Printer Friendly

The government's role in deposit insurance.

For a discussion of why these problems took so long to reveal themselves, see footnote 7.

3 This is a hypothetical laissez-faire benchmark. It is therefore not to be confused with the pre-FDIC regime, though it certainly has some similarities to that regime. The pre-FDIC regime still had various legal restrictions and interventionary agencies (for example, the Federal Reserve System) that would be absent under the laissez-faire benchmark.

4 I am aware of course that the monitoring of bank management must take place in a world where information is scarce and During the 1980s, banks and thrift institutions failed at a rate the United States has not experienced since the Great Depression. Deposits at most of these institutions were insured by the federal government, and covering the insurance liabilities has required hundreds of billions of dollars in taxpayer funds. The crisis in the banking and thrift industry has led to a reexamination of the federal deposit insurance system. This article is a collection of six essays on deposit insurance and the federal government's role in providing it. Each author is an academic or a Federal Reserve economist who has published research on deposit insurance and related topics. Steven Russell edited the collection.

The essays in the collection express a variety of different views on a broad range of important questions: Does the protection provided by deposit insurance encourage financial institutions to take excessive risks that cause them to fail? Should the federal deposit insurance system simply be abolished? Would abolishing the insurance system bring about a return of the problems of financial instability that existed before the system was established? Can we reform federal deposit insurance in a way that makes the financial system stable and competitive without encouraging risk taking and imposing large costs on taxpayers? If so, how should we go about it?

THE DIFFICULTIES OF BANKS and thrift institutions during the last decade have created a great deal of interest in U.S. banking system reform. Among the options that have been considered is restructuring the federal deposit insurance system or eliminating it entirely. The Federal Reserve Bank of St. Louis recently invited six economists who have conducted research on banking and financial regulation to write short articles on deposit insurance and the federal government's role in providing it. These six articles are collected in the following pages.(1)

Each article in this collection addresses one or both of the two basic questions that confront anyone who might contemplate reforming the deposit insurance system. The first question involves the problem of liquidity crises, or financial panics, that troubled the U.S. banking system during the 150 years before the establishment of federal deposit insurance. There have been no liquidity crises since deposit insurance was established, and most economists believe that this is not coincidental--that deposit insurance has in fact prevented liquidity crises. Any proposal for deposit insurance reform that involves limiting the coverage of insurance or eliminating it entirely must address the problem of financial panics. The second question involves the so-called moral-hazard problem of deposit insurance--the fact that it provides insured banks incentives to take excessive risks. Most economists believe that moral-hazard problems played a major role in causing the wave of bank failures that occurred during the 1980s. Proposals for deposit insurance reform that involve retaining an insurance system of approximately the current scope must find some way of solving the moral-hazard problem.

The first article in this collection describes the history of the state deposit insurance systems that preceded the federal insurance system and argues that these systems were also afflicted by moral-hazard problems. The second article argues that the problem of liquidity crises has been overblown, that an unregulated banking system would be stable, and that deposit insurance is not needed. The third article argues that though liquidity problems may have existed in the past, recent innovations in the financial system would enable banks to prevent them without relying on deposit insurance. The fourth article summarizes the theoretical basis for the claim that we need deposit insurance to solve the liquidity crisis problem and challenges the argument that adequate market solutions to this problem are now available. The fifth article presents a theoretical analysis of the prospects of solving the moral-hazard problems of deposit insurance by means of a system of risk-based insurance premiums. The sixth and final article outlines the risk-based premium system recently adopted by the Federal Deposit Insurance Corporation and questions whether it will succeed in solving the problem of bank failures.

The current debate over the role of government in deposit insurance can be adequately understood only in the context of the U.S. historical experience with monetary and financial institutions. A key feature of this experience has been a sequence of largely unsuccessful attempts to reform the financial system to solve the problems created by bank runs, bank failures and financial panics. This process, which culminated in the establishment of the federal deposit insurance system in 1933, is surveyed in the shaded insert on p. 6.

For its first 50 years of existence, federal deposit insurance seemed to succeed both in preventing financial panics and in sharply reducing the number of bank failures. The losses associated with the failures of banks and thrift institutions were easily covered by their respective insurance funds. After 1980 the failure rates of banks and particularly thrift institutions skyrocketed. The federal savings and loan insurance fund was overwhelmed, and hundreds of billions of dollars in savings and loan losses had to be converted out of general federal revenues, which is to say by federal taxpayers. The policy problem we now face is to reform our financial system to prevent a repeat of the hugely expensive bank failure problems of the period after 1980 without recreating the problem of instability and panics that existed before 1933.

State Deposit Insurance Systems

David Wheelock, an economist in the Research Department of the Federal Reserve Bank of St. Louis, wrote the first article in this collection. Wheelock has written extensively about the history of state deposit insurance systems. He begins his article by pointing out that in 1933, government deposit insurance was neither a new concept nor an unprecedented policy. During the pre-Civil War era, six states established systems to insure state bank notes; during the early twentieth century, eight states established systems to insure state bank deposits. The systems operated under a variety of different regulatory environments and financial arrangements. According to Wheelock, these differences may help explain differences in the systems' performances. For example, mutual-guarantee insurance systems, in which each insured bank could be assessed any amount necessary to cover depositors' claims against insured banks that failed, had better records than conventional systems, in which banks paid premiums that were used to create insurance funds to cover depositors' claims. Wheelock argues that mutual guarantee systems were more successful because they gave insured banks a stronger interest in monitoring the soundness of other insured banks.

Because membership in each of the state insurance systems was effectively voluntary, they were exposed to the problem of adverse selection. Risk-prone banks were more likely to join than conservatively managed banks, and well-managed banks tended to leave insurance systems at the first sign of trouble. Wheelock reports that because a bank's insurance premium was not linked to its degree of failure risk, insured banks were encouraged to increase the riskiness of their loan portfolios. (This moral-hazard problem is a recurring theme in deposit insurance literature and is discussed in each article in this collection.) This was particularly true for insured banks that found themselves in financial distress.

