed capital stock, on a per capita basis, ranged from a high of per cent 40.3 per cent in Italy to a low of 9.3 per cent in Denmark. Research and development capital stock Closely related to the physical capital stock is the concept of the research and development (R&D) capital stock. In an era of rapid technological change, expenditure on R&D is a crucial ingredient in the ability of society to innovate and create wealth. Statistical agencies do not produce R&D stock data. The Center for the Study of Living Standards has constructed a R&D stock series for 11 countries from OECD data on annual flows of total business enterprise expenditure on research and development. Between 1980 and 1996, the rate of change in the per capita real business enterprise R&D stock for business enterprises ranged from an increase of 233 per cent in Australia to a decline of 11 per cent in Germany. Value of natural resource stocks Current consumption levels could be increased by running down stocks of non-renewable natural resources or by exploiting renewable resources in a non-sustainable manner, but this would be at the cost of the consumption of future generations. A key aspect of the wealth accumulation component of economic well being is net changes in the value of natural resources. Unfortunately, there are currently no internationally comparable time series estimates of the value of natural resources for OECD countries, so this component of wealth stocks has not been included in the index of economic well being developed in this paper. Stocks of human capital The human capital accumulated by the workforce generates both current and future income. Trends in the stock of human capital, including both formal educational attainment levels and on-thejob training, are important determinants of current and future economic well being. This same approach has been followed for the OECD countries. OECD data on the educational attainment of the 25–64 population and expenditure per student (available in both local currency and U.S. dollars) for the early childhood, primary,
358
LARS OSBERG AND ANDREW SHARPE
secondary, non-university tertiary and university level education have been used to estimate the per capita stock of human capital.3 Net foreign indebtedness Since interest payments on the net foreign indebtedness of citizens of one country to residents of other countries will lower the aggregate future consumption options of those citizens, increases in the level of foreign indebtedness reduce economic well-being within a given country. These estimates have been converted to current price national currencies at market exchange rates and then deflated by the GDP deflator and adjusted for population to obtain real per capita estimates in the net international investment position, expressed in national currency units. State of the environment Like the excess depletion of natural resources, current consumption can be increased at the expense of the degradation of the environment, reducing the economic well being of future generations. Consequently, changes in the level of air and water pollution should be considered an important aspect of the wealth accumulation. Countries pass on from generation to generation both a natural and made-made national heritage. If this heritage were damaged, the economic well being of future generations would be reduced. Since it is very difficult, if not impossible, to put a monetary value on, for example, the pristine condition of national parks, or historic buildings, there will be no attempt to set an aggregate value to these assets. However, the issue of trends in well being is the change in such assets, which is easier to measure and for which indicators of environmental quality can be developed. Probably the best-known environmental change is global warning arising from increased emissions of greenhouse gases, the most common of which is carbon dioxide emissions. Fortunately, data are available on these emissions and it is possible to estimate the costs of these emissions.4 These costs can then be subtracted from the stock of wealth to obtain an environmentally adjusted stock of wealth.
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
359
Estimates of total wealth As the estimates of the physical capital stock, the R&D capital stock, net foreign debt, and environmental degradation are expressed in value terms, they can be aggregated and presented on a per capita basis. Net foreign debt per capita is a negative entry, while the social costs of emissions are subtracted from the stocks of wealth. For the 1980–96 period, estimates for the five components of the wealth stock included in this paper are available for seven countries. The rate of change for per capita real wealth stocks in national currency at constant prices for these countries ranged from an increase of 44.3 per cent in Norway to 15.6 per cent in Sweden. Inequality and Poverty The idea of a “Social Welfare Function” which is a positive function of average incomes and a negative function of the inequality of incomes has a long tradition in welfare economics. Since the economic well-being of the population is affected by inequality in the distribution of income and by the extent of poverty, there are two issues: (1) one’s perspective on the importance of inequality/poverty compared to trends in average income, and (2) one’s view of the relative weight to be placed on poverty compared to inequality. We therefore suggest that a compound sub-index to recognize explicitly these issues would place some weight on a measure of inequality in the aggregate distribution of income and some weight on a measure of poverty. The most popular measure of inequality in the distribution of income is undoubtedly the Gini index. For the purposes of the construction of the index of economic well being, we have chosen the Gini coefficient of after-tax household income. Recently, Osberg and Xu (1997) have noted that the SenShorrocks-Thon measure of poverty intensity is both theoretically attractive as a measure of poverty, and also convenient, since it can be decomposed as the product of the poverty rate, the average poverty gap ratio and the inequality of poverty gap ratios.5 The poverty rate is the proportion of persons who fall below the poverty line, defined here as one half the median equivalent after-tax family income. The poverty gap ratio is defined as the percentage gap
360
LARS OSBERG AND ANDREW SHARPE
between the poverty line and the income of those below the poverty line. The poverty rate varies greatly among the 14 countries for which LIS data are available. For the most recent year for which micro-data tapes are available for each country, it ranged from a high of 18.0 per cent in the United States to a low of 5.0 per cent in Belgium. There was much less variation across countries in the average poverty gap ratio. For the most recent year for which micro-data tapes are available for each country, it ranged from a high of 38.9 per cent in Denmark to a low of 22.4 per cent in Belgium. The overall index of equality is a weighted average of the indices of poverty intensity for all units or households and the Gini coefficient, with the weights 0.75 and 0.25 respectively.
Insecurity If individuals knew their own economic futures with certainty, their welfare would depend only on their actual incomes over their lifetimes, since there would be no reason to feel anxiety about the future. However, uncertainty about the future will decrease the economic welfare of risk averse individuals. Individuals can try to avoid risk through social and private insurance, but such mechanisms do not completely eliminate economic anxieties. Given the value that individuals place on economic security, any increase in insecurity reduces economic well being. Over fifty years ago, the United Nations’ Universal Declaration of Human Rights declared: Everyone has the right to a standard of living adequate for the health and wellbeing of himself and of his family, including food, clothing, housing and medical care and necessary social services, and the right to security in the event of unemployment, sickness, disability, widowhood, old age or other loss of livelihood in circumstances beyond his control. [Article 25]6
For this paper, we construct measures of the percentage change over time in the economic risks associated with unemployment, illness, “widowhood” (or single female parenthood) and old age. In each case, we model the risk of an economic loss associated with the event as a conditional probability, which can itself be represented as the product of a number of underlying probabilities. We weight the
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
361
prevalence of the underlying risk by the proportion of the population that it affects. The core hypothesis underlying the measure of economic insecurity proposed here is that changes in the subjective level of anxiety about a lack of economic safety are proportionate to changes in objective risk. Unemployment The economic risk associated with unemployment can be modeled as the product of the risk of unemployment in the population and the extent to which people are protected from the income risks of unemployment. We have taken as a proxy for the risk of unemployment changes in the employment rate (employment/population ratio). Changes in this ratio reflect changes in the unemployment rate and changes in the participation rate (both cyclical and structural). The extent to which people have been protected by unemployment insurance (UI) from the financial impacts of unemployment can be modeled as the product of: (1) the percentage of the unemployed who claim regular UI benefits and (2) the percentage of average weekly wages replaced by UI. Unfortunately, internationally comparable data on these two variables, particularly the first, have proven very difficult to obtain and are not included in this version of the paper as they were in earlier papers on Canada and the United States. Instead, an unpublished OECD series on the gross replacement rate for the unemployed has been used in the calculation of the risk of unemployment. Illness Viewed from a longer-term perspective, the economic insecurities associated with illness in developed economies certainly dropped considerably with the introduction of universal health insurance in many countries. Since our other data series are often difficult to obtain prior to 1971, the period covered by the present study is 1971 to 1996. This period unfortunately largely omits the improvement in economic well being that universal health care represented. Nevertheless, it is still of interest to examine how the economic insecurities associated with illness have evolved over the last quarter century. We focus on the risk of large out of pocket health care costs, with the risk directly proportional to the share of private medical
362
LARS OSBERG AND ANDREW SHARPE
care expenses in disposable income. Medical expenses as a proportion of disposable income in 1996 ranged from a high of 8.8 per cent in the United States to a low of 1.3 per cent in the United Kingdom. Single parent poverty When the UN Universal Declaration of Human Rights was drafted in 1948, the percentage of single parent families was relatively high in many countries, partly as a result of World War II. At that point in time, “widowhood” was the primary way in which women and children lost access to male earnings. Since then, divorce and separation have become the primary origins of single parent families. However, it remains true that many women and children are “one man away from poverty”, since the prevalence of poverty among single parent families is extremely high. To model trends in this aspect of economic insecurity, we multiply (the probability of divorce) * (the poverty rate among single female parent families) * (the average poverty gap ratio among single female parent families). [The product of these last two variables is proportional to the intensity of poverty.] The annual divorce rate ranged in 1996 (or the most recent year before 1996 for which data are available) from a high of 4.33 per cent of legally married couples in the United States to a low of 0.47 per cent in Italy. The poverty rate for single female parents in the most recent year (in brackets) for LIS micro-files ranged from a high of 44.0 per cent (1997) in the United States to a low of 2.8 per cent (1992) in Sweden. The average poverty gap ratio for single female parents in the most recent year (same year as for the poverty rate above) for LIS files ranged from a high of 41.6 per cent in Norway to a low of 18.2 per cent in Sweden. Poverty in old age Since income in old age is the result of a lifelong series of events and decisions, which we cannot hope to disentangle in this paper, we model the idea of “insecurity in old age” as the chance that an elderly person will be poor, and the average depth of that poverty. The poverty rate for the elderly in the most recent year (in brackets) for LIS micro-data files ranged from a high of 33.1 per cent (1994)
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
363
in Australia to a low of 2.6 per cent (1991) in the Netherlands. The average poverty gap ratio for the elderly in the most recent year for the LIS micro-data files ranged from a high of 48.7 per cent in Denmark to a low of 9.3 per cent in Norway. Overall Index of Economic Security The four risks discussed above have been aggregated into an index of economic security. The aggregation weights are the relative importance of the four groups in the population. For unemployment, the proportion of the 15–64 population in the total population. For illness, the proportion of the population at risk of illness, which is 100 per cent. For single parent poverty, the proportion of the population comprised of married women with children under 18. For old age poverty, the proportion of the population in immediate risk of poverty in old age, defined as the proportion of the 45–64 population in the total population. Estimates of Trends in the Overall Index of Economic Well-being Weighting of components Trends in any index are determined by the choice of variables that are included in the index, the trends in those variables and the weights these variables receive. Since the four main dimensions of average consumption, intergenerational bequest, inequality/poverty and insecurity are separately identified, it is easy to conduct sensitivity analyses of the impact on perceived overall trends of different weighting of these dimensions. For discussion purposes, our “standard” weighting gives consumption flows a weight of 0.4, wealth stocks a weight of 0.1, and equality and economic security have each been given weights of 0.25. As the sub-components of the consumption flows and wealth stocks are expressed in dollars, there is no need for explicit weighting. Their dollar values represent implicit weights. In terms of the inequality/poverty subcomponents, a somewhat Rawlsian perspective would assign greater importance to poverty than to overall inequality trends. On this basis,7 and a weight of 0.1877
364
LARS OSBERG AND ANDREW SHARPE
or (= 0.25 * 0.75) has therefore been given poverty intensity and 0.0625 (= 0.25 * 0.25) to the Gini coefficient. In other words, poverty is given three times the weight of inequality. The subcomponents of the economic security index are weighted by the relative importance of the specific population at risk in the total population. Table I shows the indexes for all four components of the index of economic well being and the overall index. To put all the subcomponents to a common base of 1, the constant 2 has been added to the index of inequality to convert it to an index where an increase corresponds to a rise in economic well being. Trends in the overall index of economic well-being We are acutely conscious that the data sources available to us are far from what we would like. We know that restricting ourselves to internationally comparable data series has meant that we have neglected issues (such as the decline in UI coverage in Canada) which are important for some countries. We also know the reliance on interpolation between the data points available in the Luxembourg Income Study implies, necessarily, that we cannot detect year to year fluctuations in some components of our index. However, we hope that enough data remains to give a preliminary indication of trends in economic well being, from a broader perspective than that provided by GDP accounting. Since we want to examine the sensitivity of a measure of economic well being to alternative possible weightings of accumulation, income distribution and insecurity, Figures 1 to 8 present both our “standard” and an “alternative” weighting. The “alternative” is much more heavily weighted to average consumption (0.7), has the same weighting on accumulation (0.1) and less heavily emphasizes income distribution (0.1) and insecurity (0.1). For each country, we compare trends in the “standard” and “alternative” indices with trends in GDP per capita. For all countries, consideration of bequest, inequality/poverty and insecurity reduces the measured rate of growth of economic well being, compared to use of the GDP per capita index. Generally, the more heavily current average consumption is emphasized, the closer our index comes to GDP per capita. However, in every instance the consideration of a wider range of issues than those recognized in
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
365
GDP accounting reduces the measured increase in economic well being. In some countries, the change in perception of trends in well being that a broader measure produces is striking. In the United States, GDP per capita increased by approximately 30% over the 1980 to 1997 period, but our “standard” index is essentially flat, with a total increase of 3% over the period. In the UK, increases in per capita GDP were even larger (39.8%), but our “standard” weighting (which has a heavy emphasis on economic inequality and insecurity) shows a decline of about 10%. Both the US and UK have been marked by a substantial increase in economic inequality over this
366
LARS OSBERG AND ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
367
368
LARS OSBERG AND ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
369
period and increases in money income have been limited to the top end of the income distribution. As well, increases in money income in the US have been obtained at the cost of substantial increases in working hours. Hence, this is not an unreasonable finding. For the UK and Sweden, GDP per capita rose, while our “standard” index of economic well being declined. In both cases, however, this qualitative result is quite sensitive to the relative weighting of current consumption compared to distribution and insecurity – the “alternative” index does not actually decline8 (although it is almost flat in Swedish data). Norway is a country where trends in economic well being are, more or less, scaled down versions of the trend in GDP per capita. In this case, our current estimates of trends in the Index of Economic Well Being could be said to provide relatively little “value added”, compared to trends in GDP per capita, since each index moves in much the same way over time (albeit showing much stronger growth in GDP per capita than in economic well being). However, Australia and Canada – whose economies share a relative dependence on raw materials production – are noteworthy
370
LARS OSBERG AND ANDREW SHARPE
in showing a greater cyclical sensitivity in GDP per capita than one finds in either measure of economic well being, or in GDP per capita in other countries. In Canada and Australia, the recessions of both the early 1980s and early 1990s show up clearly in per capita GDP fluctuations – to a much greater degree than Germany or Norway (the early 1980s recession is hard to find in UK or Swedish GDP per capita data). However, in both countries the trend in economic well being indices is much smoother, because changes in current income can be much more rapid than changes in wealth stocks, income distribution and insecurity. Canadian trends in economic well being are also quite similar for “standard” and “alternative” weightings of the index.9 Level comparisons of economic well-being Comparisons of the level of well being across countries are inherently much more problematic than comparisons of the trends in various components of economic well being within countries. In across country comparisons, the institutional context of economic data differs to a far greater extent than in within country, over time comparisons. Calculations of purchasing power parity equivalence across several countries have greater uncertainty than comparisons of within country consumer price levels. Statistical agencies in different countries differ in their data availability and data gathering practices to a greater degree than they change those practices over time in the same country. For all these reasons, this paper avoids direct commentary on comparative levels of economic well being. Conclusion
This paper has developed an index of economic well-being based on four dimensions or components of economic well-being for selected OECD countries, with the weight given each component in bracketsconsumption flows, stocks of wealth including physical capital and natural resources, income distribution, and economic security. We argue that providing explicit weights of these components of well being is important in enabling other observers to assess whether, by their values of what is important in economic wellbeing, they would agree with this assessment of trends in the
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
371
modern economy. Some events – like a major recession – may have adverse impacts on all four dimensions of well being, producing lower average consumption, more inequality, more insecurity and less accumulation of capital for the benefit of future generations. In such a case, values as to the relative weights to be assigned to the components of well being are of secondary concern. However, in other instances (such as environmental policy concerning global warming) the relative weights assigned to different dimensions of well being may be crucial. A major reason for being explicit about the weights to be assigned to dimensions of well being is to be clear about when there is, and when there is not, a conflict of fundamental values in the assessment of social trends. In general, however, a key finding of this paper is that economic well-being, for at least two different sets of relative weights, has increased at a much slower rate over the last 25 years than real GDP per capita, a widely-used indicator of economic well-being.
