Larry Schweikart teaches history at the University of Dayton.
Those of us old enough to have parents or grandparents who lived through the Great Depression have probably heard the remark that “Franklin Roosevelt saved the banking system with deposit insurance.” The purported value of federal deposit insurance for keeping banks solvent is assumed, and virtually no one seeks to question it anymore. While challenges to Glass-Steagall, the law that separated commercial and investment banking, surfaced fairly routinely in Washington (it has been slightly modified in the last 20 years), there has been nary a peep about ending deposit insurance.
This is all the more striking because of deposit insurance’s role in the savings and loan (S&L) debacle of the late 1970s and early 1980s. Support for deposit insurance is even more misguided now that we have some actual data on which to judge it. The bottom line? Government-mandated deposit insurance is at best useless and at worst dangerous. At the very least it contributed to the banking collapse of the 1920s that weakened the financial structure which toppled in 1932.
A number of scholars began to examine the role of deposit insurance after the S&Ls collapsed, finding that while the “interest rate mismatch” (paying market price for deposits via interest, but receiving much less from fixed, lower-rate mortgage payments) was the primary culprit, weak S&Ls were pushed over the edge by the incentives created by deposit insurance.
In a nutshell, managers and owners of S&Ls, knowing that their friends, neighbors, and customers were “protected” and “insured” by the U.S. government, engaged in riskier behavior than they normally would have if those people’s savings had indeed been at risk. Any gambler playing with house money has no problem raising the bet.
But bank deposit insurance did not originate in the 1970s or even in the New Deal. It has a longer history. The first insurance efforts came out of the private sector in the form of clearinghouse associations, in which members created markets in one another’s liabilities and used a variety of devices to keep “free riders” out. The purpose of these organizations, though, was not insurance per se, but a form of information-sharing that would preclude the need for insurance.
Nevertheless, various early insurance schemes appeared. One example is the New York Safety Fund (1829), which by the 1830s had transformed from a predominantly private-sector organization to a state-run apparatus. When the insurance systems were privately held, run, and funded, they had a good shot of surviving. But the state-directed insurance funds all flopped.
After nationwide “reforms” placed authority over the state banking systems in the hands of either state bank examiners, commissioners, or banking boards, there was no longer a question of whether the state would regulate banks. However, the principles by which a state would regulate the banks still mattered greatly, especially as the extremely prosperous period for agriculture (1914-1920) fed higher growth rates for rural banks. To preclude failures, states in these agricultural areas instituted a second wave of state insurance funds after 1907.
At the same time, several agricultural states did not have insured banks. To assess credit or blame for failures, however, it is necessary to understand that whether banks were insured or not, branch banking might exist, permitting private banks to open extension offices to make loans and collect deposits. Or branch banking might be prohibited by state law, permitting only unit banks; to put an office in a new location, a bank would have to obtain a separate state charter as though it were a new institution. Even if the same owner held a majority interest in many unit banks, the assets could not be mixed, nor could the liabilities be diversified through shared risk. In short, when permitted, the banking systems developed their own “insurance” funds through branching, and, as a backup, clearinghouse associations.
The agricultural collapse in the United States began in the 1920s, as former allies and enemies all put down their weapons of war and went back to the fields. With prices falling, the farm sector entered a death spiral, taking agricultural banks with it. Disastrous as it was, it nevertheless provided a carefully controlled experiment in the efficacy of deposit insurance. In a series of studies that debunks the long-touted benefits of deposit insurance, Columbia University economist Charles Calomiris has analyzed the capital and assets of insured and uninsured banks, and of branch and unit banks. The results establish without question that the more regulation the state imposes, the worse things get, and that markets, and structures that freely evolve from the market, provide the best defense for depositors.
Comparing the insured and uninsured states, and superimposing the data of branch bank and unit bank states, Calomiris found that “from 1921 to 1929 only 37 branching banks failed in the United States,” which constituted only a 4 percent failure rate of all branch bank facilities. On the other hand, Calomiris found that those states that had compulsory deposit insurance, especially those that had established their system early and thus provided a long gestation period for abuse, fared the worst, losing between one and five times the equity of an average state bank.
These studies make clear that in the 1920s the best alternative for protecting bank deposits was branch banking in a state with no insurance, followed by branch banking in a state with voluntary insurance. The worst scenario was to have unit banking in a state where deposit insurance was compulsory, that is, where the government regulators wielded the most power and allowed the smallest window of banking options.
The overall structure of banking within states was weakened owing to the critical role of perceptions of and information about bank stability. All banks became suspect when some banks failed, and no matter the reputation that large banks had built up over the years, public panics and “manias” can often outrun the flow of accurate information. It was precisely this understanding of banking that led virtually all of the state legislatures in the antebellum period to ignore laws that required that any bank that “suspended” during a run (refused to pay out gold or silver for its paper money) should be closed. Quite the contrary, the fact that almost all banks in a given state, then usually across the country, would “suspend” simultaneously indicated that information was getting out and that the banks were acting in concert to prevent runs from even starting.