Wheelock concludes by observing that the historical record of state deposit insurance systems has generally not been considered successful. He suggests two options the government might select if it chooses to retain deposit insurance: a mutual guarantee system modeled after the successful systems of the early 19th century or a system of limited insurance combined with continued regulatory restraints on bank risk taking.

The Free Banking Option

Historically, monetary and financial questions have occupied a special place in economics. Economists skeptical of most types of government involvement in economic activity have often been willing to make important exceptions with regard to the monetary and financial sectors. For example, Milton Friedman, a throughgoing free-marketer on most issues, has endorsed both a government monopoly over currency provision and tight government regulation of creation of deposits.(2) He has also written approvingly of federal deposit insurance.(3)

In recent years a small but growing group of economists has argued that monetary and financial institutions are not an exception to the principle of the superiority of laissez-faire. These so-called free bankers believe that banks should be allowed to operate in a truly competitive market environment--free from government regulations, such as restrictions on the nature or quantity of their assets and liabilities (including monetary liabilities), and also free from government protections, such as legal restrictions on entry and competition, Federal Reserve System last-resort lending, and federal deposit insurance.

Kevin Dowd, a reader in monetary economics at the University of Nottingham (United Kingdom) wrote the second article in this collection. Dowd is one of the leading advocates of a free banking system. He argues that government insurance, far from protecting banks, weakens them and makes them more likely to fail. Uninsured banks, he asserts, would have incentives to acquire safe assets, obtain adequate capital from investors and provide proof of their soundness to depositors. Competition would ensure that bankers struck the right balance between depositor protection and return to investors. According to Dowd, the historical record indicates that banks in relatively unregulated banking systems maintained strong capital positions and retained depositor confidence. He argues that historians have greatly overestimated the severity of the problem of runs and panics. Bank runs were usually constructive events that weeded out weak banks.

Dowd argues that deposit insurance weakens a banking system by freeing banks from depositor scrutiny, which gives them incentives to weaken their capital positions and make riskier loans. Banks that find themselves in financial distress have incentives to take even greater risks in an attempt to recover, and bank runs no longer put a stop to this process by forcing them to close. Based on their private incentives, regulators act slowly to close insolvent banks, and the resulting losses must be covered by the insurance corporation. Eventually the insurance corporation also becomes insolvent and must seek a financial bailout from taxpayers.(4)

Dowd's concluding recommendation that the federal government eliminate deposit insurance no longer seems as radical as it might have a few years ago. It must be noted, however, that his reading of the historical record regarding bank runs and financial panics is far more optimistic than that of most other economists.(5) Historically, the public seems to have believed that runs and panics constitute a serious problem whose solution requires government intervention. This belief has provided a powerful stimulus for banking reform. Relatively few economists would feel comfortable asserting that it has been entirely misguided.(6)

Market-Based Alternatives to Deposit Insurance

J. Huston McCulloch, a professor of economics at Ohio State University, wrote the third paper. Professor McCulloch has published on the role of banks as financial intermediaries. McCulloch, like Dowd, advocates the elimination of government deposit insurance. Unlike Dowd, however, he is willing to concede that banks and thrift institutions may once have had two special problems that necessitated government intervention: mismatching of the terms of their assets and liabilities and vulnerability to liquidity crises (runs). He argues that financial markets have now developed private solutions to these problems, so government solutions are no longer needed.

Most of McCulloch's article is devoted to a discussion of the problem of liquidity crises. The solution to this problem, he asserts, is the money market mutual fund (MMMF). Because the value of an MMMF's liabilities is tied directly to the value of its assets, a change in the value of the assets does not give depositors an incentive to run. If depositors run anyway, the assets, which are very liquid, can simply be sold. McCulloch notes that MMMFs have already survived runs--large, rapid declines in the total amount invested--that would have been disastrous for banks.

McCulloch argues that the risk of large changes in the value of MMMF assets is too small to discourage consumers from investing and that the risk is also small enough to allow consumers to write checks drawn on their fund balances. He concedes, however, that in a completely competitive market, traditional banks offering conventional checking accounts might coexist with checkable MMMFs. These traditional banks, he argues, should not be insured by the government.

In McCulloch's view, deposit insurance was possible only in an environment of restricted competition between banks. Consumers paid a high but indirect price for these restrictions, which permitted banks to pay artificially low interest rates to depositors. The restrictions also made bank charters very valuable, a fact that prevented bank managers from taking risks that might cause their banks to fail and charters to be forfeited. The financial deregulation of the early 1980s revived interbank competition and reduced the value of bank charters. This inevitably led to increased risk taking, huge losses and insurance fund insolvency.

McCulloch concludes by noting that the most convincing theoretical case for government provision of deposit insurance is based on a formal model developed by Diamond and Dybvig (1983). In the Diamond-Dybvig model, banks provide important risk-sharing services to depositors but can do so effectively only if the government provides deposit insurance. McCulloch contends, however, that uninsured financial institutions could provide equally effective risk-sharing services.

The Case for Retaining Deposit Insurance

Phillip Dybvig, a professor of finance at Washington University in St. Louis, wrote the fourth article. Professor Dybvig is coauthor of the Diamond-Dybvig article, a seminal work on bank runs that provided theoretical support for government deposit insurance. He opens his article by observing that the optimal scope of government regulation is one of the most difficult questions confronting economists. Deposit insurance, he comments, may be an exception to the rule that government intervention rarely improves the outcomes produced by competitive markets.

Dybvig's defense of deposit insurance is based on the Diamond-Dybvig article, which he says made three basic points. The first two points are that banks perform a key role in creating liquidity and that banks' efforts to create liquidity expose them to runs. The third point is that bank runs can be prevented in any one of the following three ways: by laws permitting banks to suspend convertibility of deposits into currency, by government deposit insurance, or by a government lender of last resort. Because suspension is potentially very costly to depositors and last-resort lenders typically suffer from credibility problems, deposit insurance seems like the natural solution. In practice, Dybvig notes, deposit insurance seemed quite successful before the 1980s.