NOTES 1
Estimates of the index of economic well-being for Canada and the United States in this paper differ somewhat from our earlier estimates because the emphasis on international comparability of estimates has meant that some data originally used to construct the index for Canada and the United States, and not available for other countries, has not been used. Examples include estimates of the underground economy, commuting, human capital, and natural resources. These omissions can affect perceived trends. In a sense, the index of economic well being presented in this paper is less sophisticated than our earlier estimates – an unfortunate tradeoff needed if more countries were to be included. 2 Unpaid work contributes to economic welfare and thus should also in principle be included in the total consumption flow component of an index of economic well being. Because of the unavailability of internationally comparable estimates of the value of unpaid work for countries other than Canada, this component has not been incorporated into the index of economic well being for OECD countries developed in this paper. 3 In the valuation of human capital across countries, a choice must be made whether to use the local cost per student of education at the various levels (and then convert to U.S,. dollars using purchasing power parities) or whether to apply the cost of education in the United States to all countries. If skills derived from educational attainment, like natural resources, are a marketable commodity because of international migration, foreign investment, and international trade,
372
LARS OSBERG AND ANDREW SHARPE
then one can argue they should be subject to the law of one price and that their value is the cost of production in the country receiving the largest migratory inflow, that is the United States. This is the approach adopted in this paper. 4 Fankhauser (1995) has estimated that the globalized social costs of emissions (with no adjustment for different national costs) at $20 US per ton in 1990. World Bank researchers (Atkinson et al, 1997) have applied this number to emissions in developed countries to estimate the value of the loss of environmental services as a proportion of output and the measure of genuine saving. According to data from the International Energy Agency, world emissions in 1997 were 22,636 millions of metric tons. Based on the $20.00 U.S. per ton cost of emissions, the world social costs of emissions was $452,720 million. This amount was allocated on the basis on a country’s share of nominal world GDP, expressed in U.S. dollars. It was then converted into national currency at the purchasing power parity exchange rate and divided by population. As these costs represent a loss in the value of the services provided by the environment, they can be considered a deduction from the total stock of wealth of the society. For example, in 1997, per capita stocks of wealth in Canada were reduced by $415 Canadian per cent because of the social costs imposed by emissions according to this methodology. 5 Furthermore, since the inequality of poverty gap ratios is essentially constant, poverty intensity can be approximated as twice the product of the poverty rate and the average poverty gap ratio. 6 In the 1990s, the gender specificity of the language of 1948 will strike many people as odd – but Article 2 makes it clear that all Articles of the Universal Declaration of Human Rights are to be guaranteed to male and female persons equally. 7 A pure Rawlsian would put all the weight on the well-being of the least well off. 8 As well, we would caution that because we have not been able to get, for this paper, estimates of the income replacement provided under unemployment insurance in these countries, we may be over estimating the importance for economic insecurity of the rise in unemployment in these countries. 9 But this paper does not capture the rise in economic insecurity produced by declining UI coverage.
REFERENCES
Cobb, C., T. Halstead and J. Rowe: 1995, The Genuine Progress Indicator: Summary of Data and Methodology (Redefining Progress, San Francisco). Fankhauser, S.: 1995, ‘Evaluating the social costs of greenhouse gas emissions’, Energy Journal 15, pp. 157–184. Osberg, L.: 1985, ‘The measurement of economic welfare’, in D. Laidler (coordinator), Approaches to Economic Weil-Being, volume 26 of the research
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
373
program of the Royal Commission of the Economic Union and Development Prospects for Canada (MacDonald Commission) (University of Toronto Press, Toronto). Osberg, L. and K. Xu: 1997, ‘International comparisons of poverty intensity: Index decomposition and bootstrap inference’, Working Paper 97-03, Department of Economics, Dalhousie University and Working Paper Number 152, Luxembourg Income Studies; forthcoming Journal of Human Resources (2000). Osberg, L. and A. Sharpe: 1998, ‘An index of economic well-being for Canada’, paper presented at the CSLS Conference on the State of Living Standards and Quality of Life in Canada, October 30–31, Ottawa, Ontario, (paper posted at www.csls.ca under conferences), Research Paper R-99-3E, Applied Research Branch, Human Resources Development Canada, December. Osberg, L. and A. Sharpe: 1999, ‘An index of economic well-being for Canada and the United States’, paper presented to the annual meeting of the American Economics Association, January 3–5, New York, New York. Waring, M.: 1988, Counting for Nothing: What Men Value and What Women Are Worth (Allen and Unwin, Wellington, New Zealand).
Department of Economics Dalhousie University 6214 University Avenue Halifax, Nova Scotia B3H 3J5 E-mail:
[email protected] Center for the Study of Living Standards 111 Sparks Street, Suite 500 Ottawa, Ontario K1P 5B5 E-mail:
[email protected] Lars Osberg
Andrew Sharpe
374
LARS OSBERG AND ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
375
376
LARS OSBERG AND ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
377
378
LARS OSBERG AND ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
379
380
LARS OSBERG AND ANDREW SHARPE
INTERNATIONAL COMPARISONS OF TRENDS IN ECONOMIC WELL-BEING
381
382
LARS OSBERG AND ANDREW SHARPE
MICHAEL R. HAGERTY
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS: REVIEW OF 13 OECD COUNTRIES
ABSTRACT. I review studies of national elections from 13 OECD countries over 30 years to examine the effect of declining or increasing Quality of Life (QOL) on citizens’ voting behavior. The results are consistent with the theory that citizens hold the incumbent party responsible for increases or decreases in QOL during their term in office, and vote against an incumbent party that fails to deliver improved QOL. Studies show that this “responsibility hypothesis” holds not only for economic variables such as GDP/capita, but for murder rates, violent crime rates, nutrition, and inequality. The platform and qualification of candidates also has an effect on voting, but the macro-variables composing QOL occur prior to and act to shape the platform and strategy of candidates. These results hold important implications for constructing indexes of QOL, because researchers can use national voting behavior to recover the weights that voters use in combining domains of QOL. By using these weights from the ‘representative voter’, researchers can create a unified index of QOL for social reports.
1. INTRODUCTION
National governments have long been exhorted to improve Quality of Life (QOL) for their citizens. “To promote the general welfare is specifically stated in the preamble to the U.S. constitution. This exhortation is praiseworthy, but most nations’ constitutions fail to specify enforcement mechanisms that can monitor QOL and penalize governing parties that fail to improve QOL. Are the OECD governments motivated to improve QOL? The research I review here reveals that voters successfully provide the enforcement mechanism – voters hold the incumbent party responsible for declining QOL, and vote out of office parties that fail to improve QOL. I will review evidence that this is true for many elements of QOL, including GDP/capita, nutrition, crime rates, unemployment, inflation, and inequality. Social Indicators Research 58: 383–402, 2002. © 2002 Kluwer Academic Publishers, Printed in the Netherlands.
384
MICHAEL R. HAGERTY
This research has important implications not only for governments, but for researchers, because it reveals the implicit weights that voters use in judging Quality of Life. For example, how do voters trade off an increase in income with an increase in the crime rate? In this paper, Section 2 reviews the political and economic theory for why election results should be influenced by changes in QOL. Section 3 describes the studies, and Section 4 reviews the accumulated evidence for this theory. Section 5 concludes with implications for proper weighting of QOL indexes.
2. THE THEORY OF WHY QOL INFLUENCES ELECTIONS
Economists and many political scientists take the view that voters in democracies attempt to maximize their “utility” or Quality of Life by (1) electing politicians who are most likely to increase their QOL, and (2) voting out of office politicians who failed to increase their QOL. Strongly stated, it assumes that individuals maximize only their own personal utility, but the same results hold if one relaxes this assumption to allow individuals to vote altruistically, taking others in their community into account. I adopt this point of view to track voters making such choices, to see whether and how they trade off various domains of QOL. Figure 1 provides an overview of this research. The variables in the left-hand column are objective inputs to the domains of QOL, and the variables in the right hand columns are the output or perceived domains of QOL. Objective variables in the input column affect citizens’ perceived QOL in the output column. For example, the first three variables on the left are GDP/capita, inflation, and income inequality, which affect the output domain, perceived “material well-being”. In a similar way, the number of calories available per capita is a determinant of the health domain. The variables murder rate and violent crime rate contribute to Personal Safety, unemployment to Work domain, and divorce to the Family and Friends domain. In turn these domains contribute to citizens’ overall evaluation of their Subjective WellBeing (SWB). This initiates the feedback cycle, where citizens compare their actual SWB to their SWB goal. To the extent that they have not achieved their goals, they attempt to change both their indi-
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
385
vidual choices (work harder, get more education, avoid unhealthy habits, etc.) They also attempt to change government policies that may have prevented them from achieving their target SWB, and will vote out of office a party that fails to achieve their QOL goals. In this paper, I concentrate on the feedback relation between the input variables and citizens’ voting behavior for new governments. I do not directly measure citizens’ perceived satisfaction (output), though many studies have done so through surveys and public opinion polling (Veenhoven, 1995; Hagerty, 1999). In this paper I make the simplest possible assumption that declining inputs cause declining outputs (satisfaction), and that the feedback loop will then cause citizens to modify their public choice in the form of voting for different government policies that will improve their QOL. Although a full model would also examine changes in people’s measured SWB and personal choices over time, it is easier to observe citizens’ public choices for governments because these votes are already carefully validated and preserved as a matter of public record.
386
MICHAEL R. HAGERTY
Why do we focus on QOL changes, in spite of the fact that most media and voters focus on the platform and qualification of candidates? We argue that QOL changes occur prior to party’s selection of candidates and platforms, and that these effects are (partly) caused by the changes in the macro-variable QOL. Figure 2 presents more detail on the dynamics of the voting decision. It shows a time line and the various factors that citizens might take into account when making their voting decision. It begins at the top in the distant past, and ends at the bottom in the future. Factors occurring in the recent past that will influence their decision prior to election day are: the campaign itself and the promises (platform) that each party makes, the experience and quality of each candidate, any possible scandals that happened during the current government, and
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
387
purely ‘exogenous events’, such as a global energy shortage. We also list (near the top) events in the distant past that may affect voting decisions, such as voting rules of the country, and party identification of the voter. The estimation model will control for these influences. The estimation model in the studies reviewed ignore the platform of the candidates, and concentrate instead on the actual QOL changes during the incumbent party’s term. To the extent that the platforms add to variance explained, the of the models will suffer. However, the omission of the platforms should not bias the estimation of the net effects of QOL, since QOL changes occur prior to and help to influence the platforms. Furthermore, there are good arguments for why voters should ignore candidates’ platforms. The first is put forward by Kramer (1971) and by Fair (1978), who assume that voters try to predict how each party will affect their QOL, then vote for the party that will most improve their QOL. Their model assumes that the best prediction of a party’s future performance is the party’s past performance while in office. Hence the party with the best recent record is voted in, regardless of stated platform. Could voters do better by considering party platforms, or unexpected events beyond the incumbent’s control (such as an oil embargo caused by external circumstances)? Downs and Rocke (1995) argue that voters are better off by ignoring this other information, and by focusing exclusively on the changes in QOL provided by the last government. They model a voter as a player in a strategic game, and show that the rational voter would distrust party platforms and promises because the elected party often ignores party platforms and cannot be compelled to keep promises afterwards. Hence, the electorate trusts only prior actions of the incumbent, and punishes with only one action: voting the incumbent out. If the electorate were to give “the benefit of the doubt” and re-elect a party after QOL declines, then this action would serve as a signal to later incumbents that they need not strive for promoting the general welfare. All of this implies that voters implicitly ask a simple question, “Are we better off now than 4 years ago?” This simple model of voters’ behavior is testable, and evidence from several countries (Paldam, 1991; Fair, 1978) confirms that satisfaction in the economic domain does follow this model: as objective economic
388
MICHAEL R. HAGERTY
circumstances improve, voters are more likely to vote for the incumbent party. However, most previous literature has considered only variables in the economic domain (e.g., growth in GDP, inflation). In this paper, I review effects on national elections of other domains of QOL, such as crime rate, inequality, nutrition, and family issues. The estimation model employed in the following studies defines the dependent variable as the percentage of votes received by the incumbent party in election year t, written The predictor variables are changes in QOL indicators prior to the election, as well as variables that conrol for events in the distant past (Figure 2) such as party identification. Specifically, this model is expressed as
where denotes the percentage change prior to the election, and is the normal error term. The lagged vote is included to capture other influences on the vote. For example, if one party enjoys consistently large support over the years (because of it has done better recruiting), the controls for this. Similarly, if longterm trends that are favorable to one party (e.g., long-term shift from Liberal to Conservative), the lagged vote term controls for this, so that we can study the pure effects of QOL. It is important to specify the duration of the difference term Under the theory above that citizens are optimally controlling their QOL, the duration should include the entire time since the last elections – 4 years in the U.S. where elections are regularly scheduled. However, early studies (Fair, 1976) found that one year prior to elections was sufficient, arguing that citizens would want to weight recent behavior more heavily. Later studies with more data have detected effects going back at least 4 years, though with more weight being placed on the most recent one year (Peltzman, 1990; Abrams and Butkiewicz, 1995). Predictions
Using the above theory that citizens increase their voting support for the governing party when QOL increases, then 2 types of predictions can be tested in empirical studies. First, we can predict the
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
389
sign of each variable contributing to QOL. The variables expected to have positive signs are: GDP/capita, Disposable Personal Income (DPI) per capita, and nutrition. The signs expected to have negative signs are: unemployment, inflation, murder rate, violent crime rate, infant mortality, and divorce rate. The second prediction concerns the duration of effects. The change in QOL should be computed over the entire term of the incumbent party, not just over one or another year. To test these predictions, we use data on QOL and national elections across countries. We next describe the data.
3. DATA
Data from these studies have been drawn from national elections since 1960 in high-income democratic nations. The dependent variable was measured as the percentage of voters who voted for the incumbent party. For example, in the U.S., national and state vote shares are reported by Statistical Abstract of the Census Bureau. For other nations, voting data were obtained from Mackie and Rose (1991), and updated from annual statistical yearbooks from each country. The incumbent party was identified as the party forming the government during the period prior to the election, documented by Katz and Mair (1992) and updated by Day, German, and Campbell (1996). Some OECD democracies (e.g., Switzerland, Denmark, and Netherlands) were excluded, consistent with the results of Powell and Whitten (1993), because they tend to be governed by multiparty coalitions, where defining incumbency and responsibility is difficult. The predictor variables were various aspects of QOL during the years prior to the election. They are reported by national or international agencies that impose quality-control standards on data collection. GDP/capita, inflation, and unemployment are documented by the OECD and the World Bank (1997). DPI is reported by the Commerce Department of the U.S. Nutrition data are reported by the Food and Agriculture Organization (1995). Homicide and major theft rates were collected from Bennett’s (1991) Correlates of Crime database, and supplemented by Interpol reports of intentional homicides for all reporting countries. Safety from war was calculated as one minus the probability of dying from war or rebellion in
390
MICHAEL R. HAGERTY
that year. Deaths from wars for each country were computed from Eckhardt (1989) and supplemented by SIPRI (1995). Deaths from war included not only military deaths, but deaths of civilians, and not only international wars, but also civil wars. Divorce rates per thousand-population and infant mortality per thousand live births were recorded from the annual U.N. Demographic Yearbook. The specific elections, countries, and time periods covered are listed in Table I for the studies cited in this paper.
