Vol. 14 No. 3

COMMUNITY BANKING, MONITORING, AND THE CLINTON PLAN

Ralph Chami and Jeffrey H. Fischer

One of President Clinton's campaign promises was to increase the availability of credit in poor urban areas by establishing 100 community development banks throughout the United States. The banks would ``assist inner-city entrepreneurs,... invest in affordable housing, and ... help mobilize private lenders'' (Washington Post 1992). While this plan has been scaled back since then (Washington Post 1993c), the Clinton administration clearly believes that community banks can both be successful and help alleviate poverty. Hyman Minsky, et al. (1992: 2) also stress the importance of community banks in eliminating what the authors perceive to be a failure in the market for credit in poor, urban areas. Community development banks, they believe, will

deliver credit, payment, and savings opportunities to communities not well served by banks and ... provide financing throughout a designated area for businesses too small to attract the interest of the investment banking and normal commercial banking communities.

The evidence suggests otherwise. Successful community banking projects, both in the United States and abroad, have occurred in communities with relatively homogeneous and geographically immobile borrowers. This has allowed lenders to monitor prospective borrowers to assess their creditworthiness. Monitoring is a substitute for a trouble-free credit history that commercial banks require. In contrast, unsuccessful projects have faced mobile, heterogeneous populations, hindering the ability of lenders to monitor prospective borrowers.

This paper begins by examining existing community banks and related institutions. We highlight the role of information in explaining differences between successful and unsuccessful experiences, and then present a model of borrowing that uses some stylized facts to explain (1) why commercial lenders are reluctant to lend in certain areas and (2) why community lending works in some of those areas but not others. Examining the Clinton proposal in light of this model, along with the work of Minsky et al. (1992), we note that the plan fails to overcome existing informational barriers that make commercial lenders reluctant to operate in certain areas, and is, therefore, unlikely to succeed.

Successful Community Banks

A number of community banking projects have been successful, both in the United States and abroad. In this section we examine some of these projects. We find that the evidence is consistent with our hypothesis that ease of monitoring borrowers, along with a geographically immobile, homogeneous borrowing population are important market characteristics.

The Grameen Bank

This Bangladesh ``bank,'' based in Dhaka, was started in 1976 by Mohammed Yunus in order to give access to capital to the rural poor. The operation targets those who would otherwise be denied credit by the commercial banking sector. Loan amounts are typically small, averaging about $100 in 1985. Repayment rates have averaged about 99 percent within two years of loan disbursement, a higher rate than United States commercial banks typically achieve (World Development Report 1989: 117).

The bank organizes potential borrowers into five-person circles. No collateral is required. Instead, each circle member ``must establish a regular pattern of weekly saving before seeking a loan'' (ibid.). We refer to this as pre-loan monitoring. In addition, ``the first two borrowers must make several weekly payments on their loans before other group members can borrow'' (ibid.). Each circle member agrees to repay the loan of any other member, if necessary. If one member defaults on his or her loan, the others are prohibited from borrowing until the loan is repaid (Wall Street Journal 1993b). The structure--small circles of borrowers with a mutual interest in seeing loans repaid--gives borrowers the ability and incentive to monitor their fellow circle members.[f1] This intra-group monitoring behavior is clearly efficient, and appears again in rotating credit associations, described below. Commercial banks are typically unable to tie the status of one loan to that of another individual--in the United States this would be a form of redlining and would therefore be illegal--and hence are unable to benefit from the practice. Bank staff also closely monitor field operations, in what can be referred to as post-loan monitoring.

In addition to using monitoring to reduce the rate of default, the bank uses the prospect of future loans to promote repayment. Borrowers know that ``the availability of future loans depends on the repayment of borrowed funds'' (World Development Report 1989: 117).

Borrowers in rural Bangladesh are unable to take the money and run. For one thing, the small size of each loan reduces the benefits of deliberately defaulting. Second, living in a small village makes disappearing within the village impossible, unlike living within a large city. Third, family ties make relocating to avoid repayment a costlier decision than in a highly mobile society such as the United States. Fourth, the skills residents of rural areas possess are typically such that rural labor is not substitutable for urban labor, again making default less likely. Thus the combination of monitoring and geographic immobility allows the bank to remain profitable despite the bank's inability to screen out high-risk borrowers as do commercial banks.