Dybvig concedes that deposit insurance systems tend to be vulnerable to the problem of moral hazard--insured banks taking excessive risks. Managing this problem by government supervision and regulation is essential to the success of any insurance system. Dybvig's reading of the historical record suggests that it may be possible for government regulatory agencies to do this job effectively--though he admits that the jury is still out on this question.

Dybvig concludes his article by commenting on three policy issues. First, he argues that the recent reduction in the maximum coverage of deposit insurance will not encourage depositors to monitor their banks more carefully. Second he asserts that 100-percent-reserves banking (the type proposed by McCulloch) can be a viable alternative to deposit insurance only if the economy has surplus liquidity and liquidity creation by banks is no longer necessary for efficient functioning of the economy. This, Dybvig writes, seems doubtful. Finally, Dybvig notes that the government's need to control the money supply is another possible reason we might need to retain the current banking system and thus federal deposit insurance.

Resolving Moral Hazard through Risk-Based Deposit Insurance Premiums

Anjan Thakor, a professor of finance at Indiana University who has written on the fair pricing of deposit insurance, contributed the fifth article. Professor Thakor begins by identifying two basic problems confronting deposit insurance systems: private information and moral hazard. The private information problem is that a bank's managers are better informed than its regulators about the risk characteristics of the bank's loans--an informational advantage that may allow them to frustrate regulators' attempts to price deposit insurance efficiently. Insurers, Thakor writes, may attempt to respond to this problem directly by auditing the banks to try to increase their information or indirectly by trying to construct an insurance pricing scheme that is incentive compatible. An incentive-compatible scheme presents a bank with a menu of different insurance contracts that is constructed so that the bank's choice of a particular contract from the menu reveals its private information.

Chan, Greenbaum and Thakor |CGT (1992)~ explore an insurance scheme that ties a bank's deposit insurance premium to the value of its equity capital. A bank with risky assets will not wish to maintain a high level of capital because the capital will be lost if the bank fails; it therefore will accept a high insurance premium. A bank with safer assets will be comfortable maintaining a higher level of capital but will desire a lower premium. CGT show that an insurance pricing system of this form can be incentive compatible. If each bank chooses the contract that maximizes its expected profits, its choice reveals the riskiness of its assets. Thakor notes, however, that such a system can work only if banks can earn economic profits from their activities--which means only if there are restrictions on interbank competition or if the government provides banks with subsidies. Economic profits, Thakor observes, can also help control the moral-hazard problem by giving banks an incentive to avoid excessive risk taking. An unfortunate implication is that the public's desire for a more competitive banking system may well be inconsistent with its desire to reform the deposit insurance system.

Thakor goes on to raise two other potential problems with deposit insurance systems: they may encourage government interference in other aspects of banking, and they may induce self-interested regulators to conceal the problems of financially distressed banks. He concludes by observing that the many problems with the current deposit insurance systems make him pessimistic about the prospects for its successful reform and goes on to present a brief discussion of more radical options for banking reform. These include a 100-percent-reserves banking system of the type discussed in the McCulloch and Dybvig articles and a system in which insured banks restricted to acquiring very safe assets would coexist with uninsured banks whose asset choices were not restricted.

The FDIC's System of Risk-Based Insurance Premiums

Mark Flood, an economist in the Research Department of the Federal Reserve Bank of St. Louis, wrote the last article. Mr. Flood has written on the history of deposit insurance and on the use of option pricing models to analyze deposit insurance. His contribution describes and evaluates the system of risk-based insurance premiums recently adopted by the Federal Deposit Insurance Corporation (FDIC). He begins by reviewing the moral-hazard problem and noting that risk-based insurance premiums are a potential solution to the problem. He goes on to describe the risk-based premium system that has been adopted by the FDIC. Under the new system, a bank's insurance premium is jointly determined by its level of capitalization and an evaluation of its financial health provided by bank regulators. The most important element of this evaluation is the bank's CAMEL rating--a five-point summary ranking of its overall soundness.

Flood identifies two potential problems with the FDIC's proposal. First, it may be possible for people to use a bank's risk premium and other publicly available information to infer its confidential CAMEL rating. This might lead to runs on banks with low CAMEL ratings.(7) Second, banks may try to use window-dressing accounting schemes or other cosmetic devices to deceive regulators about their financial health.

The most controversial aspect of Flood's article is his suggestion that we may have misidentified the cause of many of the bank failures. The moral-hazard explanation says that bank failure rates rose because competent bank managers responded to financial incentives to take increased risks. Flood proposes an alternative explanation: incompetent bank managers were unable to evaluate the risks they were taking. Financial regulation, he speculates, protected these incompetent managers from the rigors of the competitive marketplace. When regulation was rolled back in the early 1980s, they were unable to adapt and many of their banks failed.

Flood argues that we do not yet have enough evidence to determine which of these two problems--moral hazard or inferior management--was the principal cause of the banking troubles of the last decade. He concludes by noting that if inferior management caused many of the recent bank failures, risk-based insurance premiums may not solve the problem of failures and alternative regulatory responses may be needed.

The six articles in this collection present a wide range of views on the need for deposit insurance and the federal government's role in providing it. This diversity of opinions is an accurate reflection of the current state of the debate on these issues. Virtually every economist and policymaker agrees that the federal deposit insurance system as it existed in the 1980s required major reform. There is, however, no apparent consensus about whether the reforms that have already been implemented are adequate to solve the problems of the U.S. banking system, or if further changes are needed, what the nature of those changes should be.

1 Each of the six authors was a participant in "Aspects of Government Deposit Insurance: Opposing Views on the Role of Government," a symposium held December 11, 1992, at the Federal Reserve Bank of St. Louis.

