|Cato Policy Analysis No. 97||December 15, 1987|
by Steven Horwitz and G. A. Selgin
Steven Horwitz and G. A. Selgin are, respectively, a research assistant and an assistant professor of economics at the Center for the Study of Market Processes, George Mason University.
Financial deregulation has recently become one of the United States' most hotly contested political issues. Like past debates on financial reform, the current one has been spurred on by a rash of crises in the banking industry. The failures of Continental Illinois, savings and loan banks in Ohio and Maryland, and numerous small banks have brought to light fundamental weaknesses of deposit insurance schemes and, less directly, of the entire apparatus of regulations affecting banking in the United States.
Of the policies that have actually weakened the U.S. banking system, among the most important are long-standing restrictions against interstate branching. Its long history of unit banking places the United States in a unique position among nations with developed banking systems. Most other developed countries have nationwide branch banking and have had it for a long time. Both economic theory and the experience of these countries suggest that nationwide branch banking must be part of any lasting cure for our nation's financial ills.
Policy makers, including the present Congress, have made numerous proposals for reform, deregulation, and even reregulation. Some have discussed plans for nationwide branch banking. Steps were taken in that direction with the passage of the Garn-St Germain Depository Institutions Act in 1982, but the issue is far from settled. Current law, based on the McFadden Act of 1927, the Banking Act of 1933, and the Bank Holding Company Act of 1956, with their amendments, allows each state to establish its own policies concerning the ability of out-of-state banks to enter the state's market. Some 30 states and the District of Columbia have reached reciprocal interstate banking agreements with states or other groups of states. Four other states permit banks from any state to operate within their boundaries. Over three-fourths of these 34 states limit such interstate banking to purchases of in-state banks by out-of-state bank holding companies as opposed to allowing out-of-state banks to set up completely new branches. The other one-fourth permit de novo branching. The remaining 16 states allow no interstate banking at all.(1)
Why We Still Don't Have Branch Banking
Current proposals are not the first ones to recommend interstate branch banking for the United States. After the panic of 1907, branch banking was considered as a means of avoiding future crises, but it lost out in the political battle to the formation of a central bank. When the banking industry collapsed in the early 1930s, nationwide branch banking was proposed again but was rejected in favor of federal deposit insurance. On both occasions, nationwide branching was opposed by powerful interest groups that successfully promoted alternative measures. Instead of involving fundamental structural changes, however, these alternatives were but palliatives that proved less than adequate in the long run.
To understand how these crises and their solutions evolved, it is helpful to begin with the entry of the federal government into bank regulation in 1863. The National Banking Act of 1863 created the national banking system, and while it did liberalize many banking laws, it did not form a branch banking system. The states retained control over branching, and most of them prohibited it. Although branching regulations were subsequently relaxed in many states, they were never made liberal enough to contribute to the national system's stability.
The panic of 1907 was the last, and worst, of a series of financial and economic crises that plagued the national banking system following the Civil War. These crises usually included a fall in output and prices (especially on the stock market), currency shortages, and bank runs. Like its predecessors, the panic of 1907 occurred in the autumn harvest season, when crop movements caused a substantial increase in the demand for currency in the rural areas. Legal restrictions of the National Banking Act of 1863 prevented banks from expanding the total amount of bank notes in circulation to meet this demand for currency.(2) Therefore, the need for additional currency could be met only by drawing on the reserves of "country" banks and their city correspondents, which caused the total money supply to shrink.(3) Currency shortages at harvest time were chronic in the post-Civil War era, and the "inelasticity" of currency became the most critical issue in monetary reform.
The debate that followed the 1907 panic drew the battlelines that would mark all similar debates in the future. This post-panic debate really involved three distinct interest groups: the big New York city banks, the midwestern city banks, and the country banks. Of these groups, only the midwestern city banks were really interested in positive structural reform. Their proposals, put forth by the American Bankers Association (ABA), which was dominated by this group, sought reform through deregulation, including "asset currency" (note issues free of restrictive bond-deposit requirements) and interstate branching.(4)
The New York banks were mainly interested in maintaining their supremacy (the status quo) by preventing schemes that would give more importance to midwestern banks--especially in Chicago--that were threatening that supremacy.(5) The country banks, on the other hand, wanted to prevent any branch- banking scheme that would put them in direct competition with the city banks.(6) So, although the country banks were most vocal in their opposition to Wall Street, they unwittingly and ironically aligned themselves with it in blocking the reform proposals of the ABA.(7) Today, members of the Independent Bankers Association continue to believe that "Full nationwide interstate banking would give an insuperable advantage to a handful of very large banks."(8)
The power of the New York bankers, combined with the numerical superiority of the country banks, doomed any reforms involving interstate branch banking, including the proposed Baltimore Plan of 1893 and Fowler Plan of 1902. The Baltimore Plan would have eliminated bond-deposit requirements (permitting notes to be issued on banks' general assets) while allowing limited branching. The Fowler Plan had similar features and added an embryonic central clearinghouse and more extensive branching provisions.(9) These and other asset-currency plans, which sought to end banking crises by deregulating note issue, were rejected by country bankers, who saw a link between them and proposals for branch banking. Some also argued that free note issue would be inflationary. This argument, ironically, was more likely to be justified if note issue was deregulated without branch banking as a complementary reform.
Advocates of branch banking (with or without asset-currency) had some support from contemporary economists and other authors, who argued that interstate branching could have prevented the panic of 1907 and that more liberal branching laws had promoted economic growth in the South and West. These writers also pointed to the success of Canada's branch- banking system in avoiding crises.(10) Unfortunately, although academic writers were mainly sympathetic, their political influence was not great.