4. EMPIRICAL RESULTS
Table II shows a summary of all the research, Table III presents representative results from a study of U.S. voting behavior (from Hagerty, 2001), and Table IV presents representative results from an international study of voting behavior (from Hagerty, Naik and Tsai, 2000). Table II presents predicted and actual coefficients from 9 indicators of QOL that have been tested for effects on national elections. I caution that this table is not the result of a single regression on a single database, but the aggregation of multiple studies. Therefore the beta coefficients are not exactly comparable because they are based on different data with different estimation models and variables. Nonetheless, it is a useful summary of multiple studies. The first rows present results for the economic domain. In the economic domain, the variable that has consistently shown the largest effect is GDP/capita. The second column shows that its predicted effect is positive on vote share for the incumbent, and the third column confirms that its actual effect is positive. The incumbent party can expect an increase of 2% in the vote when GDP/capita has risen by 1 percent in the year prior to elections, and can expect an increase of 4% in the vote when GDP/capita has risen by 1 percent during all 3 years prior to election. The fourth column displays the standardized beta for each QOL variable. Since GDP/capita has the largest effect, we compare the standardized beta of all other effects relative to it. The next rows show the estimated effect sizes for other variables’ effects on the vote share for the incumbent party. The varibles calories/capita, murder rate, violent crime rate, and inequality are also significant and in the expected direction. In summary, Table II shows that QOL theory correctly
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
391
392
MICHAEL R. HAGERTY
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
393
predicts the sign of all variables that are significant in the 6 studies reviewed. Table III gives more detailed results from a voting study by Hagerty (2001). It shows weighted least squares (WLS) coefficients from 50 states in 7 U.S. presidential elections from 1972–1996. Each column is a separate regression, showing the effect of different lags prior to election for growth in national violent crime (VC) and growth in property crime (PC) rates in the U.S. Columns (1–4) of Table III present the regression coefficients predicting from only crime rates, with income excluded for the moment. The coefficients for national violent crime are all significant and consistently negative. This confirms the responsibility hypothesis for violent crime, consistent with the first prediction made by QOL theory. In contrast, the 4 coefficients for property crime reach significance only once, and then have the wrong sign. Therefore we cannot detect an effect of property crime on voting, but find a strong effect of violent crime. Column 5 in Table III examines the joint effect of DPI and violent crime for the 3-year lag specification. As above, both coefficients are significant and have the correct signs. Hence the information added by crime is not redundant with that for DPI. The effect of an additional 1% annual growth in DPI over the 3 years prior to the election is worth an additional 4.8 percentage points in votes to the incumbent party. This is higher than the 2.3% reported by Markus (1988) for DPI, but he reported results only for a one-year lag. The next row provides the first estimate for the effect of crime on U.S. presidential voting. The effect of an additional 1% increase in violent crime per year yields a loss of 0.19 percentage points in votes for the incumbent. Note that the effect of violent crime on voting is much smaller than that of DPI. Specifically, the relative magnitude of the violent crime coefficient to DPI is 4.0%. This appears to be a small effect, but the effect is magnified because actual variation in crime rates is four times larger than variation in income growth. Hence the standardized beta coefficient for violent crime is four times larger than the raw beta in Table III. The last row of Table III allows a test of the second prediction of QOL theory, that the duration of effects should span not just one year, but all 4 years (in the U.S.) of the incumbent’s term. Examining
394
MICHAEL R. HAGERTY
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
395
the adjusted in columns 1–4 shows a peak at the 3-year lag, with a slight decline in year 4. This shows that voters use information on QOL changes over not just one or two years, but at least 3 years. Though the theory predicts that voters should consider changes over 4 years, the actual drop in variance accounted for is small in the fourth year. It may also suggest that voters discount the first year of a governments term in office, allowing for a ‘honeymoon’ effect. The same pattern of a peak in at year 3 was replicated when national income was examined, approximating the 4-year prediction of QOL theory. Table IV gives more detailed results from an international voting study, and presents results from 13 nations, with the coefficients and t-values computed from a non-metric dimension-reduction procedure Sliced Inverse Regression (Hagerty, Naik and Tsai, 1999). The first set of rows shows the coefficients from the economic domain. Higher GDP/person, lower unemployment and lower inflation all yield higher voting support, with GDP/person very significant. These are all consistent with the predictions of the responsibility hypothesis in QOL theory. In the safety domain, coefficients are positive except for theft rate, which is not significant. The coefficient for homicide rate is almost as large as for GDP/person, and much larger than for other economic variables. This was the first time that a non-economic domain has been shown to predict voter support, providing strong evidence that extant research on voting behavior (see, e.g., Beck, 1991) must extend its focus to non-economic attributes of QOL. In the food domain, protein and calories are both positive, with calories significantly so. Finally, the coefficients for the family domain are all non-significant.
5. DISCUSSION
I outlined a framework to study how QOL affects national elections, and reviewed studies spanning 30 years in 13 OECD countries. Results show that voters reward or punish governing parties based on the increase or decrease in QOL during the party’s term in office. National elections from 13 countries support this theory for the domains of economic welfare, crime, inequality, and nutrition. In
396
MICHAEL R. HAGERTY
contrast no effect on voting was detected for changes in the family domain. Based on this review, the best estimates for the net political damage that QOL can cause on national elections is given by their beta coefficients from the regressions listed in Table II:
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
397
The political damage from a decline of one standard deviation (s.d.) in GDP/capita in the 3 years prior to election is 4% of the vote. The political damage from an increase of 1 s.d. in the murder rate is 3.6% The political damage from an increase of 1 s.d. in the violent crime rate is .8% The political damage from a decline of 1 s.d. in calories per day is 2.4%. The political damage from an increase of 1 s.d. in the ratio of income inequality (the top 10% to income of bottom 10% of households) is 1.6%. (A change of one standard deviation is computed over the change in the QOL indicator prior to election for all countries for all elections in the regressions.) Our analysis extends previous research in two ways. First, it extends the literature on election research to non-economic QOL variables. Second, it reveals the weights to create a unified QOL index for the “representative voter.” Elections are Influenced by Non-economic Indicators of QOL
Previous research in economics and political science (Nannestad and Paldam, 1994; Beck, 1991) has considered only the effect of economic variables on voting. The current studies generalize these results to other domains of QOL, including public safety, inequality, and nutrition. The conclusion is that many domains of QOL affect the outcome of national elections, and governments risk a loss of power by not monitoring and controlling QOL. Examining the relative size of the effects shows that economic variables have the largest effect. This suggests that the earlier development of economic indicators (such as National Accounts) is adaptive to preferences of voters. But as indexes of non-economic QOL are developed and publicized by the media, one might hypothesize that voters will weight them more heavily than in the past. It is important to caution that the studies reviewed here rely on correlational evidence, so that explanations of reverse causation or spurious causation by some omitted variable should be considered. Certainly, reverse causation is not possible, since the
398
MICHAEL R. HAGERTY
votes during an election cannot affect the measurements of GDP, crime, etc., which are completed months or years before the election is held. With respect to spurious causation, Figure 2 summarizes the usual causes of election results, including party identification, media attention, qualifications of candidates, and platform of candidates. The research reviewed has always controlled for party identification (by using the lagged dependent variable.) However, none of the models reviewed controls for qualification of candidates, platform of candidates, or media attention. Beck (1991) has shown that more qualified candidates are recruited to oppose incumbent congressmen in the U.S. when QOL (specifically economic conditions) decline. His results suggest that QOL changes affect the qualification of candidates, arguing for an indirect (not spurious) causal effect of QOL. Similarly, candidates often try to exploit weaknesses of the incumbent, suggesting that they would emphasize the incumbent’s “failures” by publicizing the components of QOL that have declined (for example, Clinton’s 1992 strategy “It’s the economy, stupid”). Both of these strategies would tend to magnify the effects found for changes in QOL because of the strategies of opposing (nonincumbent) candidates. On the other hand, the incumbent party would attempt the opposite “spin”, emphasizing to the media and the public the components of QOL that have improved. The incumbent party could even attempt to further improve objective QOL, by deficit spending, tax reductions, adding police, etc. These strategies would tend to reduce the size of effects found for QOL. Regardless of the net magnitude of QOL effects, all theories agree that changes in QOL are causal in selection of candidates’ platform and in qualifications of the candidate, consistent with the causal model used here. The Weights from Voter Studies Allow Computation of a Unified QOL Index for the “Representative Voter”
Just as voters appear to combine all aspects of QOL to determine whether to reward or punish the incumbent, researchers can use the same combination strategy as the “representative voter” to create a unified index of QOL. In the past, many philosophers and social scientists despaired of creating a unified QOL index for two reasons. First, they could find no single set of weights that people could agree
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
399
on as ethical “a-priori”. And second, since many people appear to have different weights for each indicator, no single index is possible. Our method avoids the first objection because we find the weights “a-posteriori” that citizens actually use in directing their democracies. (We do not claim that they are the ethically correct values, and may well vary with a nation’s level of development and other circumstances.) Our method avoids the second objection because it attempts to measure only the “representative voter”, not all voters. (Though the method could be applied to different regions that may have different voting preferences). In this way, our weights are only “central tendencies” and do not capture the entire distribution, but a central tendency is a good first step in capturing the entire distribution of citizen’s preferences. A unified QOL index is computed for the U.S. from 1968 to 1996 to demonstrate this capability. The raw beta weights estimated in Table II for the U.S., applied to the objective indicators, compute the QOL index that best captures voters’ weights from national elections reviewed here. Figure 3 shows the resulting QOL index when the three variables Disposable Income/capita, violent crime rate, and Gini coefficient are used as objective indicators for QOL for U.S. presidential elections since 1968. The best fitting raw beta weights for these indicators are, respectively, 2.40, –0.215, and –0.783. Applying these to average changes in the US during the 4 years prior to each election gives predicted change in overall QOL
400
MICHAEL R. HAGERTY
in Figure 3. It shows the U.S. beginning in 1968 with an (arbitrary) QOL of 100. In 1972 it rose to 103.6, followed by eight years of declining QOL due to little GDP growth and increasing inequality and crime. The 1980’s brought a sustained period of increase, due to DPI improvements and in spite of increases in crime and inequality. The 4 years prior to 1992 saw declining QOL, due to near-zero growth in DPI, and increases in crime and inequality. Finally, the years prior to 1996 yielded increasing QOL due to drops in violent crime and inequality, and to a small increase in DPI/capita. The final QOL index in 2000 was 10.8% higher than it was in 1968. This contrasts with a GDP/capita of almost 40% higher than in 1968, suggesting that GDP/capita overestimated improvement in QOL during this period. Limitiations
We caution that our results do not imply that all citizens weight QOL with the weights found here. Our conclusions are limited in two ways. First, they apply only to voters in national elections, who may differ from non-voters. Second, they apply only to swing voters who change their vote over the years. Our method cannot estimate other possible segments of voters who are always “loyal” to one party, but may have preferences different from swing voters. Since votes of the “loyal” segment never change, we cannot estimate their preferences from this method. In addition, the studies I have reviewed consider only national elections. Voters may react somewhat differently in provincial or local elections, because they may hold those officials responsible not for economic conditions, but for the condition of local parks, health care, etc. Nevertheless the coefficients we have identified here are very important because they come from the most politically active citizens in national elections, who can bring down governments if their QOL preferences are ignored. REFERENCES
Abrams, B.A. and J.L. Butkiewicz: 1995, ‘The influence of state-level economic conditions on the 1992 U.S. presidential election’, Public Choice 85, pp. 1–10. Beck, N.: 1991, ‘The economy and presidential approval: An information theoretic perspective’, in H. Norpoth, M. Lewis-Beck and J. Lafay (eds.), Economics
DECLINING QUALITY OF LIFE COSTS GOVERNMENTS ELECTIONS
401
and Politics: The Calculus of Support (University of Michigan Press, Ann Arbor). Bennett, R.R.: 1991, ‘Development and crime: A cross-national, time-series analysis of competing models’, The Sociological Quarterly 32(3), pp. 343–363. Bureau of Justice Statistics: 1993, Highlights from 20 Years of Surveying Crime Victims: The National Crime Victimization Survey, 1973–92 (U.S. Department of Justice, Washington, D.C.). Day, A.J., R. German and J. Campbell: 1996, Political Parties of the World, 4th ed. (Cartermill). Downs, G.W. and D.M. Rocke: 1995, Optimal Imperfection? Domestic Uncertainty and Institutions in International Relations (Princeton University Press, Princeton, NJ). Eckhardt, W.: 1989, ‘Wars and war-related deaths 1945–1989’, in R.L. Sivard (ed.), World Military and Social Expenditures (World Priorities, Washington, D.C.). Fair, R.C.: 1978, ‘The effect of economic events on votes for president’, The Review of Economics and Statistics 60(2), pp. 159–173. Food and Agriculture Organization (FAO): 1995, Production Yearbook. On computer files. (FAO, Rome). Hagerty, M.R., P. Naik and C.L. Tsai: 2000, The effects of quality of life on national elections: A multi-country analysis’, Social Indicators Research 49 (Feb). Katz, R.S. and P. Mair: 1992, Party Organizations: A Data Handbook on Party Organizations in Western Democracies, 1960–90 (Sage, Thousand Oaks, CA). Kramer, G.H.: 1971, ‘Short-term fluctuations in US voting behavior, 1896–1964’, The American Political Science Review 65 (March), pp. 131–143. Mackie, T.T. and R. Rose: 1991, The International Almanac of Electoral History (Congressional Quarterly, Washington, D.C.). Markus, G.B.: 1988, ‘The impact of personal and national economic conditions on the presidential vote: A pooled cross-sectional analysis’, American Journal of Political Science 32, pp. 137–154. Nannestad, P. and M. Paldam: 1994, ‘The VP-function: A survey of the literature on vote and popularity functions after 25 years’, Public Choice 79, pp. 213–245. Paldam, M.: 1991, ‘How robust is the vote function? A study of seventeen nations over four decades’, in H. Norpoth, M. Lewis-Beck and J. Lafay (eds.), Economics and Politics: The Calculus of Support (University of Michigan Press, Ann Arbor). Peltzman, S.: 1990, ‘How efficient is the voting market?’ Journal of Law and Economics 33 (April), pp. 27–63. Powell, G.B. Jr. and G.D. Whitten: 1993, ‘A cross-national analysis of economic voting: Taking account of political context’, American Journal of Political Science 37(2), pp. 391–414.
402
MICHAEL R. HAGERTY
Veenhoven, R.: 1993, Happiness in Nations: Subjective Appreciation of Life in 55 Nations 1946–1990 (RISBO, Erasmus University, Rotterdam. Available by FTP). World Bank: 1997, World Development Indicators (World Bank, Washington, D.C.).
Graduate School of Management University of California, Davis One Shields Avenue, A. O. B. IV Davis, CA 95616-8609 USA E-mail: mrhagerty @ ucdavis. edu
REGINA BERGER-SCHMITT
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS: CONCEPT AND MEASUREMENT
ABSTRACT. In recent years, the concept of social cohesion has received great attention in scientific research as well as in politics. It represents a central policy goal at the national and the supranational level as declared for example by many policy documents of the European Union. The present paper integrates social cohesion into the components of quality of life and distinguishes two essential goal dimensions inherent in the concept: the inequality dimension and the social capital dimension. The first dimension incorporates the goals of reducing disparities, promoting equal opportunities and combating social exclusion, while the second dimension deals with all aspects aiming at strengthening social relations, interactions and ties. A proposal is made on how to measure this conceptualisation of social cohesion within the framework of a European System of Social Indicators, and some examples of indicators of social cohesion in Europe are discussed.
