Government-provided deposit insurance is seen as necessary to protect the economy from (unjustified) bank runs and bank panics. A bank run is unjustified when the bank being run on is solvent—when its assets exceed its liabilities, so that it has more than enough to pay off all its depositors—but depositors nevertheless come to fear that the bank is insolvent. Believing that not everyone will be able to get his money out, depositors “run” to the bank to get their own out while they can.
A bank panic, as opposed to a run, is a rash of bank runs caused by a kind of contagion: A run on one bank, whether justified or not, makes depositors at other banks fear that perhaps their banks are at risk, too. If depositors at other banks don’t have reliable information that their banks are sound, they run on their banks, too, in a panic that can bring down sound banks along with the unsound ones. Deposit insurance does prevent bank panics by assuring depositors that they will be able to get their money back even if their bank fails. The FDIC has been very successful in preventing bank panics in the U.S.
This discussion raises two further questions, however: Were market forces failing to regulate banks’ safety and soundness properly? And was government-provided deposit insurance itself safe and effective?
If the first question means “Were free-market forces failing to regulate banks’ safety and soundness properly?” the answer must be “no,” because there was no free market in banking before 1933. Banking in the U.S. has been hampered by legal restrictions from early on. Instability in banking was caused not by the freedoms the banks retained but by these legal restrictions.
One important government intervention was a restriction on the freedom of banks to issue banknotes (the standard paper money we are accustomed to, but issued by individual banks—remember this was before the Federal Reserve). Between the Civil War and 1913, virtually all banknotes were issued by individual banks, but banks were allowed to issue notes only if they followed the government’s rules. By those rules, a bank in effect had to loan the federal government an amount of money roughly equal to the quantity of bank notes they issued. A bank did this by purchasing certain U.S. government bonds. A consequence of this restriction was that from time to time, typically during harvest season, banks were unable to find and buy those bonds in sufficient quantity to produce as much paper money as their customers needed to pay their bills. A shortage of money at harvest season was the trigger for the bank panics of 1893 and 1907.
Another important intervention was the prohibition on branch banking that lasted in the U.S. until the 1980s. In a free market, businesses are free to expand to wherever customers are willing to do business with them. But in the United States, banks based in one state were not allowed to open branches in other states. In some states, in fact, “unit banking” rules allowed banks no branches at all, not even in their own regions: they were allowed to have one location only!
This legal restriction greatly weakened banks by preventing them from diversifying the kinds of loans they made. A bank in farm country, for example, would lend almost exclusively to farmers, and so was as vulnerable as its farmer clients to bad weather and drops in crop prices. If banks had been allowed to branch widely around the nation, they could have made loans to a wide variety of businesses in different places, so that when one region of the country or one sector of the economy fell on hard times, the bank’s branches in healthier regions or sectors could support branches in the troubled areas. Also banks with many branches can quickly move cash to branches that are running short of reserves from branches that have plenty. George Selgin addresses the importance of freedom to branch in this way:
No episode illustrates more dramatically the weakening effect of anti-branching laws than the Great Depression. Between 1931 and 1933 several thousand US banks—mostly unit banks—failed. In contrast Canada’s branch-banking network did not suffer a single bank failure even though in other respects Canada was just as hard hit by the depression.
And Canada did not have government-provided deposit insurance until 1967.
Government-provided deposit insurance thus does not seem to be necessary: the problem of bank runs and panics it supposedly addresses would be better addressed by repealing the legal restrictions which lead to runs and panics.
In fact, government-provided deposit insurance in the United States has historically been not only unnecessary but damaging. When it was instituted in 1933, it was known to cause problems. Though he signed the Federal Deposit Insurance Act, Franklin Roosevelt himself opposed the idea in the 1930s, as did many others, because already evidence was clear that government-provided deposit insurance at the state level was harmful. According to Clifford Thies and Daniel Gerlowski, “[f]rom 1908 to 1917 eight states passed deposit guaranty legislation.… Of the eight, all but the Texas guaranty fund [had] left depositors with uninsured losses.” The general pattern was that the deposit insurance allowed a few unscrupulous and/or incompetent bankers to attract a lot of deposits which they then loaned out unwisely. The troubles of imprudent banks put the insurance fund in jeopardy; prudent banks then pulled out of the system, leaving only the worst banks participating. Thies and Gerlowski’s account of events in Kansas is representative:
In Kansas, the 1923 failure of the American State Bank of Wichita, the third largest bank in the system, embarrassed a guaranty fund already deep in debt. Interest on the fund’s outstanding debt plus interest on the debt expected to be incurred to make good on losses at the American State Bank approximated revenue from assessments, meaning no money would be available to cover future losses. This situation led to a massive exodus of banks from the system and left a remaining risk pool of banks that were uninsurable. When the guaranty law was repealed in 1929, depositors of 88 failed banks were left with nothing.