With limited country-bank support, the law approved in 1913 to deal with the recurring panics called for 12 Federal Reserve banks, with partial federal government control but with control really based on the New York Fed.(11) The country bankers, by opposing free note issue and wrecking other mid- western proposals for decentralized reform, helped give Wall Street exactly what it wanted. In brief, central banking was a compromise erected in response to the New York banks' desire for continued hegemony, plus the country bankers' opposition to branch banking. Consequently, the possibility for true structural reform was passed by in favor of second- best tinkering with a flawed system.
That the Fed was not an adequate answer to the instability of the banking system was dramatically revealed in the early 1930s, when 4,801 banks failed, leaving many depositors empty- handed, worsening the general economic collapse, and provoking President Franklin D. Roosevelt to declare a nationwide bank holiday. As the crisis deepened, reform proposals were again raised, including renewed appeals for nationwide branching. The country banks opposed the proposals once again and sought further palliative measures to block them. According to economic historian Eugene White, "The unit bankers' actions were largely defensive, but what they lacked in terms of leadership or a program they made up in brute political clout."(12) Most of this clout was due to sheer numbers, which enabled unit bankers to put heavy pressure on state legislatures and Congress. The unit bankers also dominated state bankers' associations and effectively weakened any ABA proposals promoting branch banking.(13)
The unit bankers were anxious to preserve the dual banking system in which regulatory power was shared by state and federal authorities. Proposals by some larger national banks were viewed as attempts to sabotage the state banking systems that were supposedly responsible for providing banking services to small towns. Unfortunately, such small-town banks (with capital of less than $100,000) also accounted for over 90 percent of the failures in the 1920s and 1930s.(14) Many reform bills claimed that branch banking could prevent these problems. Nonetheless, aided by state regulatory authorities jealous of the expanded power that branch banking would give to federal authorities, the unit bankers were able to block any far-reaching branch-banking reform. They were also aided by proposals for deposit insurance, which presented a quick, apparently viable alternative to branch-banking that (it was claimed) would also safeguard the banking system. The establishment of the Federal Deposit Insurance Corporation (FDIC) "weakened the previous sense of urgency to modify federal statutes regulating branching."(15)
The city bankers rightfully claimed that mandatory deposit insurance, in contrast to a branch-banking reform, would involve a tax on competent bankers, forcing them to subsidize poorly managed banks. Unit bankers responded by raising once again the specter of a "money trust." They claimed that nationwide branching would allow the big New York banks to invade the small towns, putting neighborhood bankers out of business. Deposit insurance, in contrast, was supposed to protect depositors from the capricious behavior of big banks.
The idea of deposit insurance was not new. It had been tried in several states after the panic of 1907, and in each case where it was introduced, problems quickly arose. First, by making banks pay into a fund out of their own assets, the states faced an incentive problem. Because they bore only a fraction of the added risk of failure, banks began to hold more risky portfolios. Second, deposit insurance could do little to save banks when the economy turned bad. As agricultural prices fell in the late 1920s, banks in states with insurance plans began to collapse, and the insurance funds dried up. By early 1930 all eight state funds had gone bankrupt.
Despite the poor record of state deposit insurance schemes, unit banking interests argued successfully that nationwide deposit insurance could be effective if handled properly.(16) Their appeal to the plight of the small depositor, together with fears that a money monopoly would develop, carried the day politically. As a result, the FDIC was established. The unit bankers had won another political battle against structural reform. As one modern observer put it, "Once again the small banks in the country fought off legislation that would have made branching completely free . . . by suggesting a plan of deposit insurance as an alternative way of shoring up the unit banking system."(17)
Twice--in 1933 and in 1907--legislators faced an opportunity for true structural reform, and twice they applied BandAids instead. Both central banking and deposit insurance merely treated the symptoms of a bad system, and did so poorly at that. As one writer opined in 1933,
The opportunities inherent in the dramatic circumstances of the time were lost. Emergency measures . . . were taken . . . to patch up temporarily the forty-nine broken down systems. . . . But then, instead of passing the laws necessary for the development of a single system adequate for the needs of the Nation, Congress again resorted to patchwork. The Glass Bill, for all the revolutionary implications of its provision for the insurance of bank deposits, leaves the basic elements of weakness in our banking structure essen-tially the same as before.(18)
These same weaknesses have now brought "dramatic circumstances" to bear once again, and once again the unit bankers are braced to resist nationwide branching. Although they are no less politically influential than before, the unit bankers have also run out of quick fixes. So the time is ripe to reexamine the twice neglected case for nationwide branch banking and to challenge proposals calling for the adoption of further palliatives.
The Current Legal Status of Branch Banking
The McFadden Act of 1927 and the Banking Act of 1933, which are still in effect, prohibit banks headquartered in one state from operating additional deposit-taking offices in any other state. In the past, banks got around the law by forming multi-bank holding companies that extended across state lines. Political forces--led once again by unit bankers--rallied to block this strategy by passing the Douglas Amendment to the Bank Holding Company Act of 1956, which prevented holding companies from owning a bank in another state without that state's permission. Until around 1980, this legislation effectively checked interstate operation of full-service banks or branches.
About that time, changing market conditions increased the incentive for banks to evade the law. The failure of numerous savings and loans in the late 1970s and early 1980s made clear the need for takeovers by out-of-state banks as a part of a viable deposit-protection strategy. These takeovers were not strictly legal, but both the industry and the states allowed and encouraged them to save problem banks. In 1982, the Garn-St. Germain Act officially permitted out-of-state emergency takeovers under specific conditions, while also encouraging other kinds of interstate activity. At the same time, "nonbank" banks (banks not covered by the Bank Holding Company Act of 1956 and the Douglas Amendment because they do not make commercial loans) owned by nonbank companies such as Sears, Roebuck & Co. and Merrill Lynch & Co. began to provide banking services across state lines, further eroding interstate barriers.