Finally, circle members tend to be homogeneous, as does the rural population as a whole. Lending to a homogeneous community narrows the distribution of borrowers as measured by the ex ante probability that a borrower will repay the loan. Thus screening is less important than in highly heterogeneous communities. This benefit, however, accrues to commercial lenders as well as the Grameen Bank, should a commercial bank choose to lend to such a homogeneous community of borrowers.

Rotating Credit Associations

A second type of informal, community ``bank'' that has enjoyed success in a number of forms has been the rotating credit association. These associations often share many of the characteristics that made the Grameen Bank successful: group homogeneity, pre- and post-loan monitoring, and, at times, limited mobility outside the group.

Members of a typical rotating credit association contribute a set amount on a regular basis into a common kitty. The use of the funds rotates among the members until each has had a turn. Then the rotation may begin again or the association may disband. The funds are generally used to start small businesses, although Ivan Light and Edna Bonacich (1988) note that, particularly among the Korean community, saving for weddings and funerals is also common.

Rotating credit associations are useful for purchasing objects that cannot be jointly owned, and cost more than a single member can purchase at one time. Timothy Besley, Stephen Coate, and Glenn Loury (1993) show that, for such goods, rotating credit can be an efficient use of capital. All members of the association except the last one can purchase the object earlier than if they did not belong to an association, while the last member purchases the object at the same time with or without the association. For example, suppose each of 10 members needed $50,000 to start a business and each could save $5,000 per year on his own. Without access to capital through a commercial bank or other lender, members must each wait 10 years to start the business. The opportunity cost of the $50,000 saved in each of the intervening years is quite large. By forming a rotating credit association, the individuals can collectively raise $50,000 each year, starting one member in business every year. Capital never accumulates more than a single year before being put to use, and every member but the last starts his business one or more years before he otherwise would.

To make the system work, however, members who have already had use of the funds must be persuaded to continue contributing toward the fund until each member has had a chance to use the capital. Hence family and community ties are particularly important for the success of the association. Besley, Coate, and Loury (1993: 794) describe the group characteristics that make rotating credit associations successful:

For Roscas [rotating credit associations] to operate successfully it is necessary that individuals keep their commitment to pay into the Rosca after they have won the pot. This may appear problematic since Rosca members are often not able to borrow in conventional credit markets precisely because they cannot be presumed to repay such loans. Roscas circumvent such default problems by exploiting individuals' social connectedness. This is borne out in the anthropological literature, which reveals how the incentive to defect from a Rosca is curbed by social constraints. Roscas are thus typically formed among individuals whose circumstances and characteristics are well known to each other. Defaulters are sanctioned socially as well as being prevented from any further Rosca participation.

Thus social sanctions supplement monitoring to ensure that early users of the fund continue to participate; that is, that the borrowers repay the loan. Having a small group of participants who know one another is particularly useful for this process. Relatives, residents of villages, small towns, and insular communities within larger towns or cities are bound by social contracts, which reduces the risk of default.

One well-documented form of a rotating credit association is the Korean institution of a kye. This practice dates back at least several hundred years and is as much a social institution as a way of raising capital. In the Korean immigrant community in Los Angeles, where capital is difficult to come by since recent immigrants often have no collateral or credit history, the kye has become an important source of capital for small businesses within the community. (Loans from more established immigrants, whether friends or family, also play an important part in this process.)

Members of a kye contribute money, usually weekly, into a common fund which then rotates among the participants. Social pressure to continue payments even after a member has used the funds helps ensure the viability of the kye. Light and Bonacich (1988: 252-53) present a great deal of anecdotal evidence on the prevalence of kyes in the Los Angeles Korean community. They cite an estimate from the Korean Times that there were at least 1,000 kyes in Los Angeles, with funds ranging from $500 to $100,000. Many kyes are used for the development of small businesses; one survey the authors cite found that 70 percent of regular kye practitioners were self-employed or planned to be self-employed in the future.

