Depression-Era BankFailures: The GreatContagion or the GreatShakeout?John R. WalterDeposit insurance was created, at least in part, to prevent unfoundedbank failures caused by contagion. The legislation that created theFederal Deposit Insurance Corporation (FDIC) was driven by thewidespread bank failures of the Great Depression. In the years immediatelybefore the 1934, when the FDIC began insuring bank deposits, over one-third of all extant banks failed. Many observers argue that these failuresoccurred because the banking industry is inherently fragile since it is subject tocontagion-induced runs. Fragility arises because banks gather a large portionof their funding through the issuance of liabilities that are redeemable ondemand at par, while investing in illiquid assets. Specifically, loans, whichon average account for 56 percent of bank assets, tend to be made based oninformation that is costly to convey to outsiders. As a result, if a significantsegment of bank customers run, that is, quickly require the repayment of theirdeposits, the bank is unlikely to be able to sell its assets except at a steepdiscount. Bank failure can result.But do Depression-era bank failures imply the need for government-provided deposit insurance, or is there another explanation of the failuresother than contagion and inherent fragility? Some observers question theview that banks are inherently fragile. They argue instead that the bankingindustry developed various market-based means of addressing runs such thatthe danger of failure was reduced. They also argue that the banks that failedThe author benefited greatly from comments from Tom Humphrey, Ned Prescott, John Wein-berg, and Alex Wolman. Able research assistance was provided by Fan Ding. The viewsexpressed herein are not necessarily those of the Federal Reserve Bank of Richmond or theFederal Reserve System.Federal Reserve Bank of Richmond Economic Quarterly Volume 91/1 Winter 2005 3940 Federal Reserve Bank of Richmond Economic Quarterlyin response to runs were weak and likely to fail regardless of runs (Calomirisand Mason 1997; Benston and Kaufman 1995).If not fragility, what might explain the widespread failures before 1934?One possible explanation is that the banking industry was experiencing ashakeout, not unusual in industries that have previously enjoyed significantgrowth. The number of banks had grown briskly from the mid-1880s until1921. Beginning in 1921, bank failures increased significantly, such that thenumber of banks began a precipitous decline that continued until 1934. Thereare reasons to think that the industry had become overbuilt and that macro-economic shocks, in conjunction with overbuilding, produced a retrenchmentin the industry that lasted for the next 12 years. Indeed, many authors point tothe relationship between bank failures and weakening economic conditions1.This article suggests that overbuilding could have made the banking industryall the more sensitive to macroeconomic shocks.A number of other industries provide examples of growth followed byshakeouts, the most recent of which is the telecom industry. If a large portionof Depression-era banking failures were the result of a shakeout rather thancontagion, an important argument for deposit insurance is undercut.Though the termination of bank failures and the creation of the FDICin 1934 occurred simultaneously, implying that contagion must have been atwork, other explanations are just as credible. First, deposit insurance aug-mented the profits of risky banks, protecting them from failure. Second, thecreation of deposit insurance undercut a market process that caused supervi-sors to close troubled banks quickly.1. GROWTH IN THE NUMBER OF BANKSThe number of U.S. banks grew rapidly from 1887 until 1921 (Figure 1).Much of the increase coincided with improving economic conditions. Yet,commentators also claim that a good portion of the increase resulted from astatutory change that lowered the minimum capital required to form a new bankas well as careless application of entry standards by regulators. Many of thenew banks were viewed by commentators as being ill-prepared for the businessof banking. In other words, too many banks were formed without adequatefinancial or managerial resources. The banking market was overbanked.As shown in Figure 1, the number of banks began growing rapidly in thelate 1880s. The initial run-up in the number of banks followed an economicrecovery occurring in 1885 and 1886. The increase in the number of bankswas rapid enough, and the size of new banks small enough, to drive down theU.S. average bank size fairly significantly. The average size bank shrank from1Temin (1976, 83–95) discusses banking failures that resulted from macroeconomic weaknessduring the Depression.J. R. Walter: Depression-Era Bank Failures 41Figure 1 Number of Banks35,00030,00025,00020,00015,00010,0005,00001834 1854 1874 18941844 1864 1884 1904 1914 1924 1934195419441964$1.04 million in 1886 to a low of $660,000 in 1896 and did not return to its1886 level until 1916. The growth in number of banks was much faster thanthe pace of economic growth, so that the increase in the number of banks isstill quite apparent even when the number of banks is deflated by the level ofreal GDP (Figure 2).Most commentators focus on the increase in the number of banks, es-pecially of very small banks, after the beginning of the 20th century. Figure1 shows that the growth in the number of banks was indeed rapid from 1900until 1921. An important explanation for the growth in the number of banksduring these two decades was the reduction in the minimum capital required toform a bank (Mengle 1990; Wheelock 1993). Specifically, the Currency Actof 1900 lowered from $50,000 to $25,000 the minimum capital needed frominvestors to start a national bank. In turn, over the next ten years, two-thirdsof newly formed banks were quite small, averaging capital of only slightlymore than the minimum $25,000 (Mengle 1990, 6).Beyond this reduction in minimum capital, regulatory laxity was alsothought to have contributed to the rapid increase in the number of banks. Forexample, Federal Reserve analysts concluded that during the first two decadesof the 20th century “insufficient attention was paid to the qualifications of those42 Federal Reserve Bank of Richmond Economic QuarterlyFigure 2 Number of Banks Divided by Real GDP (2000 Dollars)0.0010.0020.0030.0040.0050.0060.001850 18901930 1970to whom charters were