The federal authorities also did not interfere when many states began to permit various forms of interstate banking. These forms ranged from reciprocal, regional agreements to entry by subsidiaries of bank holding companies headquartered in any of the 50 states. As of January 1, 1987, more than 91 percent of U.S. domestic banking assets are in states that allow some form of interstate banking.(19) These actions by state banking authorities have provoked increased calls for federal legislation to establish full nationwide branching. Before we suffer yet another crisis in banking, we must once again give serious consideration to the arguments for and against such legislation.
The Advantages of Branch Banking
Nationwide branch banking would provide numerous benefits to U.S. consumers, both directly and indirectly, through economywide stability and efficiency. Branching would enable bank depositors throughout the country to deal with the same bank in different states, thereby making it easier for people to move, travel, and do business across state lines. Banks could also make use of their brand-name capital, providing a cheap source of information to potential depositors. Limited economies of scale would lower banking costs and increase service. Economywide benefits of branching include lowering of bank risk through diversification, lessening of bank failures, greater ease of dealing with insolvent banks, greater ease of interbank funds movement, and increased competition in the financial industry. These benefits are explored below.
Direct Benefits to Customers
On the microeconomic level, the most obvious advantage of interstate banking to consumers is that, when a depositor moves to a new state, he would not have to spend time and effort searching for a new bank. The individual's former bank could well have a branch located near his new home. In addition, he would not need to go through the trouble of physically transferring deposits and records. The bank would be more than happy to take care of such interbranch transfers.
A similar advantage would accrue to travelers if their banks had branches in places they visit. Local merchants would be far more likely to accept a check drawn on an out- of-state branch of a bank operating locally than on an out- of-state bank. Since banks are more certain of obtaining funds from their own branches, they can clear such checks more quickly. Indeed, information about the balance in the traveler's account at home would be immediately available to branch employees. Furthermore, the traveler could cash his checks at the local branch of his bank and use cash. In either case, branch banking would reduce the need for, and the trouble of obtaining, traveler's checks, and the extra risk of traveling with cash.
Businesses and borrowers would also benefit from the ability to use the same bank in different states. Businesses that operate in many states could simplify their accounting procedures by dealing with the various branches of a single bank, instead of with different banks in each area of operation. This would also create opportunities for banks to design programs to meet the needs of interstate firms. Borrowers would have easier access to credit as funds would move more smoothly between branches than between banks. Business borrowers could reap the same advantages as business depositors in dealing only with one bank. Many of these cost savings to businesses would be passed on to consumers in the form of lower prices and more efficient service.
Permitting branching across state lines would also give banks a chance to deploy their brand-name capital. One reason firms use brand names is that they provide a low-cost source of information on product quality to consumers. This is especially important for banks because depositors do not have the level of expertise required to determine the quality of bank management and services without actually experiencing them. In the absence of information intermediaries, potential depositors will rely on brand names.
Consider the depositor who moves but is unable to find a branch of his bank in his new town. With nationwide branch banking, he would probably recognize the names of other banks, the reputations of which would be known to him. Thus he could forgo the expense of a detailed examination. In addition, it would be much easier for banks to enter a new market where other banks were providing inferior service. Because its brand name would be recognized, a new entrant with established branches elsewhere need not incur as many start-up costs as a truly de novo entrant would. This would keep incumbent banks more efficient as the cost of entry to new competition would be lower than under the present system.
Further, consumers would benefit from nationwide branching as banks come to take advantage of economies of scale. Aside from the other benefits associated with large, diverse banks, size itself can result in lower average operating costs and bring about lower prices for services and higher interest on deposits. It is evident from experience in states that already permit branch banking that branched banks tend to be larger than unit banks, and that they require fewer managers and administrators and are less expensive to operate because all services need not be developed in-house. More important, branches of larger banks tend to employ more knowledgeable, experienced bankers who, when they do not know an answer, have greater access to expertise. In addition, McCall reports that branch banks, especially in smaller rural areas, tend to provide a wider range of services than unit banks.(20) This seems more the case when a unit bank is acquired by a larger branch bank, allowing the acquired bank to provide new services that can be easily provided only by a large bank. Finally, banks in states with more liberal branching laws use a greater proportion of their resources for loans and have lower liquidity needs than banks in unit banking states.(21)
Benefits to the Macroeconomy
Diversification. Perhaps the most important benefit of nationwide branch banking is that it would enable banks to diversify their portfolios and reduce their overall exposure to risk. When a bank's area of operation is geographically limited, it depends for its deposits and loans on the continued prosperity of a geographically limited number of specific industries. For example, branches would have been superior to deposit insurance in handling the fall in farm prices in the 1920s. If there had been nationwide branching then, the banks in the farm belt states probably would have been branches of larger banks able to shift funds from more prosperous areas to keep branches in agricultural states afloat. Or, when necessary, the banks could have closed branches that no longer generated much business, just as Canadian banks did in the 1920s. Clearly, depositors do not lose as much when a branch is closed. Either their accounts are paid in full or they are transferred to another branch in a nearby town.
Recent examples of the problems caused by a lack of diversification are evident in the experience of banks that have depended on the success of the agriculture and oil industries. A disproportionate number of bank failures in recent years have occurred in the farm belt states, where small unit banks continue to rely on the fortunes of local farmers. As farm failures have increased, bank failures have followed in their wake. In the last few years, falling oil prices have caused similar problems in Texas and Oklahoma, where numerous unit banks have failed or have suffered large reductions in the value of their equity.(22)
Branch banking would have allowed Texas banks to diversify their holdings and lower their overall risk. If Texas banks could have operated branches in California, they could easily have loaned to and taken deposits from high-technology firms. When oil prices first fell, high-tech firms were doing well, so a bank with branches in both areas could simply have shifted funds between its own branches in the two areas. Of course, the same process would work in reverse if oil prices rose and the high-tech industry became depressed. Farm belt banks could diversify by branching into locations with industries unrelated, or countercyclically related, to farming. Interstate banking and other reductions in banking regulation would enable banks to diversify their assets and liabilities, both geographically and in terms of customer service.
