Free Banking (Vol. II) History

Free Banking Volume II History

Edited by

Lawrence H. White Associate Professor Department of Economics University of Georgia, USA

An Elgar Reference Collection

The International Library of Macroeconomic and Financial History Series Editor: Forrest H. Capie Professor of Economic History and Head of the Department of Banking and Finance The City University Business School, London

1.

Major Inflations in History Forrest H. Capie

2.

Multinational and International Banking Geoffrey Jones

3.

Monetary Regime Transformations Barry Eichengreen

4.

Financing Industrialization (Volumes I and II) Rondo Cameron

5.

Financial Crises (Volumes I and II) Michael Bordo

6.

Price Controls Hugh Rockoff

7.

Protectionism in the World Economy Forrest H. Capie

8.

Commodity Monies (Volumes I and II) Anna J. Schwartz

9.

Debt and Deficits (Volumes I, IT and III) Lakis C. Kaounides and Geoffrey E. Wood

10. Central Banking in History (Volumes I, II and III) Michael Collins 11. Free Banking (Volumes I, II and III) Lawrence H. White Future titles will include: War Finance Larry Neal Stock Market Crashes and Speculative Manias

Free Banking Volume II

© Lawrence H. White 1993. For copyright of individual articles please refer to the Acknowledgements. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without the prior permission of the publisher.

Published by Edward Elgar Publishing Limited Gower House Croft Road Aldershot Hants GUll 3HR England Edward Elgar Publishing Company Old Post Road Brookfield Vermont 05036 USA

British Library Cataloguing in Publication Data Free Banking. - (International Library of Macroeconomic & Financial History;No.ll) 1. White, Lawrence H. II. Series 332.1

ISBN 1 85278 597 7 (3 volume set)

Printed in Great Britain at the University Press, Cambridge

Contents Acknowledgements

vii

An introduction to all volumes appears in Volume I.

PART I

PARTU

SECONDARY ACCOUNTS OF FREE BANKING THOUGHT 1. S.G. Checkland (1975), 'Adam Smith and the Bankers', in Andrew S. Skinner and Thomas Wilson (eds), Essays on Adam Smith, Oxford: Clarendon Press, 504-23 2. Tyler Cowen and Randall Kroszner (1987), 'The Development of the New Monetary Economics', Journal of Political Economy, 95 (3), June, 567-90 3. Scott Sumner (1990), 'The Forerunners of "New Monetary Economics" Proposals to Stabilize the Unit of Account', Journal of Money, Credit and Banking, 22 (1), February, 109-18 4. Lawrence H. White and George A. Selgin (1990), 'LaissezFaire Monetary Thought in Jacksonian America', in Donald E. Moggridge (ed.), Perspectives on the History of Economic Thought, 4, Aldershot, UK: Edward Elgar, 20-39 5. George A. Selgin and Lawrence H. White (1990), 'LaissezFaire Monetary Theorists in Late Nineteenth Century America', Southern Economic Journal, 56 (3), January, 774-87 THE 'FREE BANKING' ERA IN THE UNITED STATES 6. Fritz Redlich (1947), 'Free Banking', The Molding of American Banking: Men and Ideas, Part I, 1781-1840, New York: Hafner, 187-204, 292-7 7. Hugh Rockoff (1974), 'The Free Banking Era: A Reexamination', Journal of Money, Credit and Banking, 6 (2), May, 141-67 8. Robert G. King (1983), 'On the Economics of Private Money', Journal of Monetary Economics, 12 (1), July, 127-58 9. Arthur J. Rolnick and Warren E. Weber (1983), 'New Evidence on the Free Banking Era', American Economic Review, 73 (5), December, 1080-91 10. James A. Kahn (1985), 'Another Look at Free Banking in the United States', American Economic Review, 75 (4), September, 881-5 11. Hugh Rockoff (1985), 'New Evidence on Free Banking in the United States', American Economic Review, 75 (4), September, 886-9

3

23

47

57

77

93 119 146

178

190

195

vi

Free Banking II

PART III OTHER EXPERIENCES 12. Richard H. Timberlake (1987), 'Private Production of ScripMoney in the Isolated Community', Journal of Money, Credit, and Banking, 19 (4), November, 437-47 13. Tyler Cowen and Randall Kroszner (1989), 'Scottish Banking Before 1845: A Model for Laissez-Faire?', Journal of Money, Credit, and Banking, 21 (2), May; 221-31 14. Lawrence H. White (1990), 'Scottish Banking and the Legal Restrictions Theory: A Closer Look', Journal of Money, Credit, and Banking, 22 (4), November, 526-36 15. Eugene N. White (1990), 'Free Banking During the French Revolution', Explorations in Economic History, 27 (3), July, 251-76

201

212

223

234

PART IV PRIVATE CLEARINGHOUSES 16. Charles W. Muon (1975), 'The Origins of the Scottish Note Exchange', Three Banks Review, 107, 45-60 17. Richard H. Timberlake, Jr. (1984), 'The Central Banking Role of Clearinghouse Associations', Journal of Money, Credit, and Banking, 16 (1), February, 1-15 18. Donald J. Mullineaux (1987), 'Competitive Monies and the Suffolk Bank System: A Contractual Perspective', Southern Economic Journal, 53 (4), April, 884-98 19. George A. Selgin and Lawrence H. White (1988), 'Competitive Monies and the Suffolk Bank System: Comment', Southern Economic Journal, 55 (1), July, 215-19 20. Donald J. Mullineaux (1988), 'Competitive Monies and the Suffolk Bank System: Reply', Southern Economic Journal, 55 (1), July, 220-23 21. Gary Gorton and Donald J. Mullineaux (1987), 'The Joint Production of Confidence: Endogenous Regulation and Nineteenth Century Commercial-Bank Clearinghouses', Journal of Money, Credit, and Banking, 19 (4), November, 457-68 22. Steven Horwitz (1990), 'Competitive Currencies, Legal Restrictions, and the Origins of the Fed: Some Evidence from the Panic of 1907', Southern Economic Journal, 56 (3), January, 639-49

330

Name Index

341

263

279

294

309

314

318

Acknowledgements The editor and publishers wish to thank the following who have kindly given permission for the use of copyright material. Academic Press, Inc. for article: Eugene N. White (1990), 'Free Banking During the French Revolution', Explorations in Economic History, 27 (3), July, 251-76. American Economic Association for articles: Arthur J. Rolnick and Warren E. Weber (1983), 'New Evidence on the Free Banking Era', American Economic Review, LXXIll (5), December, 1080-91; James A. Kahn (1985), 'Another Look at Free Banking in the United States', American Economic Review, LXXV (4), September, 881-5; Hugh Rockoff (1985), 'New Evidence on Free Banking in the United States', American Economic Review, LXXV (4), September, 886-9. Elsevier Science Publishers for article: Robert G. King (1983), 'On the Economics of Private Money', Journal of Monetary Economics, 12 (1), July, 127-58. Macmillan Publishing Company for excerpt: Fritz Redlich (1947), 'Free Banking', The Molding of American Banking, Part I, 1781-1840, 187-204. Ohio State University Press for articles: Hugh Rockoff (1974), 'The Free Banking Era: A Reexamination', Journal of Money, Credit, and Banking, 6 (2), May, 141-67; Richard H. Timberlake, Jr. (1984), 'The Central Banking Role of Clearinghouse Associations', Journal of Money, Credit, and Banking, XVI (1), February, 1-15; Richard H. Timberlake (1987), 'Private Production of Scrip-Money in the Isolated Community' , Journal of Money, Credit, and Banking, 19 (4), November, 437-47; Gary Gorton and Donald J. Mullineaux (1987), 'The Joint Production of Confidence: Endogenous Regulation and Nineteenth Century Commercial-Bank Clearinghouses', Journal of Money, Credit, and Banking, 19 (4), November, 457-68; Tyler Cowen and Randall Kroszner (1989), 'Scottish Banking Before 1845: A Model for Laissez-Faire?', Journal of Money, Credit, and Banking, 21 (2), May, 221-31; Scott Sumner (1990), 'The Forerunners of "New Monetary Economics" Proposals to Stabilize the Unit of Account', Journal of Money, Credit, and Banking, 22 (1), February, 109-18; Lawrence H. White (1990), 'Scottish Banking and the Legal Restrictions Theory: A Closer Look', Journal of Money, Credit, and Banking, 22 (4), November, 526-36. Oxford University Press for excerpt: S.G. Checkland (1975), 'Adam Smith and the Bankers', Andrew S. Skinner and Thomas Wilson (eds), Essays on Adam Smith, 504-23. Royal Bank of Scotland pIc. for article: Charles W. Munn (1975), 'The Origins of the Scottish Note Exchange', Three Banks Review, 107, 45-60.

viii

Free Banking II

Southern Economic Journal for articles: Donald J. Mullineaux (1987), 'Competitive Monies and the Suffolk Bank System: A Contractual Perspective', Southern Economic Journal, 53 (4), April, 884-98; George A. Selgin and Lawrence H. White (1988), 'Competitive Monies and the Suffolk Bank System: Comment', Southern Economic Journal, 55 (1), July 215-19; Donald J. Mullineaux (1988), 'Competitive Monies and the Suffolk Bank System: Reply', Southern Economic Journal, 55 (1), July, 220-23; George A. Selgin and Lawrence H. White (1990), 'Laissez-Faire Monetary Theorists in Late Nineteenth Century America', 56 (3), January, 774-87.

University of Chicago Press for article: Tyler Cowen and Randall Kroszner (1987), 'The Development of New Monetary Economics', Journal of Political Economy, 95 (3), June, 567-90.

Every effort has been made to trace all the copyright holders but if any have been inadvertently overlooked the publishers will be pleased to make the necessary arrangement at the first opportunity . In addition the publishers wish to thank the library of the London School of Economics and Political Science and the Marshall Library, Cambridge University for their assistance in obtaining these articles.

Part I Secondary Accounts of Free Banking Thought

[1] Excerpt from Andrew S. Skinner and Thomas Wilson (eds), Essays on Adam Smith, 504-23

XII

Adam Smith and the Bankers S. G.

CHECKLAND""

S

attention has been given to Adam Smith's views on monetary questions at the level of high theory. But Smith also commented at length on the operation of banking systems both actual and ideal. It jg necessary, of course, to relate these two levels of discussion to one another. In so doing there appears the classic dichotomy of outlook between the intellectual and the practising banker. This has been present at least since William Paterson and John Law and is today as lively as ever. Smith's treatment is very largely based upon the Scottish banks,t as had been that of Sir James Steuart. 2 This, of course, is not surprising, for it was Smith's native system and had been largely formed in his lifetime. Moreover, it could fairly be claimed to be the most advanced pattern of banking in Europe, and presumably the world. OME

II

To establish a perspective on Smith's ideas about banking it is necessary to take account of the institutional evolution of the Scottish system. Great changes took place in it during his creative lifetime, beginning shortly "fter the 1745 Rebellion, when Smith was in the last two years of his tenure of the Snell Exhibition that had carried him from Glasgow College to Balliol, and continuing through his period in the Glasgow chair (1751-63), the famous continental tour (1764-6), and the years in which he wrote The Wealth of Nations, first in Kirkcaldy (1767-73) and then London (1773-6). Those parts of the book which dealt with banking underwent only the minor revisions after the first edition; nor did Smith take up the subject again in any other form. At the time of the 'Forty-five' there were two 'public' banks in Scotland: the Bank of Scotland (founded in 1695) and the Royal Bank of Scotland (founded in 1727)' These were the only institutions strictly entitled to be called banks. As public banks they enjoyed the status of limited liability, by • Professor of Economic History in the University of Glasgow. I The principal treatment of the Scottish banks is in WN II.ii.41-78. 2 Sir James Steuart, An Inquiry into the Principles of Political Oecol/omy (1767; cd. Andrew S. Skinner, 1966), Book IV, Ch. iii to xxii.

Free Banking 1J

4

Adam Smith and the Bankers

505 virtue of being state creations, the one by Act of the Scottish Parliament and the other by Royal Charter. Scotland had thus two such public banks while England had only one. This was to mean that neither of the Scottish banks could claim the central position enjoyed by the Bank of England to the south. After a period of open warfare the two banks had established a kind of truce. Each was wary of the other, keeping substantial holdings of its rival's notes in case it might be necessary to take either aggressive or defensive action. The Bank of Scotland had considerable Jacobite leanings; the Royal Bank had the opposing sympathy, having been the creation of a Whig clique led by Campbell, Duke of Argyll, part of the programme of Walpole's government for the pacification of Scotland. There were also the private bankers, mostly based in Edinburgh. They inevitably became powerful in the parlours of the public banks. using the purchase of shares as a means of entering the bank directorates, and in varying degrees dominating them. The financing of the tobacco trade was not without its significance in this connection. 3 The bills of exchange drawn by the French Farmers-General provided a very important element of liquidity in Scottish banking, passing through the hands of the private bankers. Perhaps the greatest of these merchant bankers was William Alexander: he was the first merchant member both of the directorate of the Royal Bank and of the Edinburgh Town Council (Price. i.608). The private bankers did not issue notes. but used those of the public banks, made available to them through credits. A third order of banking institution had appeared in 1747-50 when in Aberdeen and Glasgow local groups of merchants set up provincial banking companies. The two public banks composed their differences. entering into a pact of mutual support in 1752; they then set about the attempt to destroy the new provincial banking companies. The immediate incentive to do this was the issuing of notes by such companies; the public banks believed this to be lawfully their monopoly. The note issue was the basis of the business of the public banks, the private bankers having accepted the role of the public banks in this regard. The public banks succeeded in their destructive tactics in the case of Aberdeen, but the two Glasgow companies. the Ship Bank a nd the Arms Bank, fought off their attackers. Meanwhile the private bankers had further flourished. Thus, when Sir James Steuart was setting out his views on banking in his I1/quiry into the Principles of Political Oeconomy. between 1763 and 1767. the Scottish system had a kind of classic unity, with two tiers of institutions. The public banks were at its centre. They issued notes to borrowers on the basis of heritable (real) property. and on personal property. They were not 3

Jacob Price, Fra1/ce a1/d the Chesapeake, a History of the French Tobacco Monopoly, and of Its Relationship to the British alld Americall Tobacco Trades (1973), i.

I6]4-I79I

539·

Free Banking II

Essays on Adam Smith 506 discounters of inland trade bills, and so had no direct connection with commerce; to these banks alone Steuart argued should be entrusted 'the great national circulation' (Principles, 485). They provided the credits and the bank notes on which the 'exchangers' (Steuart's name for the private bankers) could operate. The exchangers in turn performed the function of bill discounters, of course charging interest on their discounts, plus a commission of per cent or more. They were the risk-bearers of the system, the link between the publicly created banks and the world of commerce. 'This set of men', said Steuart, 'are exposed to risks and losses, which they bear without complaint because of their great profits' (Principles, 484). Steuart envisaged that the private bankers could be controlled and when necessary sustained by the public banks. In any case casualties among the exchangers could never be serious for the system as a whole: 'These exchangers break from time to time; and no essential hmt is thereby occasioned to national credits.' Steuart, while delineating the resp('ctiv~ functions of public banks and private bankers, took no account of the fact that the latter were more than bankers, being typically merchants and commission agents as well, using public bank resources in their own transacti,ons, .often speculatively. Steuart does not discuss the provincial banking compantes. But though Steuart, in one sense, conceived of a kind of ideal system based upon division of function, he could criticize the way these functions were performed, cspecially by the public banks. In his discussion of thc long and difficult exchange crisis in the 1760s, when Scotland was drained of the precious metals, and bills on London rose to a high premium, Steuart condcmned the public banks for their attempt to right the situation by reducing thcir credits. Their correct course, Steuart believed, would have been to maintain credit and incomcs, by means of vigorous borrowing abmad (Principles, 507). Steuart presumably did not know that the public banks had indced sought foreign credits in Holland, but had found thc cost prohibitive. 4 Steuart's discussion of banking went beyond the structure of Scottish institutions and the way they had behaved in the recent past. He envisaged a general system of public responsibility for the money supply, discharged through a public office, in the light of criteria derived from the economy as a whole: the tax systcm should be uscd to reinforce the policics that lay bchind monctary action. 'Thc statesman', wrote Steuart, 'ought at all timcs to maintain a just proportion between the produce of industry and thc quantity of circulating (.quivalcnt in the hands of his subjects for the purchase of it' (Principles, 323). Should the circulating medium bc inadequatc

t

4 Dank of Scotland (hereinafter DS) Minlltes, 20 March 1764; DS Ollt Letter l1ooh, 6 July 1764. Dank of Scotland, Edinburgh, National Register of Archives (Scotland), Survey 945, 1/1/5.

5

6

Free Banking II

Adam Smith and the Bankers

the public office should 'supply the deficiency of the metals by such a proportion of paper credit as may abundantly supply the deficiency'. Presumably in Steuart's eyes the Scottish public banks had some such responsibility, though they were owned by shareholders and enjoyed no special position created by the state other than the limitation of their liability. Indeed he urged that the two banks should merge 'in order to form a really national bank' (Principles, 513). Scarcely had Steuart committed his two-part scheme of Scottish banking to paper, a reasonable enough description of the roles of public banks and private bankers, when it began to lose its coherence. The public banks, having long desisted from discounting inland (i.e. Scottish) bills, reentered upon that practice from 1761.5 Already certain of the private bankers had pulled ahead of the rest, creating dominating positions both in the Scottish money market and within the public banks. The Glasgow banking companies, and those formed in Dundee and Ayr in 1763, and in Perth, Aberdeen, and Dumfries (all in 1766), did not trade on the basis of Steuart's division of function, but sought to operate as autonomously as possible, offering a full range of banking services, and relying heavily on their own note issues. Finally the situation was further complicated by the taking in of deposits by the public banks on a new scale after 1766, when the promissory note replaced the old Treasurer's bond; all Scottish banking institutions then became rivals for deposits. 6 The tidy rationale of Sir James Steuart of the mid-1760s was destroyed by the growth of the economy and the banking responses these provoked, especially in the provincial towns. There was now a three-part distinction of scale and function between public banks, private bankers, and provincial banking companies. Steuart had had too little time on his return to Scotland to acquaint himself directly with the on-going developments in Scottish banking; by contrast his writings of 1777 on the working of the corn laws show that he had been able to inform himself in depth in practical terms of the real anatomy of the situation. 7 It is also probable that because the view Steuart took of banking fitted with his general economic philosophy, it cut him off from complex and changing realities. In the 1760s there had been a good deal of public debate in Scotland about banking: it culminated in the Act of 1765. This banking statute and its preliminaries were of fundamental importance to Scottish practice. There were three principal questions. Should entry to banking be free? Should the use of paper substitutes for specie continue to be permitted without regard to the smallness of the units in which it was issued? Was 5 DS Mil/utes, 27 Mar. 1761; Royal Bank of Scotland (hereinafter RBS) Mil/utes, 25 June 1761 . 6 BS, Balances taken from ledgers. 7 'Memorial on the Corn Laws', 14 Oct. 1777, in Skinner, ed., 737-8.

Free Banking II

508

Essays

011

Adam Smith

the w,e of the optional clause a legitimate banking practice? (The public banks and the Glasgow banking companies had taken to inserting in their notes a clause which, if activated by endorsement to the note, allowed the is!'ming bank to postpone redemption in specie for a period of six. month!';, interest being payable at 5 per cent in the interval; WN Il.ii.98.) All three questions centred upon the note issue. One of the responses to the attempts by the public banks to curtail credit during the long exchange crisis was the issuing of petty notes by a great many parties, who th\ls entered upon quasi-banking activities. The small-notes mania of the 1750S and 1760s produced promises to pay for 10 shillings,s shillings, and eyen I shilling. It was easy to put such notes into circulation because of the chronic shortage of silver coin, making people willing to receive payment, especially of wages, in whatever medium was available. The two public banks had made no real effort to supply this need, though the Bank of Scotland issued notes for 10 shillings from q60 R and the British Linen Company had paid its cottage spinners and weavers this way. How far the private hankers joined in this kind of issue is not clear, though Mansfield & Co., one of the largest, appears to have issued 10 shilling notes. There was, of course, no legal restriction whatever on the issue or the size of notes in Scotland. It was this great facility in issuing notes that made Scotland the most striking example of free entry into banking. Nothing whatever was required by way of capital provision or any other safeguard. To this the public banks took great exception, for they claimed with some justice that it was impossible to conduct any sort of reasonable monetary policy under such conditions. They desired, if banking companies could not be put down entirely, that they should be made to desh,t from issuing notes, acting as Sir James Steuart had envisaged his second tier of banking institutions, accepting the 'leadership' of the public banks and using their notes only. The size of notes that should be permissible related to two questions: how was a convenient hand-to-hand currency to be provided in an economy in which the precious metals had already been replaced by paper money to a unique degree, and how far should the loss of specie be allowed to go as paper money displaced it ever lower down the scale of transactions? The optional clause of course represented the fundamental question, how far was it permissible for the state to allow banking concerns, by 'agreement' with their note-holders, to depart from the principle of convertibility of notes on demand into specie? The clause had first come into being as a device necessary when there was bank war: a concern threatened by a rival with a run had only to decline, under the clause, to pay specie. But the optional clause could have a further, far-reaching effect: it could permit the bankers greatly to economize on their specie holdings, at the same time 8

BS Jl1illlltes, 25 June 1760.

7

8

Free Banking II

Adam Smith and the Bankers yet further reducing the precious metal element in the economy. Moreover, it was soon also discovered that the clause could be used by bankers in times of monetary stringency temporarily to exempt themselves from pressure and so to continue fairly generous credits. In this way short-term influences on the monetary situation could be cushioned, and in larger crises the bankers could liquidate their position in a more orderly manner. Conversely, the freedom to do such things freed the bankers from the strict discipline of the gold standard. In 1764 the government in London became aware of the need to intervene in the banking situation in Scotland. It did so through the Privy Council. The Lord Privy Seal was a Scot. He had received correspondence from Provost Ingram of Glasgow on behalf of the local banking companies there, the Ship, the Arms and the Thistle Banks. The Provost enclosed a memorial, together with 'thoughts', to both of which documents Sir James Steuart had appended lengthy and penetrating notes. 9 James Oswald, Adam Smith's boyhood friend, was a member of the Council, together with another prominent Scot and friend of Smith, Gilbert Elliott.lO The knowledge that official action was pending caused the two public banks to make a joint approach to the government. They proposed that they should have 'an exclusive privilege of banking and of issuing printed notes'.l1 If pressed, their delegates were to offer to pay an agreed annual sum to be added to the developmental funds available to the Trustees for Improving Fisheries and Manufactures in Scotland. In so doing the Scottish public banks were asking for a position stronger even that that of the Bank of England, for though the Bank had a monopoly in the sense of a prohibition of other banking companies with more than six partners, at no time had the Bank had a sole right of note issue of the kind the Scottish banks were now seeking. If awarded such a monopoly the public banks were prepared to abandon the optional clause. Alternatively, the public banks proposed that a man or partnership might not issue' notes without having ready to produce £10,000 in land or in the funds. The committee of the Privy Council that dealt with the delegates of the Scottish public banks was led by Oswald and Elliott. Adam Smith may well have been in London at this time (January 1764) for it was at the end of January that he and Henry, Duke of Buccleuch, met there preparatory to their departure for France. It is an intriguing conjecture whether Smith was consulted. In any case the answer given to the delegates was thoroughly Smithian. 'The right of banking', pronounced the Committee in the grand Enlightenment manner, P Partners in the Glasgow banking companies, Memorial, with tlotes by Sir Jallles Steuart of Coltness, 4 Feb. 1763; Thoughts concerning banks atld the paper currency of Scotla1ld, with Notes by Sir James Steuart, 1764. Both in Mitre of Caldwell. Selectiolls of the family papers preserved at Caldwell, Maitland Club part 2, vol. i (1883). 10 BS Milllltes, 1 Feb. 1764. 11 Ibid., 6 Jan. 1764.

Free Banking II

510

Essays on Adam Smith

'is a matter not of Publick favour but of Right to every Subject in Common.'12 The delegates were further informed that should the two public banks proceed at law in an attempt to put down the banking companies, the government itself would introduce new legislation to make such compani('s explicitly legal. The delegates were also told that the optional clause Illust go. Finally, it was proposed that not only should the 'small notes' cease, hut that Scotland should have no notes under £5. This of course was the view held by Adam Smith. The public banks were thus rebuffed. The only thing saved from the wreckage was the £1 and the guinea notes, for the government did not persist with the idea of a £5 minimum for Scotland. In England notes under £t were made illegal in 1775 and those under £5 were banned ill 1777. But the ending of the optional clause did not mean that the Scottish banking system was one that rested upon the instant convertibility, in full, of notes into specie on demand. The Scottish bankers had developed other devices for minimizing such a requirement, plain though it was in law. They used the award of credits against clients to oblige them to be very moderate in their specie demands; a banker typically required borrowers to draw only his notes, and, if possible, to pay in only the notes of others. In this way the Scottish system was one of more or less continuous partial suspension of cash payments. In any case businessmen operating 011 any scale knew that in times of crisis the outcome depended not on the meagre specie supply in Scotland, but the availability of London credits. Meanwhile the exchange crisis continued. In 1764 the two public banks had sought to sustain the Edinburgh-London exchange by loans from the Bank of England and from Amsterdam. Neither effort succeeded. The Bank of Scotland favoured an all-out war by the two public banks against other issuers, but the Royal Bank would not agree, partly because its protege, the British Linen Company, was a vigorous issuer of notes. IJ As the Scottish specie supply reached dangerously low levels both banks bought silver in England and carted it by wagon to Edinburgh. This was very costly. In the course of 1762 the Bank of Scotland had brought from London upwards of £100,000 at a cost of about £4·IOS. per cent, £3.4s. per cent being paid for exchange and £1.6s. per cent for carriage. 14 Bctwecn 1764 and 1769 the Royal Bank spent over £20,000 in this way, equal to the profits of a note circulation of £80,000. 15 The publie banks also moved against 'English riders', that is men who collected specie for export; moreover they adopted strong sanctions against anyone who aided such activities. 16 By mid-summer 1769 the public banks had brought about 12 IJ 14

15 16

Ibid., I Feb. 1764. Ibid., 5 Dec. 1764,2 Jan., 5, 16 June 1766; RBS Minutes, 22 May 1765. Select Committee 0/1 Banks of IsSlie (1841), 303. HOS lH.inlltcs, 4 Dec. 1771. Also WN 1.286. BS fl1il/llles, 13 Mar. 1764; HBS filiI/lites, 18 December 1767.

9

Free Banking 11

10

Adam Smith and the Bankers

5I 1

a considerable credit reduction; the exchange crisis was a good deal relieved. In this situation and in this year the firm of Douglas, Heron and Co., popularly known as the Ayr Bank, was formed, with powerful backing from the landed nobility of the south-west of Scotland, including Adam Smith's protege, the Duke of Buccleuch. It at once began to isssue notes on a large scale by making available easy credits. The two public banks knew that they could not put down the Ayr Bank by their own efforts. Nor could they expect any help from the government in disciplining the system. They were also aware by this time that they themselves were under-capitalized. Retreat was the only tactic: they greatly reduced their lending and note issue, leaving the field very largely to the Ayr Hank. By December 1771 the issue of the Royal Bank had been brought down to £22,753. 17 Between 1769 and 1772 the Scottish banking system consisted of the two public banks, the British Linen Company, the Ayr Bank, about twenty private bankers (all general merchant houses, among which two, Sir William Forbes and Co. and Mansfield and Co., greatly outdid the rest), and four banking companies operating as partnerships in Glasgow together with five others in different towns. There were initiatives afoot for further provincial banking companies. Moreover, at all levels banks and bankers had or sought London connections, either with the Bank of England as in the case of the public banks or with London bankers. All had learned the importance of London-held assets or London credits. The situation was thus a fairly complex one. The Ayr Bank ran its hectic course, collapsing in June 1772. All but four private banking houses went with it. The shareholders of the Ayr Bank were confronted with an appalling deficit. The only way to prevent legal proceedings against them, and the sale of large parts of their landed estates, was to find a means of borrowing at long term in London. This was done through the sale of annuities, authorized by Act of Parliament. In this painful matter Adam Smith was of service to the Duke of Buccleuch and others. 'My attention', he wrote, 'has been a good deal occupied about the most proper method of extricating them.'18 The annuities were a very expensive way of raising the necessary funds, but there appeared to be no alternative. By the end of 1772 Scottish banking had been through ten extraordinary years, involving a prolonged exchange crisis, a major banking statute, a proliferation of new provincial banking companies, a purging of the private bankers, and the brief life cycle of a venture of extraordinary size, daring, and ineptitude. There followed important new developments in Scottish banking, which 11 18

RBS 1I1imlfes, 4 Dec. 1771. Letter of 5 Sept. 1772. in John Rae. Life of Adam Smith (1895), 253.