1. INTRODUCTION
For the last thirty years, quality of life as a central societal goal has guided national policy in many European countries. The concept of quality of life replaced the idea of wealth as the then dominant goal of societal development. The very broad and multidimensional notion of quality of life enlarged the perspective of societal development by considering not only economic aspects but also social and ecological concerns. Nowadays, the concept of quality of life is probably the most prominent and widely used theoretical framework for assessing the living conditions in a society. As regards the empirical operationalisation of quality of life, one can distinguish various approaches which reveal different notions of the concept and thus highlight different components as relevant. But a common feature of all empirical approaches is the focus on the Social Indicators Research 58: 403–428, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
404
REGINA BERGER-SCHMITT
individual. They are based on a conceptualisation of quality of life as concerning individual characteristics. In opposition to this individuum-centred perspective of societal development, several other welfare concepts emerged during the last 10 to 15 years which put the focus on aspects concerning societal qualities such as the extent of equality, security or freedom or the quality and structure of the social relations within a society. Among those welfare concepts referring to characteristics of societies are for example sustainability, social quality or social cohesion. Besides the model of sustainable development, especially the idea of the social cohesion of a society received great political attention at the national and supranational level. At the national level for example, the Canadian Government, at the supranational level the World Bank, the OECD, the Council of Europe and the European Commission are dealing with issues of social cohesion.1 The economic and social cohesion of Europe constitutes a main policy goal of the European Union as declared first of all in the Treaty on European Union in 1993 and repeatedly confirmed in many policy documents (e.g. European Commission, 1993, 1996, 1997, 1998, 2000a). Only recently at the Lisbon Summit of the Portuguese presidency of the EU in March 2000 as well as at the Nice European Council of the French presidency of the EU in December 2000 the strengthening of social cohesion was emphasised as one of the most important European policy concerns (s. European Council, 2000a,b). Also the Council of Europe has been committed to the promotion of social cohesion. It established a European Committee for Social Cohesion in 1997 to develop and implement a new strategy for social cohesion and created a Social Cohesion Development Unit for research on social cohesion in Europe (Council of Europe, Committee of Ministers, 2000; Council of Europe, DG III-Social Cohesion, 2000). The great interest of politicians in issues of social cohesion can be attributed to at least three reasons. First of all, social cohesion is viewed as a condition of political stability and security. Inequalities and divisions within a society increase the risk of political disruption and breakdown of the political system. The former Soviet Union or the former Yugoslavia are good examples of possible consequences of deficiencies in certain aspects of social cohesion. Secondly, social cohesion is considered as a source of wealth and economic growth
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
405
and thus also strengthens a country’s international weight. There are many research results which stress the links between the social cohesion within a country and its economic performance (s. section 2.2). Thirdly, deficiencies in social cohesion such as social exclusion from the labour market, weak social ties and feelings of solidarity within private networks, or a low level of civic engagement in voluntary work may have the effect of increasing public expenditure, for example in terms of social benefits or the provision of services. Corresponding to the high attention devoted to issues of social cohesion at the political level, much conceptual and empirical research on social cohesion has been undertaken. Within this context the present paper is serving two purposes. First, an attempt is made to clarify the meaning of the concept of social cohesion and to determine its inherent dimensions by reviewing the existing theoretical approaches to this issue. This includes a reflection of the relationship to the concept of quality of life by discussing the question whether conflicting, supplementary or identical societal goals are addressed by the two concepts. Secondly, a proposal to operationalise and to measure social cohesion is presented. The suggestions refer to the efforts made in the framework of a research project whose objective is to develop a European System of Social Indicators. Based on the conceptual considerations outlined in the first part of the paper, measurement dimensions of social cohesion are derived and appropriate indicators are exemplarily defined and quantified for the EU member states and several other countries.
2. CONCEPT OF SOCIAL COHESION
Social Cohesion is viewed as a characteristic of a society dealing with the connections and relations between societal units such as individuals, groups, associations as well as territorial units (McCracken, 1998). The Social Cohesion Network of the Policy Research Initiative of the Canadian Government defined social cohesion as “the ongoing process of developing a community of shared values, shared challenges and equal opportunity within Canada, based on a sense of trust, hope and reciprocity among all Canadians” (Policy Research Committee, Government of Canada, 1999: p. 22).
406
REGINA BERGER-SCHMITT
There have been various efforts to determine the dimensions of social cohesion (Jenson, 1998; O’ Connor, 1998; Woolley, 1998). A review of this work shows a rather high agreement on the key aspects of the concept: the strength of social relations, networks and associations; a sense of belonging to the same community and the ties that bind, in terms of shared values, a common identity and trust among members; equal opportunities; the extent of disparities, social cleavages and social exclusion in a society. Thereby, explicit or implicit references to two further concepts are made which have become rather popular during the last decade, too: social exclusion and social capital. For example Dahrendorf et al. described a social cohesive society as a society preventing social exclusion (Dahrendorf et al., 1995); other scientists have emphasised the social capital of a society as an essential foundation of its social cohesion (McCracken, 1998; Maxwell, 1996). As a conclusion from this inquiry on definitions and conceptions, in the present article the proposition is made to distinguish mainly two societal goal dimensions incorporated in the concept of social cohesion: (1) The first dimension can be shortly denoted as the inequality dimension. It concerns the goal of promoting equal opportunities and reducing disparities and divisions within a society. This also includes the aspect of social exclusion. (2) The second dimension can be shortly denoted as the social capital dimension. It concerns the goal of strengthening social relations, interactions and ties and embraces all aspects which are generally considered as the social capital of a society.
The two dimensions must be viewed as independent from each other to a certain degree. In principle, strong ties within a community can be accompanied by the tendency to discriminate and exclude those people who do not belong to that community (Narayan, 1999: p. 8). The issue of a strong social cohesion within a community which itself is exclusive has lead to the question “Can social cohesion be a threat to social cohesion?” (Jenson, 1998b: p. 4) and to the conclusion “that inclusion could also mean exclusion” (Bernard, 1999: p. 18). This problem highlights the importance of considering both dimensions – the inequality dimension and the social capital
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
407
dimension – in order to get a comprehensive picture of the social cohesion within a society. 2.1. Inequality Dimension of Social Cohesion
The first dimension of social cohesion covers several inequality aspects of the welfare distribution within a society which represent a range of very prominent European policy objectives. The present approach suggests to distinguish three subdimensions of the inequality dimension. The first subdimension refers to regional disparities of living conditions. The European Union has committed itself to the goal of reducing gaps between Member States and regions which are tackled by supporting backward regions through the EU’s Structural Funds. The European Commission’s first report on economic and social cohesion revealed considerable regional disparities within Europe: for example in 1993 the average level of income per head in the most prosperous region was about 4 times that of the poorest region, or, as to unemployment rates, in 1995 the 10 worst-affected regions had an average rate nearly seven times the average rate in the 10 least-affected regions (European Commission, 1996: p. 21 resp. p. 25). Equal opportunities is another aspect of high political relevance which can be distinguished within the inequality dimension of social cohesion. This issue concerns the inequitable distribution of chances and the extent of discrimination on the basis of gender, age, social stratum, disability, nationality, ethnicity or race. The policy agenda of the European Union especially ranks high the objective of gender equality. An annual Equality Report informs on the success of strategies to improve the situation of women, for example with respect to employment or the participation in politics and other public realms. As the most recent report shows, there are still large gender gaps. Examples are the average earnings of women in the private sector, which are 28% less than men’s, or the share of women in the national parliaments of the EU countries, which on an average amounts only to 18.6% in 1999 (European Commission, 2000b: p. 3 resp. p. 19). Social exclusion is conceived to constitute a third component of the inequality dimension of social cohesion. Although this represents a very popular concept which has increasingly attracted atten-
408
REGINA BERGER-SCHMITT
tion of politicians and researchers since many years, it is nonetheless clear what is really meant by the term “social exclusion”. One can distinguish two main conceptions. Originally, the concept was used in France and defined as disruption of social ties between the individual and the society due to the failure of societal institutions, such as the democratic system, the labour market, the welfare state, or the family, to integrate individuals (Silver, 1994; Rodgers, Gore and Figueiredo, 1995; de Haan, 1999). In opposition to this perspective, the term social exclusion has also been used in the sense of a multidimensional notion of poverty and defined as a lack of resources at the disposal of individuals or households (Room, 1995, 1998; de Haan, 1999). While the first conception is directed towards the processes and causes of social exclusion, which are societal properties, the second view focuses on the outcomes of those processes which can be described by individual attributes (Gore and Figueiredo, 1997; IILS, 1998). This latter notion of social exclusion is envisaged in the conceptualisation suggested here. Thus, deficient living conditions and deprivations such as poverty or long-term unemployment are covered by the term “social exclusion” as used here. Matters of social exclusion have been addressed by European social policy since the middle of the 1980s. In the Treaty of Amsterdam (1997) the combatment of social exclusion has been explicitly mentioned as a policy concern. The high priority of this objective is also documented in the establishment of an European Observatory on National Policies to Combat Social Exclusion in 1989 and a series of research programmes launched by the European Commission in this framework. One of the most recent activities with regard to empirical research in this realm is the creation of the Eurostat Task Force on Statistics on Social Exclusion and Poverty in Spring 1998 which followed a similar conception of social exclusion as proposed here. The Task Force’s work include income poverty indicators as well as a range of non-monetary poverty (= social exclusion) indicators such as unaffordability of consumer durables, poor housing conditions, problems of health or social isolation. The most striking results are the considerable differences between the EU Member States, with Portugal and Greece in the most disadvantaged position with respect to both types of indicators, the
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
409
apparent connection between poverty and social exclusion and the rather high risk of poverty and social exclusion for unemployed and for members of single parent households (European Commission, 2000c; Mejer, 2000; Mejer and Schiepers, 2000). 2.2. The Social Capital Dimension of Social Cohesion
The conceptualisation of social cohesion as it is proposed here considers social capital as representing a second main dimension of the social cohesion of a society. In previous research, the concept of social capital has been used with differing scopes covering topics like the relations between people at the level of primary networks, associations and organisations, the structure and functioning of societal institutions, the commitment to common values and norms, a common identity, a sense of belonging and, finally, the trust between people growing out of this. These features are supposed to constitute the fundamentals of the internal social coherence of a society (McCracken, 1998; Woolley, 1998; Jenson, 1998a; O’Connor, 1998; Social Capital Initiative, 1998; Immerfall, 1999). The approach followed here distinguishes three main aspects of the social capital dimension. The first aspect to be considered are social relations and activities within primary social groups and associations. This notion of social capital as social networks is closely associated to the work of Robert Putnam who made the concept popular and represents one of the most prominent authors in this field of research. In a study of 20 regions in Italy Putnam found a positive correlation between the economic success of a region and the density of voluntary associations of people (Putnam, 1993). Putnam argues that membership and engagement in voluntary associations foster communication and dissemination of information and generate and reinforce trust in societal norms which is conducive to co-operation and economic development. This work of Putnam initiated a bulk of research on the positive impacts of the social capital of a society on it’s economic and social well-being. Not surprisingly, especially in the framework of the World Bank, an organisation striving to enhance the economic and social performance of underdeveloped and transition countries, many research activities have been undertaken in recent years.2 As an example, results from an empirical study by Narayan and Pritchett (1997) may be mentioned
410
REGINA BERGER-SCHMITT
which support the findings of Putnam: by means of survey data of Tanzanian households they found a strong relation between the level of associational activity in a village and the household incomes in that village. The second aspect of the social capital dimension is the quality of social relations in terms of shared values, feelings of affiliation, solidarity and trust. The separate treatment of these issues follows a proposition made by Woolcock (2000) to distinguish between the sources and the consequences of social capital. This distinction takes also into account that the embeddedness within a social network not necessarily has positive effects, as it is implicitly assumed by Putnam in conceiving the creation of trust as one of the factors intervening between networks of civic engagement and economic outcomes. Several authors have emphasised the possible negative consequences of social networks such as restriction of individual freedom, corruption and cronyism, the promotion of criminal activities and the enforcement of conflicts (Narayan, 1999; Grootaert, 1998; Temple, 2000). Thus, a possible conclusion would be to consider these mediating attributes of the integration into social groups separately and to investigate their impact on economic and social welfare directly. This has been done by several researchers looking at the effects of trust on economic performance (Hjerppe, 1999; Knack and Keefer, 1997; Knack, 2000; Temple, 2000). In agreement, all studies mentioned report some empirical evidence of a positive link between the level of trust in people and the economic performance of a country, but the nature of this relationship still needs to be investigated and clarified in greater detail. Interpersonal relations within social networks and associations are usually conceived as micro-level social capital which has to be differentiated from macro-level social capital covering formal institutions such as the institutions of the political system, the legal and judicial system, the bureaucratic apparatus, or the social security institutions. The quality of institutions therefore represents a third aspect of social capital which has to be considered. There is high agreement among social capital researchers that the quality of those institutions in terms of their functioning, efficiency, reliability, credibility and stability is a significant component of the social
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
411
capital of a society (Grootaert, 1998; Immerfall, 1999; Woolcock, 2000). Such “high-quality institutions” provide an enabling environment for civic engagement and co-operation and proved to be an important factor of economic growth and other aspects of social development (Ritzen, Easterly and Woolcock, 2000). Several authors not only have emphasised the significance of macro-level social capital for the welfare of a country but also the relevance of the links and interactions with micro-level social capital.3 In this context, especially the work of Narayan (1999) should be mentioned who distinguished social relations within primary groups (bonding social capital) on the one hand and the cross-cutting ties between social groups (bridging social capital) on the other hand at the micro-level and considered the functioning of the formal institutions of the state at the macro-level. His argument is that it is the special constellation of the level of cross-cutting ties and the functioning of state institutions which is crucial for the economic and social development of a society. As an example of a state with a high density of cross-cutting ties but poorly functioning formal institutions he mentioned the situation of Russia after the breakdown of the communist system. According to an analysis of this situation by Richard Rose (1995), the absence or weakness of formal institutions has been compensated by strong informal networks acting independently from the state in developing private systems of help and support bearing strong characteristics of a shadow economy. 2.3. Relation between Social Cohesion and Quality of Life: Conceptual Issues and Empirical Results
Social cohesion represents a rather new term compared to the concept of quality of life which has been used for a long time to outline the welfare development in a society. One can raise the question whether the term social cohesion really describes a new idea, to what extent the aspects covered by the two concepts overlap or differ, or what kind of relationship between the concepts exists at all. The argument brought forward here is that there are several ways of thinking about the relation between social cohesion and quality of life, depending on how quality of life is conceived.
412
REGINA BERGER-SCHMITT
Let us first of all look at the relation assuming a rather narrow conception of quality of life as denoting the individual living conditions and well-being. The individual quality of life covers several components and aspects which are either also a part of the concept of social cohesion or an outcome of social cohesion. As an example of overlapping aspects of the two concepts, social relations and ties created and maintained by individuals may be mentioned, which constitute an important component of their personal quality of life but at the same time are also a dimension of the social cohesion of a society. The impacts of social cohesion on individual quality of life have been made evident by many empirical results: The social cohesion of a society not only has consequences for wealth and economic growth as documented by the studies cited above, but also influences many other welfare aspects like the level of education, the state of health, or the extent of crime. This has been shown not only on the level of aggregate data but also on the level of individual data. Some recent results by Robert Putnam and related work on noneconomic outcomes of social cohesion should be mentioned here in greater detail. Putnam (2000) constructed for the various states of the U.S. an index of social capital composed of 13 different measures of social capital. He found a strong relationship between social capital and educational performance across the states, also after controlling for other possibly influencing variables. There is also evidence for a relation between social and human capital at the level of individual data, as has been indicated by former research of James Coleman (1988). He showed by means of a sample of 4000 high school students that social capital reduced the probability of dropping out before graduation. Furthermore, Putnam’s index of social capital proved to be a strong predictor of crime as measured by murder rates which are low in states with high social capital. Moreover, social capital has positive impacts on the state of health. This result can also be confirmed by many other studies on the state level as well as on the individual level. On the state level, research by Kawachi, Kennedy and Lochner (1997) across 39 American states revealed a rather strong relationship between social capital, measured by the extent of civic trust, and mortality rates. On the individual level, Berkman and Syme (1979) proved in a long-term study of nearly 7000 adults living in California that the probability
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
413
to die was less among people with many social ties. Finally, based on individual data Putnam (2000) also found an increasing happiness of individuals with rising levels of their own as well of their state’s social capital. Secondly, one can reflect on the relation between social cohesion and quality of life presuming a broad conception of quality of life which encompasses not only individual characteristics of the life situation but also societal qualities, that is the degree to which the societal conditions are conducive to increase individual quality of life. This notion corresponds to the idea of Veenhoven (1996) on the livability of nations or the quality of life in nations and his distinction between “outer” (societal) and “inner” (personal) quality of life (Veenhoven, 2000: p. 5). Combining both aspects in a comprehensive concept of quality of life, elements of the social cohesion of a society, such as the functioning of its institutions or the realisation of equal opportunities, can be considered as “outer” quality of life and thus as an integral part of total quality of life. In this sense, quality of life represents a common overarching policy goal with social cohesion as an important component.