In the Scottish banking system of the early 19th century, which had both free note issue and unlimited branching, failures were few and far between as local economic problems rarely threatened any individual bank. Branching also allowed full-service banking to reach even the smallest towns in Scotland. In neighboring England, which had more limited branching, economic problems caused numerous failures, most probably due to the lack of geographic diversification. For example, in 1825-26, at least 76 banks failed in England while only 4 small Scottish banks succumbed.(23)
Interbank Funds Movement. A further benefit of nationwide branch banking is that it would improve interbank relationships. Everyone has had the experience of depositing a check written on an out-of-state bank and being told it will take ten days to two weeks for it to clear. Of course, in-state and own-bank checks clear in a couple of days. In a system consisting mainly of unit banks, a check written on one bank but deposited at another, unaffiliated bank has to pass through a network of correspondent banks to get back to the first bank. To see some empirical evidence of this process, examine any canceled check that was deposited at an out-of-state bank and notice all of the endorsements on the back. With wider branch banking, banks could clear checks drawn on distant branches of rival banks through local branches of those banks; so the delay experienced by the consumer would be significantly reduced.
In addition to reducing the inconvenience to banks' customers, improvements in the check-clearing process would have macroeconomic effects as well. Inherent in interbank check clearing is an uncertainty, both about the other bank and about the check writer. In a correspondent system, the receiving bank runs a risk that it will not receive funds for a check because it does not know the financial condition of the other bank or of the check writer. It is this double uncertainty that lengthens the time that banks need to clear checks drawn on other banks.
Another problem shows up in the amount of reserves held by banks. Transfers of funds between banks clearing checks often involve the transfer of banks' reserves, since banks in different states do not conduct enough business with each other to justify simple accounting adjustments. In a correspondent banking system, unit banks therefore generally hold more reserves than they would otherwise need to, limiting the funds available to borrowers.(24)
An extensive system of branch banking would resolve many of these difficulties. A higher proportion of checks would be drawn on other branches of the same bank, necessitating only a transfer of credit between branches instead of through several correspondent banks. This also would conserve banks' reserves and leave more funds available for customer use. In addition, one of the two sources of uncertainty would be removed if a deposited check were drawn on another branch of the bank. The branch receiving the check would not need to be concerned about the financial condition of the other branch. Rather, it would have to worry only about the credit of the check writer, and that information could be readily available through the bank's computer network. Finally, branch banking systems have fewer banks generally, so check clearing becomes much more efficient, magnifying all of the other benefits.
Past experiences in Canada and Scotland demonstrate how branch banking can lead to a more efficient system for transferring funds between banks than that which has developed in the United States. Canadian banks were relatively free of government intervention until the middle of this century and were allowed to branch freely and circulate their own notes. Canada did not develop clearinghouses until much later than the United States, but not because its banking system was backward. Rather, as Kurt Schuler points out, with 13,000 banks in the United States and only 5 major banks in Canada, a far higher percentage of Canadian checks were cleared easily and quickly without multilateral clearings and transfers of reserves.(25) The Canadian system was also very effective at moving funds between the vast farmland regions of the Canadian west and the urban areas of the east. Despite the overwhelming importance of agriculture in the Canadian economy, Canada never experienced the harvest season currency shortages and banking crises or the exposure to agricultural depressions that plagued (and to some extent continue to plague) the United States.
The Scottish experience in the 18th and 19th centuries shows how the more efficient movement of interbank funds enhances macroeconomic stability. Banks failed less frequently both because of their greater diversification and because they were able to move funds rapidly between their various branch offices. The banks also dealt with each other more frequently (due to fewer numbers and common areas of involvement) and were more likely to assist each other in times of trouble through interbank loans.
Reduced Bank Failures. Taken together, diverse portfolios and easier funds movement help explain the superiority of branch banking in preventing and handling crises and panics. By allowing banks to diversify across economic and geographic boundaries, branch banking helps prevent crises associated with particular regions and industries. If problems persist, banks can more easily obtain needed funds from their own branches or on loan from the competition. A developed branch- banking system has far less need for an interbank reserve market like the federal funds market in the United States, since banks are fewer and more familiar with one another. Banks are far more likely, and much more able, to deal with each other directly, eliminating the need for middlemen and speeding up the movement of funds in crises. The Canadian branch system is a superb example of this. It was unaffected by the panic of 1907 and had no bank failures during the Great Depression, despite its close economic ties with the United States.(26)
Empirical evidence from the United States also supports these conclusions. During the period from 1930 to 1932, there were exceptional failure rates for all kinds of banks, and although the failure rate of larger banks increased more than that of smaller ones, failures were almost three times as frequent among smaller banks (that is, banks with less than $1 million in loans and investments).(27) As a group of economists recently noted, "Most of these smaller banks were unit banks and were apparently not sufficiently diversified to withstand default on their farm loans, which was their major loan category."(28) Restrictions on branching kept the bank small and prevented them from adequately diversifying, thereby increasing their risk of failure. In the absence of these restrictions, bank failures in the 1920s and 1930s would have been significantly reduced.
Another source of evidence is the different frequency of bank failures in states that permit unlimited intrastate branching compared to those that restrict such branching. One study by the Federal Reserve Bank of Atlanta indicates that since the 1920s and 1930s, failure rates have been lower among states with predominantly branched banks than in those states with more unit banks.(29) Other studies point to the fact that, during the 1960-75 period, a greater proportion of failed banks were in unit-banking states than in branch- banking ones.(30) According to a study in the Journal of Bank Research, during the 1970s "proportionately fewer multi- office than unit banks in the different size categories generally have been closed."(31) Branched banks also show no greater propensity than unit banks to appear on the FDIC's problem list.(32) While such studies are helpful, they do not capture the benefits of diversification that would be reaped with national branching. Nevertheless, these comparisons of failure rates among branched banks and unit banks in the United States do show that branched banks have significantly lower failure rates.