Given its well-understood bad record at the state level, how did deposit insurance get passed at the national level in 1933? Take a moment to anticipate the answer: recall the Incentive Principle about how government intervention tempts people to benefit themselves at others’ expense and what we said in Chapter 7 about the special interest effect. What special interest group might benefit from government-provided deposit insurance?
Federal deposit insurance, as originally passed, was supposed to be a temporary, emergency measure. It was promoted by Senator Henry Steagall of Alabama, a state whose many small banks were politically powerful. Recall that, at the outset, the maximum amount insured by FDIC was quite small, only $2,500; this is a significant clue. What interest group would benefit from insurance on such comparatively small amounts? It would be less important to larger banks whose clients typically had much larger sums on deposit. But to smaller banks with many smaller depositors, this insurance was particularly useful, because it shielded small banks from having to compete with larger banks on the basis of safety. No bigger bank, say, from Birmingham or Montgomery, could take customers away from a small Alabama bank by claiming, “Your money is safer with us,” because your money—up to the $2,500 insured by the federal government—was just as safe in the small, local bank.
Federal deposit insurance in the U.S. began, as so much intervention does, as special interest legislation. It aimed at benefiting small banks by eliminating some of the competition they would otherwise have had to face. As frequently happens, the “temporary” intervention became permanent and the amounts insured rose steadily. To protect taxpayers, bank regulations such as the Basel rules were imposed, and these interventions triggered their own ill consequences as we have seen.
The Federal Savings and Loan Insurance Corporation (FSLIC), FDIC’s sister organization which insured depositors in savings and loan associations, collapsed dramatically in the 1980s. FSLIC not only failed to protect depositors but also had the effect of encouraging risky business. Covering the deposits it insured required over $25 billion of taxpayer money on top of special assessments on healthy insured banks. The luridly colorful tale includes accounts of politicians obtaining “regulatory forbearance” for banks that
supported them politically, and “zombie” S&L’s—banks already deeply insolvent but not yet closed by the understaffed FSLIC—losing hundreds of millions of additional dollars on go-for-broke investment gambles. They were gambling with depositors’ money, of course, but the depositors were insured, so they paid little attention. In consequence, resources were misallocated badly and taxpayers paid dearly.
How much simpler, cleaner, more effective and adaptive would the regulation of banking be were market forces the regulator? We can’t be certain, of course, but there is much reason to believe it would be far better.
The Need for Regulation by Market Forces
Government regulatory agencies face no market competition; they cannot be driven out of existence by failure to attract repeat business when they do a bad job. The rules they propose for regulating business behavior cannot be refused by enterprises that prefer to do business according to other rules. The revenue stream of government regulators is captive: It originates not in satisfied clients in repeated dealings day after day, but in taxpayers’ surrender of their property to the IRS every April 15th. Government regulatory agencies can botch their jobs—as the Basel Committee surely has done—they can implement restrictions that hamper enterprises’ ability to create value for customers, and yet stay in business year after year. There is no process in government regulation for getting rid of a bad regulator, or for rejecting bad regulations.
What governmental regulatory agencies are attempting to do is central economic planning. It differs from Soviet central planning only in its extent, not in its nature. Whereas Soviet attempts to plan the entire economy prevented improvement in the standard of living of the Soviet people, the U.S. government’s partial planning of sectors of the economy merely slows down improvement of Americans’ standard of living. The damage to economic well-being is proportional to the extent of the top-down planning attempted.
And that brings us back to the fundamental reason for relying on regulation by market forces that we discussed in Chapter 6: In a world of pervasive uncertainty and frequent human error, society has no better way of discovering what standards and practices best protect the public health and safety—in this case financial safety—than entrepreneurial innovation and profit-and-loss selection of those standards and practices. We need a market process through which a variety of different regulatory approaches is tried out, with those that perform better being increasingly used, and those that perform worse being abandoned.
The decisive advantage of regulation by market forces over regulation by legislative restriction is that regulation by market forces fosters continuous experimentation and adaptation, informed by success and failure in accomplishing its purposes. Banks that manage their capital poorly in a free market will tend to make losses; banks that manage it well will tend to make profits. Through profit-and-loss selection, beneficial regulation spreads and useless or counter-productive regulation dies out.
The best we can do is to let banks and associations of banks regulate themselves on a contractual, consensual basis, aiming not at the high social goal of good systemic performance, but at the humble individual goal of creating value for their customers and thereby profiting as much as possible. Good systemic performance will emerge spontaneously from the choices free banks and free customers make under such general rules. The system and its results will not be perfect because human beings are so imperfect, but they will be the best achievable.
It is easy for me to assert that, but how might it work in practice?
How Regulation of Bank Capital by Free Market Forces Might Work
In the absence of taxpayer-funded deposit insurance and one-size-fits-all bank capital regulation, we would have … who knows what, but lots of innovation through which banking would evolve. We could expect the emergence of robust methods of making sure depositors don’t lose their money. It’s impossible to say what would evolve—one of the beauties of innovation is how often it surprises us with goods and services far better than we could have imagined—but let’s speculate.