3. MEASURING SOCIAL COHESION – EXAMPLES FOR DIMENSIONS AND INDICATORS IN THE FRAMEWORK OF A EUROPEAN SYSTEM OF SOCIAL INDICATORS
The concept of social cohesion will be operationalised and measured in the context of a European System of Social Indicators (EUSI) which is going to be developed in the framework of an ongoing European research project.4 This indicator system considers dimensions of social cohesion5 across a wide range of life domains (s. Tables I and II). As explained above, two principle goal dimensions of social cohesion and various aspects within each dimension will be distinguished and operationalised: (1) The inequality dimension covers the following issues:
regional disparities equal opportunities of women and men generations
414
REGINA BERGER-SCHMITT
social strata disabled citizenship groups social exclusion
(2) The social capital dimension includes the following components: social relations and activities within primary social groups and associations quality of social relations quality of societal institutions Furthermore, in operationalising the social capital dimension, as a transverse aspect European-specific concerns will be considered, that is aspects of the social cohesion between European countries. These two principal goal dimensions of social cohesion with the elements listed above have been operationalised within each of the 14 life domains covered by the European System of Social Indicators, as far as they are meaningful for the respective domain (s. Tables I and II). In the following sections, this operationalisation will be explained in greater detail and illustrated by some examples of indicators. 3.1. Measuring the Inequality Dimension of Social Cohesion
Concerning the inequality dimension of social cohesion, measurement dimensions for the various aspects can be derived for nearly all life domains (Table I). Regional disparities are taken into account for example with respect to access to transport, leisure and cultural facilities, educational and health care institutions, employment opportunities or the state of the environment. The great attention attached to these issues by the policy of the European Union has already been pointed out. Within the European System of Social Indicators measures of regional disparities are usually defined as the ratio of the highest to the lowest value across the regions of a country.6 An example for such an indicator is the disparity of regional unemployment rates.
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
415
416
REGINA BERGER-SCHMITT
The development of this indicator during the last five years shows a rising level of disparity in 12 of the 15 EU countries; only in Greece, Portugal and Austria there are declining or at least stable regional disparities of unemployment rates. In 1999 the largest differences between the most favourable and the most unfavourable regional situation can be stated in Finland (ratio of 7,5) and in Italy (5,3), while the smallest differences occur in Austria (1,3) and Greece (1,6). Issues of equal opportunities and inequality are considered across many life domains, too. The European System of Social Indicators will capture differences between women and men, generations, social strata, disabled and non-disabled, nationals and nonnationals in the respective countries. Equal opportunities of women and men are not yet realised in several domains: As to the domain of ‘households and families’, there are well-known discrepancies in the engagement in housework and childcare. In the domain of ‘social and political participation and integration’ women are in a less favourable position than men with respect to their political engagement. This is indicated for example by the ratio of women and men in national parliaments which is still below 1 in all European Countries, despite improvements during the last years. Furthermore, multiple disadvantages to women exist in the domain of labour which are monitored by indicators such as the ratio of employment rates and the ratio of unemployment rates of women and men, the ratio of women and men employed in leading occupational positions, or wage differentials of female and male employees. For example, some results for an indicator of gender inequality of earnings are presented below which concern the ratio of the average gross hourly earnings of female and male manual workers in the manufacturing industry (Figure 1). This indicator not only reflects the equity of remuneration of women and men but also differences in the level of vocational training. Besides the fact that women earn less than men in every European country, the indicator reveals some interesting differences between the countries as to the level of inequality and the developments during the last two decades. The slightest inequality of women’s and men’s wages can be observed in Sweden, where in 1998 women’s payment amounted to more than 90% of men’s. The situation is similar in Denmark, where as
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
417
early as in 1980 a comparable high level of equality was achieved. Also the remaining Scandinavian countries can be characterised as rather egalitarian in this respect. Contrary to this, the United Kingdom proves as one of the countries with the most pronounced and persistent gender gaps in earnings. Since about twenty years the average hourly wages of women have achieved less than 70% of men’s wages. At the beginning of the 1980s, in Luxembourg and in Greece gender inequality in earnings was even on a higher level than in the United Kingdom. However, towards the end of the 1990s, inequality in Luxembourg has diminished to the level of the United Kingdom and there has been a remarkable decrease of inequality in Greece, where women now are faced with a more favourable relative position than women in the United Kingdom. The European System of Social Indicators will monitor the development of many further issues of equal opportunities and inequality which concern other population groups. Examples are inequalities of generations with respect to the availability of social networks or with respect to employment chances; differences in the opportunities of social strata to achieve reasonable housing conditions or a good state of health; equal opportunities for disabled people in access to public transport, educational institutions or promo-
418
REGINA BERGER-SCHMITT
tion chances; differences between nationals and non-nationals concerning enrolment in higher education, employment opportunities and unemployment risks. These are examples of societal divisions which detract from the social cohesion and the quality of life in a society. Problems of social exclusion, as the third aspect of the inequality dimension of social cohesion, can be identified with respect to several life domains. The indicators developed in the framework of the European System of Social Indicators measure manifestations of the processes of social exclusion such as homelessness, social isolation, social discrimination, dropouts of school, long-term unemployment, and poverty. 3.2. Measuring the Social Capital Dimension of Social Cohesion
To think of the various aspects of the social capital dimension is meaningful for particular life domains only. Social capital is a concept which primarily makes sense with respect to the domain of ‘Social and Political Participation and Integration’ (Table II). This domain covers most of the individuals’ social relations and activities within primary groups and associations. Appropriate measurement dimensions are the availability and frequency of social relations to friends, relatives and neighbours; social support exchanged within these informal networks in case of needing help in special activities, personal or financial problems; membership and activities in political and social organisations and other forms of civic engagement in public realms. Examples of measurement dimensions referring to other life domains are the availability of social relations and social support within the own household or family or, as to the domain of working life, membership and engagement in trade unions or professional organisations. The quality of social relations is another aspect of the social capital dimension which can be mainly operationalised for the domain of ‘Social and Political Participation and Integration.’ The European System of Social Indicators distinguishes between four principal measurement dimensions within this domain: the quality of relations to members of the informal networks which relies on concrete persons; the quality of social relations to particular population groups and other people in general, such as attitudes towards
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
419
420
REGINA BERGER-SCHMITT
foreigners, conflicts between population groups, or general trust in people; regional affiliation such as national identity; organisational ties such as identification with a political party. For example, two indicators of this aspect of social capital are presented above (Figure 2). The first indicator shows the quality of social relations between ethnic groups in the former Czechoslovakia, Poland and Hungary, based on the first New Democracies Barometer study conducted in 1991.7 These relations were rather bad in the former Czechoslovakia as compared to Hungary and Poland, since only 18% of Czechs and 27% of Slovaks reported that problems between ethnic groups can be well handled or don’t exist at all, while in Hungary and in Poland the respective percentages were 53% and 47%. The second indicator measures the degree of identification with the state which was lower in former Czechoslovakia than in Poland or Hungary. Only 61 % of Czechs and 54% of Slovaks primarily feel themselves attached to the state they belong to and not to the local region or even to Europe. The comparative figures for Hungary and
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
421
Poland are 68% and 69%. Both indicators show that social cohesion in Czechoslovakia was low before the split-up in 1993. The quality of societal institutions is a component of social capital which applies to several life domains, since for example institutions of the political system, the legal system, education, labour relations, health care, or social security are addressed. Examples of indicators are trust in political parties, approval of the system of health care, confidence in labour unions, or satisfaction with social security institutions. Finally, across the aspects of social capital already mentioned, the indicator system will devote special attention to the topic of European cohesion, that is the relations and ties between the European countries. The relevant issues are related to several life domains. Linkages between European countries can be measured in terms of frequencies of journeys between the countries, dissemination of the cultural products of a country across Europe, the exchange of pupils and students, the dissemination of European language abilities, the employment of non-national Europeans in the respective countries, the quality of social relations to other EU nationals, or the identification with Europe at all. As an example of an indicator of the emergence of a European identity of citizens, the figure below shows the percentage of people conceiving themselves as Europeans (Figure 3). In 1998, Luxembourg, Italy, France, Spain and the Netherlands turned out to be the countries with the strongest identification with Europe; a majority of the respective populations of these countries saw themselves as Europeans. Among the countries with a comparatively weak identification with Europe, the United Kingdom and Denmark can be found. These countries voluntarily opted out of the eurozone and thus demonstrated their rather aloof position towards the promotion of European integration which – as to Denmark – was only recently confirmed by a referendum of the Danish population against the euro.8 However, participation in the single currency only indicates a formal commitment to Europe mostly decided by the national governments and does not necessarily give any evidence of people’s inner closeness to Europe. So, the population of several countries now belonging to the eurozone expressed a rather weak identification with Europe in 1998, the year before the euro was introduced.
422
REGINA BERGER-SCHMITT
That the degree of the formal attachment of a state to the European Union does not correspond to people’s affinity to Europe is also demonstrated by another result displayed in Figure 3. Based on a different survey question on European identity, asked in 1992 not only in Western European but also in several Central and Eastern European countries, it can be stated that the population of the five Central European applicant nations proposed for the first wave of European Union’s enlargement then identified with Europe to a similarly high or even higher extent as some of the member states. 4. SUMMARY AND CONCLUSIONS
In recent years the concept of social cohesion received great attention by social scientists as well as by politicians. Social cohesion represents a central societal goal at the national and the supranational level. For example, the European Commission has strongly
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
423
emphasised the economic and social cohesion of Europe as a main policy goal of the European Union, besides the goal of enhancing living conditions and quality of life. As to the relation between the concepts of social cohesion and quality of life, the present paper suggests to conceive quality of life as the overarching policy goal with social cohesion as one major component. From a review of the scientific and political discussion of the concept of social cohesion the conclusion was drawn that there are mainly two goal dimensions inherent in the concept: (1) the reduction of disparities, unequal opportunities and social exclusion – denoted as the inequality dimension and (2) the strengthening of social relations, interactions and ties – denoted as the social capital dimension. Both dimensions can be further divided into several aspects. In the framework of developing a European System of Social Indicators an effort to operationalise this conceptualisation of social cohesion across several life domains was described. Empirical results show considerable deficiencies of social cohesion both within and between European Countries. There is also some empirical evidence that social cohesion improves other aspects of quality of life in a society although the details and mechanisms of this relationship have still to be clarified further. This underlines the importance of developing indicators of social cohesion and analysing their impact on quality of life. Maybe it will turn out that policies aiming at strengthening the social cohesion of a society will also promote individual quality of life. NOTES 1
In the framework of the Policy Research Initiative (PRI) of the Canadian government a Social Cohesion Research Network involving several federal departments was established (http://policyresearch.schoolnet.ca/main-e.htm; http://policyresearch.schoolnet.ca/networks/cohsoc/socialco-e.htm; Jeannotte 1997; Policy Research Committee, Government of Canada 1999). As to the international level, a study of Jeannotte (2000) provides a good overview of the importance attached to the problem of social cohesion by the OECD, the Council of Europe and the European Union, the definitions used and the threats to social cohesion and factors promoting it as perceived by the three organisations. 2 The great attention the World Bank has paid to the significance of social capital is documented in the Social Capital Initiative (http://www.worldbank.org/ poverty/scapital/scindex.htm) which was launched in 1996 to support research
424
REGINA BERGER-SCHMITT
projects on social capital and the establishment of a special website on social capital issues (http://www.worldbank.org/poverty/scapital/index.htm). Also other organisations have been concerned with the subject as reflected for example in the International Symposium on The Contribution of Human and Social Capital to Sustained Economic Growth and Well-Being organised by the OECD and Human Resources Development Canada (HRDC) in Quebec City in March last year (http://www.oecd.org/cer/obj1/docs.htm). 3 The paper of Grootaert (1998) reports on several studies which yielded some empirical evidence of the positive impacts of links between governmental programmes and local associations on environmental and social development issues. 4 The project bears the title “Towards a European System of Social Reporting and Welfare Measurement” (EuReporting) and was funded by the European Commission from March 1998 until August 2001. The development of a European System of Social Indicators was the objective of a subproject located at the Social Indicators Department of the Centre for Survey Research and Methodology (ZUMA), Mannheim, Germany. The subproject was directed by Heinz-Herbert Noll, the head of the Social Indicators Department. 5 Besides dimensions of social cohesion, dimensions of quality of life and of sustainability are covered by the indicators system. See Berger-Schmitt/Noll 2000. 6 For most of the European Union countries the basis for the calculation of these ratios is a regional disaggregation of the indicators at the NUTS-1 level of territorial units developed by Eurostat. For Finland, Portugal and Sweden a regional disaggregation at the NUTS-2 level is used. The regional disaggregation for the other European countries included, Norway, Switzerland, Czech Republic, Hungary and Poland, follows the practices of the respective statistical offices. 7 The New Democracies Barometer (NDB) was created by the Paul Lazarsfeld Society, Vienna, in 1991, in order to monitor trends in attitudes and behaviour in Central and East European societies. A total of five surveys covering 10 countries were conducted between 1991 and 1998. For further information see http://www.cspp.strath.ac.uk/catalog4_0.html. 8 A referendum in September 2000 resulted in a majority of 53% of the Danish population voting against the Euro.
REFERENCES Berger-Schmitt, R. and H.-H. Noll: 2000, ‘Conceptual framework and structure of a European system of social indicators’, EuReporting Working Paper No. 9 (Centre for Survey Research and Methodology (ZUMA), Social Indicators Department, Mannheim). Berkman, L. and S.L. Syme: 1979, ‘Social networks, host resistance and mortality: A nine-year follow-up study of Alameda county residents’, American Journal of Epidemiology 109(2), pp. 186–204.
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
425
Bernard, P.: 1999, ‘Social cohesion: A critique’, CPRN Discussion Paper No. F09 (Canadian Policy Research Networks, Ottawa). Coleman, J.: 1988, ‘Social capital in the creation of human capital’, American Journal of Sociology 94, pp. 95–120. Council of Europe, Committee of Ministers: 2000, ‘European Committee for Social Cohesion (CDCS) – Strategy for Social Cohesion’, CM Documents CM(2000)92addrev (http://cm.coe.int/reports/cmdocs/2000/2000cm92addrev. htm). Council of Europe, DG III-Social Cohesion: 2000, ‘Social cohesion: Developments’, Newsletter No. 1 (http://www.social.coe.int/en/cohesion/strategy/Unit/ bulel.pdf). Dahrendorf, R. et al.: 1995, ‘Report on wealth creation and social cohesion in a free society’ (London). European Commission: 1993, ‘White paper. Growth, competitiveness and employment’, COM(93)700 (Office for Official Publication of the European Communities, Luxembourg). European Commission: 1996, ‘First cohesion report 1996’ (Office for Official Publication of the European Communities, Luxembourg, http://www.inforegio. org/wbdoc/docoffic/official/repcoh/repco_en.htm). European Commission: 1997, ‘Agenda 2000 – Volume I – communication: For a stronger and wider union’, COM(97)2000 (Strasbourg, http://europa.eu.int/ comm/dg1a/dwn/agenda2000/strong/strong-en.pdf). European Commission: 1998, ‘Social action programme 1998–2000’, COM(98)259 (Brussels, http://europa.eu.int/en/comm/dg05/news/sapen.htm). European Commission: 2000a, ‘Building an inclusive Europe’, COM(2000) 79 final (Brussels). European Commission: 2000b, ‘Equal opportunities for women and men in the European union – 1999. Annual Report from the Commission’, COM(2000)123 final (Brussels, http://europa.eu.int/comm/employment_social/ equ_opp/report_en.html). European Commission: 2000c, ‘European social statistics: Income, poverty and social exclusion’ (Luxembourg). European Council: 2000a, ‘Presidency conclusions. Lisbon European Council 23 and 24 March 2000’ (http://europa.eu.int/council/off/conclu/mar2000/ mar2000_en.pdf). European Council: 2000b, ‘Presidency conclusions. Nice European Council meeting 7, 8 and 9’ (http://europa.eu.int/council/off/conclu/dec2000/dec2000 _en.pdf). Gore, C. and J.B. Figueiredo: 1997, ‘Social exclusion and anti-poverty policy: A debate’, International Institute for Labour Studies, Research Series No. 110 (Geneva). Grootaert, Ch.: 1998, ‘Social capital: The missing link’, Social Capital Initiative Working Paper No. 3, (World Bank, Washington, D.C., http://www. worldbank.org/poverty/scapital/wkrppr/index.htm).