Besides the use of kyes, Korean immigrants rely on a network of more established friends and relatives to raise funds. Light and Bonacich describe a successful gasoline service station owner who learned how to operate a station by working at a relative's station before purchasing his own. In a telephone survey of Korean entrepreneurs, the authors found that 20.7 percent of firms employed relatives or extended kin, averaging 2.4 persons per firm. Employing friends or relatives in an apprentice-like role before lending money allows the lender to engage in a substantial amount of pre-loan monitoring. In addition, familial ties reduce the cost of post-loan monitoring. Unlike commercial banks, which must rely on documentary evidence that a borrower is a good credit risk, local lenders can observe the work habits and even the entrepreneurial skill of a potential borrower and screen out high-risk borrowers. This is particularly useful for recent immigrants, for whom documentary evidence of the sort that commercial banks would require may not be available.

The Washington Post Magazine (1993) reports an example, on a smaller scale, of patrons in an Arlington, Virginia bar who have started a fund similar to a rotating credit association. Borrowers must be known to the bar regulars, are limited to small loans, and must explain how they will repay the money. No collateral is required, but default rates nevertheless remain low. As with kyes and other forms of local lending, pre-loan monitoring reduces the riskiness of the loans, and social sanctions (being banished from one's favorite bar) promote repayment.

Besley, Coate, and Loury (1993: 793, fn. 3) note that the 19th-century tradition of a barn raising is also a form of rotating credit association. Instead of contributing money, each member contributes his labor, allowing neighbors to raise barns that otherwise would be too difficult for one family to handle. Community pressure helped ensure that neighbors whose barns were raised early in the process would continue to contribute their labor to subsequent raisings.

South Shore Bank

The institution that appears closest in spirit to the Clinton proposal is Chicago's South Shore Bank. Formed in 1973 in an effort to revitalize the South Shore community, the bank makes loans to residents for community projects. The bank started with home mortgages and small business loans, services that commercial banks were no longer willing to supply.

Lending to community residents allows the bank a greater degree of pre- and post-loan monitoring than a commercial bank would have. Once the loan has been made, for a mortgage on a local house or to start a local business, the borrower is tied to the community, strengthening what were presumably existing ties. Combined with community pressure, these ties help raise the probability of repayment.

The Washington Post (1993a: F26) credits the South Shore Bank for revitalizing its community ``by doing what a local bank is supposed to--know its neighborhood and borrowers, assist small businesses, lend to homeowners and show a profit.'' Earlier, the Post (1992: A1), quoting Richard Taub of the University of Chicago, found that the bank's success was in part because the ``bank found a community that still had people who were working and had pride. You need some strengths to build on.''

Monitoring is also important. In a profile of a resident who borrowed from the bank to keep her business going, the Post (ibid.) noted that ``the bank liked her spirit and her neighborhood reputation.'' The article continues,

In its early days, the bank made home mortgages and some business loans--services no longer being offered by anyone else in the neighborhood. At the same time, bank officers and directors talked with tenants, shopkeepers and landlords to find out what the neighborhood needed.

The bank later ``sought out and nurtured several dozen of these budding capitalists.'' Thus monitoring appears to be important to the profitability of the bank. The fact that commercial banks, which cannot monitor borrowers nearly as effectively as a local bank, were unwilling to provide the same services suggests that the expected rate of return to the commercial bank fell below its required level.

Unsuccessful Community Lenders

The previous section described some successful community banks and other lending mechanisms. Examples of unsuccessful community lending projects are harder to come by. Failures, unless in some way spectacular, generally do not attract the notice of the press, and rarely seek media coverage. Nevertheless, there is some evidence that successful community banks are the exception rather than the rule.

Southside Bank and Grand Rapids, Michigan

Inner-city residents of Grand Rapids, Michigan, appear to have much in common with residents of the South Shore community in Chicago. Commercial banks are unwilling to lend into the community, fueling talk of racism. In an effort to revitalize the community, residents are attempting to start the Southside Bank, ``a community institution that will focus on financing development'' in the area by recycling ``deposits from the area into loans to support investment there'' (Wall Street Journal 1993a: A1).

The project, however, is having difficulty attracting sufficient start-up capital. According to the city's acting community development director, drugs--particularly crack--and crime have hampered the bank's efforts. Another difficulty appears to be the way in which the bank is attempting to raise capital. Community officials have expressed disappointment with commercial banks for their reluctance to fund the Southside Bank. The Wall Street Journal (ibid.) says that ``the Grand Rapids debate is over whether to promote development by pushing existing banks to lend more in inner-city areas or to establish special community banks.''