One possibility is private deposit insurance. Depositors and/or their banks would have to pay for it themselves rather than pass the expense on to others through the FDIC. George Selgin writes:
Such insurance, provided on a competitive basis, would have a distinct advantage over present government-administered insurance. Government insurance assesses individual banks using a flat-rate schedule, charging them only according to their total deposits. This procedure subsidizes high-risk banks at the expense of low-risk ones…. In contrast, profit-maximizing, competing private insurers would attempt to charge every bank a premium reflecting the riskiness of its particular assets.
Some banks might contract with individual insurance companies; others might work with several at once, offering their depositors a choice of plans. Banks might also offer their depositors the option of going with no deposit insurance in exchange for slightly lower fees or higher interest rates on their deposits. Banks would probably advertise the benefits of their insurance options.
Unlike today, this would be real insurance: it would be a contract between insurer and insured, with freedom to exit on both sides. That freedom assures that each side is creating value for the other. By contrast, banks today are not permitted to decline the “insurance” by the FDIC, which assesses banks’ “premiums” based not on the quantity of deposits insured, but on the bank’s total assets. By this means virtually all bank customers are forced to pay for FDIC, including those with no deposits and those with deposits over the insurable maximum. The fees charged by FDIC are better understood not as premiums but as a tax.
There are various other ways of providing depositors with assurance that their deposits are safe, some of them probably better than deposit insurance as such. One such way is extended liability. In various times and places, bank shareholders have accepted double, triple, or even unlimited liability for the bank’s obligations. That means that a shareholder who had invested, say, $1000 in a bank would agree to be personally liable for $2000 or $3000 worth of bank obligations, or any amount the bank owed. Similarly, some banks’ by-laws have obliged shareholders to respond to calls for additional capital, if necessary. In these cases the wealthy individuals who own the banks in effect insure deposits out of their own property.
In other cases, notably in the bank panics of the late 19th and early 20th centuries in the U.S., sound banks have assured the public of the safety of their deposits and thereby headed off possible runs by publishing their balance sheets in the newspapers. This disclosure of the details of their assets and liabilities gave depositors the confidence they needed.
A promising kind of private deposit insurance is a system of cross-guarantees among a large number of banks. The group of banks would collectively insure the obligations of each member, so that, in effect, each bank’s obligations are backed by the capital of all. Such a policy would provide strong motivation for all the banks in the insurance pool to monitor one another’s solvency and banking practices. The group would almost certainly set rigorous capital requirements and require regular inspection of member banks to identify any problems before they grow large. Any bank that should fail inspection would run the risk of suspension or expulsion from the group; that would give it a strong incentive to behave prudently.
In a free banking system, clearinghouse associations would be natural groups for cross-guaranteeing deposits. According to George Selgin, the discipline on bank soundness exerted through clearinghouses historically was so strong that other ways of assuring the safety of deposits, such as extended liability, were neither necessary nor even particularly significant as a check on banks’ risk-taking.
Undoubtedly banks would fail on occasion. Failures would not usually mean losses for depositors, however, either because a rival bank takes over the failed bank’s assets and liabilities or because of the kinds of insurance we have been discussing. Surviving banks and their insurance companies or clearinghouse association partne
rs would study the causes of those failures and make changes to their business practices accordingly. Then those new practices would be put to the market test. Unlike in today’s one-size-fits-all regulation by a centralized bureaucracy, we would see various competing systems of regulation generated and tested by actions of many market participants as the industry evolves.
In such a setting, whoever is on the hook for deposits—the insurance companies, the depositors themselves if they choose to go uninsured, the shareholders with extended liability, or the sister banks in a cross-guarantee pool—would have a strong incentive to monitor the capital adequacy and risk levels of the banks that hold the deposits … or pay someone else to monitor them.
I would not expect depositors to do much monitoring; I certainly wouldn’t monitor my bank. Like most people, I don’t have the time or knowledge required. But most of us also don’t have the time and knowledge to evaluate the fire resistance of the insulation in our homes or the competence of our auto mechanics. We pay Underwriters Laboratories (UL) and The National Institute for Automotive Service Excellence (ASE) to certify them for us. Or rather, we pay manufacturers and service stations a bit more for the certifications they provide us, and they pay UL and ASE to do the certifying.
Similar kinds of certification (or something better not imagined yet) would evolve in a free market for banking and financial services. Many depositors would be willing to pay a bit more, or give more business, to banks that offered trusted third-party certification of their safety and soundness. Insurance companies would offer better rates to certified banks, so banks would pay for that certification. Cross-guarantee pools and clearinghouse associations might hire bank-examination specialists to conduct regular examinations in order to assure the public, and themselves, of their members’ soundness. One way or the other, unlike today, banks would have to compete on safety and soundness.
Free Our Markets Page 30