426
REGINA BERGER-SCHMITT
Haan, A. de: 1999, ‘Social exclusion: Towards a holistic understanding of deprivation’, Deutsche Stiftung für internationale Entwicklung (DSE), Villa Borsig Workshop Series (Berlin, http://www.dse.de/ef/poverty/dehaan.htm). Hjerppe, R.: 1999, ‘Social capital and economic growth’, in R. Wiman and T. Partonen (eds.), Putting People at the Center of Sustainable Development. Proceedings of the Expert Meeting on the Social Dimension in Sustainable Development. Volume 2: Contributed Papers (National Research and Development Centre for Welfare and Health (STAKES) and Ministry of Social Affairs and Health, Helsinki, pp. 87–105). Immerfall, S.: 1999, ‘Sozialkapital in der Bundesrepublik. Thesen zu Konzept und Größenordnung’, in E. Kistler, H.-H. Noll and E. Priller (eds.), Perspektiven gesellschaftlichen Zusammenhalts. Empirische Befunde, Praxiserfahrungen, Meßkonzepte (Ed. Sigma Bohn, Berlin, pp. 121–128). International Institute for Labour Studies (IILS): 1998, ‘Social exclusion and anti-poverty strategies’. Research Project on the Patterns and Causes of Social Exclusion and the Design of Policies to Promote Integration: A Synthesis of Findings, second edition (Geneva). Jeannotte, S.: 1997, ‘Social cohesion research workplan’, Department of Canadian Heritage, Strategic Research and Analysis, SRA-266 (Hull, http:// policyresearch.schoolnet.ca/networks/cohsoc/socialpapers-e.htm#Canadian Heritage). Jeannotte, S.: 2000, ‘Social cohesion around the world: An international comparison of definitions and issues’, Department of Canadian Heritage, Strategic Research and Analysis, SRA-309 (Hull, http://policyresearch.schoolnet.ca/ networks/cohsoc/socialpapers-e.htm#CanadianHeritage). Jenson, J.: 1998a, ‘Mapping social cohesion: The state of Canadian research’, Canadian Policy Research Networks, CPRN Study No. F 03 (Ottawa, ftp:// ftp.cprn.org/family/msc2_e.pdf). Jenson, J.: 1998b, ‘Mapping social cohesion’, Backgrounder Speech Presented at the Policy Research Secretariat’s Conference ‘Policy Research: Creating Linkages’, October 1, 1998, Canadian Policy Research Networks (CPRN), Ottawa. Kawachi, I., B.I. Kennedy and K. Lochner: 1997, ‘Long live community: Social capital as public health’, The American Prospect 35, pp. 56–59 (http:// www.prospect.org/archives/35/35kawafs.html). Knack, S.: 2000, ‘Trust, associational life and economic performance in the OECD’, Paper presented at the Symposium on the Contribution of Human and Social Capital to Sustained Economic Growth and Well-Being, organised by the OECD and Human Resources Development Canada (HRDC) in Quebec, March 19–21, 2000 (http://www.oecd.org/cer/obj1/docs.htm). Knack, S. and P. Keefer: 1997, ‘Does social capital have an economic payoff? A cross-country investigation’, Quarterly Journal of Economics 112, pp. 1251– 1288. Maxwell, J.: 1996, ‘Social dimensions of economic growth’, Eric John Hanson Memorial Lecture Series, Vol. VIII (University of Alberta, Alberta).
CONSIDERING SOCIAL COHESION IN QUALITY OF LIFE ASSESSMENTS
427
McCracken, M.: 1998, ‘Social cohesion and macroeconomic performance’, Paper presented at the, Conference, The State of Living Standards and the Quality of Life’, Centre for the Study of Living Standards (CSLS), October 30–31, 1998, Ottawa, Ontario/Canada. Mejer, L.: 2000, ‘Social exclusion in the EU member states’, in European Commission and Eurostat (eds.), Statistics in Focus 1. Mejer, L. and J. Schiepers: 2000, ‘Social exclusion indicators. Progress of work’. Paper presented at the Brainstorming on Social Indicators, Council of Europe, 14–15 September 2000, Strasbourg. Narayan, D.: 1999, ‘Bonds and bridges: Social capital and poverty’ (World Bank, Poverty Group, Washington, D.C., http://www.worldbank.org/poverty/ scapital/library/narayan.htm). Narayan; D. and L. Pritchett: 1997, ‘Cents and sociability: Household income and social capital in rural Tanzania’, Policy Research Working Paper 1796 (World Bank, Washington, D.C., http://www.worldbank.org/html/dec/Publications/ Workpapers/WPS1700series/wps1796/wps1796-abstract.html). O’ Connor, P.: 1998, ‘Mapping social cohesion’, Canadian Policy Research Networks, CPRN Discussion Paper No. F| 01 (Ottawa, ftp://ftp.cprn.org/ family/msc_e.pdf). Policy Research Committee, Government of Canada: 1999, ‘Sustaining growth, human development, and social cohesion in a global world’. A Report Prepared for the Policy Research Initiative (http://www.schoolnet.ca/pri-prp/ keydocs/sustain99/index-e.htm). Putnam, R.: 1993, ‘The prosperous community – Social capital and public life’, American Prospect 13, pp. 35–42. Putnam, R.: 2000, ‘Social capital: Measurement and consequences’. Paper presented at the Symposium on the Contribution of Human and Social Capital to Sustained Economic Growth and Well-Being, organised by the OECD and Human Resources Development Canada (HRDC) in Quebec, March 19–21, 2000 (http://www.oecd.org/cer/obj1/docs.htm). Ritzen, J., W. Easterly and M. Woolcock: 2000, ‘On “Good” Politicians and “Bad” Policies: Social Cohesion, Institutions, and Growth’, Policy Research Working Paper 2448 (World Bank, Washington D.C.). Rodgers, G., C. Gore and F.B. Figueiredo: 1995, ‘Social exclusion: Rhetoric, reality, responses’ (International Labour Office, Geneva). Room, G.: 1995, ‘Poverty and social exclusion: the new European agenda for policy and research’, in G. Room (ed.), Beyond the Threshold. The Measurement and Analysis of Social Exclusion (The Policy Press, Bristol, pp. 1–9). Room, G.: 1998, ‘Social quality in Europe: Perspectives on social exclusion’, in W. Beck, L. van der Maesen and A. Walker (eds.), The Social Quality of Europe (The Policy Press, Bristol, pp. 289–297). Rose, R.: 1995, ‘Russia as an hour-glass society: A constitution without citizens’, East European Constitutional Review 4(3), pp. 34–42.
428
REGINA BERGER-SCHMITT
Silver, H.: 1994, ‘Social exclusion and social solidarity: Three paradigms’, International Institute for Labour Studies (IILS), Discussion Papers Series No. 69 (Geneva). Social Capital Initiative (SCI), World Bank: 1998, The initiative on defining, monitoring and measuring social capital. Overview and program description’, Social Capital Initiative Working Paper 1 (Washington, D.C., http://www. worldbank.org/poverty/scapital/wkrppr/index.htm). Temple, J.: 2000, ‘Growth effects of education and social capital in the OECD’, Paper presented at the Symposium on the Contribution of Human and Social Capital to Sustained Economic Growth and Well-Being, organised by the OECD and Human Resources Development Canada (HRDC) in Quebec, March 19–21, 2000 (http://www.oecd.org/cer/obj1/docs.htm) Veenhoven, R.: 1996, ‘Happy life-expectancy. A comprehensive measure of quality-of-life in nations’, Social Indicators Research 39, pp. 1–58. Veenhoven, R.: 2000, ‘The four qualities of life. Ordering concepts and measures of the good life’, Journal of Happiness Studies 1, pp. 1–39. Woolcock, M.: 2000, ‘The place of social capital in understanding social and economic outcomes’, Paper presented at the Symposium on the Contribution of Human and Social Capital to Sustained Economic Growth and Well-Being, organised by the OECD and Human Resources Development Canada (HRDC) in Quebec, March 19–21, 2000 (http://www.oecd.org/cer/obj1/docs.htm). Woolley, R: 1998, ‘Social cohesion and voluntary activity: Making connections’, Paper presented at the, Conference ‘The State of Living Standards and the Quality of Life’, Centre for the Study of Living Standards (CSLS), October 30–31, 1998, Ottawa, Ontario/Canada.
Social Indicators Department Centre for Survey Research and Methodology (ZUMA) Mannheim, Germany E-mail:
[email protected] CHRISTINA BEHRENDT
DO INCOME SURVEYS OVERESTIMATE POVERTY IN WESTERN EUROPE? EVIDENCE FROM A COMPARISON WITH INSTITUTIONAL FRAMEWORKS
ABSTRACT. Comparative poverty research flourishes, especially since comparable income data are easily available through the Luxembourg Income Study. However, widely-used income surveys suffer from a number of methodological pitfalls for comparative poverty research. This research report uses the data available in the Luxembourg Income Study for three countries – Germany, Sweden and the United Kingdom – to exemplify the limited comparability of widely-used income data used in poverty research. In a simple simulation exercise, entitlements to means-tested benefits are imputed for each household in the sample, based on the institutional regulations in each country. Compared to actual poverty rates in the original sample, imputed poverty rates are markedly smaller, if not reduced to zero. Even if one accounts for an incomplete take-up of benefits, a large gap between actual and simulated poverty rates still remains, largely caused by problems in survey design. The paper concludes with a number of recommendations for improving income surveys from the perspective of comparative poverty research.
1. INTRODUCTION
Comparative poverty research flourishes, especially since comparable income data are easily available through the Luxembourg Income Study. However, the measurement of poverty is unavoidably coupled with a number of methodological choices that may influence the results. Small differences in research strategies can have a large impact on measured results (cf. Atkinson et al., 1998), and widely used measures of poverty, such as poverty rates, are not very robust. Although there is a growing body of literature on methodological problems in comparative poverty research, some methodological pitfalls are often overlooked. How does the underlying survey design affect results and cross-national comparability? Are low-income strata adequately represented in those surveys, is there a systematic bias of response rates among those groups, Social Indicators Research 58: 429–440, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
430
CHRISTINA BEHRENDT
and how do these issues vary across countries? In addition, some types of income – such as means-tested benefits, being particularly relevant for poverty research – tend to be under-reported in some surveys. In order to exemplify the limited comparability of income data widely used for poverty research and to explore the question of why is there still poverty in highly developed welfare states, this research note reports some insights from a larger research project on the relationship between social assistance and income poverty in Germany, Sweden and the United Kingdom (Behrendt, 2002). While this short research note can only briefly summarise the main results, a more fully-fledged discussion of the issues in question including more detailed references can be found in Behrendt (2000). 2. METHODOLOGICAL PUZZLES AND PITFALLS IN WIDELY USED INCOME SURVEYS
In recent years, comparative poverty research has seen a major surge, largely spurred by the availability of very useful and easily accessible data from the Luxembourg Income Study (LIS).1 The LIS project has assembled a large number of national micro-data that contain detailed information on socio-demographic characteristics and incomes of private households. However, income surveys are not an ideal source for the assessment of poverty for three reasons: 1. Undercoverage: Low-income groups tend to be poorly represented in income surveys. Sampling errors may occur at different stages of the sampling process. First, some groups of the population may be a priori formally excluded from the sampling frame, as e.g. persons living in institutions, the homeless or some categories of foreigners. Second, the procedure of selecting potential respondents from the sampling frame may also contribute to the undercoverage of some groups of the population. 2. Non-response: The poor representation of some low-income groups in surveys is reinforced by a certain non-response bias. Non-response may occur either as a failure to be included in the survey at all (unit non-response) or may be confined to certain survey questions (item non-response). The first type of non-
IS THERE POVERTY IN WESTERN EUROPE?
431
response affects low-income households in particular if their life circumstances make it unlikely to be included in the sample, especially for homeless and very mobile people. Although most surveys operate with strict rules for the selection of sample households, compliance on the part of the interviewers is hard to control. Unit non-response may occur if complex questionnaires exceed the intellectual or communicative skills of the poor, or if fear of stigmatisation or administrative sanctions prevents people to uncover their precarious life situation to interviewers. Although non-response rates cannot associated with a general “middle class bias”, there is some evidence for a correlation with the level of education and unemployment.2 3. Quality of income data: Many surveys also suffer from the problem that not all income is reported properly, especially for the lower and higher ends of the income strata (cf. Atkinson et al., 1995: 142–154). Especially means-tested benefits tend to be under-reported in income surveys; so income for poverty-prone groups of the population may be underestimated. The under- or overreporting of certain types of income is partly a function of specific patterns of general non-response (unit non-response), but given that income is considered a very sensitive issue in surveys, it is also prone to item non-response. Respondents may either not know the exact amount of their household income, or hesitate to disclose it to the interviewer. Missing data due to item non-response is not distributed equally in the population, but tends to be most marked for very low and very high incomes. These problems are supposed to be partly counterbalanced by the use of sample weights and other data editing procedures, yet some of these problems unavoidably will go uncorrected. However, the effects of these procedures are fairly limited since they can only achieve a partial correction of errors on certain variables (e.g. family type), but cannot reflect the full complexity of sampling and nonsampling errors (cf. Schnell, 1993). Sampling errors for povertyprone groups of the population may lead to underestimating poverty, yet this bias may at least be partly compensated by sampling errors for the better-off.3 Nevertheless, in spite of all flaws, there are hardly any alternatives to the use of these data for quantitative analyses of income
432
CHRISTINA BEHRENDT
poverty. In spite of all flaws, the LIS project has done a great job in harmonising data, and still offers the best available evidence for comparative studies. However, as with any data, we must be aware of the limitations of the data in order to be able to execute a thorough analysis of the data and interpret the results carefully.
3. A PRELIMINARY EVALUATION OF THE LIMITATIONS OF THE DATA: A SIMPLE SIMULATION EXERCISE
Before prematurely dismissing income as an indicator for poverty, we should carefully evaluate the impact of potential methodological flaws. A simple simulation exercise can help to shed some more light on this question, and provide some external validation for the measurement of poverty rates, albeit it is not possible to come up with a definite estimation of the effects on poverty rates. This approach combines the quantitative data from these surveys with information on national institutional arrangements. As the quality of the income data in the surveys is in question, especially for social assistance benefits and other means-tested income, this approach uses entitlements to social assistance benefits as a yardstick for the quality of income data. It starts from the assumption that most poor households will be eligible for social assistance benefits. Though, many households will not properly indicate the amount of benefits received, possibly leading to an overestimation of poverty. The simulation of these social assistance entitlements can therefore indicate some kind of minimum poverty rate, based on the assumption that every poor household received the full amount of its social assistance entitlement.4 The actual poverty rate probably lies between the results found in conventional analyses and in this simulation. The assessment of social assistance entitlements must take into account that minimum income benefits are dependent on the specific needs of the individual household. Benefits are determined by the type and the composition of the household, the age of household members, entitlements to one-off benefits, the cost of housing as well as a bundle of other characteristics that may trigger eligibility to additional premiums or other social assistance benefits (as for example disability or pregnancy of household members).5 The
IS THERE POVERTY IN WESTERN EUROPE?