The best form of market protection (and, if governments insist on regulating, of regulation) is the kind that enhances the transmission of information. Deposit insurance interferes with this by concealing the weaknesses of a bank—by postponing runs that would otherwise occur without it. A better policy, one that involves more freedom, has been under our noses for more than 150 years—branch banking. But resistance from the unit-bank lobby prevented an interstate branch-banking law from becoming a reality in 1928 (when most observers were sure it would pass). Continued unit-bank resistance nipped interstate branch-banking in the bud in the late 1980s, when again many proponents thought it would pass.
In the meantime, the S&L crisis struck. In light of that crisis, we should be clear on this: branch banking is not perfect; it couldn’t save the banking system in states such as South Carolina in the 1920s. Information transmission is an excellent means to prevent runs only if some diversification is possible. In cotton-oriented South Carolina in the mid-1920s, there were few alternative investments. A similar case existed with the famous chain-bank system (not a branch system, but the next closest variant) in the late 1920s. It also failed, because the members of the chain were overwhelmingly concentrated in sheep, and mining was about the only other potential investment.
As has been discussed often, the S&L crisis—like the California power crisis today—was initially blamed incorrectly on “deregulation.” It would have been correct to say “partial deregulation,” which often is as bad as none at all. In the case of the S&Ls, they had already operated with certain advantages over banks since the New Deal; they were allowed to offer one-half a percent more for deposits. But they also were limited by law in what they could lend on: consumer and business loans were prohibited, but mortgage loans were encouraged. Mortgages of the 1960s typically had a life of 15-20 years. But in the period of rapid inflation after 1969, almost all S&Ls found themselves paying 6-10 percent for deposits, but collecting only 5-7 percent on 15-year mortgage loans. Obviously, that could not continue for long.
Congress decided to make the S&Ls more like banks by removing “Regulation Q,” which limited the interest rates banks and S&Ls could pay. Both banks and S&Ls opposed this for obvious reasons: costs for deposits would rise. But the S&Ls had a more substantial argument against partial deregulation: if Congress did not make them entirely like banks, they would “pay out short, but take in long.” Congress allowed the S&Ls to pay competitive rates for deposits to attract more money, and it loosened some of the lending restrictions. But they remained overwhelmingly frozen into the long-term mortgages, and that “interest-rate mismatch” could not be overcome quickly enough. By the time the slow-liquidating mortgage loans could be recovered and new higher interest-rate loans made, the S&Ls’ cash would be gone. The only solution was to invest in assets that brought higher returns—mostly land, but also, to a smaller degree, junk bonds.
At that point, deposit insurance again played a critical role. An honest, effective S&L owner or manager could see that only the riskiest investments would ever “beat the clock” and offset the slow-recovering mortgages. But the key word was “risky.” The owners and managers knew that they could just as easily lose their institutions’ assets as recoup enough money to place them on a paying basis again. And most owners, despite the characterization of them in the media, were not snakes. They knew and lived with their customers. Under other circumstances, they would have filed for bankruptcy protection and done their best to pay each depositor a percentage of the proceeds. But the “magic pill” of deposit insurance made all that bankruptcy talk unnecessary. If the S&L gambled big and won, the depositors were safe. If it sank the depositors’ money in a dry hole—the depositors were still safe! As the comedian Yakov Smirnoff once said, “Is this a great country or what?”
The presence of deposit insurance unquestionably encouraged S&L owners and managers to take extreme risks in hopes of righting their financial ships. I suspect that deposit insurance did more to push otherwise ethical and well-meaning owners and managers into risky ventures than it did the out-and-out crooks, who cared little if depositors got hurt anyway.
All of this comes full circle to branching. By the 1970s many S&Ls in fact were intrastate branch operations. The largest S&L in Arizona, for example (where I kept my paltry savings), was Western Savings, which had dozens of branches. Again, however, the branch is only a means of transmitting information, and it will transmit bad information as well as good. When land values went south in the late 1970s and early 1980s, Western Savings, along with hundreds of Texas, Oklahoma, California, and Florida S&Ls heavily vested in land, went belly up. Ironically, as has historically been the case, land values do eventually return: the land held by the U.S. government eventually paid off nearly 100 cents on the dollar. But the S&Ls did not have the benefit of time.
Moreover, while intrastate branch banking is good, interstate branch banking is better. The S&Ls could not legally set up branches in different parts of the country to permit wide diversification. That and the presence of deposit insurance sealed their fate.
In nearly 100 years of active bank regulation, at both the state and national level, one reasonable policy—interstate branching—has been ignored in order to implement a poor policy—deposit insurance. The reader can draw his own conclusions as to why governments favor deposit insurance, but no one can make the argument that it is the best policy to keep banks sound and depositors’ money safe.
- These eight states were, in order of appearance of the statutes, Oklahoma, Texas, Kansas, Nebraska, South Dakota, North Dakota, Washington, and Mississippi. See Charles Calomiris, “Is Deposit Insurance Necessary? A Historical Perspective,” Journal of Economic History, June 1990, pp. 283-95, and his “Deposit Insurance: Lessons from the Record,” Economic Perspectives, May/June 1989, pp. 10-30.
- Banks in the “Wild West” were remarkably solvent and resistant to holdups due in part to these “information assets.” See my “The Non-Existent Frontier Bank Robbery,” Ideas on Liberty, February 2001.