Perhaps the most convincing evidence of the reduction in failures comes from Canada. While U.S. banks reeled from the panic of 1907, Canadians were barely aware of the trouble. In the northern United States, merchants sold their goods at bargain prices across the border to obtain much-needed currency--even though it was Canadian--to ease the currency shortage in this country. This was the only evidence the Canadians had of anything unusual going on.(33) From 1921 to 1929, Canada recorded only one bank failure, the Home Bank in 1923, which cost depositors $15 million. This compared with $565 million lost to depositors in the United States during the same period.(34) In the early 1930s, while U.S. banks failed everywhere, no Canadian bank failed, even though the Canadian economy was also depressed. When a number of Detroit banks failed in the early thirties and bank runs set in, Detroit's neighbor across the river, Windsor, Ontario, felt no ill effects and gained millions in U.S. deposits.(35)
Mergers of Failing or Troubled Banks
Mergers are one way for banks today to extend their business across state lines. An important benefit of interstate mergers is that they increase the opportunities for failing banks to be rescued by healthy ones. Regulatory authorities have three alternatives for dealing with failing banks: payoffs, bailout, or purchase and assumption.(36) Regulators can close a failing bank and pay off its depositors. This causes a fair amount of havoc, as checks in process bounce and uninsured depositors are left to the mercy of the liquidation process. In a bailout, inefficient firms are kept afloat and their managers may benefit from their own inability to perform. Although bailouts rescue depositors, they often fail to provide the proper incentives to management. Under a purchase-and-assumption agreement, a solvent bank purchases the failing bank and assumes its assets and liabilities. This is the least disruptive and least costly approach as it maintains banking services to the community while also giving desired discipline to management. This was the case in Scotland in the 19th century, for example. When the Ayr Bank failed, other banks redeemed its outstanding notes and acquired the business of its former depositors with little interruption of service.(37) Although this was not technically a purchase, the benefits to depositors were similar to those provided by purchase-and-assumption agreements.
The provisions of the 1982 Garn-St Germain Act that permit out-of-state banks to acquire troubled institutions recognize the benefits of purchase and assumption. Unfortunately, the act does not apply to small troubled banks.(38) Permitting interstate banking would increase the number of potential saviors for insolvent institutions by removing any remaining obstacles to interstate mergers. This would increase system- wide stability by reducing failures and the resulting drain on the FDIC. Under nationwide branching, depositors at banks of all sizes would be better protected against failure.
But why wait until a bank is in serious trouble, when it is a less desirable (less salable) acquisition, before approving a merger? Why not let this happen before a crisis is imminent? Mergers can provide benefits to struggling banks while simultaneously promoting banking efficiency by bringing in new managers who may be more skilled than the old managers of the acquired bank. This not only reduces the possibility of failures but also increases the value of shareholders' equity. Even attempts at mergers have such effects, in that they often lead current management to improve its performance or face the prospect of being bought out. A recent example is First Interstate's attempted takeover of Bank of America. That a bank is willing to try an acquisition often means that it perceives that the target bank is in need of better management.
Branched banks themselves are also more desirable merger partners. A branched bank in trouble is a far more attractive acquisition than a unit bank. The argument that there are few banks big enough to be able to acquire big branch banks would not hold if branch banking were the norm. In that case, there would be more large banks in the system able to acquire a failing branched bank.
Alleged Disadvantages of Nationwide Branching
A major concern of unit bankers and other opponents of branch banking is that nationwide branching would lead to large, unresponsive banks that would eliminate competition in the banking industry. That concern also gives rise to the fear that the market would be dominated by a few firms and they would conspire to fix prices and provide shoddy service. But empirical evidence and economic theory suggest a different scenario. They support the view that monopolization of the banking industry is not a likely outcome of nationwide branching.
Lack of Local Control
One argument often advanced against branch banking is that it removes banking from local control, leaving loan decisions up to bankers in faraway places with little knowledge of the local environment. This criticism ignores both theory and experience.
One of the benefits of branching is that it leads to the development of more professional bankers.(39) Many unit banks are operated by less experienced personnel and are often just a part-time occupation for the management. The problem of nonprofessional bankers came to the forefront in the Ohio savings and loan crisis. Many of the thrifts that failed began as little more than "men's clubs" for local businessmen. They were open only a couple of hours a week, usually as a place for the managers to play poker.(40) When banking becomes a hobby, problems develop, including excessive lending to insiders and a greater chance of fraudulent activities. Branch banking encourages a higher quality of management, thereby reducing the risk of crises and failure. In Canada, it should be noted, the Canadian Bankers Association has historically been ahead of its U.S. counterpart in providing the educational and training programs needed to develop the qualified bankers necessary for branch banking.(41)
Another important characteristic of branch banking is that, assertions to the contrary notwithstanding, loan decisions are usually left up to local branch managers and loan officers; only in cases of very risky or large loans do such decisions go to the head office. In branched banks in both Canada and the United States, the vast majority of consumer and small business loans have always been handled by local branch managers.(42) Local branch managers have the same knowledge about local circumstances as do unit bankers. More important, since they are usually well-trained professionals, local branch managers are less likely to submit to personal preference in making loan decisions. Unit-bank supporters fear the loss of personalized banking but fail to recognize that uneconomic loans, made for personal reasons, are an important cause of bank failures and crises, which can hurt everyone involved. In the 1982 Penn Square failure, this was the precise problem.(43) In the Ohio savings and loan crisis, Home State Bank, which triggered the problem, had lent a great deal of money to ESM Securities, which was owned by a relative of the owner of Home State. ESM turned out to be bankrupt, and the bank collapsed.(44) If a banking system is to be stable, it is essential that it be run by professional and knowledgeable bankers.