Yet for commercial banks to support either alternative would be surprising.[f2] Commercial banks have already made the business decision that lending into certain communities is not profitable, no doubt in part because the banks are unable to screen attractive candidates from unattractive ones, and in part because the banks have no alternative mechanism for lowering the probability of default, such as monitoring. Simply investing in community banks could be a way for commercial banks to engage in pre- and post-loan monitoring, but at a cost. First, investing in a local bank generates a secondary agency problem. Not only does a commercial bank have to be concerned with actions its borrowers take that might reduce the odds a loan is repaid, but the bank must now be concerned with actions the local bank might take that would be opposed to the investor's interest. Second, commercial banks do look to make profitable loans. The bank will take the upper tail of the distribution of loan applicants in any area (though the size of that tail might be small in inner-city communities). These loans are the most likely to be repaid and hence are the most profitable loans. By investing in a local lender, the commercial bank no longer makes these profitable loans itself and receives only a fraction of the profits when the loan is made through the local bank. Other loans are made that, if profitable, will partly offset the loss of those loans the commercial bank was already making, but because the remaining loans are by definition riskier, the marginal profit on these loans is likely to be smaller.[f3]

Few Other Community Banks Exist

Other evidence of the difficulty of starting and sustaining a community bank comes more from the absence of such banks than specific examples. The Washington Post (1993b) notes that ``not many banks ... have successfully invested in poor communities.'' Other than the South Shore Bank, there are but two community development banks in the United States: the Self-Help Development Bank, Durham, N.C., and the Community Capital Bank, Brooklyn, N.Y.[f4] The Post (1993b: C1) explains why,

Few bankers in the United States have the combination needed--the ability to judge and monitor small and unconventional loans, plus the entrepreneurial bent to encourage businesses and would-be homeowners to expand the investment base in their neighborhoods.

If this explanation is correct, we are unlikely to observe many successful community banks in the United States, and those we do see are likely to be subsidized.

Related Institutions

Beyond the traditional types of lending arrangements are other examples of how monitoring plays an important role in extending credit. In this section we discuss two such examples: the German practice of having bank representation on corporate boards of directors, and the inability of sharecroppers to obtain credit from sources other than their landlords.

The Link between Corporations and Banks in Germany

German banks take a more active role in monitoring their corporate borrowers than do United States banks. It is not unusual to find banks having an equity interest in companies with large lines of credit, or to see representatives from the bank on the company's board of directors. According to the Institutional Investor (1993: 35),

Only about 650 of Germany's nearly 3,000 public limit companies are listed on the stock market ... and the vast majority of these are controlled by families or banks with scant concern for short-term profits.

These dominant shareholders have access to much more complete financial information than do ordinary shareholders reliant on published figures. Dominant shareholders also tend to control the election of the shareholder representatives on supervisory boards. And the power of bank shareholders is strengthened by their additional role as custodians of bearer shares who often act as proxy for other, smaller shareholders. Furthermore, the banks frequently are important lenders--Hausbanken--to the companies they partly own. In other words, concludes Global Proxy's Lufkin, ``They're controlling the borrower to whom they're lending the money.''

One prominent example of this is the Deutsche Bank's 28 percent share in Daimler-Benz. The bank's chairman is also chairman of Daimler's supervisory board. The Deutsche Bank has more than 100 other seats on various corporate boards, including RWE (a Ruhr-based utility company), and Siemens.

Provision of Credit in Interlinked Transactions

A similar market structure is that of credit provision in interlinked transactions. Clive Bell (1988: 797-98) defines an interlinked transaction as one in which the terms of all trades between individuals are jointly determined. For example,

the landlord who finances his tenants' consumption and working capital; the employer ... who gives advances to laborers in return for a claim on their time when he needs them or as part of a contract to employ them for a continuous, extended period; and the trader or commission agent who finances cultivators on the condition that they sell to, or through, him when the crops go to market.

A common arrangement between sharecroppers and landlords is that the tenant borrows money from the landlord to survive during nongrowing seasons, and repays the loan at harvest time. The tenant is unable to obtain credit through other means because of the moral hazard arising from the loan. The tenant's output is sufficiently variable and his effort sufficiently difficult to monitor. Thus, after receiving the loan, effort falls, the probability of default rises, and the loan becomes unprofitable.[f5] Sharecropping internalizes much of the moral hazard problem because the landlord need only monitor output, rather than effort. The tenant receives a fixed proportion of output, leading to an optimal provision of effort (though perhaps at the cost of suboptimally shifting risk from the landlord to the tenant).