Free Banking II 512

Essays on Adam Smith

bring the story down to The Wealth of Nations in 1776. Under the influence of the Bank of Scotland the note exchange, briefly begun not long before the Ayr Bank failure, was revived and extended so that by 1774 the Scottish note exchange in Edinburgh unified the system: by weekly and latcr twiceweekly exchanges the argument, so often used in theoretical dcbate, that no bank could force into 'the Circle' more notes than the public would willingly hold, was given greater reality.19 The bank of Scotland now embarked upon a branch system: by 1780 it had seven branches and by 1790 eighteen. The Royal Bank, though it did not open branches except in Glasgow in 1783, became Edinburgh agent for a number of provincial banking companies, providing them with credits. Both public banks set ahout strengthening their capital position. These developments, taken together, signified a new and more constructive approach on the part of the two puhlic hanks to their role within the system. But Scottish banking at the level of the public banks continued to be highly political. There were rumours in 1774 that the Bank of Scotland, through one of the private banks (probably Mansfield and Co.) was trying to secure influence on the Court of the Royal Bank. The private bankers, especially the important ones, had large holdings of the stock of hoth public banks: this was both a good investment and a means of excrcising control through the election of directors. A proposal for the union of thc two public banks, as Steuart had suggested, came to nothing. 20 But one intrigue did succeed. Henry Dundas, in 1776, was in the early stages of building up his political control of Scotland. One path to doing so lay through the Royal Bank. Dundas persuaded William Ramsay and Patrick Miller, senior partners in Mansfield and Co., to sell a considerable part of their holding of Royal Bank stock in order to bring in new shareholders, and through them new directors. As a result of this tactic Sir Lawrence Dundas was ousted as Governor, and the Duke of Buccleuch, so lately involved in the Ayr Bank debacle, put in his place. 21 It is against this background, extending over Adam Smith's years from the age of twenty-two to the age of fifty-three, that his treatment of banking must be placed. III

The crcation of a successful banking sector requires that there be at least some degree of comprehension of the system by legislators together with reasonably adequate rules of conduct for its participants. The statesman I. BS Alil/fltes, 8 July, 9 Dec. 1771, DS Oflt Leiter Boo!I, 19, 2S Junc, 8, 10,17 July 1771; 29 J unc 1774· 20 Anon., A Letter to the Proprietors of the Balik of Scotland (1777). 21 \Villinm Ramsay to the Duke of Duccleuch, 22 July, Apr. 1790. Scottish Record Office, GD I 13/283/8.

11

12

Free Banking II

Adam Smith and the Bankers

51 3

and the theorist start their reasoning from the needs of the economy; the banker typically starts from the conditions of profitability and safety of his own business, given its demand liabilities and the soundness of the local and regional trading with which his principal assets in the form of advances lie. At the theoretical level it is necessary to identify the elements of which the system consists and the relationship of those elements to one another and, in interaction, their links with the economy as a whole. In the years in which The Wealth of Nations was forming, the fundamental theoretical problem could be expressed in the terms: how were the proper limits to the creation of bank notes, to replace or supplement the coin provided by the state, to be determined? The other elements of monetary debate, including the terms of entry into banking, related directly to this. In seeking an answer theorists had to define the role of money in the economy. Was its supply governed by automatic principles so that the issue of paper money had ouly a displacement effect? Did paper simply push sl'xie out of circulation, without altering the size of the total money supply from what it would 'naturally' have been, namely such as was required by the level of trade and incomes, and without inducing further effects? Or was the money supply capable of being positively changed upward and downward by the action of the bankers or a public authority, through expansion or contraction of lending, with far-reaching repercussions upon employment, incomes and stability? Scotland had already been through this argument, some seventy years before The Wealth of Nations, when the protagonists had been William Paterson and John Law,22 the former favouring a system of convertibility of notes into specie on demand, the latter advocating a policy of st:tte manipulation of the money supply. 'l'he banker seldom tried to answer such questions. I lis mind was rooted in his business. This was true over the miscellaneous range of public banker, private banker, and provincial banker. The basic question for all, in the state of banking in Smith's time, was: how much of one's own note issue could be got into the circle and kept there? This gave the banker his interest-free loan from the public, typically some multiple of the capital he had provided. But his note issue depended to a significant degree upon the terms upon which he was prepared to make advances in discounts or any other form. A weak banking concern or one trying to make a beginning would be tempted to accept the riskier kind of borrower, or to extend further in the upswing, thus exposing itself to losses. Under the simplified system envisaged by Sir James Steuart the public banks would act in the public interest, and would have some degree of control over other banking houses. But with the refusal by government to limit free banking in any 22 John Law, MOlley ami Trade Considered: with a Proposal for Supplying the Natioll zuith fdolley (1705.1750); William Paterson, The Occasion of Scotlalld's Decay ill Trade (1705).

Free Banking [J

Essays on Adam Smith way, at a time when a number of new banking enterprises were coming into being, this ability was much weakened, if not destroyed. IV

At the general theorelical level Adam Smith took the natural or equilibristic view of the money supply.23 It could not properly be varied by bank or state action, except perhaps within very narrow limits. Certainly it should not be used as a tool of public policy. Smith was, of course, very much aware of the temptation to use the state in monetary matters. The situation in various of the British colonies had attracted his attention, especially in America where certain colonial governments had not only issued their own notes, but had made them payable not on demand but several years after issue; to get such instruments into circulation, and keep them there, they were made legal tender and receivable in payment of taxes. Such action could only add to inAationary pressure; Smith cordially approved of the Act of the British Parliament in 1764 forcing colonial legislatures to stop such actions (WN Il.ii.IOO-103). But simple abdication by the state in monetary matters was not enough. Though the state was not itself to be a direct participant, it had a responsibility to provide an appropriate legislative environment for banking, as indeed for every economic activity. In Smith's thinking two sets of rules were thus required: one for the state, for the furtherance of an automatic monetary system, and one for the bankers, so that they might behave in a way appropriate to their own interests and yet promote an optimal banking situation. The principles the state must maintain were, in Adam Smith's view, of two kinds. One asserted the rightness of free banking, but the other imposed certain constraints on banking action. The state should assume no supervision over entry into the banking business. But it should encourage the erection of as many banking enterprises as possible, and it should give monopolies to none (WN Il.ii.I06). Thus no banking concern should be in a position dominant over others, much less empowered to control their conduct. For with a Plultiplicity of modest banking concerns the consequences of the error; or speculations of a few, resulting in their ruin, would be dispersed throughout the system so that no serious damage could result. Free entry and vigorous competition were thus the formulae. This view of Smith's is difficult to reconcile with the real situation in Scotland, with the two public banks very much larger and more powerful than any other part of the system, and bound by a pact of mutual assistance. It is perhaps surprising that the extraordinary tale of the Ayr Bank, which lJ See Hobert V. EagJy, 'Adam Smith and the Specie Flow Doctrine', Scottish JOllmal of Political Ecollomy, vol. xvii. 1970.

13

14

Free Banking II

Adam Smith and the Bankers conducted operations on a major scale with a naivete passing into knavery, involving Smith's own patron and others of his friends in heavy losses, did not cause him to amend his simplistic atomistic view of the Scottish banking structure. The competitive system, as envisaged by Smith, could have the further beneficent effect that the banks, in mutal competition, would press their note issues as far as they would go in terms of acceptability to the community, thus economizing on sterile specie holdings. Paper money could thus be an acceptable substitute for metallic currency because it replaced the more expensive instrument of commerce (gold and silver) with the less expensive (paper) (WN II.ii.26). Smith therefore approved of Scotland's abandonment of a high proportion of its stock of coin, fur when necessary, 'goods will always bring in money'. Yet certain constraints imposed by the state were necessary. There were three of these. The first had to do with the size of notes. Smith was a strong opponent of the Scottish small notes, and would, indeed, have restricted the size of notes to a minimum of £5. This he justified on the ground that in this matter of petty notes the natural liberty of a few 'beggarly bankers', which might endanger the security of the whole society, 'ought to be restrained by all governments' (WN II.ii.92-4). Such notes, by penetrating the pockets of 'many poor people' were a threat to such humble folk through the frequent bankruptcy of their issuers. Smith does not discuss the problem of shortage of the means of payment, in small denominations, which gave rise to such note issues. But there was another reason for condemning notes under £5. The general principle was held by Smith that the smaller the size of the notes in circulation the further went the displacement of the precious metals by paper. Such petty notes might cause an almost total loss of specie, forcing it out of its final repositories, tills, and pockets. At this point Smith echoed the fears of Bume for a country that had almost none of the metals, its commerce and industry 'suspended upon the Daedalian wings of paper money'.24 Indeed it is hard to see how a system of convertibility could be operated at all under such conditions. Smith believed that the Act of 1765 had had a good effect, in relieving the dangerous scarcity of gold and silver in Scotland (WN II.ii.92). Smith, indeed, would have stopped the displacement of specie by notes at the level of 'dealers' (i.e. merchants and wholesalers), and not let it reach the level of consumers (WN II.ii.89-90). Smith saw the money supply as being different in kind at the levels of wholesale dealing versus retail. But to have implemented such a view in Scotland after 1765, confining notes to a minimum size of £5, would also have acted as a severe restraint on entry 24 (WN II.ii.86); David Hume, Writillgs RBS Minlltes, 12 May 1773.

011

Economics, ed. Eugene Rotwcin (1955), 69.

Free Banking II

Essays on Adam Smith

to the banking business, for the £1 notes were an essential element in the' profitability of the proliferating provincial han king companies. The second infraction of natural liberty advocated hy Smith in connection with hanking had to do with the optional clause. For Smith, the Act of 1765 was right in denying the liberty to offer or accept a bank note, the convertibility of which was optional on the issuer. This remained true even though the acceptor was willing. For to some degree the optional clause had become yet another means of minimizing gold and silver holdings, contributing to the specie denudation of Scotland heyond the point of legitimacy. Moreover, it had also heen used as a means of disciplining borrowers: those who insisted on gold instead of notes would havc it hinted to them that if they did not abate their demands the clause would be operated against them. 2s Nor should any banking concern havc the 'liberty' of such self-defence against factitious attack. Smith, indeed, h:1, p. 74) asserled Ihal Ihe Gulvestotl News was Ihe mosl pupular and wi.ltly rt'acl daily Ilaper in Texas and had advocaled free banking fllr a num1Jcr of years . • :1 This daim is impurlilnl nol only for ils possible pulicy implicalions bUl also because il pnullises a slliluioll III Ihe nagging quesliun why individuals huld money. If IIIlIney W(I"t III disappear in lhe absence of government intervention, lhen we need nOl auempl III derive ilS continued exislence from individual oplimizing behavior in a pure markel sClling. (The origin of lIloney, however, can slill indeed be explained by an invisible hand IIIcl:hanism; see Menger (1892).) Thus Ihe absence of money from Ihe basic mmlcl of ~eneral e(luilibl'ium may be a virtue of Ihe Iheory ralher lhan a failure. This a"s~lIcc woulon', IM94 leuer 10 H. D. Macleod. "uolcd in Wiliialll..1Illb3, 303nl. 1M. Compare l'uwcll13bl and Richards 13MI. 19. See Williall,"un Ib3, t411. Fur a cmuplcle 1i,1 of Alkinson's sub,Lanlial wrilillg' un bimelallblll consul! Ihe bibliugraphy in Ihis work.

2(). For Atkin~on'~ ~cneraJ views on IIAe causes of business cycles

1101·

s~c

"Cause alld KCllh!c..Iy I(}r

Businc~s Ikpn.'s~il)n"

Free Banking II 7XO

Georg(' A. S('/gill alld L(/lI'rell(,(, H. While

currcnl'y could have the opposite erfect. ellusing a return of high-powered money to hank reserves and :m unwarranted expansion of total money supply. St:ltutory reserve requirements aggmvated thcse elleets: lJllller Ihe present l'omlitions of compulsory reserve on deposits ami investmenl of capilal in honds. the hank in!! comnnll1ity is kgally lilrhidden either to extend Mlpport 10 IIll'rl'iHlnls or ii' noll' cirl'ulution al Ihe time when hoth arc most nee,ied. On Ihe other hand. hanks nHly he. :md often ure. oppressed hy the accunllllillion of governmenl noles which eunnot he used excepl in unwholesome speculation at :1 period when there is little or no call for slllull nole circulalion II). 121. Besides favoring the repe:ll of hond-eolhlteml and st:ltutory reserve requirements, Atkinson argued for a discontinuution of the prohihitive 10% tax on state-bank note issues, which :liso interfered with desired adjustments in the currency component of the money stock IXI. In . July I~I;X, .jW-·MI. "The Surplus Rcvl.!nuc," Tht' fopular ScieJlc#! MOII/hly, June I~KH. 145··9, FiliwlCe tJlkllJtml.ing. Rcprinted from the Shoe tliU/ Lemller R('/wrlt'r. lkccmbcr HiYO. 1551· 55, "The Unil Vatue ill AU Trade." t;lIgill"rillg Mug"zill", Augu'l 1~93, 555-67. "1),) We Need Siale Ilank CurrencyT Ellgilleerillg Ma!:uzille, July tg93, 497-505. lJ .. - - - - . Tht' /Jallking Pri"niJ/e, New Y~)rk: Rdorm Club Sl)UIlU Curn:lIl'y CUlIIllIiltce, IH95. 10. - - - - . "Cau!)c and RCJlIcdy lor Businc!)s Oepression." I:.'''gillt'erillg Magazille, July 1~96, 611-20, It. - - - , "The Philosuphy uf Money." nit! Mo"i.I·I, April 1~96, 337-50. t2. Bagehol, Waller. LOII/b"rd Street. London: Henry S. King, IM73. 13. Btack, Fischer, "Banking and Inle(esl Rales ill a World Wilhoul Money: The Elfeels of Unconll'lllle Discourse on the Life and Services of G. C. Ver";l'an~ (New-Y;;-;k-;-l870). -- ---74. See above 191. 75. Paine, Willis S., £E. Cit., 25. 76. The following is taken from a letter of Mann to A. C. Flagg, reprinted in Flagg, £E. cit., 40-42. Joseph Sabin in his Dictionary of ]looks relating to America quotes a pamphlet of Abijah Mann, Correspondence ~ A. Mann, Jr. and Hon. A. C. Flagg ••• (Brooklyn, 1868), but no copy could be' located. Mann in his vanity must have told people wild stories about his share in the Free Banking Act. Bonnefoux, once a Free Banker himself, relates: "Mr. A. Mann, Jr •••• vas, I believe, the gentleman who drew up the seven or eight sections which formed the groundwork of the Act to authorize the business of banking, passed April 18, 1838. The original idea, Mr. Mann told me, sprung from the example given by Mr. Stephen Girard WhO·, when he established his bank in Philadelphia, voluntarily hypothecated his immense property for the due payment of his circulating notes." Bonnefoux, Vindication, 11. 77. A vivid description of this attempted examination is given by Benton, £E. cit •• 458 ff. Se~ also u.S. £22 Congr., 1st ~., ~ Report 481, passim. 77a.On April 3, 1834 Biddle wrote to J. G. Watmough, "Tell me something about this Mr. Mann of New York," (Biddle Letter~).

------

Free Banking II

115

THE MOLDING OF AMERICAN BANKING

78. Flagg as Comptroller of the State of New York had to carry the law into execution. His interesting report with some documents added is in U.S. 25th Congr., 2!:!! ~., House Executive Document ££r, 2Pl ff. 79. David B. Ogden (1775-1849) vas a famous New York lawyer and scion of a socially prominent family of ironmasters. He played no important part in public life, but in 1838 served a term in the Assembly. 80. Report of 1849, quoted from the reprint in U.S. 31st ~., 1st ~., ~ Executive Document 68, 128-9. The cases arising out of the Free Banking Act are discussed by Bray Hammond, passim, and Livermore, ~. cit., 287 ff. Some of these pertain to the legal character of Free Banks. 81. It was one of the doctrines of Mercantilist banking theory that capital functioned exclusively as security for the note-holder. Thence, as mentioned before, MacVickar took his suggestion that bank capital should be "vested" and "from the banker's control." Based on the same Mercantilist idea American legislatures had developed the principle of limiting the issue of notes to twice or three times the capital. MacVickar and Lord went a step rurther: they proposed the limitation of note issues to the amount of capital; and since both suggested at the same time its investment in securities, notes and securities were tied together indirectly. The amount of the notes was "never to exceed the IlIIIOunt of their pledged stock," says MacVickar (page 38}. and according to Lord (page 84) the bs nks had to use their credit alone in dIscounting "to an extent not exceeding the amount of capital invested." Ricardo, who had freed himself from Mercantilist tradition, could be direct and logical from the beginning in suggesting the deposit of securities, "In some proportion to the amount of their issues." (See page 191). 82. In consequence thereof, to the extent that note issues were based on mortgages Free Banks in fact practiced Mercantilist banking on private credit. Individuals transferred mortgages to banking associations upon the condition that they received accommodation loans immediately. They melted down real property into symbolical money, as Sir James Steuart would have expressed it. (~Report of the New ~ Bank COmmissioners, for 1840, 9.)

83.

84.

85.

86.

87. 88.

~

of 1840, chapter 363, section 6. The amendments are reprinted in bill ~, 10 ff. Horace Greeley in his Recollections mentions Willis Hall as the man to whom the state is especially indebted for the Free Banking Act. (Pages 125-26.) Dixon Ryan FOX, in The Decline of Aristocracy in the Politics of New~, Studies in H1story, EconomiCS, and Public Law, No. 198 (New York, 1919), 4-5, misquotes the passage. A thorough investigation by the present suthor as well as by the reference librar1ans of the New York Public and the Albany State Libraries brought nothing to light to corroborate this statement. Willis Hall (1801-1868) was a Whig politician, an Assembly man in 1838, Attorney General of the state in 1839. He wrote a few political pamphlets and after 1848 was close to the Free Soilers. A. C. Flagg (~. cit., 43) criticized his wild and vague banking theories. If there is anything at all to Greeley's statement, Hall must have played his part behind the stage and may thereby indeed have had a meritorious influence on the development. This, however, is mere guess work. An Act to s.uthorize the business of banking, passed 4/18, 1838, ~ of the ~ of New York, passed at the 61st session (Albany, 183~ Most important are the sections 1, 2, 3, 4, 7, and 11. The history of the Act has been written pyMillard Fillmore in the document quoted above and by Helderman, ~. cit., 21 ff. Important are the ~ Reports of the New York ~ Commissioners, especially those of 1840 and 1841. O.S. 26th ~., 1st ~., ~ Execillve Document ~, 115 ff. and 2nd ~., ~ Executive Document 111, 122 ff. For the beginnings see also Ogden, Remarks, 28 ff. ~y Trust Company of New York, One ~ ~ of Banking ~ (pr~ vstely printed, 1939), 3-10. Curtis had been ca.hier of tHe Exchange Bank in PrOVidence, had played a part in public affair·s there, and was destined to become one of the leading figures in the young New York Clearing House. The first attempt to base a Free Bank in the state of New York on exeeptionally good security was made by L. Bonnefoux, ~. ill., passim. Principles of Political Economy, Second Part (Philadelphia, 1838) 257 ff. Financial Regisber, II, 10; Treatise, 201 ff, 243 ff, 308 ff, 321 ff. The

116

Free Banking II FOOTNOTES

last quotation may be found on page 202. 89. See especially, pages 124 ff, 131, 132, 136. 90. Hildreth, R., ~ ~ to his Excellency ~ ~ .Q!l Banking and the ~ £y (Boston, 1840), 14. 91. See especially, pages 147 ff., 153 ff.. , 200 fr. 92. See also the articles on "Free Banking" in ~ ~ XII, 610 fr., XIII, (1852) 127 ff., XIV, (1853) 28 fr., 151 ff. The case for and against Free Banking is summed up in this series, the author being opposed to it. 93. Kentucky and Virginia adopted the bondsecured bank-note rather than general incorporation in 1851 and 1852, respectively, being preceded in this respect by Michigan which in 1849 incorporated this feature into bank charters. In Maryland e Free Banking bill was defeated in 1852. Helderman, £Q. cit., 149 and 97. In 1857 it was suggested by the Superintendent of the Banking Department of the State of New York to' extend the principle of bond security for banknotes to the still existing Safety Fund banks. ~' MagaZine, XII, 615 ff. 94. The Louisiana Free Banking Act stipulated a thirty-three and one-third per cent specie reserve against deposits and a restriction of at least two-thirds of all loans to ninety-day paper, thereby becoming the most modern banking act of that time. The progress is here due, as it is so often, to a combination of two different sets of idea~. 95. An Act to organize and regulate banking aSSOCiations, No. XLVII, Acts of the Legisla ture of the ~ta te of MiChigan, passed at the annual session of 1837 (DetrOit, 1837), 76 ff. 96. Felch, £Q. cit., 114. The act was praised in 1837 by a contemporary author for "divesting the state of its undue control of banking." Rafinesque,.2l2..!. cit., 37. 97. ~higan Senate Journal, (1837), 45, 61, 147; ~ of the Michigan ~ of Representatives, (1837), 144. Biographies of Ellis and Lothrop in Michigan Biographies (Lansing, 1924), I, 272 and II, 36, respectively. 98. Hammond, Bray, £Q. Cit., 196.

297

Free Banking II

117

REFERENCES Benton, Thomas Hart. Speech on the Resolutions offered by Mr. Clay ... relative to the Removal of the Public Deposites from the Bank of the United States. (Delivered In the Senate, Jan. 2, 3, 6, 7, 1834.) Washington, 1834. Benton, Thomas Hart. Thirty Years' View or A History of the Working of the American Government for Thirty Years, from 1820 to 1880. New York, 1854. Benton, Thomas Hart. "On Banks and Currency" In Bankers' Magazine, Vol. XII (1857/58). Biddle, Nicholas. Oration Delivered before the Pennsylvania State Society of Cincinnati, on the Fourth of July, MDCCCXI. Philadelphia, 1811. Biddle, Nicholas, compiler. Commercial Regulations of the Foreign Countries with which the United States have Commercial Intercourse.

Washington, 1819. Biddle, Nicholas. Address delivered before the Philadelphia Society for Promoting Agriculture, at its Annual Meeting of the 18th of January,

1822. Philadelphia, 1822. Biddle, Nicholas, Mathew carey, John Sergeant ... [and others, inclUding Richard Peters, Jr.). Address to Those on Whom Heaven has Bestowed the Goods of Fortune and What is More Valuable, Hearts to Make Proper Use of Them for the Public Benefit. Philadelphia, 1824. Biddle, Nicholas. Eulogium on Thomas Jefferson, delivered before the American Philosophical Society on the Eleventh Day of April, 1827.

Philadelphia, 1827. Biddle, Nicholas. Documents from ... , President of the Bank of the United States to Hon. G.M. Dallas. 22d Congr., 1st sess., Senate Doc. 98 (1832). Biddle, Nicholas. Account of the Proceedings on Laying the Corner Stone of the Girard College for Orphans, on the Fourth of July 18:53: Together with the Address, Pronounced on that Occasion at the Request of the Building Committee. And a Description of the Plan of the College, by the Architect. Philadelphia, 1833. Biddle, Nicholas. An Address delivered before the Alumni Association . ... September 30, 1838.

Princeton, 1835. Biddle, Nicholas. Two Letters Addressed to the Han. J. Quincy Adams concerning a History of the Recharter of the Bank of the United States.

London, 1837. [Four further letters to the same were published in the newspapers and are to be found in the Financial Register, Vol. I, pp. 396, 342,400; Vol. II, p. 391.) Biddle, Nicholas. Reply to the Report of a Committee of the Bank of the United States of Pennsylvania. Paris, 1841. [Biddle, Nicholas.) Questions Submitted to the PreSident of the Bank of the United States by Mr. Cambreleng with his Answers Thereto.

[n.p., n.d.) Biddle, Nicholas. The Correspondence of Nicholas Biddle dealing with National Affairs, 1807-1844,

ed. by Reginald C. McGrane. Boston, 1919.

[Bronson, Isa.a.c.) An Appeal to the Public on the Conduct of the Banks in the City of New York, by a citizen. New York, 1815. [Bronson, Isa.a.c.) "General Propositions Explanatory of the Elementary Principles of Banking" In The Free Trade Advocate, ed. by Condy Raguet, Vol. II (1829). Flagg, A.C. Banks and Banking in the State of New York from the Adoption of the Constitution in 1777 to 1864. Brooklyn, 1868. Gallatin, Albert. The Writings of .... , edited by Henry Adams. Philadelphia, 1879. [Ga.llatin, Albert.) "Banks and Currency" In Amer. Quart. Rev., Vol. VIII (1830). Ga.llatln, Albert. ConSiderations on the Currency and Banking System of the United States. Philadelphia, 1831. Gallatin, Albert. Suggestions on the Banks and Currency of the several United States . ... New York,l84l. Hamilton, James A. Reminiscences; or Men and Evente at Home and Abroad during Three Quarter of a Century. New York,

1869. Hammond, Bray. "Long and Short Term Credit in Early American Banking," in Quart. Jour. Econ., Vol. XLIX (1934-35). Hammond, Bray, "Free Banks and Corporations: the New York Free Banking Act of 1838," in Jour. Pol. Econ., Vol. XLIV (1936). Hildreth, Richard. History of Banks: to which is Added a Demonstration of the Advantages and Necessity of Free Competition in the Business of Banking. Boston, 1837. Hildreth, Richard. Banks, Banking, and Paper Currencies. Boston, 1840. Hildreth, Richard. A Letter to his Excellency Marcus Morton on Banking and the Currency.

Boston, 1840. Jackson, Andrew. Messages. Boston, 1837. Jackson, Andrew, "Letters and Papers of ... " in Builetin of the New York Public Library, Vol. IV (1900). Jackson, Andrew. Correspondence of ... , edited by John Spencer Bassett. Washington, 1926. Lord, Eleazar. Principles of Currency and Banking. New York, 1829. Lord, Eleazar. On Credit, Currency and Banking. [2nd edition of Principles of Currency and Banking) New York, 1834. Lord, Eleazar. A Letter on National Currency addressed to the Secretary of the Treasury. New York, 1861. Lord, Eleazar. Six Letters on the Necessity and Practicability of a National Currency and the Principles and Measures Essential to it. New York,1862. [McVickar, John.) Hints on Banking in a Letter to a Gentleman in Albany, by a New Yorker. New York,1827. [McVickar, John.) "Importance of a National Bank," in The New York Rev., Vol. VIII (1841).

118

Free Banking II REFERENCES

Mann, Abijah. Correspondence between A. Mann Jr. and Hon. A.C. Flagg in Relation to the General Bank Law of New York. Brooklyn, 1868. Parnell, Henry Brooke, 1st Baron Congleton. Observations on Paper Money, Banking, and Overtrading . .. which explain the Scotch System of Banking. 2nd edition, London, 1829. Raguet, Condy. An Inquiry into the Causes of the Present State of the Circulatory Medium of the United States. Philadelphia, 1816.

Raguet, Condy. "Report on the Renewal of Bank Charters made to the Senate of Pennsylvania by ... Chairman on the 16th of January 1821," in Examiner, Vol. II (1834/36). Raguet, Condy. "Report of the Committee of the Senate of Pennsylvania, appOinted to enquire into the extent and causes of the present general distress," in Financial Register, Vol. II (1838). Raguet, Condy. A Treatise on Currency and Banking. London, 1839. 2nd edition, Philadelphia, 1840. Ricardo, David. Proposals for an Economical and Secure Currency; with Observations .... London, 1816.

Ricardo, David. Plan for the Establishment of a National Bank. London, 1824. Smith, Adam. An Inquiry into the Nature and Causes of the Wealth of Nations-with a Life of the Author and Introductory Discourse by J.R. McCulloch. Edinburgh, 1828. Steuart, Sir James. Principles of Banks and Banking of Money, as Coin and Paper ... London, 1810. (Being Book III of his Inquiry into the Principles of Political Economy, London, 1767.) Van Buren, Martin. The Autobiography of ... Washington, 1920. Venit, Abraham H. "Isa.a.c Bronson: His Banking Theory and the Financial Controversies of the Jacksonian Period," in Jour. of Econ. Hist .. , Vol. V (1946). [Young, Samuel]. Considerations on the Bank of the United States in which its Repugnance to the Constitution Albany, 1832. Young, Samuel. Oration Delivered at the Democratic Republican Celebration of the 64th Anniversary of the Independence of the United States. July 4, 1640. New York, 1840.