Closely related to the lack-of-control argument is the complaint that large branch banks are overly bureaucratic and are insensitive to depositor needs. Proponents of this view have failed to demonstrate why large banks are more likely to be insensitive than other retail firms, and their allegation founders upon the rocks of both theory and experience. Larger firms have the same incentive as smaller ones to attract customers by providing the services and prices desired. By not doing so, a large firm loses profits like any other sized firm. If the firm grows too large, it has incentive to scale back to a more efficient operating size. If banks are strong rivals, any attempt to deviate from consumer preferences is sure to impose a significant cost on the perpetrator.
Recent developments in the industry provide evidence of bank responsiveness. Since deregulation, for example, banks have started a wide variety of specialized budget-banking programs to meet the needs of low-income depositors.(45) In general, previous deregulation has not led to bureaucratic monoliths. On the contrary, greater exposure to competition has made banks more responsive to consumer needs. Nationwide branch banking, being just a more effective way for banks to compete with each other, would do the same.
Concentration in the Industry
The last few years have seen an increase in both the number and the size of mergers in the banking industry. A common fear is that such mergers simply create paper profits and could lead to the monopolization of the industry. Millions of dollars change hands but no real services are created, and depositors are stuck with big, unresponsive banks. Such arguments demonstrate a misunderstanding of the nature of mergers and acquisitions and an ignorance of their numerous procompetitive effects.
In fact, the advantages from interstate branching are most readily obtained through merger and acquisition. Depositors benefit through more efficient, more stable banks. Shareholders benefit from the increase in equity value due to the increased effectiveness of new management. Merger activity also increases competition in local markets by expanding the options available to consumers.
Mergers can promote efficiency by allowing firms with specialized information to share it with acquired firms.(46) When an acquired firm obtains this information, it may be better able to serve depositor needs at lower cost. Mergers may also provide a low-cost way for banks to expand into consumer-demanded areas of business.(47) By combining with a partner that provides services it does not, a bank can take advantage of existing equipment and knowledge. By eliminating start-up costs in the new area, the merger enables resources to be allocated to other, higher valued uses.
When two banks combine, layers of management are eliminated, operations are consolidated and centralized, and productive efficiency is generally enhanced. Mergers may also lead to economies of scope by combining different types of services under a single firm. A simple example would be a merger between a brokerage firm and a bank, in which the merger would enable the bank to advertise its brokerage services in its monthly account statements.(48) Economies of scale and scope lower costs and increase efficiency, and mergers and acquisitions are two ways to achieve such economies.
The threat of merger also promotes competition by keeping incumbent firms on their toes. Permitting banks to branch nationwide through merger and acquisition, even if it leads to fewer banks, would provide the same competitive benefits as de novo branching.
Bank Size and Competition
The merger threat leads to the most common complaint against nationwide branch banking: its possible effect on the market structure of the banking industry. Opponents fear that market shares will become concentrated in the hands of a few large banks, raising again the specter of a money monopoly. This argument flies in the face of both evidence and theory. Econometric evidence on market shares in states that permit unrestricted branching indicates that branches do not increase market shares to any potentially troublesome degree.(49) Although the number of firms shrinks, the number of offices increases, and any possible adverse effect on prices is compensated for by increases in services and office hours.(50) The whittling away of traditional "banker's hours" in many areas is one particularly cogent bit of evidence of the increased competitiveness that branching promotes. And who has a more effective monopoly than the protected bank in a one-bank town? Restrictions on branching themselves create monopoly power.
Operation of the Canadian system confirms on a broader scale the limited evidence from states with more liberal branching laws. Despite being few in number, Canadian banks have historically been very aggressive competitors. Geographic dispersion and intense competition drive interest rates on loans to uniform levels, making funds more accessible to smaller borrowers. An observer in the 1930s commented that competition was very strong even in depressed economic conditions.(51) This counters the argument that fewer banks mean monopoly power and higher prices.
While such evidence is helpful, it goes only part way in showing that competition can thrive even as the number of banks declines. The view that economic efficiency is directly related to market shares and concentration is also disputed in economic theory. Scholars are turning away from the simple view that the number of firms determines firm behavior and efficiency. Many now view competition as a rivalrous process.(52) Instead of merely adding up market shares, the "competition as a process" approach is more concerned with entry and exit barriers, government restrictions on firm behavior, and market uncertainties. Our previous discussion of interbank relationships is a good example of how the two views differ. The old view would see a highly concentrated industry and conclude it was inefficient. The process approach would ask whether there were benefits to having fewer firms and the particular arrangements they promote. Because they look only at market shares and concentration ratios, many of the previously cited studies understate the degree of competition in a branched banking industry.
Of greatest concern to process-approach theorists are legal barriers to entry.(53) As long as firms are legally able to enter and exit a market as they please, the benefits of competition will be reaped. The concept of "potential competition" is the key here.(54) As long as there are no legal barriers to entry, the possibility of a new entrant in a market provides sufficient incentive for the incumbent firm or firms to be responsive to consumer preferences. If existing firms deviate from efficient behavior, other firms will enter the industry to discipline them. It is only when the law prevents entry and exit that existing firms can behave inefficiently and get away with it.
A recent econometric study confirmed these theoretical conclusions, finding that "branch banking does not necessarily reduce the level of competition in local banking markets, even when branching increased local market concentration."(55) In the market studied, potential competition from new entrants kept existing firms in line. Permitting interstate branching also lowers the cost of new entry by allowing banks to deploy their brand-name capital in a new market. By artificially limiting a bank's options, prohibitions on branch banking prevent the development of more efficient, lower cost ways of providing banking services. The new view of competition argues that regardless of concentration ratios, permitting banks to explore these options can leave us no worse off and is likely to improve things considerably.