433
evaluation of social assistance entitlements should therefore allow for different household need levels. A comparative analysis of social assistance entitlements can follow two strategies. A relatively easy and exact method of analysis is the use of model households. Social assistance entitlements are calculated on the basis of the legal regulations of each country for a number of pre-defined model families (e.g. Eardley et al., 1996). This methodology produces a relatively easy and exact measure of adequacy, yet its results are not necessarily representative for the entire population. More comprehensive results are yielded by a simulation of social assistance entitlements on the basis of household micro-data. This method allows us to gain a thorough picture of social assistance entitlements among the sampled population. The simulation of social assistance entitlements uses the information of the composition of the household and the age of the household members found in the LIS datsets. Combined with information on national entitlement rules, social assistance entitlements can be calculated for each household in the sample based on the institutional regulations in each country, under the assumption that all households are eligible for general social assistance.6 It is assumed that households do not have any income from employment or financial assets and do not receive any contributory social security benefits. The calculation of benefit rates combines standard benefit rates with premiums for specific claimant categories (single parents, the elderly), the value of one-off benefits for specific, irregularly occurring needs (such as large household amenities), and housing benefits. On the basis of these benefit components, social assistance entitlements can be calculated for each household in the sample. This allows us to compute a simulated income distribution that assumes that each household has an income that is equal or higher than the minimum income level as stipulated in the social assistance regulations. Thus, the disposable income of households with an yearly income (before social assistance) of less than the social assistance level has been augmented to this minimum income level. Households with a higher income obviously do not have any entitlement to social assistance, so their actual disposable income remains unchanged. By this token, the simulation produces an income
434
CHRISTINA BEHRENDT
distribution that virtually cuts off the lower part of the income distribution. Because of the complexity of social assistance regulations, a simulation can hardly ever reflect the exact social assistance entitlement of an individual household. The main sources for errors are different household definitions in the surveys and by the social assistance regulations, and payments that closely relate to the circumstances of an individual household. Generally speaking, the more discretionary the social assistance payments, the less exact are estimates of social assistance. By the same token, estimates are not able to exactly reflect social assistance payments (or the refusal of payments) if the underlying decisions relate to individual characteristics or behaviour that is not monitored in the LIS data (e.g. health problems or refused payments in the case of able-bodied claimants refusing to work). Nevertheless, this exercise can provide some guidance on social assistance benefit levels. The following exercise presents poverty rates calculated from the simulated income distribution and compares them to the original data. Poverty rates are calculated as a percentage of median equivalent income, as for the original data. The median income has been calculated from the original data, not from the simulated income distribution, in order to avoid distribution effects in the comparison of original and simulated poverty rates.7 Figure 1 presents poverty rates for the original and the simulated income distribution in Germany, Sweden and the United Kingdom. The shading of the columns depicts different intensities of poverty using the four different poverty brackets of extreme poverty (less than 30% of median equivalent income), severe poverty (between 30% and 40% of median equivalent income), moderate poverty (between 40% and 50% of median equivalent income), and near poverty (between 50% and 60% of median equivalent income). The simulation demonstrates that if all households in the sample received the exact amount of social assistance entitlements as stipulated in the national social assistance regulations, poverty would be markedly reduced or even virtually eliminated. Although there are some limitations in the simulation, the simulated social assistance entitlements should provide a good approximation of the minimum
IS THERE POVERTY IN WESTERN EUROPE?
435
436
CHRISTINA BEHRENDT
income level people would receive according to the legislature in each country. The most striking effect of the simulation of social assistance entitlements is that both extreme and severe poverty would be practically wiped out in each of the three countries. Moderate poverty is also markedly reduced in Britain and Germany, while the near poverty bracket is fairly stable. Overall, poverty rates at the 50%level would be reduced to some 2% both in Germany and Britain, given that each household would receive its full entitlement to social assistance. The effect of the simulation is thus a reduction of 7.4 and 5.5 percentage points, that is a reduction by some three quarters of the original level. The Swedish example is even more notable: poverty is virtually eradicated at the 50% level, and sharply reduced at the 60% level. Generally speaking, the poverty lines of 40% and 60% of median equivalent household income seem to delimit the scattered social assistance entitlement levels quite well – higher or lower social assistance levels appear to be fairly rare. According to these results, the evaluation of social assistance benefit levels leads to a fairly optimistic conclusion in respect to the adequacy of social assistance. In every one of the three countries – with some reservations for Britain and Germany – social assistance schemes are fairly effective in the alleviation of poverty. By and large, they provide an adequate benefit level that would allow most claimants to enjoy a decent standard of living, with only a small minority living in poverty. In turn, however, the large difference between actual and simulated poverty rates suggest a less optimistic methodological conclusion, indicating that the quality of the underlying survey data may be seriously flawed. It seems that these surveys are not able to reflect incomes of poor households in a correct way, as there is strong evidence for an underestimation of income in the survey data. However, these conclusions have to be qualified, again considering the underlying methodological assumptions of this study. The simulation was based on the assumption that all households with insufficient income are eligible for social assistance benefits, and actually receive the full amount of benefit they are entitled to. Whereas the first assumption supposedly only has a small impact on measured poverty rates, the second one is more problematic. Empir-
IS THERE POVERTY IN WESTERN EUROPE?
437
ical studies show that a marked proportion of households do not claim their social assistance benefits for a number of reasons (cf. van Oorschot, 1995). Although there is hardly any comparable evidence on the extent of non-take-up in the three countries considered, we can quite safely conclude that 15–20% of eligible households do not realize their claim in the United Kingdom, as do some 40–50% in Germany, and possibly as much (or even more) in Sweden where empirical evidence on non-take-up is hardly available.8 Yet, even if the extent of non-take-up is accounted for, a considerable difference between actual and simulated poverty rates remains. This gap can only be explained by methodological problems, largely due to sampling and non-sampling errors in the underlying surveys.
4. CONCLUSIONS FOR THE CONSTRUCTION OF INCOME SURVEYS
How could these income surveys be improved, especially in regard to improving their suitability in poverty research? The following practical suggestions may be helpful to minimise the effects of incomplete or flawed reported income, as far as they are not already realised in the construction of income surveys: 1. A detailed evaluation of sampling procedures used for income surveys with strong emphasis on the the suitability for poverty analysis could help to develop measures (such as e.g. oversampling techniques, better interviewer training and control) to reduce sampling errors for low-income groups. 2. Questions on income should be ascertained as “close to the source” as possible, which means that income components that accrue at the household level (e.g. housing benefits) should be included in the household questionnaire and be reported by the household head, whereas income components with a more individual character (e.g. earnings) should be reported by the respective household member (cf. Habich et al., 1991: 494). 3. Respondents should receive a stronger guarantee that their responses are treated confidentially and anonymously, and that their answers will not provoke any negative consequences whatsoever. By this token, it should be made clear that respond-
438
CHRISTINA BEHRENDT
ents do not have to fear stigmatisation or criminal prosecution when reporting any “sensitive” income. 4. Interviewers should urge respondents to look up income components in their documents in order to enhance the quality of the data. This does however take additional time that the interviewer may not be willing to allow. Tight restrictions in money or time for the interviewers and a lack of control by the research institute may result in “quick and dirty” interviews with poor results. 5. In addition, academics and statistical offices should think about alternative sources of income data besides the traditional questionnaire surveys. Would it be possible to use administrative data to complement the evidence gathered in traditional surveys, provided that the respondents agree, as e. g. tax files or case records of the social assistance offices? 6. Even if the large-scale matching of survey data and administrative data is not possible for whatever reason, systematical small-scale studies that compare the quality of the survey with alternative data sources could help to evaluate the quality of income surveys for poverty analysis, and derive more focused recommendations for their improvement. Most of these recommendations for the improvements of income surveys cannot be accomplished without additional cost. However, many suppliers and consumers of quantitative data on income will agree that it is preferable to have truly reliable data at slightly higher cost rather than cheaper but flawed information. Given that these survey data often form the basis for policy decisions, it may turn out that operating with imperfect information can be much more expensive in the end. ACKNOWLEDGEMENTS
For helpful comments on this piece of research I would like to thank Jens Alber, Kristina Haaf, Mike Hagerty, Frauke Kreuter, Wolfgang Lauterbach, Matthias Sacher, Ray Thomas, an anonymous reviewer and the participants of the “Special track on Assessing Quality-oflife and Living Conditions to Guide National Policy: The State of the Art” at the ISQOLS Conference in Girona.
IS THERE POVERTY IN WESTERN EUROPE?
439
NOTES 1
For a detailed description of the database, cf. Smeeding/Schmaus (1990) and Atkinson et al. (1995) and http://www.lis.ceps.lu. 2 As the Swedish data used here are partly based on tax files, they are assumed to be more reliable in this respect than the British and German survey data. 3 This can possibly explain why many surveys achieve good results in reporting average or median income. 4 The notion of entitlement refers to the amount of money a specific household should receive as stipulated in the social assistance regulations on the basis of the individual needs of the household, not the amount of money people actually have received. 5 The simulation of social assistance entitlements only focuses on the general minimum income level in the respective country, as stipulated in the social assistance regulations, but does not take into account specific minimum income schemes for subgroups of the population (e.g. asylum seekers or refugees) that may also be available in come countries. 6 In each of the three countries, some subgroups of the population are not fully eligible for social assistance. Restrictions apply notably to refugees, asylum seekers, and other foreign nationals with an insecure residency status. However, as these groups tend to be under-represented in the LIS datasets anyway, these groups can be neglected in this analysis. 7 Of course, the calculation of the median income may also be subject to a certain bias because of flawed data. However, as households with very high incomes also tend to be poorly represented in these surveys, these effects may possibly neutralise each other in the calculation of the median income. 8 It would have been interesting to be able to include evidence on the take-up of benefits in the simulation model, but the available data do not allow a solid basis for such an estimation. In order to allow any safe conclusions, take-up estimates have to be calculated from monthly data, as this usually is the time frame that is considered for the determination of claims. As LIS only offers yearly data, this exercise could not be done in a serious way.
REFERENCES
Atkinson, A.B., K. Gardiner, V. Lechêne and H. Sutherland: 1998, ‘Comparing poverty rates across countries: A case study of France and the United Kingdom’, in S.P. Jenkins, A. Kapteyn and B.M.S. van Praag (eds.), The Distribution of Welfare and Household Production: An International Perspective (Cambridge University Press, Cambridge). Atkinson, A.B., L. Rainwater and T.M. Smeeding: 1995, Income Distribution in OECD Countries: Evidence from the Luxembourg Income Study (OECD, Paris).
440
CHRISTINA BEHRENDT
Behrendt, C.: 2000, Is There Income Poverty in Western Europe? Methodological Pitfalls in the Measurement of Poverty in a Comparative Perspective, paper presented at the International Society for Quality of Life Studies (ISQOLS) Conference 2000, Girona (Spain), 20–22 July. Also published as LIS Working Paper No. 258. Behrendt, C.: 2002, At the Margins of the Welfare State: Social Assistance and the Alleviation of Poverty in Germany, Sweden and the United Kingdom (Ashgate, Aldershot, forthcoming).. Eardley, T., J. Bradshaw, J. Ditch, I. Gough and P. Whiteford: 1996, Social Assistance in OECD Countries, Volume I: Synthesis Report. Department of Social Security Research Report, Vol. 46 (HMSO, London). Habich, R., B. Headey and P. Krause: 1991, ‘Armut im Reichtum: Ist die Bundesrepublik Deutschland eine Zwei-Drittel-Gesellschaft?’ in U. Rendtel and G. G. Wagner (eds.), Zur Einkommensdynamik in Deutschland seit 1984 (Campus, Frankfurt (Main)/New York). Schnell, R.: 1993, ‘Homogenität sozialer Kategorien als Voraussetzung für “Repräsentativität” und Gewichtungsverfahren’, Zeitschrift für Soziologie 22, pp. 16-32. Smeeding, T.M. and G. Schmaus: 1990, ‘The LIS database: Technical and methodological aspects’, in T.M. Smeeding, M. O’Higgins and L. Rainwater (eds.), Poverty, Inequality and Income Distribution in Comparative Perspective (Harvester Wheatsheaf, Hemel Hempstead). van Oorschot, W.: 1995, Realizing Rights: A Multi-level Approach to Non-takeup of Social Security Benefits (Avebury, Aldershot).
Christina Behrendt E-mail: Christina.