2 See Friedman (1960), chapter 1. For a more recent statement of Friedman's views on the role of the government in monetary affairs, see Friedman (1986).

3 Friedman has written, for example, that deposit insurance "has been the most important structural change in our monetary system in the direction of greater stability since the post-Civil War tax on state bank notes" |see Friedman (1960), p. 21~ and that it is "a form of insurance that tends to reduce the contingency insured against" |see Friedman and Schwartz (1963), p. 440~.

4 Though most economists would probably agree that regulation of banks and thrift institutions (particularly the latter) has suffered from serious problems, many might disagree that government regulators are inherently incapable of monitoring and managing the problems of distressed banks. For an analysis that defends the record of bank regulators and challenges certain arguments of their critics, see Gilbert (1991 and 1992).

5 For an analytical survey of the record of U.S. banking panics between 1857 and 1933, see Dwyer and Gilbert (1989).

6 Free bankers, it should be noted, argue that the problems of the banking system have usually been caused by bad government regulation rather than by inadequate regulation.

7 Flood reports the results of his own attempt to identify banks with low CAMEL ratings, which seems to have been quite successful.

Historical Background

Before the Civil War, virtually all U.S. banks were chartered and regulated by state governments. The principal liabilities of these institutions were bank notes, which provided the economy with the hand-to-hand currency now provided by Federal Reserve notes. These notes were supposed to be convertible--redeemable in gold and silver coins, at par and on demand.

The antebellum state banking systems were afflicted by several problems, including relatively high failure rates and vulnerability to financial panics (periods when banks across the United States were confronted with runs by note holders). In most cases the banks responded to panic-induced runs by suspending convertibility, an unpopular action that reduced the acceptability of their notes and caused them to trade at discounts.(1) Financial panics were usually associated with a large number of bank failures; many banks that suspended payments proved unable to resume them and ultimately closed. In addition, panics were often followed by lengthy periods of economic depression.

The sequential link between financial panics, bank failures and economic depressions convinced many people that panics and failures caused depressions and produced political pressure for banking reform. In 1863 Congress passed the National Bank Act, which was intended to replace the state banking systems with a system of federally chartered banks. Supplementary legislation imposed a prohibitive tax on state bank notes, a move that was intended to force state banks to join the national banking system or close down. The state banks survived and prospered, however, by issuing demand deposits, which were not taxed. National banks also began to issue demand deposits, and checks drawn on these deposits soon became the dominant means of payment in the U.S. economy.

The dual banking system of the post-Civil War period--federally chartered and regulated national banks that issued both notes and deposits coexisting with state-chartered and state-regulated state banks that issued only deposits--also proved to be vulnerable to financial panics. Major panics occurred in 1873, 1884, 1890, 1893 and 1907.(2) These panics were often followed by many bank failures and prolonged periods of economic depression; the depressions following the panics of 1873 and 1893 were particularly long and severe. After the Civil War, panics came more frequently and seemed to cause more financial disruption.(3) The panic of 1907 seems to have been the last straw prompting the federal government to reform the U.S. banking system. The following year Congress established the National Monetary Commission to study reform options. The commission's report was presented in 1912 and led directly to the Federal Reserve Act of 1913, which established the Federal Reserve System. The new system created 12 federally administered Reserve Banks that were authorized to make last-resort loans to banks facing panic-induced runs.(4)

As in the aftermath of many other major U.S. banking reforms, after the Federal Reserve System was established, many people believed that the problem of banking instability had been definitively solved. The Great Depression of 1929-33 dispelled this belief in very dramatic fashion. Although the Depression was not precipitated by a short, sharp panic of the late-nineteenth century type, it was accompanied by a succession of banking crises during which many banks failed. The existence of a lender of last resort in the form of Reserve Banks did not prevent bank runs and bank failures. The banking crises culminated in the Bank Holiday of early March 1933, when newly inaugurated President Roosevelt closed all the nation's banks for a week in an effort to calm the panic atmosphere. As noted, many U.S. banks had failed before the holiday was declared. Many more did not open afterward, and others closed within a few months of the holiday. Overall, almost a third of the nation's banks failed during the Great Depression. Congress responded to this disaster by passing the Banking Act of 1933, which established the federal deposit insurance system.(5)

1 A $5 note issued by a suspended bank might, for example, trade in the open market for $4.50 in specie (a 10 percent discount).

2 See Sprague (1910). Sprague notes that the 1873 panic was not followed by many bank failures and that the panics of 1884 and 1890 were less severe than the others and did not involve suspensions.

3 Part of the problem was that during suspensions bank deposits were less readily negotiable than bank notes. See Friedman and Schwartz (1963), pp. 110, 161-63.

4 Only members of the System were eligible to receive these loans. Though national banks were required to join the System, state banks were not, and a great many state banks chose not to become members.

5 For a brief survey of U.S. monetary history up to the Civil War, see Russell (1991). For an exhaustive historical account covering the period from the Civil War to 1960, see Friedman and Schwartz (1963).

What Have We Learned about Deposit Insurance from the Historical Record?

THE INCREASE IN depository institution failures in the last dozen years and the resulting losses to the bank and thrift insurance funds have understandably generated interest in the costs and benefits of deposit insurance. Calomiris (1989a, p. 12) defines a successful deposit insurance system as "one that fully protects the payments system, without encouraging any excessive risk-taking," that is, risk taking beyond what would be optimal without insurance. The federal government's apparent willingness to guarantee deposit insurance fund liabilities reduces the probability of widespread banking panics that would threaten the payments system.(2) Providing fully credible insurance, however, may increase the likelihood of a significant deposit insurance bailout by giving depository institutions an incentive to take excessive risks. Until the 1980s, risk taking was discouraged by regulations that enhanced the charter values of depository institutions and limited competition for deposits.(3) Deregulation, however, has lowered the value of charters and provided the means for banks to increase risk.