[7] HUGH ROCKOFFo

The Free Banking Era A Reexamination INTRODUCTION

During the two decades preceding the Civil War the regulation of banking in the United States was left to the states. Various schemes were tried, but the most famous and most controversial was free banking. The term free banking meant something very specific at this time: it meant a banking system with free entry and a bond-secured note issue. Free entry provided that any potential banker who could raise a certain minimum of capital could start a bank wherever he chose. Under the older system of chartered banking, the potential banker had to secure a special grant from the state legislature. The bond security provision of the law worked in the following way. Banks were allowed to issue paper currency redeemable in gold or silver under free banking as well as under other regulatory systems. But under free banking, designated government bonds had to be deposited with a state authority as security for all circulating notes issued by a bank. The bank, so long as it remained solvent, was entitled to the interest on the bonds. But should it fail to honor its notes, the state would sell the securities and reimburse note holders out of the proceeds. Free banking was intimately related to wildcat banking, the formation 'I am indebted to the members of my Ph.D. thesis committee, Robert W. Fogel, Donald McCloskey, and Richard Zecher for potent doses of help and encouragement. I have also benefitted greatly from criticisms made by Stanley Engerman and Roger Hinderliter. They are not, of course, responsible for any remaining errors of fact, interpretation, or conclusion. This caveat applies with special force to the first three named, as this paper has changed somewhat since they were most intimately connected with it. HUGH ROCKOFF

is assistant professor of economics at Rutgers University.

120

Free Banking II 142

MONEY. CREDIT. AND RANKING

of banks with note issues of far greater volume than they could hope to continuously redeem. While something similar to wildcat banking could and did happen under the older system of chartered banking, it was less likely because it required the initial cooperation of the legislature. Judged by its ability to survive, the free banking law proved reasonably successful. Free banking laws were first passed in Michigan in 1837 and in New York and Georgia in 1838, an understandable response to the destruction of the Second Bank of the United States, and the resulting increase in the importance of state laws in regulating the banks. On the eve of the Civil War over half the states, including the most populous, had free banking laws. As late as 1858 free banking laws were passed in Iowa and Minnesota, and Pennsylvania adopted free banking in IR60. Michigan's legislature passed a free banking law in 1857, even though an earlier law produced disaster. Thus, the National Banking Act passed during the Civil War, a sort of national free banking act, was the continuation of a movement that was well underway and perhaps accelerating in the decade preceeding the Civil War. But traditionally, historians have judged the experiment, at least in the western states, a failure. However, despite the attention paid to American monetary history during the past decade, this episode has not been reexamined. The purpose of the present study is to begin such a reexamination in the light of the extant quantitative data, frail as it is, and modern monetary theory. Our primary effo.rt will be aimed at filling three omissions in traditional interpretations. The first omission is an explanation of the diversity of experiences under superficially similar free banking laws. The most striking contrast is between the free banking laws passed in Michigan and New York in the late 1830s. These laws appear very similar on the surface and their historical roots are entwined. Yet free banking in Michigan was a disaster giving rise to some of the most famous stories of wildcat banking, while the New York law (after some initial rough sledding) was the basis for one of the most successful banking systems of the ante-bellum period. No serious evaluation of free banking can be completed without an explanation of why some free banking laws exploded in "hyper-inflations" while others did not. Some writers have hinted that the source of instability was the security provisioll liS, p. 164; 19, pp. 55-56J. However, these suggesliolls do 1101 constitute an analytically complete statement. This issue is treated in section

2. A second important omission from traditional interpretations, at least from a modern point of view, is that they make no attempt to quantify the damage produced by wildcat banking. This issue is treated in sections 3 and 4. A third omission from traditional interpretations is the failure to ask whether free banking could have produced any benefits for the

Free Banking II

121 HUGH ROCKOFF

:

143

states which adopted it. We discuss one possibility, a more efficient allocation of bank capital, in section 5. 1 Before plunging into these issues however, it will be useful to examine the behavior of a bank-issued currency.

I.

THE ECONOMICS OF A BANK-IsSUED CURRENCY

Dllfing the free banking era, as we have noted, paper currency was issued by banks. The specie price of bank notes varied depending on the location and condition of the issuing bank. This state of affairs has attracted considerable attention because it appears to differ so greatly from our present system. The market's response, generally, was a system of private brokers who exchanged bank notes for specie, discounting the notes as a charge for their services. The discounts they charged were published in "bank note reporters." These were periodicals which listed each bank by state and city, and the discount (as a percentage of the face value) on its notes currently prevailing in the financial center where the reporter was published. The reporter also described any counterfeits. The specie paying banks would typically be listed at a small discount from par. Another group of banks was the "broken" banks, and the notes of these would often not be acceptable at any price. Some observers have deduced that the ante-bellum banking system was chaotic from a count of failed banks and counterfeits in bank note reporters [26, pp. 177-78]. However, any deduction about the state of the currency must be made carefully because the reporters generally listed counterfeits and bank failures even if the notes had been removed from circulation years before. The reason for this was that someone who had managed to save a stock of worthless notes might reintroduce them at some later date. A detailed comparison of a pair of bank note reporters by the same publisher issued in December of 1846 and November of 1843 revealed that slightly more than ninety percent of the banks listed as broken, failed, closed, or fraudulent in the 1846 reporter had also been listed in the earlier issue. In this particular case the danger in using bank note reporters for making inferences concerning the state of the ante-bellum banking system is similar to making inferences about the state of the banking system in the 1940s from a list of failed banks going back to the early part of the depression. Table 1 displays the nationwide pattern of discounts for selected dates. It shows that most banks were at small discounts from par, discounts that decreased over time, perhaps as a result of improvements in transportation and communication. 'The classic histories of nineteenth century banking [13,26], generally eschew 'luantitalive They tend 10 concentrate on the safety of bank notes and deposits L2S].

mea~urement.

Free Banking II

122 144

MONEY. CREDIT. AND BANKING TABLE I Discoums ON BANK NOTES AT PHILADELPHIA, SELECTED DATES, 1845-58 (Percents)" November

November

November

November

November

November

1845

1850

1855

1856

1857

1858

State

New England New York New Jersey Pennsylvania Delaware Maryland Vir~il1ia

North Carolina South Carolina Georgia Alabama Mississippi t Louisiana Tennessee Kentucky Missouri Ohio Indiana Illinois Michigan Wisconsint

1/2 3/4 3/4 0 0 1/4 i

1-1/2 1-1/2 2 5-1/2 80 2 2-1/2 1-1/2 1-3/4 2 2 70

65 ?

3/8 3/4 0 0 0 1/2 3/4 1-1/4 1 I

3-1/2 75 1 2-3/4 1-1/2 1-1/2 1-1/2 1-1/2 75 3 ?

1/4 1/2 1/4

1/4 1/2 1/4

0 0

0 0

1/2 I 1-3/4 I I

1/2 I 1-1/4 3/4 I

.5

.5

? 1/4 2-1/2 I I I 5 2 1-1/2 2

? 1-1/4 5 I I I 5 2 1-1/2 2

3/4

3/8 1/2 1/4

I

3/4 0 0

(J

0

.5

1/2 5/R

10 10 10 10 5

I

?

.5 20 10 10 10 10 20 10 20

1 3-1/2 ? 3/4 1-1/2 3/4 I 3/4 2 1-1/2 1-1/2 1-1/2

·Sources: Various iS5ue~ of Van Courl',~ Cou"terfeit Detrctor and Banle Note Lbt. tThe ? for Mio:;sissippi and Wisconsin indicates th;Jt these notes were of doubtful value, and only purchao:;ed under "recial circumstances. The tahTe gives the mooal discount.

An implication of this pattern of discounts is that it is unlikely that there were large fluctuations in the rate of exchange among regions during the free banking era except during the suspensions of specie payments in the early I 840s and late 1850s. This would seem to eliminate the possibility of regional devaluations curing incipient unemployment regularly. However, further research on this point is warranted. The devaluations that could take place during suspensions (note the large discounts in November, 1857) might have been sufficient to have reduced unemployment when the competitive position of particular regions deteriorated. Whatever the benefits from flexible exchange rates, a heterogeneous currency did make exchange less efficient. Each time a transaction took place the seller had to make some judgment about the quality of the particular set of bank notes being offered. Clearly the process of making this judgment used real resources, particularly labor, that would not have been used if the currency were homogeneous. Moreover, because of the inconvenience of the paper currency, people were led to use larger amounts of specie than otherwise. In 1859 the ratio of specie outside the treasury to the total money supply was .41. Twenty-five years later, under the National Banking system, the ratio was only .15 [12, pp. 7, 224-25; 32, p. 648]. Nevertheless, it seems unlikely that the heterogeneous nature of the

Free Banking II

123 HUGH ROCKOFF

:

145

currency was a major brake on economic growth, for in many crucial respects the system was little different from that which prevails today. Locally we use demand deposits. But these are not generally acceptable as a means of payment. Each time we wish to make a purchase by check from a businessman we force him to make some judgment about the quality of the money we are offering. Instead of having to worry about different kinds of bank notes a merchant today must worry about different kinds of deposits which could be as numerous as his customers. Counterfeiting currency is now rare, but forged checks and insufficient balances are a constant irritation. Yet no one today would argue that the heterogeneity of our deposit money is a serious impediment to the growth of national income. This analogy can be extended. Since deposit money is not generally acceptable it is more convenient when traveling to use currency or traveller's checks. During the free banking era people used bank notes locally but switched to specie (or possibly the notes of some well-known bank) when traveling. Today without much complaint we pay a small charge for travelers' checks. During the free banking era one had to pay a similar charge to a broker who exchanged "foreign" money for the local currency. We do not wish to argue that the system worked perfectly. A national currency might have been preferable. But in the light of the close similarity between the workings of the ante-bellum system, properly understood, and our present-day system, the inefficiency of a heterogeneous currency should not be exaggerated. It should also be remembered that free banking did not add to the heterogeneity of the currency except as it led to an increase in the number of banks. To the contrary, free banking greatly reduced economic heterogeneity by standardizing the assets that banks held against notes. In New York the particular kind of security backing the note was stamped on it. It was even possible under free banking for all of the free banks to use the same type of note in order to inhibit counterfeiting.

II. THE CONNECTION BETWEEN FREE BANKING AND WILDCAT BANKING 2 Wildcat banking was made possible by the free entry provision of the free banking law. But whethe~ it was profitable depended on the design of the bond security provision. Let us consider several cases. First, suppose the market value of the eligible bonds was less than the "legal value" (the value of the notes that could be issued on a bond). Under this condition it obviously paid to set up a zero reserve bank providing the notes could be placed in circulation at close to par before the enterprise went bankruot. 2This section was stimulated by [5, pp. 86-95}.

124

Free Banking II 146

:

MONEY, CREDIT, AND BANKING

The "income statement" of a wildcat bank of this sort with a circulation of $100 might look like this: Market value of notes issued -Market value of bonds (or mortgages) deposited Net Income

$100 90

$ 10

This simple model explains several important cases of wildcat banking. In Michigan (in 1837) mortgages on land were eligible and they were accepted at par [2 I, p. RO]. It was thus possible to create a mortgage on a worthless piece of property, have it certified as being valuable by some friends, and then transfer it to a wildcat bank in exchange for a mass of bank notes. This seems to be the process by which many of the Michigan wildcats issued currency. Others were simply frauds which operated in violation of the free banking law. It should be noted, however, that there was an important difference between Michigan's experience with wildcat banking (the nation's first) and later episodes: Michigan's occurred during a legal suspension of specie redemption. Thus, the situation in Michigan was unique, a group of men could issue bank notes with practically no cost to themselves and unchecked by the need to redeem the notes in specie. However, while the free banks were not required to redeem their notes in specie they were required to have 30 percent of their authorized capital on hand in the form of specie when they began operation. The often quoted complaint of the Michigan bank commissioners, "gold and silver flew about the country with the celerity of magic; its sound was heard in the depths of the forest, yet like the wind one knew not whence it came or whither it was going," [36, p. 1129] referred to attempts by the wildcat bankers to evade this requirement. 3 It is interesting to note that some 20 years after her first disastrous experiment with free banking Michigan passed a second free banking law. The eligibility requirements were far more strictly drawn than for the first [23, p. 366]. The monetary statistics of the resulting banking system clearly show that there was no wildcat banking. Michigan's two free banking laws were not part of a controlled experiment. Nevertheless, the contrasting experiences under the two laws appears to be strong evidence that the source of wildcat banking was to be found in the provisions of the free banking law rather than in something else, for example, in the conditions of frontier life. Indiana's experience with wildcat banking was perhaps another example of the simple model. According to the free banking law, bonds were to be accepted only at the minimum of the market and par values, so that there does not seem to be room for a banker to have received notes J

It has been quoted by [14, p. 28J among many others.

125

Free Banking II HUGH ROCKOFF

:

147

worth more than the bonds deposited as security [18, p. 24]. Sut according to one authority the auditor may have valued Indiana bonds at par [2, p. 172]. This could have been done because the auditor was not aware of the significance of the distinction between the par and market values. In a "currency reform" which took place in 1855, the security provision of the law was considerably rewritten and the requirement was changed so that banks received at most $91 dollars worth of notes for every $100 dollars of bonds deposited [18, p. 34]. No wildcat banking took place after this change in the law. The clearest case of the simple model occurred in Minnesota [25]. The initial version of Minnesota's free banking law was passed in 1858. In this version only bonds of the United States, Minnesota, and other states approved by the banking authority were eligible. Soon after passage, however, this part of the law was amended to extend eligibility to the Minnesota State Railroad Bonds. These were to be issued on behalf of certain railroads in exchange for mortgage bonds. When the railroads applied for the bonds due them under the law, the governor refused to issue the bonds unless a first lien on the assets of the railroads was given to the state. The railroads refused, and took their case to the Supreme Court of Minnesota. The ruling was in favor of the railroads. The nature of the ruling along with threats of repudiation made in certain newspapers led to a rapid depreciation of the bonds in the market. This need not have caused wildcat banking, because under the free banking law the banking authority was empowered to reduce the price at which he accepted depreciated bonds. Under the circumstances, however, the banking au thori ty was reluctant to take this step because it might appear to be a repudiation of the state's debt. The railroad bonds, virtually worthless in the market, were accepted at 95 percent of par. The result was wildcat banking. There is certainly one and perhaps several episodes of wildcat banking which do not seem to fit the simple model we have been using. However, a natural extension of it seems sufficient to cover these cases. Suppose that a wildcat bank could expect to survive for as long as, say, one year. Then, even if the value of the notes issued was no greater than the market value of the bonds deposited, the interest earned on the bonds might be sufficient to induce wildcat banking. The "income statement" of such a bank with a circulation of $100 might be as follows: Market value of notes issued + interest on bonds + surplus returned to shareholders

$100 6 4 $110

- Market value of bonds deposited

$104 Net income

$ 6

126

Free Banking II 148

:

MONEY. CREDIT. AND BANKING

New Jersey's episode seems to fit this model. The free banking law of New Jersey passed in 1850 provided that bonds of the United States, New Jersey, or Massachusetts could be deposited. The value of the notes that could be issued was set equal to the par price of the eligible bond with the additional provision that the market price be at or above par. There were no signs of wildcat banking during the first year of operation. However, in IR51 the free banking law was amended to permit banks to deposit bonds isslIed by New York, Ohio, Pennsylvania, and Kentucky. These were all somewhat closer to par although above it. Finally, the law was altered in 1852 to permit banks to deposit Virginia's bonds [10, pp. 52-64]. The latter amendment soon provided a bitter lesson in the danger a state faced if it linked its currency to the debt of another state. About one year earlier Virginia had passed a law which permitted her government to issue debt from time to time. In subsequent years Virginia greatly increased her debt. The largest addition was made in 1853 when some $4.6 million was issued, surely one of the largest deficits of any state in the ante-bellum period [34, p. 554]. Part of this debt was absorbed by New Jersey's free banks during a rapid expansion of the free banking system. Wildcat banking in New Jersey was similar in style to that in the western states [8, pp. 76-77]. It appears that wildcat banking could have been prevented if the laws had required a free bank that failed to pay damages to note holders in addition to the face value of the note, and if this protection had been assured by demanding that the value of the bonds deposited exceeded the value of the notes issued by an amount sufficient to pay slIch damages. A damage rate which suggests itself on a priori grounds would be the normal rate of interest, since this is what the note holder could have earned had he invested in some safe alternative. So far we have discussed wildcat banking in a very different way from what is usual. We have emphasized the bond security provision and state debt policies. We have not even mentioned the remote areas in which wildcat banks were located, or the various tricks used to keep notes in circulation. Of course wildcat banks used these tactics. The more notes they could get into circulation, the greater their profit. However. it is not usually recognized that in at least two ways the bond security provision contributed to the tendency of banks to locate in hard-to-reach places. First, under free banking, the state banking authority was empowered to sell all of the securities of the bank if a single note holder was refused specie. Thus, the banker who did not want to honor demands for specie had few alternatives to locating his bank as far as possible from the principle area of circulation. A second factor was the knowledge each note holder possessed that should a bank fail all of the notes would be redeemed on a pro rata basis out of the proceeds from the sale of the deposited bonds, and usually in the banking authority's office in the state capital

127

Free Banking II HUGH ROCKOFF

:

149

or in the state's major commercial center. Thus each note holder separately had an incentive to let someone else bear the costs of seeking out the wildcat banker, and, being refused specie, of informing the banking authority. The framers of the free banking laws were apparently walking a tightrope. If they made the gap between the market price of bonds and the "legal price" small or negative they risked, as we have seen, wildcat banking. If they made the gap slightly larger they would produce a "sound" free banking system as was the case in New York, Ohio, and Louisiana. If they made the gap still larger the free banking law could prove abortive. To take one example of the latter phenomena, Massachusett's free banking law was passed in 1851 but no banks were organized under it until 1859. Massachusetts limited the bond holdings of free banks to bonds of states in New England (which had small debts), and to bonds issued by towns in Massachusetts. The state securities sold at substantial premiums in the 1850s. Yet the par price of the bonds was the legal maximum. Moreover, the bond had to be made equal to one yielding 6 percent, so its par value, for the purposes of the law, was reduced by one-sixth if it yielded 5 percent. As the decade drew to a close a number of free banks were started on the basis of securities issued by towns in Massachusetts. These securities were, presumably, somewhat cheaper since they were more liable to default.

III.

LOSSES ON BANK NOTES

A bank failure was, in the first instance, a transfer from the note holder to the wildcat banker that left the net wealth of the community unchanged. But the uncertainty and inconvenience caused by wildcat banking could have produced decreases in total real income. In this section we examine the redistribution of wealth, while the efficiency effects are dealt with in section 4. Table 2 presents an estimate of the losses ultimately suffered by holders of free bank notes through the year 1860. It includes losses due to wildcat banking as well as losses due to ordinary mismanagement and bad luck. For the most part the estimates were compiled from standard secondary sources. When, as sometimes happened, somewhat arbitrary adjustments had to be made, a procedure which tended to bias the estimate upward was followed. In interpreting Table 2 it is useful to distinguish between the states which had "sound" free banking laws, that is, with adequate bond security provisions, and those which did not. In the former, with the exception of New York, losses were mild. In New York most of the losses came in the early years of the law when the security provision allowed the bonds of states besides New York to be used.

128

Free Banking II

150

MONEY. CREDIT. AND BANKING TABLE 2 LoSSF.s SUFI'F.RF.D BY HOI.DF.RS OF FRF.F. BANK NOTF_~ FROM TIIF. FIRST YEAR OF FREE BANKING THROUGH 1860*

State

Vermont Massachusetts Connecticut New York New Jersey Pennsylvania Ohio Indiana Illinois Michigan

First Yea,

1851 1851 1852 1838 1850 1860 1851 1852 1851 1837

Loss (dollars)

24,500 0 0

394,700 6,000 0 77,600

227,900 21,300 1,000,000

State

Michigan Wisconsin Minnesota Iowa Georgia Florida Tennessee Alabama Louisiana

First

LO!li~

Year

(dollars)

1857 1852 1858 1858 1838 1853 1852 1849 1853

-t

Total

1,851,900

0

96,900 3.01K) 0

0

·Each estimare is rounded off tJ hundred •. The dates refer to the years in which the free banking laws were p'"ssed and are laken from lhe Slalule. of lhe various slale •• In some stales additional losses occurred during Ihe Civil War due. generally, to the depreciation of southern bonds. For a list of leading references and assumptions used in estimating lhe losse•• see Ihe Appendi •. t-signifie. Ihal hllie or no banking was done under lhe fr.. banking law.

The experience under wildcat banking was quite varied. In Michigan the notes in many cases, although not all, became worthless. If the estimate in Table 2 were correct, the total volume of wildcat money would have amounted to about II percent of the annual income of Michigan in 1840 [9, p. 98]. However, while the bank commissioners refer to $1,000,000 as a low estimate, it is likely that they had in mind the face value of the wildcat issues. Even from the first, many of these notes may have borne heavy discounts for it appears that the public caught on rather quickly to the condition of the wildcat banks. The bank commissioners had already closed a number of banks and had officially reported their findings in March of 1838. Disenchantment was sufficiently widespread by April of 1838 to induce the legislature to suspend the free banking law [22, pp. 246-47]. Thus, the true condition of the wildcats was surely common knowledge sometime early in 1838. Since few of the wildcats were organized until after the general suspension of specie payments in June of 1837, the life span of the typical Michigan wildcat was about six months at the most. The episodes of wildcat banking which occurred in the 1850s were apparently not as costly to the note holder. In Indiana, for example, it appears that losses were frequently less than 5 percent. Moreover, many of the notes were probably accepted initially at some discount from their face value. It is even conceivable that some note holders made money from the wildcat bankers by taking the notes from the banker at a larger discount than prevailed when the notes were finally redeemed by the state. Seventeen free banks, a substantial number for a frontier state such as Indiana, survived to the Civil War. Today, after nearly three decades with almost no bank failures, we might

Free Banking II

129 HUGH ROCKOFF

:

151

regard the failures that occurred in Indiana as catastrophic, but contemporaries took a more tolerant view. To illustrate this we quote at length the state auditor's view of the damage: The experiment of free banking in Indiana, disastrous as it has been in some particulars, has demonstrated most conclusively the safety and wisdom of the system. The original bill was crude and imperfect, admitting of such construction as held out to irresponsible men inducement and facilities for embarking largely in the business of banking, without the ability to sustain themselves in a period of revulsion. That revulsion came. . . and yet the loss to which the billholder was necessarily subjected. in many cases, did not exceed five percent, and in no case exceeded twenty percent of the amount in his hands. [37, pp. 183-84.]

The other episodes of wildcat banking were more similar to Indiana's experience than to Michigan's. Few reliable estimates comparable to those in Table 2 exist for the non-free banking sectors of the. banking system, or for deposits.4 The upshot is that our estimate must be judged as it stands. Nevertheless, it seems to be a rather small number. It means that by 1860 note holders had probably lost less through the failure of free banks, including the wildcats, than they stood to lose in that year from a 2 percent inflation [12, p. 225]. Undoubtedly, this estimate could be considerably refined. But it seems unlikely that the unearthing of new data will require a substantial upward revision.

IV.

TOWARD A MEASURE OF THE EFFECT OF WILDCAT BANKING ON THE EFFICIENCY OF EXCHANGE

We would like to know whether in addition to the redistribution of wealth to which Table 2 is addressed, there was also a decrease in the income of the community as a whole. Our analysis follows Bailey's examination of the cost of anticipated inflation [1]. To explicate the argument we will use Figure 1. Here the cost of holding bank money in cents per dollar per year is measured along the vertical axis, while the amount held is measured along the horizontal axis. The vertical distance between the demand curve, DD and the horizontal axis is a measure of the marginal productivity of money. The area under the DD curve is thus the total value of monetary services. This relationship permits a deduction of the efficiency costs of wildcat banking. Suppose that the cost of holding money given a "sound" banking system would be Ca' This cost might be simply the interest on U.S. treasury bonds, a safe alternative to money. Under 4 An estimate by Jay Cooke placed losses on all bank nOles al $50 million per year [24 vol. I, p. 327]; also quoted in 6, p. 21].

r

130

Free Banking II 152

MONEY, CREDIT, AND BANKING

Cost of

Holding

o

o Real Cash Balances

FIG. I. The Demand for Money

wildcat banking, costs would be higher, say C b , because from time to time a money holder would expect to discover that part of the money he held had been issued by wildcat banks. For example, if the yield on U.S. bonds were 5 percent, and if a money holder expected that over the course of a year 5 dollars out of a 100 would turn out to be worthless because they were issued by wildcat banks, then the total cost of holding money would be 10 percent per year. Returning to Figure I, people would wish to hold only M b of real money balances under a regime of wildcat banking. The cross hatched area in Figure I, the expected loss rate (C b - C a) multiplied by the total amount of money held, M b' is the amount of wealth that people anticipate will be lost due to wildcat banking. We can consider the cross hatched area a pure transfer from note holders to bankers leaving the wealth of society as a whole unchanged. However, this transfer produces a loss in efficiency. The value of the services produced by the money which is not held under wildcat banking but which would be held under a "sound" banking system is the entire area between the DD curve and the horizontal axis bounded by M band M a' The social cost of these services is the product in alternative uses of the resources used in producing this amount of money.

131

Free Banking II HUGH ROCKOFF

:

IS3

For simplicity, assume that the social cost is simply the amount of specie held by banks multiplied by the interest rate-hence it is represented by the line C xr where r is the reserve ratio. A more complete analysis would " notice of the use of other forms of physical capital (bank have to take buildings and furnishings) and labor in producing money. The efficiency cost of wildcat banking, then, is the stippled area in Figure 1, the total value of the services of money that society loses, less the resources saved when the production of monetary services is reduced. It is important to emphasize that the stippled area is a measure of the amount of resources saved by the use of money which would otherwise be used in precisely those activities which historians have designated as having been effected by the quality of the currency in the free banking era. The labor used by merchants in carefully examining the currency offered to them is the most frequent example. Another is the inconvenience experienced by travellers who had to convert their home bank notes into notes circulating in the region they were visiting. Perhaps most important is the return from productive capital which is foregone because people must hold larger balances of specie. To directly apply the logic embodied in Figure 1 we would need to know the stock of bank money, the expected loss from wildcat banking, and the elasticity of the demand for money with respect to the cost of holding it. Unfortunately, direct estimates of the latter two are out of the question. Forming an estimate of the expected loss from the data fragments we possess would involve making extremely arbitrary assumptions about how expectations were formed. Moreover, the elasticity of demand cannot be estimated because regional interest rates have not been developed for the ante-bellum period. However, we were able to estimate a rough substitute for the true demand function which allows one to obtain an idea, albeit an imprecise idea, of the losses from wildcat banking. This substitute is a cross-state demand function which includes a dummy variable that takes the value 1 in states which had experiences with wildcat banking and 0 in states which did not. The coefficient on this variable will be a product of the elasticity of demand and the cost of holding money, provided that we have taken account of the other important variables. The actual regression that was run was an ordinary least squares regression of the following form (I)

where M W

the stock of bank money per capita, wealth per capita,

132

Free Banking II

154

MONEY. CREDIT. AND BANKING

U FJ

E

n, ~J' ~2' ~3' and ~4

urbanization, a dummy variable which state experienced wildcat not, a dummy variable which state had a "sound" free it did not, a random error term,

takes the value 1 if the banking and 0 if it did takes the value 1 if the banking system and 0 if

the coefficients to be estimated.

The use of wealth was suggested by demand for money studies based on modern data which indicate that wealth or income is an important variable. Urbanization was considered a good proxy for the demands for money generated by commercial and industrial activity. However, this interpretation is open to challenge, urbanization may be a proxy for variables effecting supply. For example, a greater density of banks may induce people to hold more money. Or it may even be a proxy for our unobserved regional interest rates. But even if one of these alternative interpretations apply we can still use the reduced form equation to estimate the monetary deficit in the wildcat banking states. Experimentation with a number of other variables failed to reveal a strong relationship with per capita money holdings. The F2 dummy was added so that we could compare wildcat banking directly with chartered banking. The equation was estimated for the year 1860. This was the only year in the ante-bellum period that met the twin requirements that it be a census year so that wealth estimates would be available, and that it follow the major wildcat banking episodes. In fact, the latter requirement was only partially met for Wisconsin and Illinois. Money was defined as the sum of deposits and currency (net of currency held by banks) for the year 1860. The monetary data was taken from the summaries given in the Report of the Comptroller of the Currency for 1876. Since the Report generally gives the data for the end of the preceding year, the comptroller's series were backdated one year. After deleting those states for which the data was obviously incomplete, we had a sample of 27 observations. The nominal money holdings were then deflated by the number of free people [32, pp. 12-13; 33, p. 7]. The wealth index was constructed by summing the state by state estimates of real and personal property given in the census of 1860 [30, p. 319]. This index was then deflated in the same way as the money index. The urbanization variable is simply the percentage of the total population living in communities with population greater than 2,000 [31, pp. (1-30)-(1-37)]. Two specifications of the basic equation are presented below. One uses a broad definition of the states in which free banking produced wildcat

Free Banking II

133 HUGH ROCKOFF

:

155

banking, viz. Florida, Tennessee, Indiana, Illinois, Michigan, and New Jersey, (variable F I ) and a broad definition of the states with sound free banking systems, viz. New York, Ohio, Louisiana, and Virginia, which adopted a conservative variant of free banking (variable F1 ). A second specification used a narrow definition of wildcat banking including only those states which experienced intense episodes, viz. Indiana and Michigan (variable Fi), and a narrow and more proper definition of those states with sound free banking systems, excluding Virginia (variable These regressions are reported below along with the t statistics in parentheses and the R2 statistics corrected for degrees of freedom.