The Role of Signals in Community Banking

The previous sections described several successful community banks and similar institutions and one unsuccessful project. Certain market characteristics, such as the ability of local lenders to monitor prospective borrowers or otherwise screen out excessively risky loans, appear pervasive in successful projects but are lacking in the Grand Rapids failure.

This section uses a simple model of banking to formalize the importance of monitoring. We use a variant of the model found in Gruben, Neuberger, and Schmidt (1990). Their concern was with the effects of the Community Reinvestment Act on commercial bank lending in poor neighborhoods. They found that while the CRA had the effect of increasing lending in these neighborhoods, while imposing higher costs of investigating the riskiness of potential borrowers than is optimal for the bank. Our focus is on circumstances under which commercial banks find lending in some neighborhoods to be unprofitable while local lenders may find lending in the same neighborhoods to be either profitable or unprofitable.[f6]

We assume there are two types of neighborhoods: poor (P) and rich (R). A commercial bank already lending in neighborhood R must decide whether to open a branch in neighborhood P. Because the lending market in R is competitive, the equilibrium interest rate i is such that the bank earns zero profits in R after accounting for its cost of funds rd, defaults, and information-gathering costs. Let C_R(I_R) be the cost of gathering information I_R in neighborhood R. Let p_R(I_R) be the probability a loan is repaid, given information I_R. Then competition implies that, in equilibrium,

(1) ¶_R = p_R (I_R) i - C_R (I_R) - r_d = 0.

As I_R increases, p_R and C_R both rise; we assume p_R is concave while C_R is convex in I_R. Thus more information about the neighborhood and prospective borrowers decreases defaults, though at a decreasing rate, while the costs of gathering information increase at an increasing rate.

When considering entry into neighborhood P, the bank is constrained to offering i, but chooses I_P to maximize profits. However, default rates in P are higher than in R because incomes are lower. Given a common level of information I, p_P(I) < p_R(I) for all I. If the bank's information-gathering cost is the same in both neighborhoods, so C_P(o) = C_R(o) = C(o), then it is straightforward to show that

(2) ¶_P = p_P (I_P) i - C (I_P) - r_d < 0,

so the commercial bank will not operate in P.[f7]

A local lender, such as a community bank, faces different costs. Its cost of funds, d_d, is likely to be higher than r_d; as a small, undiversified lender, it faces greater risks than a commercial bank. On the other hand, its information-gathering costs G(I_P) are likely to be lower, so G(I) < C(I) for all I.[f8] In such a case the local bank will be profitable if there exists an I_p such that

(3) ¶¹ = p_P (I_P) i - G (I_P) - d > 0.

If the functions G(o) and C(o) are sufficiently close to one another, the inequality in (3) will not be satisfied, so the community bank will fail. If, however, [G(o) is sufficiently below C(o), so that the local lender has a significant advantage over commercial banks in acquiring information, then (3) may be satisfied. Homogeneity of community members gives rise to economies of scope in gathering information. Borrowers with ties to the community allow the lender to gather information before committing to a loan, and are more likely to remain in the community after receiving a loan. Hence the same characteristics that appear to be present in successful examples of community banking--community homogeneity and low mobility--are those that are necessary in the model to improve monitoring enough to make local lending profitable.

This argument is echoed in Gruben, Neuberger, and Schmidt (1990), and in Besley, Coate, and Loury (1993). The latter note that ethnic groups and recent immigrants in the United States are more likely to be intimidated by banks and discriminated against and hence less likely to have access to external credit markets. Applicants from those groups often have unreliable income streams, which makes them unable to repay loans on a steady basis, or unwilling to repay, and commercial banks lack effective sanctions. As a consequence, commercial banks perceive the default risk to be too high. Rotating credit associations provide societal sanctions which reduce the default rate, and better ex ante information about the reliability of borrowers reduces risks. Besley, Coate, and Loury (1993) describe this as a situation where borrowers offer ``social collateral.''