Federal deposit insurance was enacted in 1933 as a response to the bank failures of the Great Depression. Deposit insurance was not, however, a new policy at that time. During the 19th and early 20th centuries, many states experimented with deposit insurance systems. The state systems were funded entirely by insured banks, and the states did not guarantee the liabilities of the insurance funds. Recently, researchers have been studying these systems to gain insights to the effects of deposit insurance in different regulatory environments. This article reviews the historical record and attempts to draw useful lessons for the current debate.


Six states operated insurance systems before the Civil War. The Vermont, Michigan and Indiana systems, and the New York system before 1842, insured both bank notes and deposits. In Ohio and Iowa, and in New York after 1842, only bank notes were insured. The performance of the different systems varied considerably. The Michigan system opened on the eve of the Panic of 1837 and subsequently closed without reimbursing any depositors or note holders of failed banks. The New York and Vermont systems were more successful because their insurance funds had time to accumulate assets before significant failures occurred. As a result, note holders and depositors of insured banks that failed in these states received at least some reimbursement for lost funds. In Indiana, Ohio and Iowa, no depositor of an insured bank lost any money. The success of deposit insurance in these states has been traced to the mutual-guarantee form of their insurance systems.

In mutual-guarantee systems, insured banks that were still solvent could be assessed any amount to cover the obligations of an insured bank that had failed.

With the exception of the Michigan system, the antebellum deposit insurance systems remained open until the Civil War, though not all of them still had active members. The tax imposed on state bank notes under the National Banking Act of 1863 caused many state banks to reincorporate as national banks. National banks were permitted to issue notes valued at up to 90 percent (later 100 percent) of the face value of the U.S. government bonds they deposited with the Comptroller of the Currency. The notes in turn were guaranteed by the federal government.(4)

During the last two decades of the 19th century, the expanded use of deposits (which were not taxed) and the liberal chartering requirements that many states adopted caused a resurgence of state-chartered banking. By the mid-1880s, Congress and several state legislatures began to consider proposals for deposit insurance. None of these was accepted until 1907, when a surge of bank failures led Oklahoma to establish a deposit insurance system for its state banks. Kansas, Nebraska, South Dakota and Texas followed within two years. Mississippi enacted insurance in 1914, as did North Dakota and Washington in 1917.(5)


The absence of significant bank and savings and loan failures between 1934, when federal deposit insurance began, and 1980 suggests that regulations limiting competition for deposits and maintaining charter values effectively discouraged excessive risk taking. The performance of the 19th and early 20th century state insurance systems also shows that the effects of deposit insurance depend largely on the regulatory environment. For example, each of the states with deposit insurance in the 19th century permitted banks to avoid the insurance system by incorporating as "free banks." In some states, such as New York, weaknesses in the insurance system caused conservative banks to exit and adopt free bank charters. Indiana, on the other hand, had a notorious free banking law that tended to attract risky banks; conservative banks chose to belong to the insurance system.(6) This may be one reason why insured banks in New York had a higher failure rate than those in Indiana.

Each of the states that had a deposit insurance system in the early 20th century prohibited branch banking but set low minimum capital requirements and permitted relatively free entry. Although the states imposed various regulations to limit risk taking by insured banks, such as minimum capital/deposit ratios and deposit interest rate ceilings, supervision tended to be cursory. In each of these states, deposit insurance is generally believed to have encouraged excessive risk taking by banks. Whether this was due to inadequate regulation and supervision or to inherent flaws in the insurance systems is not clear. It seems likely that both the insurance systems and the regulatory environment were to blame.

Alternative Funding Methods

The 19th century insurance systems of New York, Vermont and Michigan, and all eight of the early 20th century state insurance systems, required insured banks to pay into a fund for reimbursing depositors of failed banks. The insurance premiums the banks paid were unrelated to risk of failure, and upper limits were set on the assessments that could be imposed in any year. In each state but one, the liabilities of the insurance system eventually exceeded its assets and depositors of failed banks had to absorb some of the losses.(7)

The mutual-guarantee feature of the 19th century deposit insurance systems in Indiana, Ohio and Iowa ensured that there were ample funds to reimburse depositors and note holders and discouraged the excessive risk taking that appears to have characterized banks in the other state insurance systems. In mutual-guarantee systems, insured banks could be assessed any amount necessary to reimburse insured depositors or note holders. Insured banks consequently had a strong interest in the behavior of other members of the system--an interest that the state harnessed by giving members considerable supervisory authority over one another. The relatively small number of insured banks operating in each of these states further enhanced regulatory control.(8)

Voluntary vs. Mandatory Insurance

If insurance premiums are inadequately related to risk, then risk-prone banks tend to gain more from deposit insurance than banks that are managed conservatively. In the absence of insurance, depositors will demand risk premiums on deposit interest rates and will withdraw their funds from banks that take unacceptable risks. Deposit insurance removes the incentive for depositors to monitor bank risk and may "create a sense of false security in the public mind and a lack of discrimination between reliable and unreliable banks and bankers."(9) Because risky banks gain the most in terms of increased public acceptance and reduced deposit costs, they will be more likely than conservative banks to join voluntary insurance systems. Historically, this adverse selection problem has hampered the funding of deposit insurance systems.

Because banks in the early 19th century could incorporate as free banks, bank insurance systems of this time were in essence voluntary. As noted above, the performance of each state's insurance system depended on its funding method and on the incentives provided to both insured and free banks. Insurance worked better in states like Indiana, where conservative banks were attracted to the insurance system.

Because a ruling by the Comptroller of the Currency prevented federally chartered banks from participating in state deposit insurance systems, all of the state systems of the early 20th century were also essentially voluntary. Even where insurance was mandatory for state-chartered banks, a bank could opt out by switching to a federal charter. Doing this was costly, however, because national banks were subject to different regulations, including generally more restrictive limits on their lending than were imposed on state banks.