Fn.

M =

-S.56 + .017W + 54.02U - 6.55F. + 10.32 F2 (-1.97) (4.79) (5.64) (--I.S6) (2.56)

= .72 M = -9.47 + .01S W + 55.91 U - 7.S1 Fi + IO.SSF!

(2)

R2

(- 2.14)

R2

(4.62)

(5.63)

(-1.67)

(4.S0)

(3)

= .69

Other definitions of F. and Fl produce similar results. If we use the coefficient on F. or Fi to estimate the loss from wildcat banking we get a rather surprising result. Taking a linear approximation, the stippled area in Figure 1 is given by the following formula,

where L stands for the loss from wildcat banking. If we assume, for example, that the risk free rate of interest, c", was 10 percent, that the expected loss from wildcat banking, c b - c", was 10 percent and that the reserve ratio r was also 10 percent-three assumptions which bias the estimate upwards-and use the coefficient on F. as an estimate of Mb - Ma' then we get the following results:

L

= 6.55 [.10(1 - .10) + .5(.10)] = $.92

If we use the coefficient

(5)

Ft the result is L

= $1.09

(6)

Thus, the residual effect of wildcat banking was to lower income per capita in the effected regions by about $1.00. This finding, however, can be subjected to a number of substantial

134

Free Banking 11 156

:

MONEY. CREDIT. AND BANKING

criticisms. For example, wildcat banking may have affected the reserve ratios of the banks-banks may have been forced to hold greater reserves because people were more likely to return their notes if signs of trouble developed. In terms of Figure I wildcat banking may have shifted the Cnxrcurve upward. Thus, a calculation of the loss that included the cost of holding greater bank reserves might lead to a different conclusion. And, indeed, there did exist substantial interregional differences in bank reserve ratios of a sort that would lead one to believe that western banks had to hold large specie reserves because the public was worried about wildcat hanking. Introducing banking reserve ratios explicitly into the analysis, however, involves two vexatious problems. First, the effect of other economic variables on bank reserve ratios, such as interest rates or the distance between banks, must be taken into account. Second, and more important, the relevant ratios are not known. Gold and silver are clearly what count for the economy as a whole, since the economy must give up real resources to secure these and only these reserves. But when we examine separate states, notes or other assets issued in other states have the same property. In general we do not have information on the net balance of bank notes or other assets between anyone state and the rest of the economy. For these reasons we have not been able to isolate the effect of wildcat banking on the reserve ratios of banks in the effected regions. Perhaps the most serious objection to the analysis is that it involves comparing one ante-bellum banking system with another. The real point at issue, it might be argued, is whether the heterogeneity of the currency led to losses in comparison with the centrally-directed monetary system that existed before 1840 and the federally regulated system that existed after the Civil War. We can still use Figure I. But, now the horizontal axis records bank money balances for the economy as a whole. The cross-hatched area would represent the expected loss due to the discounting of bank notes used in interregional trade. However, this area could no longer be interpreted as a pure transfer, since the discounts would in part reflect the cost of the resources used in handling uncurrent money. The situation is similar to the analysis of a tax on an ordinary commodity when the proceeds of the tax are spent on socially wasteful activities. Stanley Engerman has attempted to measure the effect of the destruction of the Bank of the United States on the economy by determining what national income would have been in the 1850s had the economy been able to operate on the same specie-money ratio that prevailed in the 1830s, while maintaining the stock of money that existed in the 1850s [II J. U si ng this framework, Engerman showed that the cost of destroying the Bank of the United States, orfrom our point of view the cost of the late ante-bellum banking system, was small, about .15 of I percent of national income on an annual basis. It is true that per capita holdings of bank money

135

Free Banking II

157

HUGH ROCKOFF

TABLE 3 BANK PROFITS IN NEW YORK BOSTON. AND PHILADELPIIIA.

1849-59·:r DIVIUENOS AS A FRAcrlON O~ Ncr WOK"H

DI YIOENDS AS A FucnON Of THE PAR VALUE Of' CAPITAL

Year

1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859

New York

Boston

Philadelphia

New York

Boston

Philadelphia

8.79 8.70 9.38 9.03 8.85 8.87 9.08 8.61 7.73 6.91 7.56

8.06 8.36 7.82 7.78 8.08 8.65 7.98 7.81 7.73 7.43 7.31

9.79 10.60 10.30 10.27 11.13 11.40

7.63 7.33 7.51 7.69 7.84 7.86 8.()8 7.72 6.91 6.20 6.78

7.33 7.33 6.88 6.87 7.32 7.71 7.09 7.10 7.()0 6.76 6.70

8.14 9.04 8.61 8.70 9.30 9.54 9.52 9.()() 6.111 6.74 6.74

II.!K)

10.26 7.12 8.03 8.03

·The first three columns are unweigtued averages for all bunks in the cily. The second columns are the corresponding column to the left multiplied by the aggregate ralia of par capital to par capital plus surplus. tSources: I. Dividends. Boston; Joseph G. Marton. A Century 0/ FilWnce (New York: Greenwood Press. 1969) pp. 99-101. lOB-II. New York. 1849-51: Hunt', Merchant,' Magaline. vol. XXIIl. p. 89; and vol. XXVII. p. 92; 1852-59: Banker's Magazine, vol. XIV. pp. 556-57. Philadelphia, 1849-55: Banker's Muga1.ine, vol. X, p. 978; 1856-59: Communicatiun. (J{ the Auditur General 0/ Penn5ylvania Rdafive 10 Banks tmd Sewings Institutions, 1856, 1857, 1858, and 1859, passim. 2. Cup;tal and Surplus. Boston and New York; Annual Report of the Set'rttary of fhe Treasury on tile Condition of lhe SUIre lJanks (in Iht: Uouse Congressional Documents), !H49-18.59. Philadelphia, annual Report of the Sec:rdury 0/ .he TrOC:umenl 122. 32nd Congre... lSI Se.. ion. pp. 7R. 79. "R-127. 1%-20.1. U.S. Congre.. Ho"se Documenl 77. 361h Congr.... 2nd Se •• ion. pp. 9R. I~O. 159. 171-1R4. Stock price.

are from the Banlc:tr',.. MORad"t, I"Opulation from the Cen~u~e~1 and the dividend~ from the ~otJrce~ Ii!illcd in Table

J.

tThe dividrnd price rnli(l!ll nre fur IK~I Rnd IK~9. l'he percent !lyrnhol re'en to the annuAl perer-nlnge rnle of chAnge.

by the unwillingness of the legislature to charter new banks. Contemporary observers were certainly aware of the lack of bank capital in Philadelphia. Consider the following item from the Philadelphia North American which was reprinted in the Banker's Magazine: The Southwark [bank] was again refused an increase of capital at last session of the legislature, though, as will be seen [from its dividend rate] it is one of the best banks in the city, and were the legal sanctions granted, might easily obtain subscriptions to any desired amount of new capital. [3, p. 996.]

It is conceivable that Philadelphia's banks held riskier portfolio's than did Boston's, but this argument is less plausible for New York. Some evidence is presented in Table 4. The dividend price ratios for several years are given. They do not show a significantly higher rate for Philadelphia in comparison with that which would be expected for New York if its banks held riskier portfolios. However, the dividend price ratios are lower for Boston, indicating that these banks may have had a more conservative investment policy. Perhaps, the key factor here was the high capitalization rates in Boston. 6 The dramatic fall in bank capital per capita in Philadelphia-the product of a stable amount of capital and rapidly growing population-and the otherwise similar nature of its portfolio tend to confirm the diagnosis based on contemporary observations. The problem of allocating bank capital was more difficult in the areas of new settlement. For here the rate of growth of population, and changes in the distribution of population were more intense, so that errors in the allocation of bank capital were more costly. There is abundant qualitative evidence that allocation by state governments on the frontier was unsatisfactory. 6

An econometric study of bank portfolio behavior [16] produced similar results.

137

Free Banking II HUGH ROCKOFF

:

159

In Indiana the state bank, an often praised monopoly partly owned by the state, was set up before the railroads were built. The railroads, of course, completely altered the importance of various communities making the allocation of capital among the branches of the state bank uneconomic. People felt that it was time for a "new shuffle and deal." The upshot was that the legislature allowed the charter of the first state bank to expire and adopted free banking in its place [20, pp. 124-25]. In Tennessee, a state owned bank was formed as a relief measure dming the depression of the late HOOs. At its very inception the decision on where to locate branches involved the directors in a statewide controversy. While commercial demands for banking facilities were taken into consideration in the location of the branches, there was continual dissatisfaction. The location of the branches was a subject of debate each time the legislature met and in each gubernatorial campaign during the life of the bank [7, pp. 96-97]. In Missouri the chief source of controversy was the contention by St. Louis businessmen that the state bank did not allocate sufficient capital to satisfy the rapidly growing needs of St. Louis. In the late 1850s the legislature answered this criticism by chartering a new set of privately owned banks. While this law was not, strictly speaking, a free banking law, it did include a bond security provision [4, pp. 241-48, 253-56]. Free banking in Ohio was not preceded by a state owned bank, but rather by a system of chartered banks with standard provisions. Nonetheless, the belief was widespread in 1850 that Ohio was suffering from a lack of banks. Her newspapers pointed out that Ohio was third in population but far from third in bank capital. Papers in Cincinnati and Cleveland complained that these rapidly growing cities were being denied bank capital under the charter system. These arguments helped overcome the opposition and secure passage of the free banking law in 1851 [17, p. 208]. Examples of this sort, evidence of a widespread belief that free banking would improve the allocation of bank capital on the frontier, do not prove that in fact that free banking could or did improve the allocation of bank capital. But together they seem to establish a presumption that something was wrong with the older system of chartered banking. 7 There exists little quantitative evidence by which one can gauge the impact or potential impact of free banking legislation on western banking systems. The best evidence is for Ohio, where we can examine bank profits before and after the free banking law of 1851. This information is presented in Table 5, along with bank profits in certain other states for comparison. It does appear that in Ohio profits were high before free banking and were lowered as a result of it. While the low initial profits of the free banks could be attributed to a "start up" period, they could not be attributed 7 An alternative interpretation of these complaints is that they represent the traditional plea of the farmer for "easy" credit.

138

Free Banking II 160

MONEY, CREDIT, AND BANKING TABLE 5 BANK PROFITS IN OHIO BEFORE AND AFTER FREE BANKINO'

Banks

I.

2. 3.

4. 5. 6. 7.

R.

Ohio, free banks Ohio, state banks Ohio, independent banks Ohio, old banks State Blink of Indiana Hank of Kentucky Banks of New York City New England Municipal Bonds

1850

1851

15.0 13.6 12.1 10.0 10.5 9.2 5.1

15.3 14.1 12.4 9.4 9.0 9.6 5.1

t 1852

18S3

8.2 13.7 10.1 9.4 13.0

9.6 10.0 8.9

5.0

5.0

·Sollrce~: Dividtnds wtre obtRined from the !ources listed below the appropriate balance sheet information where necessitry was obtained from the annual reports on the condition of the stale banks made by the secretary of the treasury and printed as House documents. (1-4) Charles Clifford Hunlinron. A Hislory of Ba.kl'R a.d C,m,.cy I. Ohio B,fort Ihe Ch'U War (Columbus, Ohio: Hoer Prinling Co., 1964). pp. 21 .213.278-79, and 293-94; (5) William F. Harding, "n.e Slale Bank of Indiana,"' lournal of Polilical "co"omy. 3 (December. 1895), p. 23; (6) Gen. Basil W. Duke. Hblory of Ihe Ba.k of Kenlucky (Louisville: John P. Morlon & Company. 1895). p. 140; (7) Table 3; (8) Sidney Homer. A Hblory of ,.It,.sl RaltS (New Brunswick. New Jersey: Rutgers University Press, 1963), p. 287. t Prorits are defined as the ratio (as 8 percentage) of dividends to nominal capital. These rates are similar to more sophisticated ratios. The old banks were individually chartered. The state banks were similar except that each contributed to a common "safety fund" for the relief of note holders of failed banks. The independent banks were bond secured but entry was limited.

to the restrictions imposed by the bond security system since the independent banks also faced these restrictions. The preceding discussion relies on intraregional comparisons. To gauge the impact of free banking it is also worthwhile examining measures of efficiency available for all states. One measure is the density of incorporated banks with respect to popUlation. This is a rough index of the number of banks a potential customer faces when he enters the market as a depositor or borrower. Table 6 gives the number of incorporated banks per hundred thousand in 1840, 1850, and 1860. The most striking feature of the table is the smaller ratio of banks to population in the free banking states in 1860, the opposite of what one might expect from stories about wildcat banking, although the high ratio for Wisconsin in 1860 is an exception. A second supplement.try measure which suggests itself is the number of unincorporated banks. If the state authorities were restrictive in the issue of charters, or if incorporated banks were taxed heavily, then we would expect to see private banks developing as substitutes. We know little of the activities of private banks in the ante-bellum period. They were not limited liability institutions, nor could they issue bank notes. But it seems reasonable to suppose that they could offer some competition to the incorporated banks in the issue of deposits and the making of loans. Table 7 presents data on the private banks from the Banker's Almanac; 1859 and 1860 are among the best years for our purposes because the list of private bankers was gradually lengthened as users of the Almanac noted omissions. The higher ratio for the free banking states tends to contradict the hypothesis that free banking reduced the private banking sector, confirming the results from Table 6. However, it appears in both

139

Free Banking II 161

HUGH ROCKOFF TABLE 6 INCORPORATED BANKS PER 100,000 INHABITANTS BY STATE, 1840, 1850, and 1860· BANKS PER

100,000 IN 1840

Sial.

BANKSPEk

100,000 IN 1850

BANKS PER

tOO,OOOIN 1860

Sial •• wilh Opcraliv. Fr•• Baw... Sy.lcm. in 1860t

New York Louisiana Ohio Indiana 1I1inois Wisconsin Minnesota Avg.

3.91 13.35 2.43 1.90 1.89 3.23

6.52

4.45

4.12

5.60

2.93 1.42

7.89 1.84 2.35 2.89 5.47 14.18 1.74 5.19

All Olh.r Sial ••

Maine New Hampshire Massachusens Vermont Rhode Island Connecticut New Jersey Pennsylvania Delaware Maryland District of Columbia Virginia North Carolina South Carolina Georgia Florida Alabama Mississippi Arkansas Kentucky Tennessee Michigan Iowa Missouri Kansas Nebraska Avg.

9.36 9.12 15.45 5.82 56.88 10.00 6.97 2.85 4.47 13.64 2.18 1.33 2.36 4.20 9.26 1.18

5.49 6.92 13.16 9.87 46.62 9.97 5.31 2.29 6.52 3.94

lUI 15.95 14.88 12.70 51.43 16.09 7.44

2.32 2.07 2.09 1.99

4.12 3.12 2.84 1.70 1.43 .83

.26 .16

10.20 2.18 2.77 1.89

1.93 2.20

.78

.88

8.23

6.26

1.26

3.06

10.71 4.51

3.71 2.97 .27 1.93 3.56 1.87 3.45 7.82

·Sources: Banks: U.S. Complrolh:r ol.he Currency. Rtpo" 1876. pp. xcvi-ccxi. Populalion: U.S. Bureau of .he Census, Hislolk.1 S,.,;',ic. uflh. Ulli,." 5,., .., Culuni.1 Ti"'t"ClI1I57(W•• hinlllon, D.C., 19611), pp. 12-1). tin SQme: of these slale. ulher

kind~

of banks were important.

cases that the free banking effect is obscured by regional factors, making a simple point-in-time comparison insufficient. The potential gains from an improved allocation of bank capital were small when viewed in relation to the economy as a whole. The par value of bank capital in 1859 was $422 million [35, p. xcv]. Thus, if banks earned a monopoly profit of, say, 2 percent on the average, the total transfer would amount to an annual flow of $8.44 million, and this is an overestimate-many states had free banking laws. A calculation of

140

Free Banking II 162

MONEY, CREDIT, AND BANKING TABLE 7 INCIDENCE OF PRIVATE BANKS BY STATE,

No. OF STATE

1859 AND 1860*

RATIO OF

No. OF

PRIVATE

PRIVATE TO

PRIVATe

RATIO OF PRIVATF. TO

BANKS 1859

INC. BANkS 1859

BANKS 1860

INC. BANKS 1860

Slal.s wilh Operative Fre. Banking Systems in 1860

New York State New York City Louisiana Ohio Indiana Illinois Wisconsin Minnesota Avg.t

35 79 14 143 46 124 37 33

.14 1.46 1.17 2.70 1.24 2.58 .38 16.50 3.27

33 78 10 148 36 136 17 25

.04

3 n.a. 15 n.a. 4 I n.a. 82 2 II 6 18 5 3 10 9 19 12 2 31 9 50 76 32 7 4

.13 1.42

.77 2.85 .97 1.84 .16 10.00 2.27

All Olher Slates

Maine New Hampshire Massachusetts Vermont Rhode Island Connecticut New Jersey Pennsylvania Delaware Maryland District of Columbia Virginia North Carolina South Carolina Georgia Florida Alabama Mississippi Arkansas Kentucky Tennessee Michigan Iowa Missouri Kansas Nebraska Avg.

3 n.a. 18 n.a. 7 I n.a. 86 2 18 8 20 6 2 13 7 14 15 I 33 18 58 100 31 4 7

n.a. .10 n.a. .08 .01 n.a.

.99 .17 .56 .32 .21 .10 .46 n.a. 2.33 n.a. n.a. .89 .46 19.33 .84 4.0 3.5 1.91

.04 n.a. .08 n.a.

.04 .01 n.a. .91 .17 .35 .28 .17 .15 .34 4.50 2.38 n.B. n.a. .69 .26 12.5 6.33 1.07 4.67 2.67 1.88

'Source" Prival. Bank. 1859. 1860: Merc\~nt', and Banker', Reg"t.,. 1859, pp. 26-40; 1860, pp. 28-42. Incorporaled Banks, Stal ... 1859. 1860: U.S. Com~roller of Ihe Currenc~. Rtport, '876, ~.•cvl-cui. Incorporaled Bank•• N.w York Cily. 1859. 1860: U.S. Congr.... ouse Documenl 112.3 th Congr••• , 2nd .. ion (1859) pp. 116-17; U.S. Congress, House Documeol49. 361h Congre ... 151 Session (1860) pp. 113-14. t In some of Ihese slales olher kind. of banks w.re importanl.

Free Banking II

141 H UGH ROCKOFF

:

163

the pure efficiency loss might well produce an even smaller flow. A flow of $8.44 million in 1859 would have been only 8.28 per capita [32, p. 7]. Thus, the main import of this section is f07 an understanding of the development of the banking system rather than the economy as a whole.

v (. CONCLUSIONS The following facts and conclusions can be drawn from our reexamination of the free banking era: 1. Free banking laws were passed in eighteen of the thirty-two states, with Michigan passing two laws, one in 1837 which was subsequently repealed, and a second law in 1857. In nine states, little or no banking was done under the law, or it was given only a brief trial before the Civil War. In three states, New York, Ohio and Louisiana, some of the "soundest" banking of the era was accomplished under free banking laws. (n Tennessee and later in some of the states which initially experienced wildcat banking the system was a more modest success. Six states, Michigan (after 1837), Indiana, Illinois, Wisconsin, Minnesota, and New Jersey, experienced wildcat banking. In one of the latter, Minnesota, this was clearly due to efforts by the state to force its bonds to a higher price than they were currently bringing in the market. In three states, Wisconsin, Illinois, and New Jersey, wildcat banking can be traced to the linking of the supply of currency with the debt of another state. In only one of the 19 free banking experiments, Michigan (after 1837), did wildcat banking result from a system of currency backed by privately issued securities. 2. Wildcat bank notes lost most of their value only in the first episode in Michigan. In other cases the losses were much less. More typical of free banking was the average loss of 15¢ on the dollar for failed banks in New York. 3. Simple regressions suggest that episodes of wildcat banking had a mild long-run impact on the services people derived from holding money in the effected regions. 4. The evidence for or against the proposition that free banking improved the allocation of bank capital is too slender to support firm conclusions. However, it does appear that New York City benefitted from free banking in its competition with Philadelphia for financial leadership, and that in the West free banking was, or at least appeared to be, a way of solving the vexatious problem of how to allocate bank capital in a region of new and rapid settlement.

142

Free Banking 11

164

MONEY, CREDIT, AND BANKING

LITERATURE CITED

I. BAILEY, MARTIN J. "The Welfare Cost of Inflationary Finance." Joumal of

Political Economy, 64 (April 1956),93-110. 2. BAKER, HENRY F. "An Historical Sketch of Banking in the State of Indiana." Banker's Magazine, 7, New Series (September 1857), 161-79. 3. Banker's Magazine, 8, New Series (June 1859). 4. CARLE, JOHN RAY. The Bank of the State of Missouri, Studies in History, Economics and Public Law, Vol. 102. New York: Columbia University, 1923. 5. CAGAN, PHil.!". Determinants and Effects of Changes in the Stock of Money, 1875-1960. New York: Columbia University Press, 1965. 6. _ _ . "The First Fifty Years of the National Banking Act-An Historical Appraisal," in Banking and Monetary Studies, edited by Deane Carson. Homewood, Illinois: Richard D. Irwin, Inc., 1963. 7. CAMPRELL, CLAUDE A. The Development of Banking in Tennessee, Nashville, 1932. 8. DILl.IST1N, Wn..LlAM H. Bank Note Reporters and Counterfeit Detectors, 1826-66. Numismatic Notes and Monographs, 114. New York: American Numismatic Society, 1949. 9. EASTERLIN, RICHARD A. "Interregional Differences in Per Capita Income, Population and Total Income, 1840-1950," in Trends in the American Economy in the Nineteenth Century (The Conference on Research in Income and Wealth. Studies in Income and Wealth, vol. 24). Princeton: Princeton University Press, 1960. 10. ELMER, LUCIUS Q. C. A Digest of the Laws of New Jersey, 1709-1861. 3rd ed. Bridgeton, New Jersey: Elmer and Nixon, 1861. 11. ENGERMAN, STANLEY L. "A Note on the Economic Consequences of the Second Bank of the United States." Jountal of Political Economy, 78 (July / August 1970) 725-28. 12. FRIEDMAN, MILTON, and ANNA JACOBSON SCHWARTZ. Monetary Statistics of the United States: Estimates, Sources, Methods. New York: National Bureau of Economic Research, 1970. 13. HAMMOND, BRAY. Banks and Politics in America: From the Revolution to the Civil War. Princeton: Princeton University Press, 1957. 14. HELDERMAN, LEONARD C. National and State Banks: A Study of Their Origins. Boston: Houghton Mifflin Company, 1931. 15. HEPRURN. A. BARTON. A History of Currency in the United States. New York: The MacMillan Company, 1915. 16. HINDERLITER, ROOER and HUGH ROCKOFF. "The Management of Reserves by Ante-Bellum Banks in Eastern Financial Centers." Explorations in Economic History, II (Fall 1973), 37-54. 17. HUNTINGTON, CHARLES CLIFFORD. A History of Banking and Currellcy In Ohio Before the Civil War. Columbus, Ohio: Heer Printing Co., 1964. 18. INI)IANA. Laws of Indiana (1855). 19. LASIlON, OSCAR. "The Early Ways and Crazy Days of Banking." Banker's Magazine, 154 (spring 1971) 49-58.

Free Banking II

143 HUGH ROCKOFF

:

165

20. MCCULLOCH, HUGH. Men and Measures of Half a Century. New York: Charles Scribner's Sons, 1889. 21. MICHIGAN. Laws of Michigan (1837). 22. _ _ . Laws of Michigan (1838).

23. _ _ . Laws of Michigan (1857). 24. OBERHOLTZER, E. P. Jay Cooke, Financier of the Civil War. Philadelphia: George W. Jacobs & Co., 1907. 25. PATCHIN, SYDNEY A. "The Development of Banking in Minnesota." Minnesota History Bulletin, 2 (August 1917), 111-68. 26. REDLICII, FRITZ. The Molding of American Banking. New York: The Johnson Reprint Company, 1968. 27. ROIIERTSON, Ross M. History of the American Economy. 2nd ed. New York: Harcourt, Brace and World Inc., 1964. 28. SYI.LA, RICIIARD. "American Banking and Growth in the Nineteenth Century: A Partial View of the Terrain." Explorations in Economic History, 9 (winter 1971-72), 197-227. 29. _ _ . "Federal Policy, Banking Market Structure, and Capital Mobilization in the United States, 1863-1913." Journal of Economic History, 29 (December 1969),657-86. 30. U.S. BUREAU OF THE CENSUS. Eighth Census of the United States, 1860, Mortality

and Miscellaneous Statistics. 3 I. _ _ . Eighteenth Census of the United States, 1960, vol. I. 32. _ _ . Historical Statistics of the United States, Colonial Times to 1957. Washington D.C.: Government Printing Office, 1960. 33. _ _ . Ninth Census of the United States, 1870, vol. I. 34. _ _ . Tenth Census of the United States, 1880, vol. VII. "History of the State Debts," pp. 523-645. 35. U.S. COMl'"rROLLER Of' Currency, 1876.

THE

CURRENCY. Annual Report of the Comptroller of the

36. U.S. CONGRESS; HOUSE. House Document 172. 26th Cong., 1st Sess., 1840. 37. _ _ . House Ex. Document 102. 34th Cong., 1st Sess., 1856.

ApPENDIX

The following list gives the leading references and assumptions used in estimating the losses in Table 2. In a number of cases, other sources had to be consulted. I. Vermont. We have discovered only one free bank which failed: the South Royalton. Its notes were secured by mortgages and Virginia bonds. A. W. Kenney, "The Banks," in History of Royaltoll Vermont, ed. by Evelyn M. Wood Lovejoy (Burlington, Vermont: Free Press Printing Co. 1911), pp. 502-506. The mortgages may have produced considerable losses. Total losses were estimated as an arbitrary 20 percent of the South Royalton's circulation of $122,570 in August 1856. U.S. Congress, House Doc. 87, 34th Congress, 3rd Session (1857), p. 27.