The process view also recognizes that cooperation between firms can be efficiency-enhancing, whereas the market-structure view sees all cooperation as anticompetitive per se.(56) According to the process view, cooperation is one way of handling uncertainty or information problems in certain industries. In banking, a clearinghouse is an example of a cooperative venture that aids individual banks by serving as an information intermediary.(57) Yet the old, structural view of competition must perceive clearinghouses as evidence of inefficiency and as candidates for antitrust actions!
Finally, there is a related, and more fundamental, criticism of the market-structure view. It is that the actual process of competition itself is the only way to discover the "best" way to organize an industry.(58) Economists and policy makers do not have sufficient knowledge to determine the ideal structure of any industry. Another way to see this is to ask why, if we knew the best structure, we would allow competition at all. (If we know which team will win the Super Bowl, why play the game?) We could simply mandate that structure and enforce it. The problem, of course, is that we do not know and can find out only by allowing competition to operate. Market shares are virtually irrelevant, as we cannot know whether a high concentration ratio is proper or not. The only gauge we have is how well such industries meet consumer preferences. As Nobel laureate James Buchanan has said, "[Economic] order is itself defined as the outcome of the process that generates it."(59) It is only by eliminating legal barriers to entry and permitting the process of competition to proceed that the industry itself can discover the right size and number of firms. It is not just that bigger banks are better; it is that only a process of competition itself can "discover" the best structure for an industry.
By enforcing limits on branch banking, legislators are claiming knowledge they cannot have. Lacking any proper theoretical justification, they have restricted the market's ability to discover the highest valued uses of certain resources. The issue is whether we prefer a situation with a large number of little, local monopolies, unaffected by competitive pressure, or a smaller number of larger, branched banks constantly under the competitive gun and better at meeting the needs and demands of consumers. As John McGee has said, "Competition between the old and new ways of doing things proved to be more important, and more valuable, than was 'competition' among large numbers of the old style firms. That the new style firm may make large profits under these circumstances surely does not justify either dissolving it or preventing its emergence by prohibiting the new techniques. For consumers (and the surviving producer) benefit greatly in the process."(60)
Incidentally, to the extent that competition (through increased branching) has been permitted in banking, the result has not always been harmful to unit banks. Evidence gathered by Donald Savage and Elinor Solomon on new entry into previously restricted markets indicates relatively few negative effects on existing banks and general gains in service for the banking public.(61) This result essentially confirmed a study by McCall and Peterson.(62) When branch banking was permitted in New York state, many feared that the upstate unit banks would be gobbled up by the big New York City banks. The actual result, however, was that many local banks continued to thrive side-by-side with the large branch banks by providing specialized services that the bigger banks could not offer, including programs tailored to local needs. According to Savage and Solomon, "the evidence does not suggest that permitting more entry into banking markets, through either interstate branching or interstate bank holding company expansion, would endanger the safety and soundness of existing banks."(63)
Although periodically proposed as a solution to financial ills, and despite the meagerness of arguments against it, true branch banking has consistently been prevented by its politically powerful opponents. Twice in U.S. history the opportunity to enact nationwide branch banking has been at our feet and twice we have trampled it. The day of reckoning is here again. The financial sector is fast approaching another major crisis that only true branch banking can prevent. The time is right to overrule the McFadden Act of 1927 and Banking Act of 1933 by passing legislation to permit full and complete interstate branch banking. Besides helping to prevent an immediate crisis, this reform would provide high-quality banking and lasting macroeconomic stability. History has shown that patchwork measures, like the Fed and deposit insurance, are at best only temporary palliatives for a system in dire need of structural reform.
(1) See Federal Reserve Bank of Boston Economic Review (July 11, 1986); and Marsha E. Simonson, "Full-Service Interstate Banking Statutes: The Conditions of Entry," Oklahoma City University Law Review 11, no. 3 (Fall 1986): 752-53.
(2) For a good summary and discussion of the history and issues of the era of the national banking systems, see Oliver Mitchell Wentworth Sprague, History of Crises under the National Banking System (Fairfield, N.J.: Augustus M. Kelley, 1977); and Eugene Nelson White, The Regulation and Reform of the American Banking System 1900-1929 (Princeton: Princeton University Press, 1983).
(3) A. Piatt Andrew, "The Influence of the Crops upon Business in America," Quarterly Journal of Economics 20, no. 2 (May 1906): 323-53. (4) White, Regulation and Reform, p. 93.
(5) Gabriel Kolko, The Triumph of Conservatism (New York: Free Press, 1963), p. 141.
(6) Ibid., p. 149.
(7) Robert H. Wiebe, Businessmen and Reform: A Study of the Progressive Movement (Chicago: Quadrangle Books, 1968), p. 13; and Eugene Nelson White, "The Political Economy of Banking Regulation, 1864-1933," Journal of Economic History 42, no. 1 (March 1982): 33-40. (8) Kenneth A. Guenther, in Interstate Banking: How Much, How Fast? (Washington: Heritage Foundation, 1985), p. 5.
(9) White, Regulation and Reform, pp. 84-85.
(10) H. M. P. Eckardt, "If the United States Had Branch Banks," Atlantic Monthly, vol. 101, no. 4 (April 1908): 507-12; and Thornton Cooke, "Branch Banking for the West and South," Quarterly Journal of Economics 18 (February 1903): 97-113.
(11) Wiebe, p. 74.
(12) White, "Political Economy," p. 40.
(13) Helen M. Burns, The American Banking Community and New Deal Banking Reforms: 1933-1935 (Westport, Conn.: Greenwood Press, 1974), p. 59.
(14) George J. Benston et al., Perspectives on Safe and Sound Banking (Washington: American Bankers Association, 1986), pp. 53-72.