Similarly, George Borjas (1992) argues that ethnic differences in skills and earnings are transmitted across generations through investments in human capital and what he terms ``neighborhood externalities.'' That is, both family-specific human capital and ``ethnic capital''--the average quality of the ethnic environment--are important factors affecting entrepreneurial skill, work habits, and earning potential. Thus the ability of obtaining a loan may differ across ethnic boundaries. A community bank that is successful in a Korean neighborhood may not be successful in a Hispanic neighborhood because the probability of loan repayment is perceived to be different.

The Clinton Plan Ignores the Conditions Necessary for Success

One of President Clinton's campaign promises was to increase the amount of credit available in poor, urban areas by establishing a system of 100 community banks. While the Clinton administration has scaled back this proposal, it has not disappeared. A variant of the Clinton plan is embodied in the Community Development Banking and Financial Institutions Act of 1993 (HR 3474, approved by the House in November 1993). The act establishes the means to fund community banks in order to increase the amount of capital available in traditionally undercapitalized areas of the country. The bill includes funding of $382 million between fiscal years 1994 and 1997.

Minsky et al. (1992) present a positive view of the Clinton plan. First, the authors argue that there is need for community banks to provide access to credit in poor areas because the private market fails to do so. Second, community banks will be able to provide the services that commercial banks do not, while third, community banks will be profitable, so no ongoing subsidy is needed. HR 3474 also assumes that the banks will be profitable. The act requires that proposals for funds demonstrate that the project will become ``an entity that will not be dependent upon assistance from the Fund for continued viability.''

Leaving aside the issue of whether there exists a market failure,[f9] it is unrealistic to believe that the majority of community banks will be profitable. First, the fact that commercial banks choose not to establish branches in these areas indicates that such branches are not profitable.[f10] Minsky et al. (1992) explain this by asserting that commercial banks find the size of loans and deposits in these areas too small to be profitable. This would appear to require an underlying notion of fixed costs of branch-banking, which may be plausible. However, if so, community banks must somehow be able to provide the same services more cheaply than do commercial banks. Since a commercial bank can, presumably, be of any size, the argument relies on the belief that government-funded institutions are somehow more efficient than privately funded institutions. Based on experience with other government-funded institutions, this is unlikely to be the case.

Second, and related to the first point, is that none of these proposals examines why commercial banks do not find lending in these areas to be profitable. Lenders need some way of assessing the riskiness of a particular loan, and of separating high-risk borrowers from low-risk borrowers. We have argued that situations in which lenders can monitor prospective borrowers can act as a substitute for commercial banks' reliance on credit histories to assess risks. Thus the Grameen Bank, various rotating credit associations, and, to some extent, the South Shore Bank, are all able to achieve low default rates through monitoring.[f11] The conditions necessary for a local lender to engage in pre- and post-loan monitoring include a relatively homogeneous community and relative immobile community members.[f12] These conditions are often lacking in urban areas of the United States, suggesting that areas commercial banks find unprofitable may also be unprofitable to community banks. As a result, the Clinton plan will either require ongoing subsidies to many community banks or will fail to provide additional capital to poor areas in the long run.

If community banks in fact require ongoing subsidies to maintain their viability, then two policy questions become relevant. First, what are the social gains from subsidizing community banks; and second, is subsidization the most efficient way to achieve those gains? Poor areas that gain access to capital may use much of it to revitalize neighborhoods or to start successful businesses where the social payoff exceeds the private return. Reducing poverty may lower the crime rate, reduce welfare rolls, and decrease health-related problems. If so, costs related to crime prevention, incarceration, unemployment compensation, and Medicaid would fall, and society would benefit. But the cost of community banking, already budgeted for $382 million over the next four years, is substantial, and the ongoing costs are likely to drive the total even higher. Without further study, it is not clear that the benefits of subsidization outweigh the costs.

Turning to the second issue, the most efficient way to encourage neighborhood revitalization may not be to start community banks, but to directly subsidize the desired activities. As with the question of society's gain from community banking, this issue becomes irrelevant if community banks are self-sufficient. If continuing funding is necessary, however, it becomes important to address this issue.

Conclusion

Government sponsorship of community banks is becoming a reality. The motivation for this policy is apparently threefold: first, that poor communities, particularly in the inner city, have excess demand for capital and other banking services; second, that commercial banks do not find it profitable to provide such services; and third, that it has been proven that community banks and related institutions can be successful in providing capital to these areas.