144

Free Banking II 166

MONEY. CREDIT. AND BANKING

2. Massachusetts. U.S. Congress, House Ex. Doc. 25, 37th Congo 3rd Sess. (1862), p. 41. 3. Connecticut. None of Connecticut's free banks failed during the life of the free banking law which was repealed in 1855. For a list of the Connecticut free banks see Forrest Morgan, ed. in chief, Connecticut as a Colony and as a State, orOne of the Original Thirteen (Hartford Connecticut The Publishing Society of Connecticut, 1904), vol. 3, p. 21 J. 4. New York. Carroll Root, "New York Bank Currency," Sound Currency 2 (February, 1895), p. 19. 5. New Jersey. Losses were compiled from the summary sheets to the annual statements of the banks appearing in the legislative documents of New Jersey. fl. Pellll.~ylvml;a. John Tom Holdsworth, Financing an Empire: History of Banking in Pennsylvania (Philadelphia: S. J. Clarke Publishing Co., 1938), pp. 583-585. 7. Ohio. According to Huntington only 7 bond security banks had failed through 1857. Charles Clifford Huntington, A History of Banking and Currency in Ohio Before tlte Civil War (Columbus, Ohio: F. J. Heer Printing Co., 1964), p. 249. Assuming that each was a free bank and had the average circulation of a bond secured bank in 1857, and that the rate of loss was, say, 10%, we get an estimate of 68,000. In addition I have added a ten percent loss for two banks existing in 1858 and 1860 but which failed to report in February 1861. 8. Indiana. U.S. Congress, House Ex. Doc. 102, 34th Congo 1st Sess. (1856), pp. 181, 182, 185, and similar reports on other dates. 9. Illinois. George William Dowrie, Tlte Development of Banking in Illinois, 1817-1863, (Urbana, Illinois: University of Illinois Press, 1913), pp. 152-153 and U.S. Congress, House Doc. 76, 26th Cong., 2nd Sess. (1861), pp. 220-221. 10.-11. Michigan. U.S. Congress, House Doc. 172, 26th Congo 1st Sess. (1839), p. 1129. It was deduced from other evidence that there were no failures under the law of 1857 through 1860. 12. Wisconsin. Leonard Bayliss Krueger, History of Bankillg in Wisconsir1, Studies in the Social Sciences and History, No. 18 (Madison, Wisconsin: the University of Wisconsin, 1933), p. 69. 13. Minnesota. Discounts: Sidney Patchin, "The Development of Banking in Minnesota," Minnesota History Bulletin 2 (August, 1917), p. 160. Circulation: U.S. Congress, House Doc. 49, 36th Congo 1st. Sess. (1860), p.296. 14. Iowa. Howard H. Preston, History of Banking In Iowa (Iowa City: State Historical Society of Iowa, 1922), p. 75. 15. Georgia. Bank note reporters published in the 1840's listed two free banks in Georgia: the Ruckersville Banking Company and the Exchange Bank. See, for example, Bicknell's Counterfeit Detector and Bank Note

145

Free Banking II HUGH ROCKOFF

:

167

List, November 1, 1843, p. 27. The former bank apparently redeemed all of its notes; Laws of Georgia, 1853-1854, p. 192. For the Exchange Bank we have used its maximum reported circulation; U.S. Congress, House Doc. 226, 29th Congo 1st. Sess. (1846), p. 680. 16. Florida. J. E. Dovell, History of Banking in Florida, 1828-1954, (Orlando, Florida: Florida Bankers Association, 1955) pp. 44-46, and passim. This estimate is uncertain. 17. Tennessee. Claude A. Campbell, The Development of Banking in Tennessee (Nashville, 1932) pp. 150-151. U.S. Congress, House Doc. 49, 36th Cong., 1st Sess. (1860), p. 169. 18. Alabama. Theodore William Mathews, Statutory Protection of Bank Creditors Prior to the Civil War (unpublished Master's thesis, University of Chicago, 1930), p. 242. 19. Louisiana. George D. Green, Finance and Economic Development in the Old South (Stanford, California: Stanford University Press, 1972), p.23.

[8] Journal of Monetary Economics 12 (1983) 127-158. North-Holland

ON THE ECONOMICS OF PRIVATE MONEY Robert G. KING* Ulliversity of ROc/lest"', /sued by privately owned commercial banks (liurst 1973). In addition. "Nothing in the Con!>titution barred private manufacture of coin. and through the first half of the nineteenth century Congress did not nct against private coinage. . . . General contract law allowed any contractor to issue his notes and coins and circulate them so far as the market would take them" (Hurst 1973).

Free Banking II RICHARD H. TIMBERLAKE

203 : 439

Free enterprise in the issue of common tender money was accidentally encouraged in practice by the federal government's ineptness in establishing a useful denominational spectrum of fractional currency during much of the nineteenth century (Carothers 1967). Private transportation companies-canals, turnpike companies, and railroads-issued significant amounts of such currency between IX20 and IX75. Municipal and state governments did likewise. Redemption of transportation currency when called for was in services rendered, while state and local government currency was redeemed as tax payments (Timberlake 19X I). The paucity of government-issued fractional currency was catastrophically aggravated by the first issues of greenbacks during the Civil War. The metallic v,liues of subsidiary coins rose rapidly above their monetary values in the summer of I X62. and the coins disappeared from circulation. These circumstances provoked not only the ill-conceived issue of postage stamp currency, but also extensive private issues of minor coin (Carothers 1967, Faulkner 190 I). The act that authorized postage stamps as currency in 1862 also outlawed the private issue of notes, memoranda, tokens, or other obligations "for a less sum than one dollar intended to circulate as money or to be received or used in lieu of lawful money of the United States" (Act of Congress, 12 Statutes at Large, 592. July 17, 1862). Then in 1864. even the private issue of gold and silver coin was forbidden, again. "when the coins were intended for use as current money" (Hurst 1973).

3. TilE APPEARANCE OF SCRIP AS AN ECONOMIZING MEDIUM

The lack of adequate denominations in government-produced money was not the only factor that stimulated the private production of money. Shortly after fractional coinage was stabilized around 1885, coal mining and lumbering became major industries. Both coal mining and lumbering enterprises had to be organized in the vicinity of the contributory resources, so were often located in isolated areas with low popUlation densities significantly distant from commercial centers. Coal-producing regions were hilly or mountainous areas where agriculture had been marginal and other commercial development had lagged. "The 'Main Street,' "noted one observer in describing a coal mining community "was often railroad tracks" (Brown 1978). Coal mining entrepreneurs, therefore, had unique problems to contend with in organizing their enterprises. Their common problem was what is known today as a lack of infr:astructure-· no strel!ts, no churches, no schools, no residl!nces, no utilities, and no banks or financial intermediaries. The specialized industries that might otherwise have provided these services were dissuaded from doing so by the high start-up costs "nd the enduring uncertainties of dealing with low-income communities that might be there today and gone tomorrow. Alternatively, the coal mining comp"nies could deal with such conditions because they were in a better strategic position to change uncalculable uncertainties into calculable risks. (Fishback I'JX6, Johnson 1952). Mining companies, therefore, built residences, churches, schools, and water works, and opened company stores or commissaries. In so

204

Free Banking II 440

:

MONEY, CREDIT, AND BANKING

doing, they became both buyers of labor from, and sellers of commodities to, the coal miners and their households. This kind of organi7.ation invited an economy in the community's payments system-the use of scrip in lieu of ordinary money. "Scrip" has become a generic term for the issue of a localized medium of exchange that is redeemable for goods or services sold by the issuer. Originally printed cards or "scraps" of paper. scrip evolved into metallic tokens with many of the physical attributes of official coins. Indeed. scrip in the very beginning was more in the nature of a trade credit. or demand deposit. at the single local general store. Ledger credit scrip. however, gave way to scrip coupon hooks. which "eliminated the tedious bookkeeping chores that were incident to over:"thccountcr credit (day book or journal entries followed by ledger entries)" (Brown 1978). The use of scrip not only implied an issuer-the mining company-and a demander-the miner, it also required a supplying industry. The institutions that supplied coupon scrip were companies already in business printing tickets. tokens, and metal tags for various other kinds of enterprise. They advertised extensively in mining catalogues during the first half of the twentieth century touting the advantages of their own scrip systems. The Allison Company of Indianapolis, for example. noted that when one of its coupon books was issued to an employee. "He signs for it on the form provided on the first leaf of the book, which the storekeeper tears out and retains for the [company] time-keeper, who deducts the amount from the man's next time-check." Then, when the employee buys goods from the company store, "he pays in coupons, just as he would pay in cash. and the coupons are kept and counted the same as cash . . . . The coupon hook is a medium of exchange between the company employees and the company store" (from 1916 Mining Catalog, Brown 1978). Other scrip-producing ticket companies emphasi7.ed the safety of the scrip coupon system in coal mining communities "where little or no police protection is afforded" (adv. of the International Ticket Co. in the Keystone Catalog of 1925, Brown 1978). The Arcus Ticket Company of Chicago advertised a list of advantagcs of scrip to both the employer and employee, one of which for the employer was the fostcring of employee good-will by avoiding misunderstandings on chargc accounts (sic). The advantages to the employee included keeping the" 'head of the house' better informed as to the purchases made by his family from day to day . . . . This frequently puts a check to extravagance and debt" (Keystone Catalog. 1925 in Brown 1978). Local scrip of this type was very similar to modern day travelers checks. The costs of travelers checks were also the costs of coupon scrip: each unit could be used only once. It had to be signed out when it was issued and signed when it was spcnt (nrown 1978).1 IThi~ comparison must he qualified. Many travelers checks. as well as other II.S. currency. are currently lIsed as hand-to-hand media in foreign markets. Sometimes travelers checks rctllrn from ahrnad with more than a dOlen endorscmcnts on thcm. They are called "checks". hut likc food "stamps". they are a quasi currcncy.

Free Banking II RICHARD H. TIMBERLAKE

205 : 441

The transactions costs of coupon scrip eventually encouraged the increased use of metal scrip. This medium became cheaper overall than coupon scrip, in spite of metal's higher initial cost, largely due to the invention and development of the cash register after 1880. Pantographic machines also were instrumental in reducing the unit costs of metal tokens (Brown 1978). Instead of rccdving cash, the scrip-issuing "cash registers" paid out metaltokens, made a record of the pay-out and to whom it had gone, and kept a grand total of the amount issued. The scrip registers would eject a specified "dollar" amount of scrip when a lever like that on a slot machine was pulled. In a 1'.l27 advertisement, the Osborne Register Company (ORCa) of Cincinnati pictured a IO-year-old child who, in a demonstration, issued $600 worth of metal scrip in various amounts to 200 hypothetical employees in 55 minutes, implying an average emission of $3 per employee every 16.5 seconds (Brown 1978).

4. THE POSITIVE-SUM BENEFITS OF SCRIP

The economics of scrip issue, as with all exchange between economic agents, required that both the issuer (the coal mining company) and the acceptor (the employee) benefit from the transaction. The company necessarily had contact with the outside world. It bought machinery and other resources and sold coal in a national market. All these activities required the use of standard money. Scrip was used essentially as a working balance of money with which the coal operator could make advances to his impecunious employees between paydays. It was issued at the request of the miner to the extent of the wages he had already earned, and it was redeemable in standard money on the next payday. The amounts were usually small-five or ten dollars, or even less. To the worker it amounted to an interest-free, small-sum loan that he could get with almost no effort. It enabled him to buy ordinary household goods at the company store. To those workers who had "gone out and got drunk" on the previous weekend, or who had suffered some kind of household emergency, scrip was a blessing only measurable by the cost of its common alternative (Clark 1980, Johnson 1952). Its alternative in a conventional urban setting without scrip was the pawn shop, loan shark, or installment peddler (Johnson 1952). An industrial worker in the same unfortunate position in, say, Detroit, Pittsburgh, or Chicago, had access to money between paydays only by borrowing against his household capital at a pawn shop where he paid exorbitant interest rates ifhe reclaimed his pawned goods. The scrip system could be abused in such a way that a discount would also appear in some scrip transactions. Since the company store did not sell liquorfor the obvious reason that its sale would encourage absenteeism and worker inefficiency, workers would at times obtain scrip from the company clerk and sell it for conventional currency in order to buy liquor. The bootlegger (during Prohibition) or other liquor vendor, whose shop was not likely in the neighborhood

Free Banking II

206 442

:

MONEY. CREDIT. AND BANKING

of the company store, faced significant costs in redeeming the scrip for conventional money. thus giving rise to a discount (Brown 1978. Caldwell 19(9).2 In spite of the obvious advantages of the scrip system to both worker and mine owner. scrip, the company store, and the company town have been universally bemeaned (Brown 1978). The accounts of their operations include contradictions that appear sometimes in the same paragraph. (For example, sec quote of Sayre used as an epigraph. p. I. Brown 1978.) All accounts. while critical of the scrip system. acknowledge. first. that it was issued at the behest of the miner; second, that its issue cost the miner nothing; and, third, that it was r~deemable in standard money on payday. The dogma of scrip's critics was that the company store, in which the scrip had to be spent, raised prices to monopolistic levels and there~ by exploited the defenseless miner (Dodrill 1971). Fishback's and Johnson's studies of prices in company stores versus those in independent stores refute this popular prejudice. Prices were four to seven percent higher, but so were costs. (Fishback 1986; Johnson 1952). The advantage of scrip issue to the mine operator was that it was one worker perquisite he could offer to attract labor into a somewhat unattractive environment. He already offered housing and mercantile services; by issuing scrip against future wages he also provided commercial credit with virtually no interest charges to the borrowers (Johnson 1952). The practice, indeed, was so widespread that it can only be viewed as a traditional perquisite of the trade. A company that did not offer the scrip privilege would have been at a competitive disadvantage. The mine operator thus became a quasi banker. His cost for metal scrip during the 1920s varied from slightly less than I cent to 5 cents a unit for scrip tokem of simple design made in aluminum. In brass or nickel silver and with scalloped edges and more intricate designs, costs could run as high as II cents a piece. (A II these values are unit costs in thousand-unit lots. and are from advertisements of several different scrip manufacturers between 1925 and 1940, in Brown 1978). Scrip sales information from the Ingle Company sales journal of 1928 reveals that the average denomination issued was about $.25 (Brown 1978). Since the average cost per token was only about J cents and could have been even less. an investment by the coal company bank in, say, 5,000 pieces cost it about $150 for the scrip coin, and perhaps $100 more for a scrip-issuing machine. To carry out this same banking function with regular U.S. currency would have required an investment in cash alone of $1,250, as well as substantially greater security costs to protect the money. One observer noted. "The mining company could pay almost its entire payroll in company scrip, disturbing only a few dolla rs of actual working capital" (Sayre, in Brown 1978). Of course, paying out scrip gave workers some additional claims on the working capital oflhe company stores. So

2Scrip wa~ frequently advertised a~ redeemahle only to the worker to whol11 it was originally i~~ued. This condition applied in ~ome mines. "owever. for metallic ~erip. it could hardly have heen enforced. and would have detracted from the utility of any scrip if it were enforced.

207

Free Banking II RICHARD H. TIMBERLAKE

: 443

the monetary economy of using scrip was in part offset by higher costs of merchandising goods.) The difference between the payment system costs of scrip and of real money was a form of seigniorage revenue the coal mine operator realized and shared with his employees. They received interest-free loans; he was able to offer a fringe benefit that tended to reduce what would have been a higher working capital re4uirement. While scrip was usually specialized to one company in a particular community, many coal mining companies had mines in different regions. Their scrip was good in all the different locations where their mines operated. As the scrip-using communities gradually came to experience more extensive commercial relations with each other, their localized scrips became interchangeable. Even some independent stores accepted coal company scrip (Brown 1978). Given the proscriptions against the private printing or coining of money by the Acts of 1862 and 1864, one may wonder how scrip could have been issued and used legally. The key is the word "intended" in the proscriptive laws. The courts ruled that scrip was not intended to circulate as money: first, because it was redeemable only in merchandise until payday; and, second, because it resembled money only superficially and was clearly distinguishable from standard money. (The coin under the court's scrutiny was a 50-cent token, but weighed only onefifth as much as a standard 50-cent piece.) Any token that was redeemable in lawful money on demone/was construed to be illegal, and whether the token in question was coin or pasteboard did not matter (Brown 1978). 5. TIlE ENVIRONMENTS IN WHICH SCRIP APPEARED

The extent of scrip use has many dimensions-temporal, geographical, and industrial. Its most notable occurrence in the twentieth century was in the coal mining regions of West Virginia, in part because the state government passed a "wide open" scrip law some time before 1925. However, it was extensively used in other states as well. The Tennessee Coal Iron and Railway Company, for example, ordered 547,500 pieces between 1933 and 1937 from the Ingle-Schierloh Company of Dayton, Ohio (Brown 1978). Another source lists 20,000 coal company stores in the United States, Canada, and Mexico all of which used scrip between 1903 and 1958 (Dodrill 1971). Numismatic records indicate that scrip was also used extensively in several other industries-fishing canneries, agriculture (to pay crop-pickers), fruit canneries, logging and lumbering companies, and paper companies (Brown 1978, Trantow 197H. Trantow's index lists over 1,100 companies that issued scrip currency in 40 states). One scrip numismatist cites a Chicago newspaper of 1845 that regularly 4uoted the discounted prices of coal scrip, city scrip, canal scrip, railII am indebted to Huston McCulloch for this observation.

Free Banking II

208 444

:

MONEY. CREDIT. AND BANKING

road scrip. Michigan scrip. Indiana State scrip. and Indiana land scrip. as well as the notes of private and chartered banks. Private businesses issuing such scrip numbered in the thousands (Harper 1948). Furthermore. as Brown observed. "The use of paper scrip was much wider than the use of [coin] scrip . . . [but] only a comparatively small amount [of the paper] has survived." Therefore. the extent of scrip use must have been much greater than the vestiges in metallic collections would indicate (see also Caldwell 1969). Just as Brown in his work seemed unaware of scrip that had preceded the issues bycoal companies. Harper in his study of Scrip and Other Forms of Local Money thought that intensive use of scrip only appeared in the United States during the dcpression years. 1932-1935. \lis research uncovered several sources of "depression" scrip: (I) issucs by local governments due to decreases in tax revenues; (2) issues by chambers of commerce after local bank failures as a means of "corralling as large a proportion of the depression diminished volume of business as possible for their membership"; (3) issues by "home-owned stores as a weapon against . . . chain-store competition"; (4) issues by "barter groups as a means by which the unemployed could more conveniently exchange services"; and (5) issues by charitable organizations to needy persons as "commodity orders" for foodstuffs. "Local money in some form." he concluded. "is likely to recur in response to a public demand under substantially similar circumstances." Most of this "depression" scrip had appeared in earlier times-for example. municipal scrip that was redeemable as tax payments. The depression scrip. however. was usually linked to a dated stamp scheme that required the holder to fix low denomination (2- or 3-cent) stamps to the scrip at specified times. The stamps were to provide the revenue to redeem the scrip and to encourage spending. but they added an undesirable burden that greatly reduced the efficacy of the scrip's use. They also detracted from the scrip's effectiveness as an addition to the existing stock of ordinary money (Harper 1948).

6. IMPLICATIONS OF THE SCRIP EPISODE

The phenomenology of scrip issue has significant implications. First. no one had any incentive to leave scrip behind for monetary researchers to count or to analyze. Demanders of such currency would not regard it as a store of value for any time longer than the period between paydays. Suppliers. to whom the scrip was an outstanding demand obligation. would redeem it first if they liquidated. merged. or closed down their enterprises. In addition. everyone who used it and benefited from it was aware of its questionable legality. Archival records of its outstanding quantities. therefore. are almost nonexistent (Timberlake 1981). Scrip's unrecorded existence is emphasized as well by the research that has uncovered its former use. Each scholar who has unearthed one of the diverse scrip appearances has treated the phenomenon as unique. and with good reason. Each one was widely separated in time. place. and circumstance from the others. Yet. each one had characteristics similar to the others. All episodes combined

209

Free Banking II RICHARD H. TIMBERLAKE

: 445

emphasize the feasibility of the spontaneous production of money in the private sector. The coal mining scrip episode adds significantly to the total scrip experience for a number of reasons. First, it lasted for over 50 years, so it was not just a temporary happenstance. Second, it appeared in a wide range of independent communities. In West Virginia alone, almost 900 coal mining companies employing about 120,000 miners issued scrip in one form or another. I n other areas of Appalachia-southern Virginia, eastern Kentucky, eastern Tennessee and southwestern Pennsylvania-the experience was similar. Third, scrip's tenure was not dependent on the previous existence of standard legal tender money. True, the coal company was bound to redeem the scrip on payday, but this guarantee was only a flourish that enabled scrip issuers to avoid violating the proscriptive laws against the issue of private moneys. As it was, many children living in coal mining communities did not see a dollar of "real" money until they grew up and left the area (Caldwell 1969). The self-sustaining nature of the scrip system without recourse to standard money, stemmed from the fact that both the demander and supplier of scrip were active participants in both the labor market and the household goods market at the company store. This intimacy in two markets by both participants enabled them to evaluate wages paid and received in real terms, that is, by the quantity of household goods that the scrip wages could purchase. A decline in the purchasing power of scrip at the company store would simply have indicated to the miner that the real value of his services to the company had declined. He thereupon would have moved to another location or occupation. lfthedecline in real wages was due to an industrial depression or the competitive decline of the coal industry, as occurred simultaneously in the 1930s, both mine workers and mine operators would realize reduced real returns in the mode of any resource owners under similar circumstances. A fourth important result of the scrip system was its reflective emphasis on the returns to the capital structure of the payments system. I n the scrip system the money was supplied endogenously: the coal company banks, the borrowing miners, and the scrip suppliers were all parts of an economy of private ownership. Scrip money was not dependent on any outside money, but was produced under the same conditions and incentives as any common commodity. The mining companies rather than the workers produced the scrip because in working without wages until payday, the workers were implicitly extending credit to the company. Scrip issue was a means of clearing this debt before the regular payday. In addition, the coal mining company had the collateral value of the mined coal to secure the "loan. "4 110th the companies and the workers realized the seigniorage returns from its existence. While the scrip system was small-scale and had a low profile, the government could ignore it because it posed no threat to the government's monopoly over the production of money. However, if scrip issue had shown any tendency to 'I am indcbl' patriotiques under the supervision of local authorities, they failed. When the National Assembly finally passed legislation governing the caisses in May of 1791, it decreed that all privately issued notes of 25 livres or less ami exchangeable on sight against assignats or copper coin were to be exempt from the stamp tax. The assembly reaffirmed that there was no government guarantee for the solvency of the caisses and that the hillt'ls were not legal tender (Bouchary, 1941, pp. 11-14). This gave entrepreneurs and municipalities a green light to organize (·ai.l'.\'e.\· patrioliqlle.\'. II. THE CHARACTER OF THE CAISSES PATRIOTIQUES.

The pent-up demand for small notes led to a rapid growth in the number of caisse.\' palrio/iqlles. Unfortunately, any description of their activities 2 A 5·livrc note was far too large fur most of the popUlation's (hlily transactions. According to Rude (1959, p. 21), the daily wage for an unskilled worker was 20 to 30 sous and for a skilled worker 50 sous. Hence, a 5-livre note represented 2 to 5 days wages for the vast nuuority of the workers. One livre was equal to 20 sous and each sou was worth 12 deniers. J Th'!se local initiatives had a parallel in the lucal political initiatives known as the revoilltion mlllliciptlie. Early in the revolution, new local guvernments and militias were formed to fill the power vacuum created by the cullapse uf the tlllciell regime. Like the cai.l'St's patriotiC/lies, these new entities were recognized asjilitl· accomplis by the National Assembly (Lefebvre, 1962, Vol. I, pp. 125-127).

Free Banking II 254

EUGENE N. WHITE

has been hampered by a lack of data. The government's withdrawal from regulation meant that it did not collect information on the caisses. Only later when it began to have second thoughts about free banking did official data collection begin. The most complete survey (Bloch, 1910) was ordered by the Minister of the Interior in June of 1792. Although the minister had asked the directors of the departements for detailed reports on the assets and note issue of each caisse. he received very little information as banks refused to cooperate with local officials and local officials ignored instructions. In spite of the absence of official records, the number of caisses in each departement can be fairly accurately determined thanks to the painstaking research of French numismatists interested in tracking down all known billets de cOl/fiance. The first and foremost of these works is Achille Colson's article in the Revue Numismatique (1852). This survey has been extended and corrected by later numismatists. Using their figures and information from various regional history journals, the number of caisses in each departement and some data on note issue has been collected (E. N. White, 1987a) and is presented in Table I. The assembled figures are for the summer of 1792. Their distribution reveals how quickly banks appeared in all parts of France. 4 While the exact number of billets issued will probably never be known, the peak total issue is estimated at approximately 140 million livres, significantly higher than previous guesses of less than 100 million. 5 Many of the first caisses were established by local governments. The caisse patriotique of Tours is representative of these municipally sponsored banks. The operation of the caisse in Tours was supervised by eight commissioners appointed by departmental, district, and town authorities (Bloch, 1910, pp. 158-161). The articles of organization specified • One interesting feature of Table I is the large number of misses in relatively isolatcd departemellts-the Ardeche, Aveyron, Dordogne, Dr6me, Gard, Lot, and Lot ct Garonne. These areas were not known for their financial sophistication. The establishment of a large number of small caisses. including many in very small towns. reveals thc extent of thc severe shortage of small denomination media of exchange. Even towns without mllch commerce with the olltside world were forced to establish a caisse as a consequence of the sharp rise in transactions costs. , The most widely cited earlier estimate is found in Harris (1930, p. 26). The estimate of 140 million was ohtaincd in the following fashion: In most cieparte",ellts it was common for there to be onc large, dominant caisse and many slTIall competitors. In the 36 dt'pmtemellls where there was information on the large cai.~se and many of the small cais.I'('s. the issue of any caisse without data was estimated by the average size of all known small ('aines. For the remaining 46 deplirtemellts where therc was no information on the large (,lIi,fses. an estimatc of their issue was ohtained applying a rcgression of thc note issllc of the largest caisses on demographic variahles for the 36 liepmtemellts. The avcrage isslle of all known small cai.fses was also used to estimate the issues of the small unknown misses in these departemellls (sce E. N. White, 1987a).

237

238

Free Banking II FREE BANKING

255

the details of the design and printing of the billel.\· de cOlljiallce and required them to be signed by four commissioners. The billels and assignats of the caisse were stored in a strongbox in the possession of the receveur (tax collector) of the Tours district. The three keys necessary to open the lock were held by secretaries of the depariemelll, the district, and the town. The cashier was paid 1200 Iivres per year, and he was required to post a surety bond of 15,000 livres. The caisse was open three days a week for business. Its expenses were met by a sliding scale of fees for exchanging assignats for billels, from 4% on notes of 500 livres to Y2% on notes of 5 livres. Any profits from the operation were to be distributed to the poor. This carefully designed organization, aimed at protecting the integrity of the billels de cOlljiatlCe, was typical of many misses pairiotic/lles. However, if all caisses had been like the cah'se of Tours, they would have provided an essential service but they would not have given any reason for the growing chorus of complaints about fraud and speculation. As 100% reserve banks, the caisses palrioliqlles would merely have offered transactions services for a fee to the general pUblic. They appear to have been much like the 100% reserve banks proposed by Friedman (1959, pp. 65-75). Like his banks, creating deposits as they receive reserves, these caisses simply exchanged billels for assignats that were held in strongboxes under the watchful eyes of local officials. If anything, the wisses would have contributed to a slight contraction of the money supply by running small surpluses from accumulated fees. Nevertheless, although most wisses palrioliqlles began as 100% reserve banks, many quickly became fractional reserve banks.!> The freedom to open a bank proved to be irresistible to many citizens. In addition to the caisses organized by the towns, manufacturers, theaters, bakers, printers, grocers, and many others opened their own caisses. Even many municipal caisses were transformed into fractional reserve banks. In Caen, this happened when the government of the cieparlemelll requested a loan (Beranger, 1910, p. 389). The caisse in Rouen (Beranger, 1908, p. 3(4) was begun as a joint stock bank with an initial capital of 270,0()() livres held by 216 stockholders. Although a private institution, its charter included safety precautions to protect its note issue similar to those of the caisse in Tours. The balance sheets of this bank Crable 2) and the Caisse Patriotique de Paris, one of the two largest banks in France (Table 3) show that the caisses operated much like banks of the period in Britain and America. Most of their investments were in bills of exchange and securities. These liquid assets were what most banking theorists believed were appropriate • Hayek (1978, p. 45) has suggested that laissez-faire banking could be introduced today by beginning with 100% reserve banks and then letting the system develop of its own accord.

239

Free Banking II EUGENE N. WHITE

256

TABLE I The Number of Cainn Palrioliqlln and Their Known Note Issue -_._---"---

Deparlemenl

-----~-----------.--------.--

Total caisses

Caisses with known issue

..--- .. _--, .. _......_----_.

Total known note issue (livres) • 0 - _ _ _ _ _ _ _ _ _• _ _ _ _ _

Ain Aisne Allier Alpes Basses Alpes Hautes Ardeche Ardennes Ariege Aube Aude Aveyron Bouches du Rhone Calvados Can tal Charente Charente Inrerieure Cher Correze Cote d'Or Cotes du Nord Creuse Dordogne Doubs Drome Eure Eure et Loir Finistere Gard Garonne Haute Gers Gironde Herault lIIe et Vilaine Indre Indre et Loire Isere Jura Landes Loir ct Cher Loire lIaute Loire Inrericure Loiret Lot Lot et Garonne Lozere Maine et Loire

22 32 14 10

13 74 II 7 2 7 58

41 26 33 4 29 9 9 5 7 12 68 5 82 18 IJ 13

46 43 41 18 37

12 17 7 24 6 4 12 12 2 9 53 51 28 6

18 0 0 I 0 74 I 0 2 4 0 I

22R,O(){)

1,2()()

194,51>5 444,512 722,()()() 6(',O()() 38,340 1,326,9RO

0 0 ()

I

60,000

()

4 3 0 5 0 0 0 0 0 0 0 II 0 0 I

7 I

1,057,426 36,(J()(J 305,250 95,O()()

9,285,692

IO,O()() 830,(J()(J 31O,O(J()

0 0 0 ()

0 3

I,052,70()

()

0 ()

0 Continued overleaf

Free Banking II

240

257

FREE BANKING TABI.E I-Continued

IJepariemetit Manche Marne Marne Haule Mayelllle Mcurlhc Meu,e Morbihan Muselle Nievre Noru Ois..: Orne Paris Pas ue Calais Puy uu Dume Pyrenees Basses Pyr..:nees Haules Pyr..:nees Ori..:nlales Rhin Bas Rhin lIaul Rhon el Loir..: Saon..: ..:1 Loire Saon..: Haule SarI he Seine el Marne Seine elOise Seine Infer-ieure Sevres Deux Somme Tarn Var Venuee Vienne Vienna Haule Vosges Yonnc TOTALS

Source. E. N. While (l9117a).