[email protected] INDEX
Aasland, A., 8,188 Abrams, B.A., 388, 400 Adamski, W., 222, 227 Africa, C., 271,291 African National Congress, 275, 290, 293,311 aggregation, limits to, 43 Allardt, E., 51, 82, 166, 175, 191, 201, 202,214,227 American Institute for Cancer Research, 242, 263 Andersen, A., 112 Andersen, A.S., 181, 188 Andersen, T., 8, 193, 194, 196, 198, 201,206,207,214,215 Andress, H.-J., 171, 175 Andrews, R, 2, 34,44 Annie E. Casey Foundation, 251, 252, 263 Antonovsky, A., 209, 214 Arctic, 191–216 Argyle, M., 50, 82 Assaf, S.,81,86 Atkinson, A.B., 429,431,439 average consumption flows, 354 Baltic countries, 177–189 Barg,T.,59,81,84 Barstadt, A., 181,188 Bauer, R.A., 117,140,231, 263 Baulch,R., 311 Beck, N., 397,400 Beck, W., 62, 63, 82 Becker, E., 82 Behrendt,C., 10,430,440 Bell, D., 58, 82 Bennett, R.R., 389,401 Berger-Schmitt, R., 10, 55, 81, 82, 166, 168, 175, 219, 220, 227, 424 Berghman, J., 55,56, 83
Berkman,L.,412, 424 Berman, M., 195,214 Bernard, P., 406,425 Biderman, A.D., 231, 263 Billharz, S., 81, 86 Binswanger, H.P., 294, 312 Boehnke, P., 174,175 Boelhouwer, J., 7, 34, 44, 26, 134, 135, 140 Böhnke, P., 8 Bradshaw, J., 433, 440 Brown, B.V., 268, 290 Butkiewicz, J.L., 388, 401 Campbell, A., 34, 44, 50, 83, 169, 175, 220, 227, 389,401 Carley, M., 111 Carter, M.R., 291, 305, 311 Cassidy, J., 229, 263 CDCP, 234, 238,263,264 Central Statistical Service (South Africa), 269,270,290 Cesnuityte, V., 188 Chambers, R., 299, 311 Cobb, C., 372 Coleman, J.S., 57, 83, 200, 214, 412, 425 collective ignorance, 157 Commager, H.S., 231, 236, 239, 264 command over resources, 13 comparative survey research, 220,221 welfare research, 163–176 contrasting subgroups, 107 Converse, P.E., 34, 44,50, 83 Cooney, S., 245, 264 Corbett, T., 268,290 Cornett, L.M., 245, 264 Cornia, G.A., 300, 311 Creech, J.D., 243, 264 cross-sectional
441
442
INDEX
surveys of social welfare, 221 vs. longitudinal trend analysis, 98 Dahl, J., 204, 214 Dahrendorf, R., 83, 406, 425 Danielsen, M., 206, 214 Davidsson, U., 113 Day, A.J., 389,401 de Haan, A., 55, 56, 84, 408, 426 Deininger, K., 307, 309,312 Delhey, J., 8, 174, 175 democratic system, epistemology of, 27 Department of Health, Education and Welfare, 75, 83 Deze, J., 299, 312 Dickow, H., 9, 273, 290, 291 Diener, E., 44, 169, 175 disposable personal income, 394 Ditch, J., 433, 440 Doraid, M., 62, 83 Dutch ‘welfare index’, 34 Dynamic Information Centre, 218 Eardley, T., 433, 440 Easterly, W., 411, 427 Eckersley, R., 273, 291 Eckhardt, W., 390, 401 economic growth, 61 security, 363 well-being, international comparison of trends, 349–382 education, 64 Educational Goals for America, 233 elections, 383–402 Elite Survey (South Africa), 275 Ellis, F., 294, 312 employment, 64 empowerment, 61 environment state of the, 358 protection of the, 66 quality of the quality, 321 epistemology of the democratic system, 27 equal opportunities, 407, 416 for disabled people, 67 for women and men, 67 Erikson, R., 50, 83, 175,192, 193, 215, 220, 227
Estes, R.J., 9, 347 Estonia, 177–189 EuReporting, 217–228 EUROMODULE, 163–176 European cohesion, 421 Commission, 83, 404, 407,409, 425 Communities, 83, 84 Community Household Panel, 152, 167 Council, 404, 425 identity, 68 System of Social Indicators, 47–87, 165, 403–428 Welfare Survey, 164 Fair, R.C., 387, 401 family solidarity, 325 Ferriss, A.L., 9, 231, 264 Figueiredo, F.B., 408, 427 Figueiredo,J.B., 55, 84, 86 foreign indebtedness, 358 Foster, J., 311, 312 Frank, R., 41,44 Frankhauser, S., 372 free flow of information, 230 Galtzer, W., 176 Gardiner, K., 429, 431, 439 generational differences, 108 German Welfare Survey, 34, 169 German, R., 389, 401 Glatzer, W., 34, 44 goal dimensions for the life domain, 74 Goedhart, Th., 34, 45 Gordon, D., 171, 175 Gore, C., 55, 84, 86, 408, Gough, I., 433, 440 government elections, 383–402 services, 355 Greenland, 191–216 Greenland Living Conditions Study, 192 Greenwald, P., 242, 263 Greer, J., 311, 312 Grøgaard, J., 188 Grootaert, Ch., 57, 84, 410, 411, 425 Gross, B.M., 231, 264 Habich, R., 8, 166, 169, 174, 175, 437, 440
INDEX
Haddad, L., 291 Hagerty, M.R., 10, 264, 390, 401 Häll, L., 110, 112,113. 189 Halstead, T., 372 Halucha, P., 58, 82 Hambleton, R.K., 218, 226, 227 Hanna, V., 203, 215 Hansen, E.E., 194, 215 Hansen.E.J., 188, 191,192,215 Hardi, P.,59, 81, 84 Hardin.G., 213, 215 Harris, M., 273, 291 Hart, M., 60, 84 Headey, B., 437, 440 Healthy People 2000, 233, 236, 251 heart disease death rates, 242 Hernes, G., 182, 188 Hertz, O., 205, 215 Hjerppe, R., 410, 426 Hoddinott, J., 311 Hodge, T., 59, 81, 84 Hong Kong Commission on Strategic Development, 315, 347 Hopkins, M., 58, 82 Horn, R.V., 96, 111, 117, 140 household empowerment, 302 income, 301 land rights, security of, 302 production model, 207 wealth, 302 housing, 324 Hudler, M., 8, 221, 222, 226, 227 human capital, 59 stocks, 357 human development, 60 freedom, 61 Index, 81, 117, 128 Report, 61 security, 61 IILS, 56, 84, 408, 426 IISD, 60, 84 illness, 361 Immerfall, S., 57, 84, 409, 411, 426 income surveys, 429–440 Index of Economic Well-being (Canada), 355 indicators, their use to set goals, 229– 265
443
individual and collective resources, 204 quality of life, 412 welfare, 4 inequality, 3589 dimension, 107, 406–09, 414–418 infant mortality rates, 242 information and communication technology, 65 Inglehart, R., 34, 44, 232, 264 insecurity, 360 institutionalization of social indicators, 3 institutions, quality of, 410 intergenerational bequest, 356 internationalization, 320 Inuit, 191–216 Jahn, T., 82 Jankowitsch, B., 82, 166, 175, 219, 220, 227 Jeannotte, S., 423, 426 Jensen, J. 54, 56, 84, 193, 214, 406, 409,426 Johansson, S., 5, 34, 44, 102, 111, 191, 215 Johnson, R.W., 291 Johnston, D., 2 Kacapyr, E., 43, 44 Kamen, C., 7 Kanahan, D., 44 Kapel, F.O., 214, 215 Kapteyn, A., 34, 45 Katona, G., 34, 44 Katz,R.S., 389, 401 Kawachi, I., 412, 426 Keefer, P., 410, 426 Kennedy, B.I., 412, 426 Kids Count programme, 250–263 Kirmayer,L., 195, 215 Kirsten, J., 294, 300, 312 Klasen,S., 305, 312 Knack, S., 410, 426 Knudsen, K., 182, 188, 189 Knudsen, T., 206, 214 Konings-van der Snoek, M., 140 Kotzé, H., 278, 291 Kramer, G.H., 387, 401 Krause, P., 437, 440 Kruse, J., 194, 195, 203, 215
444
INDEX
Kutsar, D., 188 Labour Force Survey (Israel), 142 land reform (South Africa), 293–312 Lane, R., 41, 45 Latvia, 177–189 Lau, S.K., 314, 347 Leask, L., 195, 214 Lechêne,V., 429, 431, 439 Lecht, L.A.,231,264 Lee, M.K., 314, 347 level of living, 94 as command over resources, 24 as degree of need satisfaction, 24 Levin, D., 294, 312 life expectancy, 354 Lipton,M., 294, 301,312 Lithuania, 177–189 liveability, 53 living conditions, 1–11, 115–140 and well-being, concepts of, 202 dimensions of, 204 index, 115,121 indicators of, 203, 205 survey, 177 Living Standards Measurement Surveys, 298 Lochner, K., 412, 426 Luxembourg Income Study, 352, 430 Lyster, P., 207, 215 MacGregor, K., 271, 283, 291 Mackie, T.T., 389, 401 Mair, P., 389, 401 Maluccio, J.A., 291 Manning-Gibbs, R.A., 37, 45 Mapeta,W., 111 Matravers, R., 81, 86 Mattes, R., 271, 291 Maxwell, J., 406, 426 May, J., 9, 269, 272, 291, 305, 309, 311,312 McCracken, M., 54, 56, 84, 405, 406, 409, 427 McNeely, J., 59, 85 measurement dimensions for the life domain, 74 Mega, V., 58, 84 Mejer, L., 409, 427 mental matters, misgivings about, 36 Miles, I., 60, 84
Mohatt, G., 209, 215 Moldan, B., 81, 85 Møller, V., 9, 269, 269, 272, 273, 290, 291 monopoly, 56 Montoya, V.C., 209, 215 Mootz, M., 140 multicultural approach to social indicators, 147 Munasinghe, M., 59, 85 Naik, P., 390, 401 Nannestad, P., 397,401 Narayan, D., 57, 85, 406, 409, 410, 411,427 National Education Goals Panel, 243–250, 264 quality of life, 1–11 natural capital, 59 resource stocks, 357 Nicholson, J., 272, 291 Nielsen, O., 206, 214 Noll, H.-H., 6, 38, 45, 50, 54, 55, 58, 60, 81, 82, 85, 166, 168, 169, 175, 176, 219, 220, 227, 424 non-economic indicators of quality of life, 397 NORBALT project, 177–189 Nordic Living Conditions Surveys, 13 North, D., 57, 85 NOU, 191, 215 Nuttall, M., 205, 215 Nzimande, B., 290, 291 O’Connor, P., 57, 84, 406, 409, 427 objective indicators, shortcomings of, 42 living conditions, 51, 135, 168, 173 resources, 180 -subjective distinction of indicators, 35 vs. subjective questions, 155 vs. subjective welfare concepts, 170 OECD, 58, 59, 60, 81, 85, 96, 123, 140, 166, 169, 176 Social Indicator Program, 2 Ontario Social Development Council, 140 Ophuls,W., 213,215
INDEX
Orkin, M, 270 Osberg, L., 10, 349, 351, 355, 372, 359, 373 Ouweneel, P., 38, 45 pace of change, 99 Paldam, M., 387, 397,401 panel attrition, 100 Pantazis, C., 171, 175 Patsula, L., 218, 226, 227 Pearce, D.W., 59, 85 Pedersen, J., 58, 84 Pelczynska-Nalecz, W., 222, 227 Peltzman, S., 388, 401 personal safety, 66, 324 Petersen, R., 214, 215 physical capital stock, 356 Pinter, L., 59, 81, 84 policy relevance of social surveys, 159 Research Committee (Canada), 405, 423, 427 political goals of the European Union, 63–69 participation, 319 planning in the Arctic, 212 stability, social cohesion as a condition of, 404 Poppel, B., 8, 207, 215 poverty, 359 in old age, 362 Powell, G.B. Jr., 389, 401 preservation of social capital, 60 Pritchett, L., 409, 427 produced/man-made capital, 59 property crime, 394 public expenditure, 405 health, 66 safety and crime, 325 support, 42 Putnam, R., 57, 84, 409, 412, 413, 427 quality of institutions, 410 of life 50, 71, 94, 115–140 of life, conceptualization and measurement for national policy, 13– 32 of life in Europe, cross-national comparison of, 217–228
445
of life indicators, economic concepts, 15 of nations, 53 of social relations, 410 of society, 52,170 Rainwater, L., 439 Ravallion,M., 301, 312 Reality Check survey, 283 Reconstruction and Development Programme (South Africa), 268, 292 redistribution of land, 295 regional differences, 109 disaggregation, 78 disparities, 66, 414 region-specific social goals, 204 reliability of self-reports, 39 representative voter, unified quality of life index for, 398 research and development capital stock, 357 restitution of land, 295 Richter, R., 8, 221, 222, 226, 227 Ringen, S., 200, 215 Ritzen, J., 411, 427 Rodgers, G., 55, 86, 408, 427 Rodgers, W.L., 34, 44, 50, 83 Room, G., 408, 427 Rose, R., 389, 401, 411, 427 Rossing Feldman, T., 81, 86 Rowe, J., 372 Saami, 191–216 Samaritan Befrienders, 347 Saris, W.E., 38, 45, 226, 227 Sauli, H., 112 Scandinavian living conditions model, 179 Schäfers, B., 86 Scherpenzeel, A.C., 226, 227 Schiepers, J., 409, 427 Schlemmer, L., 268, 272, 291 Schmaus, G., 439, 440 Schnell, R., 431, 440 Schoenmakers, A., 43, 45 Schwartz, N., 40, 44, 45 sectoral analyses, 105 self-reports, validity and reliability of, 39 Sen, A., 52, 86, 227, 299, 312
446
INDEX
Sharpe, A., 10, 349, 355, 373 Sheldon, D.M., 239, 265 Shikosana, M., 290, 291 Silver, H., 55, 86, 408, 428 single-parent poverty, 362 Smeeding, T.M., 439, 440 smoking reduction programme, 238 social accounting, 3, 89 and Cultural Planning Office, Dutch, 115–140 and demographic statistics, 89 capital dimension, 406, 409–411, 418–422 Capital Initiative (World Bank), 57, 86, 409, 428, 523 capital, 53, 591, 59 change, 71, 199 cohesion, 53, 59, 71 cohesion in quality of life assessments, 403–428 Cohesion Network, 54 Cohesion Research Network (Canada), 423 concerns, universality of, 25 Development Index (Hong Kong), 313–347 engagement, process of, 313–347 exclusion, 53, 59, 406, 407, 408, 418 exclusion, combat of, 67 indicators, 89, 94, 117, 141–162 indicators, European system of, 47– 87 indicators, multicultural approach to, 147 indicators, their role in social planning, 95 model, 138 policy, 33–45 progress, 332 quality, 62 relations and activities, 409 relations, quality of, 410, 418 reporting, 89, 101, 104, 106 security, 66 structure, 71 surveys, 89, 141–162, 166 welfare, cross-sectional surveys of, 221
welfare, subjective and objective information on, 219 societal development, 49 goals, 146 institutions, quality of, 421 qualities, 413 welfare, 4 solidarity, 55 between people, 68 Sondlk, E.J., 242, 263 South African Quality of Life Project, 267–292 specialization, 55 Starck, F., 40, 45 Statistika Centralbyrån, 193, 215 Stats South Africa, 311, 312 Stewart, F., 300, 311 Stiess, I., 82 Stoop, I., 34, 44, 126, 134, 135, 140 strength of social society, 319 structural and individual change, 208 subjective appraisals, 44 health, 136 indicators, 33–45 satisfaction, 135 well-being, 51, 173, 384 Sun, E., 169, 175 Survey of Living Conditions in the Arctic, 191–216 surveys of living conditions, Swedish, 89 sustainability, 58, 61, 71, 356 Sutherland, H., 429, 431, 439 Swedish Level of Living Surveys, 13, 34 Swedish social reporting, 89–113 Swedish survey system, 90 Syme, S.L.,412, 424 System of National Accounts (USA), 89, 350 system of social and demographic statistics, 3 Taylor, H., 271, 291 telesis, 229–265 Temple, J., 410, 428 tenure reform, 295 Thorbecke, E., 311, 312
INDEX
Titmuss, R.A., 112, 189, 215 Townsend, P., 171, 176 Trapenciere, I., 188 Truth and Reconciliation Commission, 275, 292 Tsai, C.L., 390, 401 Tyldum.G., 8, 188 ULF, 89 UNDP, 81, 86, 220, 227, 314, 347 unemployment, 361 United Nations, 96, 112 Development Programme, 140 Usher, P.J., 207, 208, 216 Uusitalo, H., 50, 86, 192, 193, 215 validity of self-reports, 39 values and attitudes, 71 van der Maesen, L., 62, 63, 82 van Oorschot, W., 437, 440 van Seventer, D.E., 289, 292 van Wijk, T., 226, 227 Van Zyl, J., 294, 300, 312 VanPraag, B.M., 34, 45 Veenhoven, R., 5, 34, 38, 39, 41, 43, 45, 53, 87, 176, 193, 219, 227, 413,428 Vile, M.J., 38, 45 violent crime, 394 Vogel, J., 7, 110, 111, 112, 113, 166, 176, 189, 193,216 voting behaviour, 383–402 vulnerable population groups, 330, 331,342, 181 analyses of the situation of, 105 Walker, A., 62, 63, 82 Wallberg, K., 1l1 Wan, P.S., 314, 347
447
Ward, L.F., 231, 236, 239, 265 Warford, J.J., 59, 85 Waring, M., 350, 373 wealth and economic growth, social cohesion as a source of, 404 Wehling, P., 82 Weiner, D., 294, 312 welfare a wider concept of, 21, 23 concepts and conceptualization, 168 measurement, 71 well-being and living conditions, concepts of, 200 conceptual apparatus for measuring, 202 Welsh, J.M., 87 Western Europe, 429–440 Whiteford, A., 289, 292 Whiteford, P., 433, 440 Whitten, G.D., 389, 401 Williams, R.M. Jr., 231, 265 Wiman, R., 58, 60, 87 Withey, S., 34, 44 Wong, S.L., 314, 347 Woolcock, M., 410, 411, 428 Woolley, F, 54, 56, 87, 406, 409, 428 Woolsey, T.D., 230, 233, 265 World Bank, 58, 87, 134, 347 World Commission on Environment and Development, 58, 87 Wurtz, E., 244, 265 Xu, K., 359, 373 Zapf, W, 8, 34, 44, 50, 51, 52, 87, 85, 166, 169, 174, 175, 176, 219, 227
This page intentionally left blank
Social Indicators Research Series 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
V. Møller (ed.): Quality of Life in South Africa. 1997 ISBN 0-7923-4797-8 G. Baechler: Violence Through Environmental Discrimination. Causes, Rwanda Arena, and Conflict Model. 1999 ISBN 0-7923-5495-8 P. Bowles and L.T. Woods (eds.): Japan after the Economic Miracle. In Search of New Directories. 1999 ISBN 0-7923-6031-1 E. Diener and D.R. Rahtz (eds.): Advances in Quality of Life Theory and Research. Volume I. 1999 ISBN 0-7923-6060-5 Kwong-leung Tang (ed.): Social Development in Asia. 2000 ISBN 0-7923-6256-X M.M. Beyerlein (ed.): Work Teams: Past, Present and Future. 2000 ISBN 0-7923-6699-9 A. Ben-Arieh, N.H. Kaufman, A.B. Andrews, R. Goerge, B.J. Lee, J.L. Aber (eds.): Measuring and Monitoring Children’s Well-Being. 2001 ISBN 0-7923-6789-8 M.J. Sirgy: Handbook of Quality-of-Life Research. An Ethical Marketing Perspective. 2001 ISBN 1-4020-0172-X G. Preyer and M. Bös (eds.): Borderlines in a Globalized World. New Perspectives in a Sociology of the World-System. 2002 ISBN 1 -4020-0515-6 V. Nikolic-Ristanovic: Social Change, Gender and Violence: Post-communist and war-affected societies. 2002 ISBN 1-4020-0726-4 M.R. Hagerty, J. Vogel and V. Møller: Assessing Quality of Life and Living Conditions to Guide National Policy. 2002 ISBN 1-4020-0727-2
KLUWER ACADEMIC PUBLISHERS – DORDRECHT / LONDON / BOSTON