However, as we have shown, successful community banks, such as the South Shore Bank and the Grameen Bank, and successful rotating credit associations, such as the Korean kyes, occur in communities with special characteristics. Pre-loan monitoring allows lenders to reduce default risks, while post-loan monitoring, close-knit communities, and a relatively immobile population all help ensure that loans will be repaid. Under these circumstances lenders can ameliorate the adverse selection and moral hazard problems inherent in agency relationships under uncertain information and with heterogeneous agents. Monitoring functions as a substitute for the more conventional use of favorable credit histories as a prerequisite for obtaining a loan.

When the Clinton plan is viewed within this framework, community banking in the United States on a large scale is unlikely to be profitable. Inner cities may need capital, but the monitoring mechanism that works so well in rural Bangladesh is not available in urban America. Minsky et al. (1992) recommend community banks as a means of correcting the market ``failure'' of commercial banks in inner cities, but do not provide a compelling explanation for how the market has failed. Before committing several hundred million dollars for another federal program, we should make certain that we are not merely providing subsidies in disguise.

References

Arnott, R., and Stiglitz, J.E. (1991) ``Moral Hazard and Nonmarket Institutions: Dysfunctional Crowding Out or Peer Monitoring?'' American Economic Review 81 (March): 179-89.

Bell, C. (1988) ``Credit Markets and Interlinked Transactions.'' In Chenery, H., and Srinivasan, T.N. (eds.) Handbook of Development Economics, Vol. 1, 763-830. Amsterdam: Elsevier Science Publishers.

Besley, T.; Coate, S.; and Loury, G. (1993) ``The Economics of Rotating Savings and Credit Associations.'' American Economic Review 83 (September): 792-810.

Borjas, G.J. (1992) ``Ethnic Capital and Intergenerational Mobility.'' Quarterly Journal of Economics 106 (February): 123-50.

Fischer, J.H., and Chami, R. (1995) ``Transfers and Incentives: Some Microeconomic Consequences of German Reunification.'' Journal of Comparative Economics, forthcoming.

Gruben, W.C.; Neuberger, J.A.; and Schmidt, R.H. (1990) ``Imperfect Information and the Community Reinvestment Act.'' Federal Reserve Bank of San Francisco Economic Review (Summer): 27-46.

Institutional Investor (1993) ``Germany Discovers the Foreign Investor,'' July: 34-40.

Jaffee, D.M., and Russell, T. (1976) ``Imperfect Information, Uncertainty, and Credit Rationing.'' Quarterly Journal of Economics 90 (November): 651-66.

Light, I., and Bonacich, E. (1988) Immigrant Entrepreneurs: Koreans in Los Angeles 1965-1982. Berkeley: University of California Press.

Minsky, H.P.; Papadimitriou, D.B.; Phillips, R.J.; and Wray, L.R. (1992) ``Community Development Banks.'' The Jerome Levy Lectures in Economics. Bard College Institute, Annandale-on-Hudson, Working Paper No. 83.

Stiglitz, J.E., and Weiss, A. (1981) ``Credit Rationing in Markets with Imperfect Information.'' American Economic Review 73 (June): 393-410.

Wall Street Journal (1993a) ``Inner-City Capitalists Push to Start a Bank for Their Community.'' 19 January: A1.

Wall Street Journal (1993b) ``Clinton Looks to a Bank in Bangladesh for Model to Help U.S. Poor Get Loans.'' 27 August: A2.

Washington Post (1992) ``Banking on Community Development in Chicago, Clinton Finds a Model to Boost Inner-City Lending.'' 20 December: A1.

Washington Post (1993a) ``Clinton's Risky Plan for Development Banks.'' 16 January: F26.

Washington Post (1993b) ``White House Eyes Community Lending Aid.'' 15 May: C1.

Washington Post (1993c) ``White House Scales Back Community Lending Plan.'' 19 May: F1.

Washington Post Magazine (1993) ``The Loan Arranger.'' 3 October: 8-9.

World Development Report (1989) Washington, D.C.: The World Bank.

Cato Journal, Vol. 14, No. 3 (Winter 1995). Copyright ;cW Cato Institute. All rights reserved.