Tolal caisses

Caisses wilh known issue

III 6 II 3 12 9 12 3 14 34 26 94 64 25 27 6 5 I 2

0 4

2 13 19

()

0

30 15 7 21 9 III 33 13 14 13 4 15

()

Tolal known nole issue (Iivres)

1,1165,141

()

2 5

1,355,000 264,509 79,000

() ()

IIO,O()O 1,431,161 1,629,412

()

3

31,159,1112

() ()

0 ()

0 1411,000

I

o 224,310

()

4 3 3 2 II

72,529 40,000 !!5,411 11,666,626 117,000 2,735,510

0 94,000 106,000

J()

0

16 13

699,665

II

3 6 2

1,666

227

67 ,(H)7 ,251

70,(H)O 50,5(H)

241

Free Banking II

25R

EUGENE N. WHITE TABLE 2

Balance Sheet of the Cai.fu Pa/rio/ique of Rouen June 6, 1792 (in Iivres) Assets Assignats held as reserves Bills of exchange Unissued and returned notes Other

Liahilil ies 2,203,576 6,126.629 681,706 247.961

Stockholders' capital Surplus Chief officer's bond Notes printed

9,259.872

527,850 218,582 20,()()() 8,493,440 9,259.872

Source. Adapted from Beranger (1908), p. 311.

investments. Their reserve ratios of assignats to billets, 26 and 29%, respectively, suggest conservative management that could handle any demand for note redemption, short of a panic. While balance sheets are quite scarce, the remaining ones provide a picture of traditional banking, not one of speculative or predatory financial activities. The fractional reserve banks operated very freely. They competed keenly with the 100% reserve banks, frequently providing superior services. In one town, the municipally sponsored caisse was open three days a week and charged 1% on all exchanges. Its competition was a merchant's caisse that offered to exchange notes for free and was open every day (Beranger, 1910, pp. 342-343). It is thus not surprising that the fractional reserve banks grew much faster and soon dominated the 100% reserve banks. The liberty enjoyed by the banks was virtually complete. Perhaps the most extraordinary development was the appearance of private issuers of coin. Just as the shortage of low denomination paper money led to private issue of notes, so did the shortage of coin open up opportunities for private enterprise. There were nine private issuers of coin in Paris and two in Lyon (Colson, 1852, pp. 447-448). The most well-known of these were the Monneron brothers (Bouchary, 1941, pp. 194-205). Their caisse exchanged 2 and 5 sou copper coins for assignats. These high TABLE 3 Balancc Shect of thc Ct/i.ul' /'a/rio/ique ele /'aris April 3, 1792 (in livrcs) 5,062.5IK) Assignats hcld as rcscrvcs 8,390,150 Bills of exchange Securities on deposit with the city 4,670,614 4,926,715 Securities 1,445,800 Other 24,495,779

Capitlll lind surpills Bank notes in circulation

6,104,1«.7 17,660.812

24,495,679

Source. Adapted from Bouchary (1941). p. 70. The discrepancy betwecn assets lind liabilities is not IIccounted for by Bouchary.

Free Banking II

242

FREE BANKING

259

quality, non-full-bodied coins circulated widely. Other caisses struck a wide variety of copper, billon, and silver coin. During the first year of their appearance, all types of caisses patl"iotiqlles were quite successful. Although they were not legal tender, the billets de confiance often moved beyond the towns and departemellIs of origin. Yet, in spite of the considerable success of the ClIisses patriotiques in solving the shortage of low denomination media of exchange and in gaining the confidence of the public, they gradually came under a cloud of suspicion. III. THE ATTACK ON FREE BANKING

Popular discontent focused on both the issues of fraud and price stability and eventually turned the public against free banking. When the cai........ es patriotiqlles were closed many people concluded, as did the Societe des Nomophiles, that "in a well-run state, the issue of notes must not be shared by the sovereign government and private individuals.,,7 Historians have accepted without dissent the arguments of these critics. Achille Colson (1852) blamed the great variety and overissue of notes for causing capital flight, counterfeiting, and depreciation of the exchange rate. Marcel Marion, author of the standard French financial history (1914), concluded that the caisses patriotiqlles were engaged in speculation and defrauding the pUblic. In 1'he A ........ ignats (1930), the classic study of the revolution's finances, Harris seconded these criticisms. Although Bouchary (1941), Horsin-Deon (1958), and Carraz (1975) are not as critical, they still blame excessive note issue for the rise in prices. Most recently Aftalion (1987) has described the experiment as a national plague fraught with deceit and scandal. While a consensus exists among historians about the evils of free banking during the revolution, there is scant evidence to support their conviction. It is important to note that the National Assembly and Legislative Assembly and their finance committees heard all of the complaints cited by historians but found little merit in them for a long time. Only after political upheavals and rising inflation gave the upper hand to the radicals did the Legislative Assembly and the Convention move to end the experiment in free banking. An assessment of the merits of free banking thus requires a close examination of the twin problems of enforcing contracts to pay and price stability.

A. Counterfeiting, Fraud, and Wildcatting The breakdown of law and order made countelfeiting during the French Revolution a persistent problem. The central government waged a long struggle against counterfeit assignats, but in 1791 and 1792 the problems 1

Quuted in Bouchary (1941, pp. 16-17). Translation of the author.

Free Banking 1/

260

EUGENE N. WHITE

confronting the caisses patriotique.~ were graver. Although impossible to gauge its extent, the tales of bold and brazen counterfeiting suggest that it was widespread. The billet.~ de cOlljiallce were easier to counterfeit because they were often less well-designed and printed than the assignats. The large variety of notes in circulation apparently made detection difficult. Counterfeiting was not limited to producing reasonable facsimiles of real notes. Sometimes the numbers would be altered so that 2 would become 20 or 3 would become 30 (Colson, 1852, p. 373). The low level of literacy, where only 47% of married men and 27% of women (Furet and Ozouf, 1982, p. 26) were able to sign their names, may have permitted crude methods of deception. In one instance (Bouchary, 1946, p. 82), low-denomination notes printed on white paper were dyed red, the color of a higher denomination. Counterfeiters also printed notes of nonexistent caisse.v and circulated them in distant departements. These notes were passed to unwary illiterate peasants, and officials testified (Bloch, 1910, p. 180181) that it was often dangerous to refuse receipt of a note. The caisses patriotique.y did not, however, stand by idly. The Caisse Patriotique de Paris published a table of all its billets and advice on how to avoid accepting counterfeits in the Parisian newspapers (Bouchary, 1946, p. 82). The caisse patriotique in Rouen printed tables of its notes, explaining the types of paper and ink used and giving a list of all the signatures (Beranger, 1908, pp. 288-300). Many other cai.~ses took similar measures to assist and protect the public. These methods seem to have deterred some counterfeiting. More extensive anticounterfeiting measures, like the antebellum American bank note reporters (Rockoff, 1974), did not appear, perhaps because of the brevity of the experiment. In contrast to the French episode, Scotland's experience with free banking seems to have been more successful in avoiding counterfeiting. This may be partly explained by a higher level of literacy and by the exclusion of the less literate classes from using bank notes because of the I-pound minimum denomination. 8 The Scottish prohibition oflow-denomination notes was a consequence of the "small note mania" of the early 1760s when numerous little banks appeared issuing very low denomination notes. There were many complaints about the excessive and varied note issue. Among the opponents of small bank notes was Adam Smith who believed that no bank notes should be issued for less than 5 pounds. Describing problems in Scotland that sound remarkably like those in revolutionary France, Smith argued that:

• Adult male literacy in Scotland is thought to have been about 90% at the end of the 18th century (see GralT, 1987, p. 246).

243

244

Free Banking II FREE BANKING

261

Where the issuing of bank notes for such very small sums is allowed and commonly practised. many mean people are both enabled and encouraged to become bankers. A person whose promissory note for five pounds. or even for twenty shillings. would be rejected by every body. will get it to be received without scruple wllt!n it is issued for so small a sum as a sixpence. But the frequent bankruptcies to which such beggarly banks must be liable. may occasion a very considerable inconvenience. and sometimes even a very great calamity. to many poorest people who had received their notes in payment. (Smith. 1776. p. 3(7)

In 1765, parliament listened to this and other complaints and forbade Ihe issue of notes below I pound (Checkland, 1975, pp. 194-105; L. H. White, 19M, pp. 29-30). Limiting the issue of bank notl!s to the educated and financially sophisticated is one, albeit extreme, way to solve this information problem. It is certainly difficult to justify this solution as thl! extl!nt of fraud committed by "beggarly bankers" in Scotland remains unknown. Only in the United States where fraudulent banks were known as "wildcats" is there some reliable evidence. The practice by banks of issuing mon: notes than they could redeem was thought to be relatively common in antebellum America. However, recently Rolnick and Wl!ber (l9H4) havl! shown that wildcatting was a relatively rare phenomenon. Rockoff (1974) and Rolnick and Weber (l9H3) have measured the losses to noteholders of failed banks and found that they were smal1. 9 The beggarly bankers were very numerous in France, but, surprisingly, there were only three alleged instances of wildcatting banks (Colson, IH52, p. 273; Harris, 1930, p. 26). It is not possible, as in the American case, to measure the losses to noteholders. Yet, the negligible' number of failurl!s up to mid-1792, out of the over 160() cain'es patrioliqlles established, suggests that actual losses to the public were quite small. When the clIisses were forced to close in 1793, only a handful of the non-Parisian banks could not fully repay their noteholders. This was true even when liquidators received more billets-presumably counterfeitsthan had been originally issued. By this measure, the problems of fraud and counterfeiting were relatively minor. The same cannot bl! said about Paris where in 1792 there was one spectacular bank failure and other banks found themselves in trouble. Noteholders of two banks would have lost substantial sums if the government had not intervened to cover the banks' liabililies. However, the lroubles of these Parisian banks, as argued later, were principally the result of government actions that panicked the public rather than any malfeasance by bankers. • According to Rockoff (1974. p. 150). losses to note holders from failed banks for the whole free banking period amounted to $1.9 million. the equivalent of a 2% inflation in 11!60. Rolnick and Weber (19113. p. 10119) calculate total losses at $1.6 million.

Free Banking II

262

EUGENE N. WHITE

B. Price Stability The other great fear was that an excessive issue of billets de cOlljiallce was driving up prices. Many believed that the caisses pnlrioliqlles were directly buying up and hoarding essential commodities.1O Although the assembly generally found no substance in these complaints, some in the government eventually began to listen. In June of 1792, the Minister of the Interior blamed the caisses patriotiques for the rising price level: ' Historians have accepted this charge against the caisses. but there is little supporting evidence. One bank that was singled out for attack was the Caisse Patriotique de Paris, one of the two largest caisus. Initially its operations had been widely praised (Bouchary, 1941, pp. 22-25), but by late 1791 it came under attack in the popular press for speculating in commodities. Its managers took the charge quite seriously. They placed a notice in newspapers denouncing these charges and affirming that the bank's principal business was discounting bills of exchange. The caisse also offered a reward of 100,000 livres to anyone who could prove these allegations. No one came forward to claim the reward, and the bank's reputation remained intact. Likewise, solid evidence to substantiate the charges of speculation by other caisses has never been presented. As this notable example suggests, the rising price level may not have been the fault of the caisses patriotiques. Proponents of free banking Hayek (1978) and L. H. White (1984) have argued that laissez-faire banking should actually help to stabilize the price level. In White's model of free banking under a gold standard, a bank may in the short run issue an excessive quantity of notes, but it will be unable to keep them in circulation in the long run. When an individual finds himself with more notes of a particular bank than he desires he will reduce his holdings by (I) redeeming notes directly with the issuing bank, (2) depositing notes with his preferred bank, or (3) passing them on to other individuals. The overexpansive bank will be forced sooner or later to contract its note issue. This will happen directly if the notes are redeemed either by the public or by the banks at which they have been deposited. The overissue may increase spending, as the third case implies, but this will cause the region where its notes circulate to lose specie as imports from outside the region increase. The overexpansive bank will bear the brunt of the drain in reserves. It is possible for all the banks in a region to overissue. but this would eventually be checked by a loss in reserves. 10 In addition to blaming the caisses for speculation, one pamphleteer accused them of buying up foreign bills with billets, thereby driving up the exchange rate (see Botlchary. 1941. pp. 18-19). 11 Between January 1791 and June 1792, prices increased approximately 50% in Paris (E. N. White, 1989. Appendix I).

245

246

Free Banking II FREE BANKING

263

France was on a fiat standard, but White's analysis can be applied as the assignats served as outside money instead of gold. An inflationary impulse from the overissue of billels is ruled out in the long run because the assignats would be redeemed, but it is entirely possible in the short run. That is precisely what the enemies of the clIisses plIlriolhllll'S claimed was happening. They were afraid that in the short run the clIisse.\· were defrauding the public, overissuing notes to buy up scarce goods, and driving up the price level. Whether the caisses would be checked in the long run was immaterial to them and irrelevant to the political debate. The question is, thus, whether the clIisse.\' palriuliqlles quickly expanded their billets to gain some quick profits in a politically unstable climate or controlled their note issue in the short run in order to maintain their reputation (like the Caisse Patriotique de Paris), ensuring their success in the longer run. If the former is true, then the radical revolutiollaries were correct and the episode provides strong evidence against the desirability of frt:e banking. If the latter is true, the French experiment shows that even in the most adverse conditions market incentives not only produce price stability in the long run but also the short run. Measuring the intlationary effects of the billels de cO/lfill/lce presents many problems. Inflation in revolutionary France varied considerably over time and from region to region. Ideally, one needs cross-section time-series data on the issue and circulation of the billels, assignats, bank reserve ratios, prices, and other economic variables. Only some of these data were collected and preserved. Fortunately, the central government ordered the local authorities to create price indices, allowing intlation to be tracked in each depllrlemelll. 12 Information on the issue and circulation of currency is much more limited and presents a major problem for an analysis of the inflationary role of the ('(/i.\'ses /JlIlrioli£/lies. It is no simple task to determine the quantity of money in circulation in each (/epllriemelli. While the billels de cUllfillllce tended to circulate in their areas of origin, the assignats were a national currency, moving freely across economic regions. The quantity of assignats in circulation had been rapidly growing. In January of 1791 there were 524 million livres of assignats in circulation. By the end of the year, this had grown to 1,369 million livres. 13 The total issue of billcis dl' cO/lfilll/cC was much more modest, reaching a maximum of approximately 140 million livres. Given that this is small relative to the total stock of money, it might be tempting to dismiss the inflationary effects of the billels immediately. 14 .2 These regional price indices are presented and analyzed in E. N. White (19119). " For estimates for the stock of assignats in circulation see E. N. White (19117b) . • 4 The bank notes issued during the free banking era in the United States formed a larger fradion of the lIloncy stock, In IK45, total specie in the United Statcs was $'17 million,

Free Banking II 264

EUGENE N. WHITE

But. they still might have had a significant effect on prices if there was a difference in the velocity between the bilfets and the assignats. This should not be ruled out as the hillets de c01~fiance were the dominant currency the public used in its daily transactions. There are no data for the quantity of billet.~ de conjiallce in circulation in each economic region or departement. What is available is the estimated total note issue in some departements derived from the actual note issue in Table I. The data presented in Table 4, along with the total number of caisses, represent the estimated peak issue of billets, attained during the summer of 1792. 15 The figures are only for the 38 departemellts where there is sufficient information to put together a good estimate of note issue. Using the quantity of billets issued in a departement as a proxy for the quantity in circulation in that departement might be considered a heroic leap of faith. Yet. contemporary sources suggest that most billets were used only locally and did not travel very far. This is not surprising as small caisses would be unlikely to have reputations that would encourage a wide geographic acceptance of their notes. The obvious exceptions were the notes of the Caisse Patriotiqlle de Paris and the Maisoll de Secours, the two biggest Parisian caisses, that had national reputations. However, even the billets of these caisses tended to remain in their departement of origin. When the Maison de Secours was liquidated. only 7.3% of the notes redeemed after January 9, 1793. came from departements other than Paris (Bouchary, 1941, p. 13\). This was relatively late in the redemption process and it is likely that a greater proportion of these notes came from afar compared to those that were redeemed earlier. Thus, this is probably an upward bound on the percentage of notes moving beyond a caisse's home departemellt. The quantity of billets issued within a departemelll may thus closely approximate the quantity in circulation. What is unknown is the quantity of assignats circulating in each deplIrtement at any time. Nevertheless, this may not be vital to determining the effects the billets had on the price level. The assignats and the billets while bank notes amounted to $105 million and deposits $39 million (Friedman and Schwartz, I97(), Table 13). Banking in France was less developed than in the United States. The nneil'n rcll!illll' had only chartered one bank, and free banking had only 2 years of growth. Although the hil/l'ls were only a small component of the money stock, they were controversial as are many financial innovations. " The data in Table 4 is derived from Table I. Only the deparlemenls where there was complete or nearly complete data on issue were used. When a few small C(li.Ul'S in a depnrlemen' had no reported issue, 5()()() livres per missing caisse were added. Thus. the Ain with 18 of 22 misses reporting has its total issue raised from 228,O()() to 24R.()()() livres. Alternative adjustments did not alter the results, apparently because of the small quantity of missing data (see E. N. White, 1987a).

247

Free Banking II

248

265

FREE BANKING TABLE 4 Caisses Pulrioliques and Their Estimated Note Issue by Deparlement

Deparlement Ain Ardeche Ardennes Aube Aude Calvados Cher Cote d'Or Cotes du Nord nordogne Doubs Gironde Indre Indre et Loire Isere Loire! Marne Mayenne Meurthe Meuse Nievre Nord Oise Paris Rhin Bas Rhin HaUl Saone el Loire Sarlhe Seine et Marne Seine elOise Seine Inferieure Sevres Deux Somme Var Vendee Vienna Haute Vosges Yonne Totals

Source. E. N. White (l987a).

Total caisses 22 74 II 2 7 26 9 5 7 68 5 18 17 7 24 9 6 3 12 9 14 34 26 64 2 0 15 21 9 18 33

Estimate I

II

248,000 194,565 582,742 722,000 107,469 1,672,555 170,584 1,(162,426 91,292 1,176,099 150,292 9,320,692 231,168 830,000 627,929 1,082,700 1,892,787 1,360,000 299,509 189,584 245,876 1,721,444 1,664,412 32,003,015 153,000 0 417,832 307,520 150,584 292,756 9,081,316 239,053 2,750,510 109,000 299,522 879,364 105,000 174,907

663

72,607,504

13 14 4 15 16 13

Free Banking II

266

249

EUGENE N. WHITE

de cOIzjiance were not perfect substitutes because of their denominatiorwl differences. The absence of low-denomination assignats was, of course, what had called the billets into existence. Omitting the assignats, whose departmental circulation cannot be measured, the demand for billets can be examined as a component of the demand for the media of exchange. It; The per capita demand for real billets balances in the ith departe111ellt (M1) was specified as

M/ = flTAXES

i ,

UP i , LlTi ) +

ei'

(I)

For the variables explaining the differences in demand across ciepartements, one expects that the higher the level of income as proxied by the total real direct taxes per capita in 1791 (TAXES;), and the higher the percentage of urban population (UP;), the greater the demand for billets. The literacy rate for married men (LIT;) was included to test whether the illiterates had a greater willingness to accept and hold billets de cOIzjiance, as critics claimed. 17 It would have been preferable to have had cross-section time-series data for all (!epartements rather than just partial cross-section data on note issue. This would have permitted the use of regression techniques to control for the spillover of notes from one departemelll to another. Thus, the estimation of the demand for billets and their inflationary effects are simple approximations. Three. versions of the demand for real billets balances model were fitted. The first and second excluded Paris, while the third included it. IR In the ordinary least-squares results reported in Table 5, the demand for money was fairly well identified. Both the proxy for income (TAXES) and the urbanization variable had coefficients with an appropriate sign and magnitude. Although it was not possible to reject the hypothesis that most coefficients on TAXES and UP were different from zero at the 10% level, the variable for literacy appears to be far weaker. The very small positive coefficient offers no support to the idea that illiterates accepted billets de cOIzjia1lCe more readily than literates as many historians have presumed. As the nation's financial and demographic center, Paris is an outlier for all variables. The inclusion of Paris in the third equation cau~;es a marked rise in the R2 but does not alter the estimates for the demalld for real balances.

I. The model used is similar to the one employed by Rockoff (1975). 17 Reinhard provided data on population (1941, pp. 26--28, Table IV) and the percentage of urban population and persons per square league (pp. 48-49). Total taxes was obtained from the Moniteur Universel (May 29, 1791). The figures on literacy come from the Ministere de I'lnstruction Publique (1880, pp. CLXVIII-CLXXI). " There were no data for literacy in five departements. including Paris (see Ministere de I'Instruction publique, 1880).

Free Banking II

250

267

FREE BANKING TABLE 5 The Demand for Billels de COl/fiunce

Constant TAXES

UP

Eq. one

Eq. two

Eq. three

-0.014 (-1.25) 0.127 ( 1.22) 0.213 (3.29)

- (1.(114 (- 0.97) 0.10'.1 (0.8'.1) 0.225 0.13) 0.0001 (0.01)

- 0.022 ( -2.30) 0.133 (1.26) O.2H4 (9.47)

1.1'1' Number of observations F

33 0.287 3.11'.1

37 0.286 6.112

H' Stati~tic

38 0.767 57.55

Nol.:. The numbers in parentlll:ses are I-statistics.

With a stable demand function for billets de confiance, an excessive issue of notes would have led to upward pressure on prices. In Table 6, some simple regressions of price level on note issue are reported. The department totals for note issue represent whatever data were available sometime in the summer months, and thus it is necessary to use more than one month's price level to determine whether there is any relalionship. In the first two regressions where Paris is omitted, the quantity of notes issued has no effect on inflation. The second pair of regressions, which include Paris, suggest that greater note issues may have raised price levels. This may be attributed to the fact that Paris had the caisse.l· issuing the most billets. However, Paris was also the center for the issue of outside money, the assignats. TABLE 6 Prices and Billets de Confiance

Price level Paris excluded Dependellt Varial1le Constant Note issue (thousands of livres)

R' F-statistic Numl1er of observations

Paris included

June 1792

Sept 17'.12

Jlllle 17'.12

Sept 1792

131.H (4.04) O'(lO2 ( 1.06) 0.03 1.13 37

135.t (3.59) 0.002 (1.15)

131.9 (3.69) 0.002 (2.74) 0.17 7.53 3H

135.6 (3.23) 0.002 (2.06) 0.1/ 4.25 311

O.()4

1.33 37

NOI.:. The numbers in parentheses are I-statistics.

251

Free Banking II

268

EUGENE N. WHITE TABLE 7 The Effect of the Billets de Confiance on the Price Level (Dependent variable = Price Level June 1792)

- - - - - - - - - - - - - - - - - - - - - - - - - - - _ . _ - - - .... -. Constant M -

Md

Number of observations RZ F-Statistic

Eq. one

Eq. two

Eq. three

133.8 (36.9) -55.31 ( -0.34)

133.2 (36.8) -35.55 ( -0.23)

135.3 (35.4) -7R.77 ( -0.46)

37 0.003 0.12

33 0.001 0.05

3R 0.006 0.22

Note. The numbers in parentheses are I-statistics.

A more sophisticated means of examining the inflationary effects of the billets de conjiallce is to take into account the factors affecting the demand for money. If a departemellt's note issue was excessive relative to its demand for money, then there should have been upward pressure on its price level. Table 7 reports regression results of the price level in the ith departemellt (PRICE j ) on the difference between the actual notes supplied (M j ) and the predicted demand for notes (M'J) obtained from the equations in Table 5. PRICE j = g(Mj

-

M'J)

+ ej.

(2)

There is no evidence from this exercise to support the argument that the billets de conjiallce were inflationary. Not only do the coefficients for the independent variable have the wrong sign, but it is impossible to reject the hypothesis that they are different from zero at the 10% level '9 • While there are generally no time series for the issue of billets, the record for the preeminent caisse in the departemellt of Calvados has been preserved. The monthly issue of this caisse in Caen and the price level in Calvados are presented in Table 8. A simple regression of the log of the price level (PRICE) on the log of the quantity of hillets (BI LLETS), reported as Eq. (I) in Table 9, reveals why the cllisse.f were held responsible for inflation. There was a close correlation between note issue and the rising price level. However, the growth of hillets could easily have been the result of the increase in assignats that served as reserves. The reserve ratios and other important information are unknown, but there are monthly data on the total assignats issued in France. As some of the total issue flowed into the departemellt of Calvados, the 19 These regressions used the June 1792 prices. The results lIsing the September 1792 prices are even less favorable to the idea that the billets were responsible for innation.

252

Free Banking II

269

FREE BANKING TABLE 8 Money and Prices in Caen

Date June 17'.11 July 1791 August 1791 September 1791 October 1791 November 1791 December 1791 January 1792 February 1792 March 1792 April 1792 May 1792 June 1792 July 1792 August 1792 September 1792 October 1792 November 1792 SOl/reI"'.

Assignats in circulation nationally (million livres)

Caen billets de contiance issued (livres)

Calvados price indl:x (l)ec. 1790 = 100)

983 1062 1114 1164 1265 1332 1369 1433 1502 1549 1598 1667 1706 1748 1821 1972 2082 2192

10,246 I!I ,200 26,220 65,520 155,342 192,130 322,578 487,130 772,530 1,038,419 1,154,080 1,189,543 1,226,943 1,278,643 1,295,643 I,Z95,643 1,31Z,043 1,325,958

109.0 1111.2 113.3

113.3 114.9 114.6 116.3 119.0 IZ3.1 128.2 123.8 128.6 133.8 130.3 134.2 12!1.2 12l!.Z 134.7

E. N. White (19l!7b), Beranger (1910), E. N. White (19119).

local cai.!..se~' would have expanded their note issue like any other fractional reserve banks. To separate the inflationary effects of the assignats and the billets de COl/fiance, the log of the billets of Caen were regressed on the log of the stock of assignats (ASSIGNATS) in circulation nationally.20 Eq. (2) shows that a substantial fraction of the growth of bank notes can be explained by the increasing supply of assignats. The price level was then regressed on the residuals of this equation, the unexplained issue of billets (UN EXPLAINED). These results of Eq. (3) indicate the hypothesis, that the unpredicted component of billets had no effect on the price level, cannot be rejected at the 1% level. There is, however, a distllrbingly high level of ulltowrrclation in Eqs. (2) and (3). This is partly a consequence of a dramatic slowdown in the issue of bank notes after March 1792. As discussed below, this change in behavior was the result of a crisis that caused banks like the caisse in Caen and the Caisse PlItriotiqlle de Paris to pursue more conservative 20 There is no measure of the assignats in circulation in Calvados. It is assumed that any national increase in assignats leads to a proportional increase in the assignats in circulatioll in the departmellt.

Free Banking II

270

EUGENE N. WHITE TABLE 9 Billets de Conliance and Innation in Caen June 1791 to November 1792 (I)

In(PRICE) = 4.31 + 0.039In(BILLETS) (74.1) (8.6) R Z = 0.823 N = 18

F-statistic = 74.8 Durbin-Watson = 0.90

(2) In(BILLETS) = -35.19 + 6.56In(ASSIGNATS) ( - 6.99) (9.53) RZ = 0.850 N = 18

F-statistic = 90.85 Durbin-Watson = 0.19

(3) In(PRICE) = 4.81 + 0.015UNEXPLAINED (279.6) (0.53) R Z = 0.02 N = 18

F-statistic = 0.3 Durbin-Watson = 0.17

(4) In(BILLETS) = -61.71 + 10.28In(ASSIGNATS) ( - 26.5) (31.6) R Z = 0.99 N = II

F-statistic = 998.4 Durbin-Watson = 1.95

(5) In(BlLLETS) = 11.67 (17.8) R Z = 0.73 N = 7

+ 0.32In(ASSIGNATS) (3.7)

Durbin-Watson = 0.95 F-statistic = 13.4

(6) In(PRICE) = 4.76 - O.OIUNEXPLAINED (294.8) ( - 0.10) R Z = 0.01 N=II

Durbin-Watson = 0.23 F-statistic = 0.01

(7) In(PRICE) = 4.87 + 0.28UNEXPLAINED (535.1) (0.59) R Z = 0.06 N = 7 Noll'.