Deposit insurance was optional for state-chartered banks in Kansas, Texas and Washington. Though all eight of the early 20th century deposit insurance systems ultimately collapsed, their survival does not seem to have depended on whether insurance was mandatory. Freedom to exit did cause the Washington system to have the shortest life. When the state's largest insured bank failed in 1921, all other insured banks withdrew from the insurance system, thus ending bank deposit insurance in Washington.

Kansas also permitted insured banks to withdraw from its insurance system, though a withdrawing bank was held liable for funds needed to reimburse depositors of institutions that failed within six months of the bank announcing its intention to drop out. Despite a large number of failures and increasing insurance premiums, banks did not leave the Kansas system en masse until 1926, when the state supreme court ruled that a bank could withdraw simply by forfeiting the bonds it had deposited with the state as a guarantee of insurance premium payments. Most insured banks then decided to withdraw, and state deposit insurance in Kansas effectively ended.

Texas banks were given the option of joining the state deposit insurance system or purchasing a private bond to guarantee their deposits. Before 1920 most banks chose to join the state insurance system. Like other commodity-producing states, Texas suffered many bank failures for several years after commodity prices collapsed in mid-1920. In 1925 the state permitted banks to drop out of the insurance system. Membership then fell off dramatically, from 896 banks holding $302 million of deposits in 1924 to 34 banks with just $3 million of deposits by the end of 1926.(10)

The histories of the mandatory deposit insurance systems are not qualitatively different from the history of the Texas system. In each case, insurance fund liabilities eventually exceeded assets and the state legislature simply repealed the insurance law instead of raising insurance premiums to cover the shortfall. Only Mississippi required taxpayers to bail out insured depositors.


Empirical investigation of the effects of insurance on the behavior of banks today is hampered by the fact that virtually all U.S. bank deposits are insured by the Federal Deposit Insurance Corporation. Comparing the behavior of insured and uninsured banks in the states that had optional insurance systems during the early 20th century is possible, however. In a study of Kansas banks, Wheelock (1992) found that members of the state's deposit insurance system had a greater likelihood of failure than their uninsured competitors and that insurance had its greatest effect on banks that were near failure. Like many banks and thrifts in the 1980s, Kansas banks often took extreme risks as they neared insolvency. Wheelock (1992) found that for banks within one year of failure, insurance system membership was an especially good predictor of failure. Wheelock and Kumbhakar (1991) also show that risky banks were more likely to join the Kansas insurance system and that insurance led banks to reduce their capital/assets ratios over time.

The early history of federal insurance of deposits at thrift institutions provides a similar opportunity to examine the effects of deposit insurance. Although insurance was mandatory for federally chartered thrifts, it was optional for state-chartered institutions and many chose not to become insured. In a study of Chicago and Milwaukee thrifts during the 1930s, Grossman (1992) found that when institutions first acquired insurance, they were less prone to risk than uninsured thrifts. Insured thrifts increased their exposure to risk over time, however, and after being insured for five years were more risky than uninsured thrifts. Grossman attributes the delay in the emergence of excess risk to the examinations the institutions underwent before their deposits were insured.


By the mid-1920s, many observers viewed deposit insurance as "an experiment that failed."(11) Despite regulations intended to contain risk taking, the state systems appear to have suffered from adverse selection and moral-hazard problems and in most instances did not fully reimburse depositors of failed banks. Bank failure rates were high in states with deposit insurance systems, and insured state banks had higher failure rates than uninsured state and national banks in the same states.(12)

Deposit insurance also appears to have created greater losses for failed institutions than there might have been otherwise. Without insurance, depositors have an incentive to withdraw their funds once a bank becomes insolvent. Deposit insurance removes this incentive, making it possible for insolvent banks to continue to operate unless closed by regulators. As in the 1980s, regulators during the 1920s sometimes permitted insolvent banks to remain open, hoping that they would regain solvency. Forbearance seems to have been unsuccessful, however, because the average liquidation value of insured state banks that closed was less than that of uninsured state banks that failed.(13)

Although the historical record of deposit insurance is not favorable, it seems unlikely that deposit insurance will be eliminated, or even significantly scaled back, in the near future. Two non-mutually-exclusive options for reform seem available. A mutual-guarantee system like those of 19th century Indiana, Ohio and Iowa could be adopted. Mutual guarantee seems to have discouraged excessive risk taking and ensured ample funds to protect depositors from losses. To operate effectively, however, such a system might require a considerable consolidation of the U.S. banking industry. Any privately funded insurance system, moreover, could be vulnerable if depositors lose confidence in the entire banking system. For insurance to be fully credible, the Federal Reserve must thus be willing to act as lender of last resort--a role it failed to perform during the Great Depression.

An alternative option that would combine some limits on insurance coverage with regulatory constraints on risk taking seems more politically feasible. Recent moves to increase capital standards for insured banks and thrift institutions, and to bring about the early closure of troubled institutions, are steps in the right direction. Historical evidence, including the events of the 1980s, however, illustrates the difficulty of limiting excessive risk taking when there are thousands of institutions to supervise. If monitoring and supervision are left primarily to public officials, moreover, there is likely to be continued political pressure for forbearance.

The United States has paid a high price for forgetting the historical lessons of deposit insurance. When the federal deposit insurance system was set up in the early 1930s, its leading Congressional proponents understood many of these lessons, and implemented regulations that checked excessive risk taking.(14) The United States should not try to restore the post-Depression bank regulatory system. Repeal of New Deal restrictions on branch banking and the securities-related activities of banks would reduce risk through diversification and economies of scope. But if federal deposit insurance is to remain, policies that prevent excessive risk taking will be required.

1 Senior economist, Federal Reserve Bank of St. Louis. Kevin Dowd, Mark Flood and Steve Russell made helpful comments on a previous draft.

2 When the FDIC was established, there was no explicit statement that the federal government would bail out the insurance fund if it became insolvent. See Flood (1992).