Ralph Chami is Assistant Professor of Economics at the University of Notre Dame. Jeffrey H. Fischer is Deputy Assistant Director in the Bureau of Economics at the Federal Trade Commission. The authors would like to thank Vandana Chandra for helpful comments. The views expressed in this paper do not necessarily reflect those of the Federal Trade Commission or any individual commissioner.

Notes

[f1]For a model in which monitoring can ameliorate moral hazard, see Arnott and Stiglitz (1991).

[f2]For example, under the Clinton administration, the Civil Rights Division of the Department of Justice has targeted certain banks, such as Chevy Chase Federal Savings Bank in Maryland, for having an insufficient number of branches in poor urban areas. The DOJ's claim is that by operating branches in profitable areas only, banks deny the poor access to inexpensive credit. Residents of poor neighborhoods rely on pawnshops and check-cashing shops to provide loans at interest rates above the rates commercial banks charge.

[f3]Part of the difference may be made up in the ability of the local bank to better monitor its borrowers and hence reduce the riskiness of any particular loan. Whether the net effect is to raise or lower expected profits to the commercial bank is then unclear.

[f4]According to the Wall Street Journal (1993a: A1), the number is five, including the South Shore Bank.

[f5]As a case in point, Jeffrey Fischer and Ralph Chami (1995) demonstrate that the moral hazard problem arising from government transfers in the context of German reunification leads to lower equilibrium effort.

[f6]Dwight Jaffee and Thomas Russell (1976), and Joseph Stiglitz and Andrew Weiss (1981) take a different approach. They ask why banks would set an interest rate and a quantity of loans that generates excess demand for credit. Both papers argue that adverse selection makes the market-clearing interest rate unprofitable. This is because low interest rates cause high-risk borrowers to masquerade as low-risk borrowers, inducing the bank to raise its rate; but high interest rates cause low-risk borrowers to opt out of the loan market, leaving only high-risk borrowers, which is again unprofitable for the bank. Our focus, as well as that of William Gruben, Jonathan Neuberger, and Ronald Schmidt (1990), is not on credit rationing as such, but rather on how informational asymmetries may cause lenders to withdraw from a particular market completely.

[f7]Compare P and R; i and rd are common to both expressions. At IP = IR = I, P - R = i(I) - pR(I)] < 0 by assumption. Suppose IP > IR. Then pP rises but C(IP) rises even faster, so profits are lower still. Suppose IP < IR. If P(IP) > 0 for some IP < IR, then the bank could reduce IR and earn positive profits at i in neighborhood R, contradicting IR as an equilibrium value.

[f8]One might imagine that i would change as well, since the local lender, unlike the commercial bank, does not lend in neighborhood R and is therefore not compelled to charge i in both neighborhoods.

[f9]Economists typically define a market failure as a situation in which a product is not offered despite the existence of a profitable point on the market's demand curve. Transactions costs may prevent the market from reaching this equilibrium. In the case of banking services, the cost of providing these services presumably exceeds the willingness of the poor to pay for the services. Hence while the ``market'' for credit in poor areas may not exist, it is not market failure in the traditional sense.

[f10]Banking regulators often use the Community Reinvestment Act as a stick with which to force commercial banks to establish branches and engage in some lending in poor areas. Regulatory approval for bank activities, such as mergers, is often dependent on how the bank is perceived to have satisfied the requirements of the Community Reinvestment Act. Commercial banks, however, find these branches to be unprofitable, and view them as a cost of doing business in profitable areas.

[f11]Presumably, commercial banks find using credit histories to be a lower-cost method of screening borrowers than monitoring. In the case of people living in poor areas, or recent immigrants, credit histories are often unavailable or unsatisfactory.

[f12]Commercial banks are less concerned with the mobility of their borrowers. Credit reports are nationwide in scope, so a borrower in one area who defaults on a loan cannot simply move to another area and expect to receive future credit.

The Cato Journal is published in the spring/summer, fall, and winter by the Cato Institute, 1000 Massachusetts Ave., NW, Washington, D.C. 20001-5403. The Views expressed by the authors of the articles are their own and are not attributable to the editor, editorial board, or the Cato Institute. Printed copies of the Cato Journal may be ordered by calling 1-800-767-1241. Back issues are also available on the Cato Institute Web site: http://www.cato.org. Email comments or suggestions to subscriptions@cato.org.