Durbin-Watson = 2.53 F-statistic = 0.34

The numbers in parentheses are I-statistics.

policies. Breaking the sample at this point and estimating Eqs. (4) and (5) eliminates some of the autocorrelation. The change in the coefficients from Eq. (4) to (5) shows how the increase in billets in response to an increase in assignats was substantially reduced after the crisis. Yet. as seen in Eqs. (6) and (7), the unexplained issue of billet.f still has no recognizable effect on inflation in either subperiod. In spite of the paucity of observations, it seems clear that inflation was being driven not by unscrupulous local bankers but the excessive issue of assignats. The caiS,fes palrioliques were not wildcat operations,

253

254

Free Banking II FREE BANKING

271

their cautious note issuing practices suggest that they were primarily concerned with soundness and their future reputation.

IV. WHY THE EXPERIMENT FAILED The story of the demise of the caisses patriotiques has always been told as a consequence of their disruptive activities; however, there is scant evidence for fraudulent or inflationary banking practices. It thus remains to be explained why the government and the public were convinced that the experiment in free banking should be abruptly terminated. A close inspection of the events reveals that sudden action by government undermined public confidence in the billets de confiance, paving the way for their abolition. Traditional exposes of the evils of the caissej' patriotiques have focused their criticism on the Parisian caisses, especially the Caisse Patriotique de Paris and the Maison de Secours.21 Yet, although these two caisses were assailed by critics, the public continued to use their notes with confidence for a long time. Serious problems did begin in early February 1792 when the notes of these caisses lost acceptability in the departemellt of the Eure-et-Loir. This crisis abated when the cais:!;es quickly dispatched agents to freely exchange assignats for their billets (Bouchary, 1941, pp. 32-33; Horsin-Deon, 1958, pp. 267-268). While excessive note issue was blamed for this crisis, the immediate reason was that the government's receveurs de district (district tax collectors) suddenly refused to accept the billets for taxes. The origins of this decision have never been adequately explained, but fear of accepting counterfeits may have precipitated this action. In the previous month, a large cache of counterfeit billets of the Maison de Secours were discovered by the police in Paris (Bouchary, 1946, pp. 83-84). News that many counterfeits had already been put into circulation probably prompted the receveurs to take this perfectly legal step. Even though confidence was restored by the caisses' prompt action, the opponents of the caisses took this opportunity to press their attack. In Paris, the Legislative Assembly's Comite des finances agreed to reexamine the question of regulating the caisses patriotiques and produced a report on February 25, 1792. While initially conceding that the billets lie confiwu:e might have contributed some to inflation and speculation, the committee reported that most fears were ill-founded. The report (Archives Nationales, AD IX 495) then proceeded to argue that any government regulation would be ineffective and an infringement on the rights of property and the principles uf liberty.

" Harris (1930, pp. 24-26), whose treatment of the cail.se~· ptllri(}liC/lle~' is brief and dismissive, mistakenly believed that these two caisses were one.

Free Banking II

272

EUGENE N. WHITE

Although the report reaffirmed the legal standing of the hillets de cOIllimlce, the debate grew more bitter and partisan. As war seemed likely ami economic dislocations mounted, more memhcrs of the assclllbly listened sympathetically to the critics of the caisse.v. On March 2M, a proposal to require the caisse.v to submit to government regulation and deposit their reserves, printing plates, paper, and ink with the local authorities was presented to the assembly (Horsin-Deon, 1958, p. 270). This extreme act was not adopted by the assembly, but a panic broke out the day after it was seriously debated. The hardest hit bank was the Mai.voll de Secours. When it closed it doors, noteholders rioted and had to be driven off by the national guard. The cashier ned the bank with some funds. leaving the owners with only 60,000 livres in reserves. The noteholders advised the Mayor of Paris of the bank's imminent ruin. He appealed to the assembly. On March 30, the assemhly grudgingly voted the city a loan of 3 million Iivres to assist any troubled caisses. The assembly's quick response averted an incipient nationwide panic. but the assembly was now convinced that regulation was necessary. On the same day as it granted the loan, it ordered the municipalities to examine all caisses and forbade any further note issue except by government-supervised caisses. The disintegration of the central government's authority meant that this decree went largely ignored. In fact. there was a boom in the number of new nominally government-supervised caisses. It is important to note that many caisses weathered these troubled times quite well. The Caisse Patriotique de Paris trimmed its sails. reduced its note issue from 27 to 17 million livres. and published its balance sheet. The government might have wished to terminate the cai.vses patriotique.v, but it ,:ould not do so during the summer of 1792 as few lowdenomination assignats had been printed and coins were still hoarded. What the assembly failed to learn from the events of March was the effect of any of their actions on public confidence. On August 22. the assembly decreed that all bearer notes were subject to the registration tax (droit d' enregistrement) (Horsin-Deon, 1958, pp. 281-282). Although it was uncertain whether this applied to the billets de cOI!fiallce, it prodllced a panic in Paris until the assemhly made it clear that bilkts were not included. Popular unrest in Paris erupted in the jOllrnee.v of August and September. pushing the assembly and. after September 20. the Convention in a more radical direction. On September 3, the private issue of coin was forbidden. 22 On September 14. the assembly accepted a plan to retire the Parisian billets de conjiance. The public was put to the great inconvenience of long lines; but apart from the Maison de Secollrs, only one 22

Some privately issued coins continued to circulate afterward (Caron. 1912. p. 215).

255

256

Free Banking II FREE BANKING

273

other caisse required government assistance to reimburse the public. The Caisse Patriotique de Paris finished its exemplary record by quickly retiring its entire issue, leaving a large surplus that was distributed among the shareholders (Bouchary, 1941, pp. 72-73). The convention took the final step on November 8 and voted to liquidate all caisses. Examination of the ca;sses and retirement of the notes was to take place by January 1, 1793. While the gradual note redemption had worked moderately well in Paris where a supply of low-denomination assignats was ensured, the provinces bitterly complained about the new redemption plan. There was a dearth of assignats of lesser value and the public had considerable difficulty in returning the variety of notes they held to the issuing caisses. This forced the convention to continually grant delays before all billets de culljiallce were finally declared worthless. Note redemption dragged out for months. In Caen, for example, only 71% of the notes that had been in circulation on November 8 had been retired by mid-August 1793 (Beranger, 1910, pp. 336). Most billets were redeemed for assignats by the end of the year, but some notes continued to be exchanged well into 1794. This huge clearing operation had many problems, some counterfeits were accepted and notes coming from other departemellts clogged the mail. But, in this undertaking, it is remarkable that very few of the over 1600 caisses required government assistance to liquidate. This achievement offers further evidence for the essential soundness of the (:a;sses patr;otiques. Most historians have assumed that when the government closed down the ca;sses patriotiques it acted in what it believed to be the best interests of the public-protecting it from bank notes of dubious quality. Yet the government actions may not have been so high-minded. By late 1792, well over half the government's expenditures were being covered by the issue of assignats. The success of such an inflation tax depended on the willingness of the public to hold money as prices rose. If the government could find some way to increase the willingness of the public to hold assignats, the inflation tax could be increased. 23 Forcing the retirement of the billets de confiance caused people to increase their holdings of assignats, thereby promoting the government's financial strategy. While this was not the stated goal, the benefits to government coffers were quite real. A simple measure of the tax is the increase in the stuck of assignats divided by the average price level: (M, - M, 1)/O.5(PI + PI_I). This index, using Paris prices, is shown in Fig. I for July 1791 to

2J One principle of "infiationary finance" is that "the infiatillll that results from a given government deficit will be lower than otherwise if the availability of close substitutes for government money is curtailed" (Nichols, 1974, p. 423).

257

Free Banking II

274

EUGENE N. WHITE 120,--------------------------------------------

90

•.e

~

'0

80 70

•c

80

E

50

~

40 30 20 December

July

1791

1791

FIG. I.

JUly

1792

December

1792

Revenue from inflation: July 1791-December 1792.

December 1792. 24 The big jump in government revenue occurs in late August and early September. This is when the plans for closing the caisses in Paris are first revealed. The forced shift from hillet.f de COIIfiallce to assignats, thus, produced a significant increase in the innation tax for the government. V. CONCLUSION

The experiment with free banking during the French Revolution, although brief, provides additional evidence on the success offree banking, France welcomed both large sophisticated banks in Paris and the "beggarly bankers" in the countryside. Counterfeiting may have been a problem, but however numerous counterfeits were, they did not frustrate the final reimbursement of noteholders. The French experience also yields some evidence on the eITects of free banking on price stability. noth cross-sectional and time-series data provide little support for the traditional view that the clIis.H'.v pat,.iotiC/lIl·.\' overissued bank notes. The vast majority of cais.ves were not wildcat banks attempting to make a quick fortune. They were sound banks interested in establishing a reputation for prudence. Although the government was loath to admit it. the principal generator of inflation was the rapid issue of assignats. The French experiment does not oITer a definitive 24 The price index is found in E. N. White (1989). Using other regional price indice~ reveals similar increases in the inflalion tax. The slock of a~signats in circulalion is found in Table 8.

258

Free Banking II FREE BANKING

275

answer to the desirability of unregulated banking. It does show that free banking was reasonably successful even in the most trying political times. Unfortunately, while free banking was viable economically, it could not survive the radicalization of revolutionary politics that brought an end to laissez-faire economic policies.

REFERENCES Aftalion, F. (19117), I. economie de la Revolution franraise. Paris: Pluriel. An:hives Nationales, AD VIII 4, Letter from Valenciennes, and AD IX 495. Beranger, J. (19011), "Billets de Confiance de la Societe Patriotique de Rouen." GC4Z/!IIe Nllmimwtiqlle Fran~'aise 12,2113-317. Bcrangcr, J. (1910), "Bons des communes de la Periode Revohllionnaire: la Societe I'atrioti'lue de Caen." Gazelle Numisllwtique Fralll;aise 14, 335-412. Bloch, C. (1910), "La Verification des Caisses Palriotiques." Billie/in d'Histoire EcO/wmiqlle de la IUvolution, 134--197. Bruneau, M. (1902/1977), Les Debllts de la Revollllion dans les DepartemenU" dll CII", et cie /"Incire. Paris: Slatkine-Megariolis. Bouchary, J. (1941), Les compagnies jinancierel' a Paris cllI XVIII siede. Paris: Marcel Riviere, Vol. 2, Bouchary, J, (1946), Les FClllx-McJllnayellrs de la Revolution Frcmraise, Paris: Marcel Riviere, Caron, p, (1912), I.e mOll/wie et Ie papier-IIuJllnaie. Paris: Imprimerie Nationale. Carraz, 1c distinguished from clearinghouse /0ulling losses frum the issues in proportion to the participating banks' capital and surplus or by the pruportiunal amuunt uf their clearings during the previous year.

288

Free Banking II 10

:

MONEY. CREDIT. AND BANKING

c1earinghousc l:UlTCIH:Y would return to the hank~ liS paYllIent for liquidating the loans that gave rise to the original issues of notes. Gilman also relt that the interest charge of 6 percent to hanks who took out clearinghouse currency "would act as a check upon their issue. and they would not be taken so much for profit as for protection and necessity" (p. 157). Gilman's plan was introduced as a hill in the House of Representatives by Ben L. Fairchild (Rep. - N. Y.) in lanuary I R96. It was then referred to the House Committee on Banking and Currency, hut it never appeared again, and no popular movement ever developed in support of his scheme. One Treasury official who pointedly objected to the clearinghouse strategy was Leslie M. Shaw. Secretary of the Treasury from 1902 to 1907 (Timherlake 1l}7R. pp. 171-85). The clearinghouse loan certificates. he stated in an address delivered in 1905. are a "plea of guilty to an indictment charging bad management locally or bad legislation nationally" (Shaw 1908, p. 279). Yet, Shaw's prescription for appropriate policy during a panic logically supported the clearinghouse method. First, he wrote, the supplemental currency should be identical with the legitimate currency that already existed. Second, he continued. the relief should he capahle of immediate and local application. Finally, he concluded, the additional currency should be retired promptly when "the demand therefore ceases" (p. 293). In point of fact, the clearinghouse issues fit Shaw's criteria reasonably well. Their use as currency was ubiquitous; they were issued in a Illultitude of localities on the perceptions and judgments of thousands of bankers, and the notes never stayed out more than a few months. Nevertheless, Shaw regarded thelll as undesirable because in his opinion they raised an alarm that something was wrong and thereby caused "hoarding" (p. 294).8 The Comptroller of the Currency. lames H. Eckels. in his Rcport for IR93 was more sympathetic. He lauded the clearinghouse associations for their actions during the previous summer. "The service rendered by them." he wrote "was invaluable •... and to their timely issuance ... is due the fact that the year's record of suspensions and failures is not greatly augmented" (U .S. Comptroller of the Currency 1893. p. 15). Eckels denied that the issues were currency: hut he had tongue in cheek. "If they had been used as currency." he argued, "the banks issuing thelll would have heen fined" (p. 16). Since the banks were not fined hy the governlllent. the law must not have been violated. This syllogism is more reasonahly viewed as pragmatic accommodation to events than logical defense of the law. The comptroller in 1907. William Ridgely. proved to be more of a legal constructionist. lie noted that the law. under Section 5192 of the Revised Statutes, declared that clearinghouse certificates ;11 tire /JOJJeJS;OIl of allY ballk belollg;IIg to the clcal';lIglroliJe associat;oll were lawful money (U.S. Comptroller of the Currency IY07. "Furthcr on in his articlc. Shaw contradictcd his own implication that thc governmcnt ~1](JUld ovcrsee a supplcmcntal. hcavily taxed, national bank note currency. "Governmcnl officials." he stressed. "arc eXlremcly cautious. The rejection of a proposition never causes trouble. Atlirnmtive acts only arc investigated and censured. Technical objections are as good as valid ones with the average oureau official" (p. 295). Given this fact of bureaucratism. a governmcnt agency could hardly oc expeclcd 10 furnish relicf for a typical crisis following thc norms that Shaw had specificd.

Free Banking II RICHARD H. TIMBERLAKE, JR.

289 :

II

p. 64). He begged Ihe question of those issues that were used universally as media of exchange outside of banks. Later in his Report he discussed the possibility of legalizing the clearinghouse issues as an "emergency circulation." He found "merit" in this suggestion, but rejected it as a "half-way measure." The full development of the clearinghouse idea, he argued, "should carry us further ... to the inevitable and logical conclusion ... , which is that we should have a national central bank of issue and reserve lacting under governmental authority)." This institution would be "more systematic and efficient" and would have none of the "disadvantages" of the clearinghouse system. Ridgely did not specify how or why the government central bank would be more systematic and efficient; he simply offered unsubstantiated assertions. For example, .. if we had had such a Igovernment centmll hank in opemtion in 1907, no such hallk panic as we have h,ld would have hecn possihle" (p. 75). Olad Ridgely been ahle to witness the monetary and banking events of 1929-33, he might have experienced a fair amount of intellectual indigestion.) Most amazingly, his next paragraph is a description of how well the clearinghouse system generally operated.

9.

RESTRICTION AND SUSPENSION OF CASH PAYMENTS

The most characteristic feature of a panic-the attempt to convert demand deposits into legal tender-stimulated the restriction of cash payments by the banks. Restriction took many forms: for example, limitation of cash payments to a nominal amount ($25 to $200) per transaction; or this amount per day or per customer. Often payments were "discretionary" with the banks and therefore negotiable with the banks' managers. Its purpose was to induce as much friction as possible into any transaction that would provoke an "internal drain" of bank reserves. Sprague argued that the clearinghouse banks in New York assumed the indulgence of restriction too readily (1910, pp. 273-77). They faced reserve requirements of 25 percent and were prohibited from making new loans when reserve ratios reached this minimum. However, the law did not restrict them from paying out reserves even though they were in a reserve-deficit position. In his article, William Dewald (1972) has reargued Sprague's analysis and found it valid. The New York banks, Dewald notes, "had made a fetish out of their twenty-five percent reserve ratio and held closely to it. ... Reserve deficits were small during crises, and New York banks in fact held enormous reserves on every occasion when they suspended payments. It was this policy that Sprague attacked and it was this policy that the legal authorization to issue clearing house 'currency' promised to change" (p. 940). Sprague, Dewald notes, recommended that reserve banks be required to maintain payments as long as they held any reserves. This principle had been repeatedly urged as well by the Comptroller of the Currency (p. 939). The conclusion of the Sprague-Dewald thesis on restriction-of-payments policy is that, far from being complementary to the issue of c1e'lringhouse currency, it increased uncertuinty and generally upset the payments systcm. It witnessed the

290

. Free Banking II 12

:

MONEY. CREDIT. AND BANKING

substitution of less efficient mcdia for conventional money and also cxtcnded the pcriod of disequilihriulll (pp. 914. 941 -42). 'I A contrary view on restrict ion comes from Friedman and Sd1\vart/. ( I WI]). and also from Margaret Myers in her earlier study (1931. p. 143). Given a bank panic. Friedman and Schwartz argue. "the fairly prompt restriction of payments was a therapeutic measure that almost surely kept the contraction from being even more severe and much more protracted than it was . . . . (Tlhe failure of one (unsound) bank did not set in train a chain reaction. Restriction of payments thus protected the banking system and gave time for the immediate panic to wear off. as well as for additional currency to be made available" (1963. pp. 163. 1(7). The alternative was for the fractional reserve banking system to experience a "black hole" type of collapse in the face of the increased demand for money with the m:companying declines of prices. wages, income. and business activity (pp. 167--6XI. Friedman and Schwartz correctly focus on the inherent instability of the fractiomll reserve banking system as the primary danger to the monetary system and thc economy. Nonetheless. given that clearinghouse issues were already in progress. the I'l'Oml" restriction of payments may have heen ill-advised. Why not require the clearinghouse banks to nm down their reserves to a very small minimlllll with graduated rate penalties on their reserve deficiencies'! A law that provided explicitly for such an action would have countered the timidity of the hanks and. in conjunction with the clearinghouse issues. encouraged them to maintain the level of their current portfolios as well as their cash payments. Ultimately, they might have hecn forced to suspend: but before they did so, their reserves would have heen largely utilized in meeting internal and external demands for liquidity.

10.

MONETARY REFORM LEGISLATION

The immediate legislative reaction to the events of 1907 was the Aldrich-Vrceland Act of 1908, which called for the grouping of ten or more national hanks into a National Currency Association. The associations anticipated from this act wcre not to operitte itS c1caringhouses. hut were to issue "emergency currency" just the way the clearinghouses had done in the past. The only difference was tlmt the new institutions' currency issues were to be under the administration of the Secretary of the Treasury, who had some discretion in allocating the notes to different sections of the country. The currency outstanding was to be taxed at an annual rate of 5 percent during the first month. and at an additional I percent each month thereafter until the annual rate reached 10 percent (Friedman and Schwartz 1963. p. 170). The Aldrich-Vreeland Act marked a critical turning point in United States policy toward the monetary system. As Redlich observed, the act "turned its back on the clearing-house loan certificate instead of legalizing it. . . . [The I suh-committee "Spntgue also favored the "pooling" of reserves in order tfl reduce the tOfl-ready tendcnl'y of Ihe banks to restrict or suspend. Dewatd apparently agrees (p. 941). However. this provision would have vitiated an important market constraint (see above. p. 6). Sprague's principal remedy requiring banks to use their reserves. in conjunction with clearinghouse issues. would have been sufficient to ,top 'my p,mic.

Free Banking II RICHARD H. TIMBERLAKE, JR.

291 :

13

[Report) of the Commillee of Banking and Currency declared the power to issue loan certificates dangerous as long as clearing-house associations were not under government control" (1951, p. 168). The allegation of "dangerous" flies in the face of the clearinghouse system's actual performance. The most extraordinary fact associated with the several clearinghouse episodes between 1857 and 1907 is that the losses from all the various note issues, spurious and otherwise, were negligible! The only loss reported in any of the accounts here considered was $170,000 in Philadelphia in 1890 out of an issue of $9.7 million-1.8 percent (Sprague 1910, p. 392). Few of the economists who analyzed clearinghouse operations even noted in passing this astonishing record, and none used it as an argument for continuing the system. When the Federal Reserve Act was debated in Congress in 1913, the sentiment again reflected the notion that the government should have some control over the currency. In a typical statement, Congressman William A. Cullop of Indiana COIltrasted the new Federal Reserve Board with the clearinghouse executive commillee: "Every fair-minded man," he declared, "would prefer that the control over the currency be vested in seven men [on the Federal Reserve Board) selected from different parts of the country than to have it remain in the hands of the five managers of the New Yark clearing house . . . . To this self-constituted and unauthorized organization [the Federal Reserve Act) . . . deals a deadly blow in order to secure for us industrial and commercial freedom" (U.S. Congress 1913a, p. 332). Cullop's hyperbole demonstrated the decided influence populism still exerted in monetary affairs. Hepburn's comment on the passage of the Federal Reserve Act was that this "measure ... , while open to criticism, is a vast improvement upon the old system and as good as could have been fairly expected under all the circumstances" (1924, p. 397). Margaret Myers was even more positive: "The passage of the Federal Reserve Act," she wrote, "ended the long struggle with monetary problems which had harassed the New York money market and the rest of the country throughout history" (1931, p. 421).

tt. RECAPITULATION A fractional reserve banking system by its very nature contains an element of instability. An increased demand for currency can so deplete bank reserves that the supply of bank-created money (deposits) declines. This possible reaction means that the supply of money is at times unstable: an increased demand for one kind of money may result in a decline in the total quantity of money supplied. The banking industry recognized this problem at an early stage. Since both the money-supplying industry-the banks-and the users of money-firms and households-were harmed by the results of bank-aggravated crises, everyone had an interest in reducing or eliminating this structural inslilhility. The solution, arrived at through force of circumstances and some financiul innovation, was clearinghouse creation of temporary currency. The banking industry

Free Banking II

292 14

:

MONEY, CREDIT, AND BANKING

simply reinstituted itself as an ad hoc central bank. and through its clearinghouse associations issued more currency. Since the laws proscribed private issues of money. clearinghouse currency had to be issued discreetly. It thcn worked so wcll that no one had any incentive to prosecute its issuers. It was the private-moncyproducing industry's answer to a pronounced need. It was, as well. constrained by market factors-interest rate charges on its issue, and the real stake the clearinghouse banks' directors had in seeing to it that the clearinghouse association did not make costly mistakes. In addition, the gold standard system regulated and constrained the monetary base. Bankers, who knew the rules of the gold standard game. realized that clearinghouse issues were only a temporizing device. They knew they had to restrain bank credit appropriately when their reserves were threatened. Nonetheless. the issue of clearinghouse currency put the brakes on the development of an unstable bank credit contraction. It did not prevent the dcmise of inefficient banks; it only stopped the fractional reserve collapse that might otherwise have occurred. One may then wonder why this system was rejected. Why was a government central bank superimposed on the banking industry when the clearinghouse method had proven so effective? The accounts of the times suggest several answers. First. the clearinghouse issues by 1907 had become recognizably illegal. This awareness could have meant simply that the laws were bad. Instead, the realization endowed the clearinghouse currency with a hucksterish quality. Because clearinghouse currency was illegal. it was makeshift funny money-never mind the logic of the laws prohibiting it. Second, since clearinghouse currency was market-regulated and because it was created and accepted voluntarily, many contemporary observers did not recognize its utility and integrity. They wanted an "official" currency similar to national bank notes or greenbacks. Third, the clearinghouse currency seemed to arise out of nothing. Everyone could understand the principle of a quasi-public central bank standing by with a reserve obtained from contributions by participating banks and ready to be applied when a crisis threatened. But few could understand how a clearinghouse system could create emergency currency without sceming to havc an emergency reserve on which to base it (Sprague 1910. p. 234). It was altogether too mystifying an operation for common understanding and acceptance. It smacked of Wall Street legerdemain. Last. the clearinghouse issues were associated with the restriction or suspension of cash payments. Sprague clearly del11onstrated that the two actions were not at all necessarily related (1910. pp. 260-77). Nonetheless, the popular mind was only too willing to apply here the fallacy that correlation implies causation. Many advocates of a central bank saw the Federal Reserve System as an evolutionary development of the clearinghouse associations. "This bill. for the most part," said Robert Owen, the sponsor of the Federal Reserve bill in the Senate," is merely putting into legal shape that which hitherto has been illegally done" (U .S. Congress 1913b. p. 904; see also Willis 1922, p: 251). The Federal Reserve alternative, however, was critically different from the clearinghouse system. It introduced a discretionary political element into monetary deci-

293

Free Banking 1/

RICHARD H. TIMBERLAKE, JR.

15

sion making and thereby divorced the authority for determining the system's behavior from those who had a self-interest in maintaining its integrity. LITERATURE CITED

Andrew, Abram Pialt. "Hoarding in the Panic of 1907." Qllarter/yJoIITlla/ o/EclIIwmics 22 (February 1908, 290-99 (a). ___ . "Substitutes for Cash in the Panic of 1907." Qlltlrterl.v )mmlll/ 0/ EnJl/omi/'.v 22 (August 19(8),497-516 (b). Cannon, Jallles Graham. C/elITillg HUII.ve.I·. Sellate OOl'lImelll Nllmb,'r -191. 61 st Congress, 2nd Session, National Monetary Commission, 19 Hl. Dewald, William G. "The National Monetary Commission: A Look Back." )mmlll/ uf MUlley, Credit, alld Btlllkil/g 4 (November 1972),930-56. Frangul, Ramzi. Book Review of Origi/ls ofCelllm/ Btlllki/lg i/l the Ullited SWtes: JOIITlla/ 0/ Fi/ltlllce 35 (March 1980), 212-14. Friedman, Milton, and Anna J. Schwartz. A MOl/etary Hil'lOry 0/ The Ullited SUItes, UJ67-1960. Princeton: Princeton University Press for the National Bureau of Economic Research, 1963. Gilman. Theodore. A Grmled BlIl/killg System. Boston and New York: Houghton Mimin, 1898. Hayek. Freidrich A. The De-NlIIilllllllizlIIioll uf MO/ley. 2nd ed. London: Institute of Economic Affairs, 1978. Hepburn. Alonzo Barton. History 0/ Coi/lage 'Illd Cllrrem'Y il/ tile UI/ited SllItes 'Illd the Perell/lial CUlllest for SOlllld MOIII!)'. rev. ed. New York: Macmillan, 1924, Klein. Benjamin. "The Competitive Supply of Money." )ollTllal of MOlley. Credit. tlml lJallkillg 6 (November 1974),423-54. Mints, Lloyd W. A History of Ballkillg Theory i/l Gre'" 8,.ilai/l I/I/el the Ullited Slates. Chicago: University of Chicago Press, 1945. Myers, Margaret. The New York MUI/ey Market. Origil/s 'Illd Del'elopmellt. Vul. I. New York: Columbia University Press, 1931. Redlich, Fritz. The Moldil/g ufAmericclII BI/I/kil/g, Mel/ 'Illelle/etls, Part II, 18-10-19/0. New York: Hafner, 195 J. Shaw, Leslie M. Cllrrellllsslles. New York: Appleton, 1908. Spragul!, Oliver M. W. History .uf Crises UI/der tile NlIIiollal 81/l/kil/g System. Semlle Doclllllelll Number 538. 61st Congress, 2nd Session, National Monetary COlllmission, 1910. Timberlake, Richard H., Jr. Origil/s of Cel/tml B'lIIkillg ill the Ullited SlIltes. Cambridge: Harvard University Press, 1978. United States Comptroller of the Currency. AIII/utli Report. Washington, D.C., 1893. _ _ . AI/III"" Report. Washington, D.C., 1907. United States Congress. CIJIIHrel'siCIIUlI Rt·l'lml. 63rd Congress, 1st Sessiun, Appendix, II) 13 (a).

___ . COl/gressiolltl/ Record. 63rd Congress, 2nd Sessiun, 1913 (b).

White, Horace. MOl/ey (/lui 8al/kil/g, Revised and enlarged by Clmrles S. Tippens "nd Lewis A. Froman. New York: Ginn, 1935. Willis, lIenry Parker. "Feder,,1 Reserve Ulll1ks." In TI,,' 1'/"'0')' (l1It111i.~w(l· ,!f U(llIkillg. edited by C::harleli F. Dunbar. New York and Lundon: G. I). I)utnam's Sons, 11)22.