3 See Keeley (1990).

4 National banks also had to contribute 5 percent of the value of their notes to a cash redemption fund maintained by the U.S. Treasury. |See Friedman and Schwartz (1963, pp. 20-21)~. The backing provisions for national bank notes were similar to those of most antebellum state free banking laws. See Dowd (1992) for analysis of free banking in this era.

5 White (1981) investigates the characteristics of states adopting deposit insurance and concludes that rural farming states were the most likely to adopt insurance and eschew branch banking.

6 See Calomiris (1989a).

7 The one exception was Texas, where insured deposits at failed banks were paid off in full. In Mississippi, the insurance fund was unable to pay off all depositors, but the state issued bonds to settle all claims eventually.

8 Calomiris (1989a) compares the two types of insurance systems in greater detail and notes that in many large cities bank clearinghouse members often jointly guaranteed the liabilities of each member during financial panics.

9 See American Bankers Association (1933, p. 39).

10 See Federal Deposit Insurance Corporation (1957, pp. 66-7).

11 See Harger (1926).

12 Alston, Grove and Wheelock (1992), after controlling for the extent of agricultural distress and other possible causes of bank failures, found that states with deposit insurance had systematically higher bank failure rates during the 1920s.

13 See Calomiris (1989b).

14 See Flood (1992). Kareken (1983) was prescient when he argued that deregulation without deposit insurance reform was like "putting the cart before the horse."

Deposit Insurance: A Skeptical View

Federal Deposit Insurance is a classic case of the wrong solution offered for the wrong problem. It seeks to protect banks against the runs to which they would be prone under laissez-faire, in which they would not have the protection deposit insurance gives them. I argue that this solution is based on a false premise: under laissez-faire, banks would not in fact be prone to runs and would therefore have no need for protection against them. The real problem is not how to protect banks against runs but how to maintain their financial strength. There are good reasons to believe that the market would provide banks with appropriate incentives to solve this problem without any need for government regulatory intervention. Deposit insurance is the wrong solution because it undermines market incentives and thereby weakens the banking system.

The paradox of deposit insurance is that trying to make banks more stable by protecting them against runs only weakens them and makes them more likely to fail. Deposit insurance transforms a perfectly healthy banking system into a chronic invalid that can be kept alive only by ever-increasing doses of public funds. It is the most effective means yet devised to destroy a nation's banking system. The U.S. experience suggests that it has done so at a staggering cost to the long-suffering federal taxpayer.(2)

The Stability of Laissez-Faire in Banking

Suppose we had a competitive banking system with no deposit insurance or lender of last resort.(3) Depositors in a bank would be aware that they would stand to lose their funds if their bank failed. They would therefore want reassurance that their funds were safe and would soon close their accounts if they felt any danger of losing their funds. Bank managers would be acutely aware of this possibility. They would understand that their long-term survival depended on their ability to retain the confidence of their depositors.

Bank managers might try to keep depositors' confidence in various ways. They might reassure depositors that they were not taking excessive risks with their funds by pursuing relatively conservative lending policies and exposing these policies to outside scrutiny. The underlying principle is that a bank that is "good" will want to signal its goodness to its customers and will also want to distance itself from "bad" banks the public wants to avoid. A good bank might, for example, hire an independent auditor from time to time to examine its books and issue a report on its financial soundness. The auditor's report would be credible because in the long run the auditor's ability to attract business would depend on the reliability of his reports. Similarly, a good bank would be able to encourage people of proved ability and integrity to sit on its board of directors; their presence on the board would in turn send a strong signal to the public that the bank was in safe hands. The same desire to send out credible signals would also encourage good banks to do things such as develop reliable and accurate accounting conventions to demonstrate their soundness and publish validated accounts of their financial health.

A bank's management would also reassure depositors that their funds were safe by maintaining adequate capital. One function of capital is to give bank shareholders an interest in the safe management of the bank. The shareholders of a well-capitalized bank have a lot to lose if the bank incurs losses. This potential loss gives shareholders an incentive to monitor the bank to ensure that its managers do not take excessive risks at their expense. This incentive is strong because the shareholders are residual claimants to the bank's assets and must therefore bear all the marginal losses the bank might take. The residual nature of their claim also means that the shareholders provide some protection to depositors. If the bank has sufficient equity capital, any losses it takes are borne entirely by the shareholders and the depositors lose nothing. Bank capital thus provides a buffer that absorbs losses and maintains the value of deposits. The bank will be unable to honor all its deposit liabilities only if its losses are so large that they exceed the value of its capital (that is, if its net worth becomes negative and it becomes insolvent).(4)

Under a laissez-faire system, a bank's capital strength--the amount of capital it maintains relative to its potential losses--is determined by market forces. The better capitalized the bank is, the more reassurance it provides depositors, and other things being equal, the more attractive it will be to them. But capital is also costly. Greater capitalization results in lower return on equity, which displeases existing investors and discourages potential new ones. There is consequently a tradeoff between reassuring depositors on one hand and discouraging shareholders on the other. The optimal capital position gives depositors the right amount of protection, given the cost of providing it.

There is also a presumption that the market will in fact produce the right amount of protection. If customers want safe banks, they will not patronize banks they consider weak, and these banks will attract no business. If a bank is too strongly capitalized, however, it will be able to attract capital only by passing on its higher capital costs to its customers and its services will be too expensive to be competitive. If bank customers want safe banks, as they presumably do, a competitive market will ensure that they get them. Indeed, banks will be exactly as safe as their customers demand.