(15) White, Regulation and Reform, p. 40.
(16) Ibid., pp. 210-11.
(17) Ross Robertson, in Policies for a More Competitive Financial System (Federal Reserve Bank of Boston, conference series no. 8, June 1972), p. 43.
(18) Guy Greer, "Wanted: Real Banking Reform," Harper's Magazine, October 1933, p. 583.
(19) Donald T. Savage and Elaine J. Peterson, "Interstate Banking Developments," Federal Reserve Bulletin 73, no. 2 (February 1987): 81.
(20) Alan S. McCall, "The Impact of Bank Structure on Bank Service to Local Communities," Journal of Bank Research 11, no. 2 (Summer 1980): 104.
(22) For example, Allied Bancshares fell 35 percent, RepublicBank Corp. fell 29 percent, and MCorp fell 30 percent between July 1985 and August 1986, paralleling the fall in oil prices. (Percentages are from the Wall Street Journal.)
(23) Lawrence H. White, Free Banking in Britain (Cambridge: Cambridge University Press, 1984), p. 47.
(24) For a very good discussion of the movement of interbank funds, see Kurt Schuler, "The Economic Significance of Branch Banking," George Mason University, 1986 unpublished manuscript, pp. 20-21. (25) Ibid., pp 20-21; and Gaines T. Cartinhour, "Branch Banks versus Unit Banks," The Annals of the American Academy (January 1934): 37.
(26) See Eckardt; Cartinhour; Schuler; and Bert Ely, "The Big Bust: The 1930-33 Banking Collapse--Its Causes, Its Lessons." In The Financial Services Revolution: Policy Directions for the Future, ed. Catherine England and Thomas Huertas (Boston: Kluwer Academic Press, 1988).
(28) Benston et al., p. 61.
(29) Robert A. Eisenbeis, "Interstate Banking's Impact upon Financial System Risk," Federal Reserve Bank of Atlanta Economic Review (March 1985): 31.
(30) Alan S. McCall and John T. Lane, "Multi-Office Banking and the Safety and Soundness of Commercial Banks," Journal of Bank Research 11, no. 2 (Summer 1980): 89.
(32) Ibid., p. 92.
(33) Eckardt, p. 510.
(35) Cartinhour, p. 37.
(36) Irvine H. Sprague, Bailout (New York: Basic Books, 1986), p. 22.
(37) Lawrence H. White, p. 27.
(38) B. Frank King, "Interstate Banking: Issues and Evidence," Federal Reserve Bank of Atlanta Economic Review (April 1984): 43.
(39) See Cartinhour; Guy Greer, "Why Canadian Banks Don't Fail," Harper's Magazine (May 1933): 722-34; and Eckardt.
(40) "Tremors from Ohio's Bank Run," Business Week, April 1, 1985, p. 29. (41) Cartinhour, p. 44.
(42) Cartinhour, p. 38; Eckardt; Greer, "Why Canadian Banks Don't Fail"; and personal interview with John Doherty, Chase Manhattan Bank, May 8, 1987.
(43) Irvine H. Sprague, p. 114ff.
(44) "Tremors from Ohio's Bank Run," p. 29.
(45) For a number of examples, see the American Bankers Association Journal (November 1985): 69.
(46) Paul Calen, "Interstate Bank Mergers and Competition in Banking," Federal Reserve Bank of Philadelphia Business Review (January-February 1987): 9.
(48) Ibid. Such a merger would require, besides liberal branching laws, repeal of the Glass-Steagall Act.
(49) Donald T. Savage and Elinor H. Solomon, "Branch Banking: The Competitive Issues," Journal of Bank Research 11, no. 2 (Summer 1980): 111-13; McCall; and George J. Benston, "Economies of Scale of Financial Institutions," Journal of Money, Credit and Banking 2, no. 4 (May 1972): 317-41.
(50) Savage and Solomon, p. 118.
(51) Greer, "Why Canadian Banks Don't Fail," p. 733.
(52) The classics in the literature are Ronald H. Coase, "The Nature of the Firm," Economica n.s. 4 (1937): 386-405; Ronald H. Coase, "The Problem of Social Cost," Journal of Law and Economics 3 (October 1960): 1-44; Israel Kirzner, Competition and Entrepreneurship (Chicago: University of Chicago Press, 1973); Harold Demsetz, "Barriers to Entry," American Economic Review 72, no. 1 (March 1982): 47-57; and Oliver E. Williamson, The Economic Institutions of Capitalism (New York: Free Press, 1985).
(53) See Demsetz.
(54) For a summary, see W. J. Baumol, John Panzer, and Robert Willig, Contestable Markets (New York: Harcourt Brace Jovanovich, 1982). (55) Christopher R. Thomas and Richard J. Rivard, "Statewide Branch Banking and De Novo Entry: The Evidence from Florida," paper presented at the Southern Economics Association meetings, November 1986, p. 10.
(56) See Williamson; and Donald Dewey, "Information, Entry, and Welfare: The Case for Collusion," American Economic Review 69, no. 4 (December 1979): 587-94.
(57) Gary Gorton, "Clearinghouses and the Origin of Central Banking in the United States," Journal of Economic History 45, no. 2 (June 1985): 277-83.
(58) For a concise and modern version of this argument, see Don Lavoie, National Economic Planning: What Is Left? (Cambridge, Mass.: Ballinger Publishing Co., 1985) and the literature cited therein.
(59) James M. Buchanan, "Order Defined in the Process of Its Emergence," Literature of Liberty 5, no. 4 (Winter 1982), p. 5.
(60) John S. McGee, In Defense of Industrial Concentration (New York: Praeger Publishing, 1971), p. 22.
(61) Savage and Solomon, p. 117.
© 1987 The Cato Institute
Please send comments to webmaster