[18] Competitive Monies and the Suffolk Bank System: A Contractual Perspective* DONALD J. MULLINEAUX UI/iversity (!f" l\el/llldy

Lexington, Ke1llucky

J. Introduction In recent years, a number of monetary economists have begun to focus research on an old topic: the economics of private money or "free banking." Much of this literature is motivated by a sense that competitive markets may well prove superior to government regulation in coping with the problems inherent in a "laissez-faire" monetary system. Noting this, Robert King, in his recent survey paper, suggests that: "The central focus of future research in this area must be the types of monetary institutions that the private market will deliver" [10,123]. By monetary institutions, King presumably means the contractual mechanisms utili7,ed in production and exchange relationships involving private money. This paper examines one mechanism -the so-called Suffolk Bank System--which "the market" did deliver in New England from around 1820 to just before the Civil War. While most historians have looked favorably on the Suffolk mechanism I, almost all have mischaracterized the underlying rationale for the system. I n particular, most analysts have interpreted the establishment of the Suffolk Bank's note redemption program as a response to Gresham's Law ("Bad money drives out good money"). However, the conditions of 19th century banking in New England (and elsewhere as well) prior to the establishment of the Sulfolk system rule out any application of Gresham's Law. Prior studies have also neglected the importance of evolutionary changes in the Suffolk Bank system. Over the /irst five years of its existence, what we will label Suffolk I was largely unsuccessful both in terms of its profitability and its capacity to control the behavior of the banks participating in the system. Several contract adaptations introduced in 1824 accounted for a sharp reversal in the fortunes of the Suffolk Bank system; it is Suffolk II which has gained the accolades of bank historians. I ronically, Suffolk II iI/creased the prospects for a Gresham's Law kind of outcome (of the sort emphasized by Akerlof [I]). However, the contract structure of Suffolk II involved properties sufficient to allow that Suffolk "The author wishes to thank James Moser and Arthur Rolnick for helpful comments and suggestions, I. lIorace White's view is typical: ", , , it brought banking in New England to a state of responsibilities, order, and solvency unknown before" [20,302]. A more recent analysis of the SulTolk system by Van Fenstermaker and Filer [18] finds that the SulTolk mechanism had no signilicant elfect on the growth rate of money, hut that it did illcrease the bank nole to deposit ratio. This finding is consistent with our interpretation of the Suffolk contract as a mechanism ror enhancing the quality of bank notes.

884

Free Banking II COMPETITIVE MONIES AND TilE SUFFOLK IlANK SYSTEM

management 10 monitor and control the behavior of the participating banks. Consequently, bad banks did not drive good banks out of t:xistt:nce. Analysis of tht: Sullolk arrangement from a contractual perspt:ctive focuses on the issue most nitical to the discussion of "free banking": the ability of "the markt:t" to control the behavior of bank managers. Hayek [9] and White [21] have argued that market forces arc capab\t: of controlling banks, provided that bank liabilities are convertible into some more "basic" money originating outside the banking system. Milton Friedman [5] exemplifies the counter view: ,\ liduci"ry currellcy ostellsibly cOllvertible illto the IllOlictary COllllllodity is thercforc likely to he overissued from time to time and convertibility is tikely to become impossible .... A liduciary currency would thus probably tend through increased issue to degellerate into a wlllmodity currency ·into a literal paper standard - there being riO stable e4uilibriulll price level short of thul at which the money vulue of currency is no gl. caler Ihan tlial of Ihe paper it contains. And in view of Ihe negligible cost of adding lCroS, il is nol clear Ihal Ihere is uny Jinite price level for which this is the case.

Most bank historians agree with Friedman's viewpoint. Bray Hammond [ll] argues that "... the Jacksonians glorified in what was a triumph of laisser faire in a field where lainer faire had no place. Sovereign and unified control of the monetary system is needed in any economy, whatever freedoms may be proper otherwise." We shall argue that analysis of the Suffolk Bank system suggests that the truth lies "somewhere in the middle." That is, the Sulfolk system shows that markets--in the sense of exchange relationships characterized by the complete ahsence {~lhierarchy- were unable to control and discipline the behavior of 19th century bank managers. Nevertheless, in New England the private market did deliver an alternative "governance structure"~ (the Suffolk Bank system) which was largely effective in achieving what "the market" could not. This mechanism involved many of the features of a more modern contractual device-the franchise system. Since a franchise has been recognized as a hybrid of "the market" and "the lirm," we argue that the Suffolk experience shows that banking involves transaction characteristics which require some form of hierarchical "controL" Consequently, Lawrence White's [21] claim that a simple note-exchange device provides sullicient control of free bank managers is questionable. Nevertheless, the Suffolk Bank systcm was a systcm of private rather than government control, and it was reasonably successful. However, the Sulfolk system was: (I) relatively unique to New England; and (2) brought down by a blend of market and extra-market "competition." The Sulfolk's failure to survive in New England may provide some hints concerning the necessity of government intervention in banking markets. The organization of this paper is as follows. Section II provides an overview of the market for bank notes during the early 19th century as background to the analysis of the Sulfolk Bank system. Section III demonstrates that Gresham's Law, ill the traditional sense of the term, cannot provide a rationale for the Sulfolk system. Section I V describes till: evolution of the Sulfolk system, emphasizing the contractual elements of the Sullolk II arrangement. Section V argues that the post-I 1124 Suffolk system can be interpreted as a fram:hisc contract through which the Sulfolk Bank provided a "trademark" or brand name 2. In Oliva Williamsun's work 12J] 011 economic institutions, this lerm is used 10 describe medlanisms which arc: innul!Il~C and l:ontroilfansacliulls costs. A contract is a well·ddincd type of gO\lc:rnance structure, but other IIH;l·h.U1i~lll~ (implicit i.lIId lIunslamJard CUUlral:ls, for examplt:) playa critical rok in aJkclilig IranS.1Clions costs.

employed (0

295

Free Banking II

296 886

DOl/aid J. Mullineaux

backing to notes issued by participating banks in New England. The final section analyzes the demise of the Suffolk Bank system and draws some tentative conclusions.

II. The 19th Century Hank-Note Market The typical bank note contract in 19th century U.S. banking involved a promise by the issuing bank to pay the bearer on demand a specified sum of "lawful money." Lawful money consisted of specie (gold or silvcr). The engraved note instrument contained the name of the issuing hank, the amount of the claim, and the signature of an ofliccr of the bank. Notes were issued by both the First and Second Banks of the United States and by state-chartered banks in various denominations and were the primary media of exchange in small exchange transactions. Notes were placed into circulation when banks made loans or when they were paid over the counter for specie deposits. Rank notes served as a medium of exchange because they were convertible into specie. Notes did not necessarily exchange at par (face value), however, for several reasons. First, notes had to be transported to the issuing bank in order to collect specie. This cost drove a wedge between the value of a note in exchange outside the bank and its specie value at the bank. Thus, while notes normally exchanged at par at or near the location of the issuing banks, the notes of geographically distant banks (labelled "foreign notes" in the 19th century) typically traded at a discount which increased with distance [20]. Second, the promise of a bank to redeem notes at par was subject to some risk of default. Such defaults were not infrequent, and, on several occasions, involved practically all commercial hanks (the socalled "suspension" periods). Consequently, the notes of riskier banks traded at larger discounts than the notes of banks judged liquid and secure enough to redeem at par. A third factor accounting for discounts on certain bank notes was the cost of authenticating the note contract. Counterfeit notes were prevalent during the 19th century prior to the formation of the national banking system, in part because of the large number of different bank note issues. The ease of "passing" counterfeit bills no doubt increased with the geographic distance to the falsely claimed bank of issue, since the signatures of bank oflicers and engraving structures on bills were less likely to be familiar over substantial distances. The presence of discounts on bank notes provided incentives for the establishment of firms which traded bank notes in a secondary market at prices suflicient to cover the costs (including a normal profit) of holding risky assets and of transporting them to the issuing banks to collect specie. Dillistin [3] describes how these note brokers (or "shavers") often published "bank-note reporters" which detailed bid quotes on various hank notes and also contained information concerning outstanding counterfeit bills. Newspapers also puhlished quotations of hank-note discounts and counterfeit-bill information. Redlich [ 14J notes that by the 1820s, a numher of commercial hanks had entered the note-hrokerage business and consequently also quoted discounts on the notes of other hanks. The hank note market ofthc 19th century essentially functioned as a floating exchangerate system, with each note convertihle into a dominant money (specie) at a rate (discount) determined by transportation costs, the perceived riskiness of the operating bank, and the costs of authenticating the note's validity. If the market was efficient, the sizc of the discount on a particular bank's notes provided some information concerning the "quality" of the note in terms of its purchasing power. Thus, a bank which pursued a systematic policy of depre-

297

Free Banking II 887

COMPETITIVE MONIES AND TilE SUFFOLK IlANK SYSTEM

ciating the value of its notes (via overissue) would find that the market "revealed" the bank's strategy via a larger discount on its notes. White [20] notes, for example, that the notes of banks judged riskier than their competition traded at larger discounts than the notes of safe banks. This is not surprising, since note brokers possessed strong incentives to monitor the riskiness of the banks whose notes they bought and sold. But if the market signaled changes in the 'Iuality of certain bank notes via the price systelll, then (,reslwlII's I.aw could not havc applkd to the note market.

III. The Suffolk Hank System and Gresham's Law The Suffolk Bank system has been interpreted by notable historians, such as White [20], Hammond [8], Redlich [14], and Dewey [2], as a private-market response to Gresham's Law. In particular, each of these authors claims that the notes of Boston banks were driven out of circulation by the "inferior" notes of the country banks. Hammond asserts, for example, that "Everywhere, of course, out-of-town notes had to be accepted by the city merchants, though they were of varying and uncertain value, dirty, and hard to be redeemed. Being inferior to city bank notes, they drove the latter out of existence" [8,549]. Dewey notes that: "The consequence was that the bills of country banks obtained the entire circulation even in Boston" [2,79]. The source of this view appears to be Whitney's The Suffulk Balik which notes that: At this time (I !!24) the city was flooded with country money. The circulation consisted almost entirely of the notes of banks outside Boston. With more than one halhhe banking capital of New England, the Boston banks supplied only one twenty-fifth of the bills in use [22,10].

The figures quoted by Whitney were in fact an estimate of the total amount of country bank notes in circulation (based on an assumption of note/capital ratios at country banks) and an estimale of the so-called "permanent" circulation of Boston banks. No details were provided by Whitney concerning the method used to identify "permanent" notes. In fact, the ratio of Boston to country-bank notes was considerably larger than Whitney and many bank historians have suggested. From 1812 to 1844 the ratio of Boston to country notes averaged 77 percent. In 1824, when the Suffolk joined the other Boston banks in a joint venture arrangement for redeeming "foreign" notes, the ratio was 76.3 percent, or roughly equal to the average for the entire period. The notes of Boston banks, indeed, never approached extinction. Nor would Gresham's Law have applied in any case. A necess,lry condition for Gresham's Law is that tht: competing monies t:xchange at a fixed price relative to specit: or some "base" money.' Country bank notes exchangt:d at variabk rates, so that price adjustments could preclude "bad money from driving out good." In n:ality, the institution of the Suffolk Bank system improved the prospects for a versioll of Gresham's Law to come into play since the effect of the (Suffolk II) system was that wuntry bank notes traded at par. That is, the Suffolk system created conditions such that bank notes wen: priced in exchange as if they were perfect substitutes for each other. As 3. Friedman and Sl.:hwartz{4], among others, hay!! cmphasilt:d the: necessity of a lixcu cXl.:h:.mgt! ment fur the upcr~tion of (jn:,slldm's Law.

nilt!

cllviron-

Free Banking 11

298 888

Donald J. Mullineaux

Klein [12) has emphasized, if note holders were unable to differentiate the quality of one bank note from another by some mechanism other than price, each bank would possess incentives to depreciate the quality of its notes. The costs of such a strategy would be borne principally by other banks. Compctition would, via a form of Gresham's Law emphasizcd by Akerlof [I) in his classic "lemons" paper, drive those banks which did not depreciate the quality of their currency out of business. Since all banks would possess the same incentives, the market would, like Akerlors used car market, break down. Ranks would complctely depreciate the quality of their notes, producing the infinite price level predicted by Friedman in a free banking regime. The deterioration of product quality is the predicted outcome in any industry where consumers, because of high information costs, cannot distinguish among products in an industry. Money is a product the quality (in terms of purchasing power) of which cannot be determined by direct inspection, e.g., the si7,c or color of a currency docs not inl1uellce its purchasing power. One mechanism which does allow differentiation among similar commodities is the establishment of a "trademark" or "brand name." We shall argue that the Suffolk system evolved into a mechanism through which the Suffolk Rank "franchised" a trademark to the participating banks. As the next section shows, this was hardly the conception of the Suffolk maintained either by its management or by the participating banks, at least in the early stages of the system.

IV. The Suffolk Bank SY!item: Version I versus Version" The Suffolk Rank was founded in Roston in 1818; it was the seventh hank to he chartered in that city. Roughly one year after opening, the management decided to enter the "foreignmoney" (country-bank note) business. The only other bank purchasing country hank notes at the time was the New England Rank. In Suffolk I, the bank purchased country-hank notes from merchants. individuals and other banks at a discount. It allowed the issuing bank to purchase these notes from the Suffolk at this same rate of discount, provided the issuing bank maintained a permanent deposit with the Suffolk of $S,O()() and agreed to make additional deposits to cover those notes redeemed at the Suffolk. If a hank chose not to participate in the arrangement, the Suffolk purchased its notes at a discount and presented them promptly to the issuing bank for redemption at par. The source of the profitability of the Suffolk I scheme apparently stemmed exclusively from its ability to invest the permanent deposits supplied by the patticipating banks (on which the Suffolk paid no interest). The discount was passed on to the participating banks, although it is unclear how the cost of returning redeemed notes from the Suffolk to these banks were apportioned. Apparently, the Suffolk absorbed these costs·. Whitney notes that competition with the New England was "lively" and that the typical discount declined from one percent to less than one-half percent soon after the implementation of the program. At the end of 1820, the Suffolk halted all purchases of the notes of nonparticipating banks since this aspect of the business had become unprofitable. lIe also reports that by 1822, "the profits of the (foreignmoney) business had hecome so small that the directors felt obliged to reduce the expenses of the bank" [22,8]. 4, Otherwise. there was no incentive to participate --the Sull'olk would recapture the discount through deliverycost rees.

299

Free Banking II COMrETITIVE MONIES AND TilE SUFFOLK IlANK SYSTEM T~blt

I. Circulalion uf NUlr.!S of Ma""achu,Clh Hanks (Thousands)

Dale

(I) Ih)Slon Banks

IH2.\ IX24 I Xl:; IH2h

$1,.\54 1,7')7 1,91H 2,206 2,103 2,067 2,07H 2,171 3,464 .,,060 2,K24

I~n

IH2H 1~2Y

IH30 IX.!I IH.l2 IH.lJ

(2) Non-Bo,wn Ballks

$1,775 2,046 2,17.\ 2,344 2,K33 2,KIK 2,670 2,'153 3.675 4.06.1 5,065

Ralio of NOII-Boslllll 10 BmwlI Banks

1..11 1.14 1.1.\ 1.01>

US 1..16 1.2H

1.36 1.06 1..1.1 1.7'J

Source: I.ake [iJ,IMX).

The management of the Suffolk Bank was unsatisfied not only by the lack of profits in the "foreign money" business, but also by a sense that country bank notes claimed an inordinate share of the circulation in the Boston market. Two of the directors, in a letter to the other Boston banks, noted that "it might be expedient for them in common (the Boston banks) to adopt, with a view of checking the enormous issues (~f' coull/ry, and especially Eastern (Maine), paper and of securing to the bills of Boston banks ajllst proportion of the circulation ..." ([8, II], emphasis added). j The Suffolk management attributed "the prodigious credit enjoyed by the country banks" not to "any superior confidence in the stability of these institutions", but rather to the discount on their notes which was "founded on the very dilliculty and uncertainty of means of enforcing this payment." The Suffolk's proposal to its sister banks was to cut the size of the discount and to generate a fund, "to be assessed in proportion to their respective capitals, ... for the purpose of sending home the bills of the banks ..." such that " ... the profit or loss shall be in common ...." The proposal (Suffolk II) was approved, a fund of $300,000 was established (payable in the notes of the so-called Associated Banks), the Suffolk was appointed as the agent bank for the association; and the discount was reduced to 1/4 of one percent. Within a year, the discount was eliminated and the notes of country banks "in good standing" were accepted at par at the Sufli.)lk. As noted in the previous section, the management's rationale for establishing. such a scheme was flawed. The note circulation of country banks was not "inordinately" large, but rather rellected the equilibrium decisions of households and businesses given prevailing payment habits and exchange rates." And a reduetion in the size of the discount on country notes should have been expected to increase their circulation, since Boston prices in terms of country notes accordingly would be reduced. 7 I ndeed, when they traded continuously at 5. Thus, the alfinity uf many bankin~ historians for a Gresham's l.aw inll!rpn:latiun may havt: been kindled by Ihe lIIallagcnll!lIt'!i view that the Un:sham prim.:iple was indeed at work in Hoston in HS25. O. The Suiti.>lk's management apparently failed to appreciate the releVath':C: of the silabk dt'po}'il balances of the Hustoll banks, buth absolutely and relative tu the coulilry bauks. Over the period 1M 12··IM44, fllr cXiJlllple, depu~ib of Uu~tOIl b"nks aVl.!faged $5.2 Inilliun, while Massal.:huSodts country bank deposits avclaged $1.7 milliull. A!; a r~llil) hI Ci1pit~J, Iotai ballk liabiltth':S (nutes plus lkpo~its) ill liOslOn avt:ragcd 62.6 pt:n:clll OWl this period, whik Ihl.: Si.lIlle filHo fur cuuntry banb averaged 64.3 pacellt. rhus, cuuntry banks wac han.lly much kss slahl!.: lhall thcir (ilY cuunJcrpans, by this lIu.:asuclo: at least. 7. iJuslOll bank mallagers were not alonc in misjUdging 1/11: likely elli.:cls uf this sysh:lll. Cuuntry bankers vigorously up posed the system, often in colorful language, siw.:e they, lUo, believed their circulatioll would be curlaikJ.

300

Free Banking II 890

Donald.!. Mullineaux

par. country bank notes were perfect substitutes for Boston bank notes. As Table I slHlws. a relative increase in country bank circulation was apparently what resulted. at I('a~t after IR2o. While the ratio of Boston to country bilnk notes did increase in the initial ~tilge of Suffolk II, Lake [IJI argues this probilbly reflects the fact that the Boston banks paid interest on some bank notes during the years I R25'-27 in yet another attempt to hoost their local circulation. Whatever the rationille for the ilrrangement, the new system WilS considemhly more successful in other respects than its predecessor. The volume of country hank notes received at the Suffolk rose from $300,000 a month in June, I R24 to over $2 million a month hy the end of 1825. Steady growth in note redemption continued over the life of the Sufl'olk svstem; by 1858, its monthly redemptions exceeded $30 million per month. Whitney [221 reports that the husiness was "very remunerative"; from 1825 until the end of the SuflC)lk systelll (1858), the Suffolk Bank was invariahly the highest priced hank stock in Boston. Whitney reports average annual dividends of a little over 8 percent from 1847-IR46 and 10 percent from IR34-1858, versus 5.7 percent from IR19-182J, the period ofSufl'olk I. The mechanics of the Sufl'olk II system estahlished in I R24 were similar in part to Suffolk I." A permanent, interest free. deposit was required of participating hanks (although the size of the deposit was made to vary with the capital of the bank. with $2,O()O the minimum size deposit) and additional balances again were to be deposited to cover note redemptions. The Suffolk redeemed the notes of participating hanks at par (after IH25) and dehited their deposit balances accordingly. There were two significant innovations to the Suffolk 11 scheme, however. First, the Suffolk permitted a bank to borrow temporarily (via an overdraft) against its permanent halance, reserving the right to send home that bank's hills for specic redemption should its borrowings exceed the permanent deposit. In addition, the Suffolk accepted as deposits (or loan repayments) the notes of any participating hank judged in good slanding, i.e., banks were not required to deposit specie to their clearing accounts to replenish temporary balances. The latter provision of the new contract was especially important si lice it allowed thc Suffolk to offset thc notes of one participating bank against those of anothcr on its halancc sheet, i.e., to act as a clearinghouse for bank notes. In effect, the Sufl'olk 11 arrangcment involved "collective net settlement" of bank notes, whereas Suffolk I involved "bi-Iateral nct settlement"! In a collective settlement arrangement, a central agent (the Suffolk) maintains bookkeeping information on the net credits and debits of all the participants and effects the necessary transfers of funds as debits and credits on its balance sheet. In a bi-Iateral scttlement, each pair of banks illdividually calculates its net credit or debit position and the necessary transfer is eflected either in specie or as a credit or debit on their respective books. The Sufl'olk I arrangement forced each New England bank to form a correspondcnt relationship with the Suffolk or settle net debit balances in specie. But this system did not allow participants to present the notes of other country banks as deposits and consequcntly did 8. The joint venture aspect of the Suffolk II system wa' ,hort-lived, cnding in 1826. Apparently. the A"ociate Banks believed their deposit requirements were too large for the services rendered and profits obtained. (Director's Records of the Suffolk Bank. March 14, 1826). The Suffolk reduced the permanent depo,it requirements of the Ilmton hanks. all of which continued to participate in the system except the Massachusetts Hank, in a series of steps from $30,000 or more in 1825 to S5.000 by 18,15. 9. (j;uhade and Silocr [6J discuss the critical differences hctween these two kinds of clearing arrangemcnts.

301

Free Banking II COMPETITIVE MONIES AND TilE SUFFOLK BANK SYSTEM

891

not result in .IIlY netting out of claims on the books of the Sulrolk. This kind of bi-lateral scttil:mcnt system economizes som.:what on the shipment of speci.:, but not as Illuch as a collective system. lndced, the enhanced prolitability of the post··1824 SulliJlk mechanism proh'lhly rellccts the reduced costs of sp.:cic shipm.:nts und.:r this syst':llI. Walk.:r cOlllmcnts on th.: "popular fallacy" that the Sulfolk system increastd the amount of th.: Sulrolk's sp.:ci.:, noting that the notion was "a fiction."J() A well-run collective net settlement system economizes on specie, which n(l doubt accounts for the grudging acceptance of the system by the New England banks and the rapid increase in the volume of note clearings following the implementation of the second Suffolk system. The efrectiveness of the Sulfolk's note clearing and redemption mechanism also depended on the ability of the Suffolk to influence and control the risk-taking activities of the participating banks. The elimination of the discount on bank notes meant that "the market" was no longer capable of signaling a change in the riskiness of a particular bank through a change in the specie price of its notes. This increased the incentives for each bank to overissue, i.e., depreciate the value of its notes. A systematic policy of overissue by Bank XYZ would be promptly revealed to the Suffolk, however, in the form of adverse clearings against XYZ. The Sulfolk could respond by presenting XYZ's notes for specie once its balance dropped below its required permanent balance (reserves), forcing a contraction in note issue. In the extreme, the Suffolk could suspend or expel a bank from the system which it judged as excessively risk. Dewey notes that a bank "ceased to be in good standing when thrown out of the Suffolk" [2,91]. The Suffolk also engaged in "jawboning" or "moral suasion" with its member banks. Hammond [8,553] reports, for example, that the Sulfolk "scolded its country correspondents like bad boys and admonished them as if their souls and not merely its earnings were at stake." It often commented to the members concerning "pressures on our money market", sometimes backing its advice with policy changes. Whitney makes note of a circular distributed to participating banks "informing them that .1Il account of tht: scarcity of money and in order to have some control over its own funds, overdrafts must be limitt:d to $10,000" [22,24-25]. Tht: Suffolk was even sympathetic on occasion, telling one bank: "We regret the necessity of these measures (elimination of overdrafts), but the deranged state of mon.:y matters makes them unavoidable". In essence, the Sullolk employed many of the risk control techniques (reserve requirements, discretionary lending, and moral suasion) utilized by other regulatory organizations such as clearinghouses in the latter 19th century [7] and by modern central banks. II The basic similarities of these mechanisms for monitoring and IU. Walk« noleS Ihal: .. rhe whole circulalion of New England banks oul of Boslon in OClob«. IH56, was SJ4.J,5(J2,J~b.

This, then. is the amount the Sulfolk is to cl!deem at sigill, as fast as il makc;:s its appli!arance in the city of what is the propurtion of its specie: to this sum of $39 millions'! Why it's less than O1i~ ('('''' ull/he dol/ur!" (ilalie> ongillal) [I ~.161. J l. ·1 hl!~e dc\'il.:c~ afC aililed at "\.fuality assurancli!" in the sense of pccdictablt: purchasing power. The Sulrulk I1lcl:hanislll alsu addressed the problem of countt:rfcit notes in an interesting and apparently successful fashion, Till! Sutlolk CllllLCacted out this aspect of 4uality assurance to the uUicer in charge of the foreign money dc:partment, ptJying him a lixcd sum ($4,25U in 1~26) to conduct tJlI aspects orlhe nott-exchange business (including hiring his own statt). lie was rcyuin.:d tu inspect I!a~h note's authenticity and to persunally assume THE ORIGINS OF TilE I'EI>

647

banks evaded these legal restrictions on note issue and profitably provided the desired liabilities. The currency-deposit problem was not a demand for base money in exchange for notes, but a demand for notes in exchange for deposits 143, II J. The extra time and trouble involved in the issue of currency substitutes (instead of conventional bank notes) were a major cost of the era's restrictive banking laws. The risk of legal penalties was also a consideration; shortages and runs had to get fairly severe before banks became convinced that the authorities would look the other way. Had blmks been able to issue notes backed by their general assets, they could have easily and quickly made the switch from deposits to notes simply by printing new notes. Such new notes could also serve to meet the demands from the banks where confidence was lost, thus preventing a ripple effect on outside money due to those runs. Since both notes and deposits would be backed by the same assets, no new assets need be purchased to meet the switch in liability preference, and the currency runs could have been avoided.'· Actual experience with the issue of currency substitutes gives important testimony to these conclusions about the viability of note issue deregulation. To the extent that concern over the elasticity of the money supply was an important factor in the debate over the creation of the Federal Reserve System, this last point comes into play. Friedman and Schwartz [9,190] argue that elastic meant "subject to substantial change in quantity over short periods for reasons other than immediate profit to the issuer." Phrasing the question this way is misleading because there should not be any reason why elasticity and bank profits are contradictory. Given the legal restrictions of the National Banking System, the two were incompatible, but if the national banks had been able to issue currency without the bond collateral requirements, then their own profit maximizing behavior would have led to an elastic currency." Friedman and Schwartz's [9,190] three conditions for elasticity," and implicitly the rationale for the Fed, overlook a fourth possibility-removing the bond collateral requirements.'" Recent literature on the origins of the Fed, for example Miron 118], emphasizes the seasonality of interest rates under the National Banking System. The justification for the Fed was the need to smooth interest rates. However, it can be argued that legal restrictions were one likely cause of such seasonality. Because the bond collateral requirements led to currency shortages, and because the shortages led to reserve drains, bank balance sheets were subject to heavy fluctuation. The harvest season drain on currency and reserves may well have led to the interest rate movements that the current literature is concerned with. In the absence of legal restrictions, shifts in the currency-deposit ratio would have posed no problem, and the resulting drains on reserves and interest rate seasonality would not have occurred. Further deregulation of note issue could have been a possible alternative to the Federal Reserve System as a way to smooth out seasonal interest rates. 20 t6. Sclgin (341 also notes how the currency·deposit ratio can play havoc with the total supply of money under central banking. Since currency (in the U. S.) is also ballk reserves, increases in the public'S currency-depo.it ratio mean a loss in bank reserves and a mulliplicd cOlllraclion in thc 10lal supply uf muncy. Such fluctuations would nol occur with unregulated banks, sinc.:c currency would be inside, and nul outside, mUllcy. 17. Kleill (161 and L. White 1471 oller preliminary sketches of the prulit maximizing comlitions tilr compelilive note issue. Selgio 1351 provides a more complele picture of lhu~e cundiliulls. i8. These conditions are: I) some body or budies to supervise ur cuntrol the crcalion and rctircl1lc.!nt uf fcdc.!ral Reserve money and (0 haudle the mechanical details; 2) some mcans fur crca,ing and retiring Federal Kcscrvl! moncy; 3) some crileria to replace profit in determining the amounllu be ncated or retired. 19. Admittedly, friedman and Schwanz wrote lhose liul!s UVl!r 25 years agu when cenlrali7.ed conlrol over Ihl.!' money supply appeared to be lhe only way 10 solve apparently endemic claslicily problcms. However, rc,-=cnt advancc~ in the theory of cOlnpetitivc banking now Dlake the passage SCCIII sUlllcwhal dated. 20. or course, any cornplele explanalion or Ihe origins of Ihe Fed would have lu ,leal wilh a myriad of ulher historical factors in addition 10 cunsideratiuns of economic theory slrklly dclincd. II

Free Banking /I 648

SI('I'('/1 Horwi/Z

VII. Conclusion Empirical cases of unregulated note issue are difficult to find, since instances of such hehavior have hcen fcw ;111