Historical evidence supports the claim that banks have been strong and stable in the absence of deposit insurance. Recent research into historical free banking systems and the U.S. banking system before the introduction of federal deposit insurance indicates that banks typically maintained strong capital positions and were able to keep the confidence of the public despite the absence of deposit insurance or an official lender of last resort.(5) Banks that were not considered sufficiently sound would lose depositors, and competition for market share would force them to maintain the margins of safety and soundness their customers demanded.(6) The evidence from the pre-FDIC period also indicates that bank runs were not the problem that later generations perceived them to have been--that later generations exaggerated the problem--and that the runs that did occur were normally restricted to problem banks whose financial positions were perceived as weak anyway. A typical run was a flight to quality in which depositors would withdraw their funds from weak banks and redeposit them in stronger banks in which they had confidence. Runs were not contagious panics in which depositors withdrew their funds from any bank they could. In short, the evidence indicates clearly that strong banks did not need deposit insurance to protect them from runs.

The Destabilizing Effects of Deposit Insurance

Suppose that we introduce deposit insurance into our hypothetical system of laissez-faire. Under laissez-faire, banks were forced to maintain their capital strength because they needed capital to reassure depositors and discourage them from running. But once we introduce deposit insurance we take away depositors' incentive to run and relieve bank managers of the need to maintain capital to keep depositors' confidence. Deposit insurance thus reduces the marginal benefit of maintaining capital. Because deposit insurance has little effect on the cost of capital, banks with insurance therefore tend to reduce their capital/assets ratios. (In other words, given that a bank can reduce its capital/assets ratio without facing a run, its rational response is to do so to increase the return it can pay on shareholder equity.) Note also that a bank would be under pressure to reduce its capital/assets ratio even if it wanted to maintain it because other banks that took advantage of insurance protection to economize on capital would be able to outcompete it by offering higher deposit interest rates. The fight for market share would then pressure the responsible bank to follow suit, whether it wanted to or not. Deposit insurance thus makes a strong capital position a liability, putting well-capitalized banks at a competitive disadvantage. The banking system now has a weaker--possibly much weaker--capital position, which means that banks are less able to absorb losses while maintaining their net worth. Deposit insurance thus weakens the banks and makes them more liable to fail.

Deposit insurance also encourages banks to take more lending risks. If a bank adopts a riskier lending policy, it can expect to keep the higher returns it will earn in the event the risks pay off. If the risks do not pay off and the bank becomes insolvent, some of the loss is passed back to the insurance corporation. Deposit insurance thus gives banks some protection against downside risk. The amount of that protection, and hence the incentive to take excessive risks, increases as the bank's capital position worsens. In the end, a bank with zero or negative net worth might face no downside risk at all. It would have everything to gain and nothing more to lose from irresponsible, shoot-for-the-moon lending policies that are almost certain not to pay off. The losses, of course, are then passed back to the insurance corporation and to the other banks or taxpayers that are forced to pay into the insurance fund. To make matters worse, deposit insurance also removes the market mechanism--a run--that would otherwise have put a weak bank out of business and stopped its irresponsible gambling. A zombie institution can always get funds simply by raising its deposit rates and can keep gambling at other people's expense until the regulatory authorities finally get around to closing it.

A point eventually comes when the insurance corporation itself has accumulated so many bad debts that it too has become a zombie with no realistic hope of ever paying its debts. The deposit insurance crisis then escalates out of control. Because the insurance corporation no longer has the resources to close its problem institutions--that is, because it no longer has the funds to pay off their depositors if it closes institutions--it simply allows these institutions to continue operating and run up debts at what is now clearly the expense of the federal taxpayer.(7) The deposit insurance corporation now plays the same game with Congress that zombie insured institutions have been playing with the corporation. It seeks federal bailouts, ostensibly to put itself back on its feet, that just throw good money after bad and only postpone the day of reckoning. The insurance corporation seeks to hide its problems by watering down accounting and capital standards so that weak institutions can meet the regulatory requirements. It therefore replaces the relatively lax Generally Accepted Accounting Principles (GAAP) with the even laxer system of Regulatory Accounting Practices (RAP), which allows expected gains from future transactions, accounting forbearances and even (incredibly) unrecognized losses to count as capital for regulatory purposes.(8) As if that is not enough, the insurance corporation then exempts many institutions that fail to meet these requirements by allowing them to continue operating anyway. It seeks to justify itself by inventing elaborate theories of regulatory forbearance that are little more than smokescreens to cover its own failure to close problem institutions. If all else fails, the insurance corporation blames its difficulties on scapegoats like fraudulent or incompetent management in its problem institutions, oil price shocks, deregulation or just plain bad luck. Congress and the Administration go along with this game for political reasons, and nothing substantial is done to stop it. In the meantime, what might have been a relatively mundane public finance disaster of perhaps a few billion dollars is transformed into a catastrophe that will cost hundreds of billions of dollars to clean up (with the cost still rising).

Policy Implications

Conventional wisdom holds that deposit insurance is sound in principle but flawed in practice. Advocates of this view maintain that correcting deficiencies in the implementation of deposit insurance will resolve the problems insurance creates. I believe that this view is profoundly mistaken. It is the very principle of deposit insurance that is flawed, and no amount of patching will put the problem right. Deposit insurance is fundamentally incompatible with a safe and sound banking system because it tempts insured banks to play Russian roulette. The U.S. Congress therefore faces a simple but unpleasant choice. On one hand, Congress can make the economically sensible but politically difficult decision to come to grips with the problem at last by introducing a program to dismantle deposit insurance and let market forces rebuild the U.S. banking system. On the other hand, it can take the easy way out, as it has done so many times in the past, by doing nothing or by making cosmetic changes that amount to nothing. How much longer will Congress fiddle while the banking system burns?

1 Department of Economics, University of Nottingham (United Kingdom).

2 A brief discussion inevitably leaves out many important issues. To prevent any misunderstanding, I do not suggest that all was well with the pre-FDIC regime or that the abolition of deposit insurance alone would solve all U.S. banking problems.
COPYRIGHT 1993 Federal Reserve Bank of St. Louis
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Russell, Steven
Publication:Federal Reserve Bank of St. Louis Review
Date:Jan 1, 1993
Previous Article:Ted Balbach: an appreciation.
Next Article:Implications of annual examinations for the bank insurance fund.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters