Community reinvestment act: review of empirical evidence.
This paper reviews the empirical evidence on the community reinvestment act (CRA) in light of three major different perspectives: (1) The lending market is efficient, (2) The lending market is inefficient due to illegal "discrimination," and (3) The lending market is socially inefficient as caused by "externalities." The empirical evidence is reviewed for the evaluation of the relative merit of these perspectives.
CRA AND EFFICIENT MARKETS DEBATE
The Community Reinvestment Act (CRA) was enacted in 1977 based on the concern that commercial banks and savings associations were engaging in "redlining" practices that were accelerating the decline of many inner-city urban areas. Redlining referred to the practice whereby depository institutions literally or figuratively drew a red line around certain neighborhoods on the basis of the racial composition, age of housing stock, or other factors regardless of the creditworthiness of individual loan applicants, and declined to make loans in those neighborhoods. The perception was that these practices were resulting in the disinvestment and decline of many older, central city, and typically low-income and minority neighborhoods and a shift of jobs to suburban areas. The CRA addressed this problem by requiring the banking regulators to encourage the institutions to help meet the credit needs of the communities in which they are chartered to do business. The hope was that by encouraging depository institutions to look for profitable lending opportunities in their local communities, the CRA would be helpful in revitalizing inner-cities at a time when investment was moving to distant money centers or to more affluent and outlying communities.
The CRA was enacted to curb what was believed to be a lack of adequate lending in low-and moderate-income neighborhoods. It was meant to ensure that bankers did not ignore good lending opportunities to creditworthy borrowers in their communities. The CRA was not intended to force high-risk lending, instead the safety and soundness was to remain the overriding factor in loan decisions.
The debate preceding the enactment of CRA has continued to this date. There are three major different perspectives: (1) The lending market is efficient, (2) The lending market is inefficient due to illegal "discrimination," and (3) The lending market is socially inefficient as caused by "externalities." The efficient markets view regards the CRA as a "tax" on the banking system, whereas the latter two views mostly support the idea that the CRA benefits lenders as well as low-and moderate-income borrowers and their neighborhoods.
According to the efficient markets view, as long as mortgage credit is extended in a competitive manner the market is best suited to determine which lenders and how many are needed to serve the borrowers. A good source of information on the EMH is Shiller (2002) and the references therein.
In the efficient markets view, depository institutions have private incentive to seek all profitable lending opportunities; therefore CRA should have little effect on lending because depository institutions already perform the tasks that the CRA intends to encourage them to do. However, if the CRA forces lenders to make unprofitable loans, then the efficient markets view would regard the CRA as a burden on the banking system. For instance, the CRA may impose substantial compliance costs, such as the costs of training staff to become familiar with the requirements of the CRA and the costs of maintaining records of actions taken to comply with the regulation to be shown to regulators.
The competing views, in turn, note that the CRA itself established and helped maintain the conditions that enabled independent mortgage companies to succeed in reaching low- and moderate-income borrowers and neighborhoods. The CRA accomplished this first by demonstrating that market opportunities existed in serving previously neglected low- and moderate-income borrowers and neighborhoods, and second by enforcing ongoing credit access in low- and moderate-income neighborhoods ensuring that housing markets in these areas remain viable.
The competing views note that as the mortgage lending to low- and moderate-income borrowers and neighborhoods has become more prevalent and more competitive, many lenders introduced products for the CRA-eligible market and used them on a regular basis. Thus, while their CRA lending is mostly extended in their assessment areas, introduction of new products to better serve these areas have likely had positive spillover effects on lending outside of assessment areas, as well as on the lending of non-CRA regulated companies. The fact that many large independent non-CRA mortgage companies have been successful at serving the low- and moderate-income market is an evidence of this process and that a reasonable portion of the CRA-eligible market is being served economically.
The author (2005a, 2005b) provides a review of the literature on the CRA and a review of the theoretical debate on the CRA, respectively.
The empirical evidence is reviewed for the evaluation of the relative merit of the different perspectives. The empirical evidence is divided into two categories: (A) The evidence on discrimination and (B) The evidence on CRA impacts.
Benston (1979), Canner (1982), Canner and Gabriel (1992), Cityscape (1999), Galster (1991, 1992), Goering (1986), Goering and Wienk (1996), Ladd (1998), Lake (1986), Schill and Wachter (1993), and Yinger et al. (1979) discuss aspects of empirical evidence on mortgage lending. Litan, Retsinas, Belsky, and White Haag (2000) in their Appendix B discuss the strengths and limitations of the Home Mortgage Disclosure Act (HMDA) data. The present study has greatly benefitted from Canner and Passmore (1997), Canner, Passmore, Cook, and Kirch (1995), Dahl, Evanoff, and Spivey (2000), Ford Founation (2002), Galster (1992), General Accounting Office (1995), Goering and Wienk (1996), Lacker (1995), Ladd (1998), Laderman (2004), Schafer and Ladd (1981), White Haag (2000), and Yinger (1996).
The legal definition of discrimination is reflected in existing laws against discrimination in mortgage lending, namely the Fair Housing Act of 1968, the Equal Credit Opportunity Act (ECOA) of 1974, and the Civil Rights Act of 1866. The principal law against this type of discrimination, the ECOA, reads in part: "It shall be unlawful for any creditor to discriminate against any applicant, with respect to any aspect of a credit transaction ... on the basis of race, color, religion, national origin, sex or marital status, or age, provided the applicant has the capacity to contract."
The 1977 Community Reinvestment Act defines discrimination in terms of the flow of funds to different types of neighborhoods, as well as in terms of decisions on individual loans. Barth, Cordes, and Yezer (1979) and Bradbury, Case, and Dunham (1989) thoroughly discuss this topic.
In the past, mortgage lenders discriminated against some groups of borrowers based on their policy guidelines. For instance, before the passage of the Equal Credit Opportunity Act in 1974, banks often followed explicit policies to treat women less favorably than men. Lenders often discounted a woman's income by 50 percent or more when evaluating mortgage applications and lenders were more likely to discount the woman's income if she was of a child-bearing age or had pre-school children. Since the Equal Credit Opportunity Act of 1974 prohibited sex-based classifications and income discounting, the change seems to have had a noticeable effect on bank lending policies toward women.
Listokin and Casey (1980) note that discrimination in mortgage lending is the most dramatic type of discrimination, with the most obvious consequence for access to housing. In their defense, lenders stress the hegemony of economic criteria, while community activists argue that lenders are guided by racial as well as economic policies.
The empirical research on discrimination may be divided into three categories (1) Earlier redlining studies, (2) Later redlining studies, (3) Default rate studies, (4) Discrimination in other aspects, and (5) Discrimination by other participants.
Earlier Redlining Studies
These redlining studies use the "old" HMDA data. The 1975 Home Mortgage Disclosure Act (HMDA) initially required a variety of depository institutions to disclose the number and dollar amount of both home mortgage and home improvement loans on a calendar year basis and by census tract or county name. The Act has exempted depository institutions with less than $10 million in assets, or covered by equivalent state laws, or located outside metropolitan statistical areas.
The redlining studies analyze the distribution of lending within metropolitan areas over time, focusing on the relative allocation of credit to whites versus blacks in suburban versus central city neighborhoods. Examples of such studies include: Stegman (1972), Greeston, MacRae, and Pedone (1975), Leven, Little, Nourse, and Read (1976), Hammer, Siler, George Associates (1979), Weicher (1979), Canner (1982), and Morrow-Jones (1986). Evanoff and Segal (1996) examines studies that modeled lending flows at the census tract level: Ahlbrant (1977), Hutchinson, Ostas, and Reed (1977), Avery and Buynak (1981), Bradbury, Case, and Dunham (1989), Schlay (1988, 1989), Perle, Lynch, and Horner (1993), and Holmes and Horvitz (1994).
In these studies, the most common dependent variable is the dollar amount of mortgage lending, and some times standardized by the number of owner-occupied homes in the tract or metropolitan area to control for variations in the level of mortgage demand. The independent variables account for influential factors such as economic (median household income), demographic (shares of population or households classified by race, family type, age of household head, and median household size), housing demand/supply (the number of building permits issued, vacancy rates, and the share of owner occupiers), and mortgage supply (number of branch offices and total amount of deposits). These studies consistently have found that each independent variable influenced mortgage lending flows, and therefore should be part of any analysis attempting to identify the effect of the CRA on these flows. This line of research has not produced conclusive results regarding the influence of the CRA on mortgage lending. Some studies have found negative disparities in mortgage credit flows to areas with lower median incomes and higher minority concentrations, and others have concluded that there was insufficient evidence to support such a claim. The typical result of studies conducted by community groups is that there has been redlining or disinvestments in a particular area or metropolitan center. That is, they find confirmation of the basic premise of the redlining hypothesis that banks curtail the supply of credit to a neighborhood for non-economic reasons such as racial composition.
The most famous analysis of "old" HMDA data is the report entitled "The Color of Money" by Bill Dedman, which was published in a four-part series in May 1988 by the Atlanta Journal--Constitution. The study used the data for Atlanta and found that mortgage loans were made in predominantly white middle-income census tracts at approximately five times the rate in predominantly black middle-income neighborhoods.
Lenders and regulators critically analyzed Dedman's (1988). According to Dedman (1988): "Senior bank executives stated that any lending differences most likely are caused by factors beyond their control, including poor quality housing and lack of home sales in black neighborhoods, fewer applications from blacks, and limitations in federal lending data. They also pointed out that lending patterns are influenced by real estate agents, appraisers, and federal loan programs."
Canner (1982) argues that these studies failed "to incorporate information on mortgage demand, loan risk, and mortgage contract characteristics, which substantially weakens their usefulness for assessing the redlining issues." Others argue that a variety of supply and demand factors should have been incorporated in these studies including lack of demand for mortgage loans in the area; adequate measures of equity; external risks that threatened property value or increased the chance of nonpayment; and decisions by entities outside the control of lenders, such as real estate agents and appraisers. Jaffey (1976) and Listokin and Casey (1980) discuss these issues.
More specifically, they argue that these redlining studies failed to show that supply rather than demand caused the mortgage lending disparities. Many factors that affect the demand for mortgage credit also vary across neighborhoods. These factors include: income and wealth levels, owner-occupancy rates, and housing turnover rates. Moreover, many of these factors are correlated with the racial composition of a neighborhood. The redlining studies fail to control for differences in the demand for credit and only concentrate on the supply of mortgage credit to minority neighborhoods. In partial reply, some sociological analysts, such as Schlay (1989), respond that there can be no ideal measure of the housing demand, since the real or perceived lack of mortgage money from a lender may in turn reduce demand for mortgage credit.
Early HMDA studies also did not analyze the extent to which home purchase credit needs in minority and low- and moderate-income neighborhoods were met by non-HMDA covered lenders. Later Redlining Studies: Financial Institution Reform, Recovery, and Enforcement Act of 1989 (FIRREA) amended the HMDA in a major way. It required that information be reported on the race, gender, and income of mortgage loan applications completed after December 31, 1989. It also required expanded coverage of large mortgage lenders not affiliated with depository institutions, such as mortgage companies.
The new HMDA data permit the calculation of loan acceptance and rejection rates by racial grouping. It resulted in a surge of interest in the analysis of the behavior of mortgage lenders, especially regarding the possibility that lenders discriminated against African Americans, Hispanics, and other minority applicants. The premise of this line of research was that acceptance and rejection of loan applications by minorities can be used as an indicator of discrimination. Some studies, including Canner and Smith (1991, 1992), revealed that the rejection rates for African Americans were twice as high as those for whites.
In response, Galster (1991, 1992), on behalf of the American Bankers Association, issued a report. He concluded: "It must be emphasized that studies based on the new HMDA data will not, and cannot, be definitive about whether any lending patterns represent illegal discrimination. The reasons that this assertion can be made so strongly is that the new data do not include information on certain factors that are critical in credit decisions, including ... characteristics of the property in question and the financial record and condition of the applicant. The data are omitted from HMDA are precisely those that, if taken into account statistically, would narrow (if not completely eliminate) any differences in dispositions of loan applications that might appear to be related to race and/or gender." Benston (1981) and Canner, Gabriel, and Woolley (1991) also discuss this issue.
The major methodological problem that any study of loan acceptance confronts is not having adequate controls. Studies with inadequate controls may conclude that there is discrimination or redlining where in fact there is none or may overstate the magnitude of the discrimination. Galster (1991, 1992) argues that the HMDA studies cannot establish evidence of mortgage credit discrimination without having substantial additional information on the lending or underwriting criteria of the bank; the wealth/debt levels of applicant; the credit history of the loan applicant; and the property serving as collateral.
Proponents of the Efficient Markets Hypothesis (EMH) acknowledge that this line of research tried to remedy the omitted variables problem by incorporating information on the economic characteristics of neighborhoods and individual loan applicants. However, they argue that when such information is taken into account, mortgage loan flows and loan approval rates appear unrelated to neighborhood racial composition. For example, Schill and Wachter (1993) estimated models of banks' loan approval decisions. In their simplest model, loan approval rate is significantly related to the neighborhood racial composition. But when neighborhood characteristics such as median income, vacancy rate, and age of the housing stock are included in the model, neighborhood racial composition is no longer significant. Similarly, Canner, Gabriel, and Woolley (1991) find that after including individual and neighborhood measures of default risk, there is no evidence of discrimination based on the neighborhood racial composition. Several other studies confirm these findings: Avery and Buynak (1981), Black, Schweitzer, and Mandell (1978), Holmes and Horvitz (1994), King (1981), Munnell et al (1992), Schafer and Ladd (1980), Schill and Wachter (1994), Warner and Ingram (1982), and Wiginton (1980). The proponents of the EMH argue that those studies that have reported evidence of discrimination either have limited or no control variables for individual characteristics. For instance, Bradbury, Case, and Dunham (1989) and Calem and Stutzer (1994) use neighborhood-level data, and so do not control for individual economic characteristics. Avery, Beeson, and Sniderman (1996) rely on HMDA data and census tract information, and so are unable to control for applicant wealth or creditworthiness. Similarly, Art (1987) has charged that banks redline older and lower-income neighborhoods. But both the age of the housing stock and borrower income affect the lending risk. The proponents of the EMH thus conclude that the statistical research of the type referred to above has failed to prove that banks discriminate against neighborhoods on the basis of racial composition.
The FIRREA amendments only partly corrected the limitations of the original HMDA statute. This is because the new HMDA data provide no information on the extent of discrimination prior to submission of a written application. That is, no information is available from HMDA data or bank records on the extent to which lenders may actively try to discourage a minority individual from filing an application. This is not to say that the HMDA data are useless. As several studies, including Canner and Gabriel (1992) and Avery, Beeson, and Sniderman (1996), have shown, careful examination of these data provide a variety of insights into lender behavior and possible discrimination. For instance, Avery et al. (1996) show that after controlling for the neighborhood, the lender, and economic characteristics of the applicant, blacks are far more likely than whites to be denied home purchase, refinancing, and home improvement loans. They point out that the black-white differentials for the latter two types of loans are unlikely to be due to omitted credit variables, because all applicants for such loans already received a home purchase loan, thereby demonstrating their creditworthiness. Nevertheless, the possibility of omitted variables cannot be ruled out and these data cannot be used to test statistically hypothesis about discrimination.
The most well-known study in this line of research is the 1992 Federal Reserve Bank of Boston research known as the Boston Fed study on mortgage lending (Munnell et al. 1996, which originally appeared in 1992). This study improved on earlier ones by including a more comprehensive list of applicants and property characteristics and by expanding the coverage to all types of lenders, rather than a restricted set.
Munnell et al. (1996) supplemented the new HMDA data with 38 additional variables to account for the omitted-variable bias. The additional variables were selected based on numerous conversations with academics, lenders, underwriters, and others familiar with the lending process about what they believed were important. Tootell (1996) notes that "It is difficult to find a variable systematically related to the mortgage lending decision and collected by the lender that the Boston Fed did not gather." Their list of controls is more extensive than that in any earlier study. Their data set includes variables in four categories: probability of default, costs of default, loan characteristics, and personal characteristics. (See Table 1).
For their data collection, Munnell et al. (1996) asked banks and mortgage companies for detailed information from the loan applicant files for a sample of Boston HMDA data for 1990. They obtained data on housing expenses, total debt payments, net worth, credit and mortgage payment histories, appraised property values, whether properties were single- or multi-family dwellings, whether applicants were self-employed, and whether applicants were denied private mortgage insurance. Combining this information for a sample of 3,062 individual applicants with applicant race and the unemployment rate in the applicant's industry, they estimated the probability of a particular mortgage loan application being denied.
Munnell et al. (1996) found that the probability of loan denial is 8.2 percentage points higher for blacks and Hispanics than for whites, controlling for the probability of and costs of default and for loan and personal characteristics. This result is statistically significant. Their finding is consistent with the view that lending discrimination is common in housing markets.
Boston Fed study suggests that discrimination takes place in a subtle form. Mortgage loan applicants of all races with good credentials were almost certain to be approved. However, the study found that the vast majority of applicants had some imperfection. Consequently, lenders have had considerable discretion to weigh different factors in evaluating creditworthiness. The Boston Fed researchers suggest that mortgage lenders seem more apt to overlook deficiencies for white applicants than for minority applicants.
Munnell et al. (1996) also estimated various specifications of their model to determine the robustness of their result. These alternative specifications had remarkably little impact on their result that discrimination exists. Carr and Megbolugbe (1993) and Glennon and Stengel (1994) also estimated various model specifications using the Munnell et al. (1996) data. None of these alternative specifications altered the conclusion that minority applicants face discrimination.
The Munnell et al. (1996) study has received much scrutiny. Liebowitz (1993) and Horne (1994) have claimed that the data used for the study are full of errors and that the results are not reliable. Browne (1993) and Browne and Tootell (1995) provide detailed explanations of the data-checking procedures conducted by Munnell et al. (1996), along with a response to Liebowitz (1993). Munnell et al. (1996) and Browne and Tootell (1995) provide a detailed response to Horne (1994). They pointed out that virtually all of the errors discovered by Horne (1994) are either misinterpretations by Horne, are small in magnitude, or are applicable to variables not included in their regression analyses. Moreover, they use a statistical procedure to show that possible (but unverified) errors in the dependent variable (loan rejection) do not influence their results. Moreover, Carr and Megbolugbe (1993) and Glennon and Stengel (1994) provide independent examinations of the Munnell et al. (1996) data that include detailed error-checking procedures. Both of these studies support the principal conclusions of the Munnell et al. (1996) study. Glennon and Stengel (1994) emphasize that some data and specification issues in the Munnell et al. (1996) study cannot be resolved without further research. Overall, Munnell et al. (1996) paid a great deal of attention to their data and no one has provided credible evidence that the results of the study are influenced by data errors.
Zandi (1993) claims that Munnell et al.'s (1996) finding of discrimination is nullified if one more variable is added, namely whether the applicant conforms to the lender's credit guidelines. That is, even though the set of control variables used by Munnell et al. (1996), which includes every factor that lenders mentioned to them, appears to be complete, according to Zandi (1993) perhaps some variables that influence the return on a mortgage and that are correlated with minority status are still omitted. Additional studies with additional control variables would strengthen Zandi's confidence that the Munnell et al. (1996) results are not influenced by omitted-variable bias. As Carr and Megbolugbe (1993) note, lenders who discriminate are likely to rationalize their behavior by claiming that minority applicants do not meet their guidelines. As it appears, Zandi (1993) simply estimated a model in which discrimination appears in his credit variable rather than in the minority status variable.
Lacker (1995) argues that variables in the Boston Fed study measuring creditworthiness are imprecise or incomplete and therefore do not capture completely the judgment of an unbiased loan officer. Measurement error is a serious problem in statistical inference. If true creditworthiness is associated with applicant's race, the model will indicate that race affects the probability of loan denial, even if race plays no direct causal role. In regression analysis, if the true explanatory variable is measured with noise, its regression coefficient will be biased toward zero. In that case, any other variable correlated with the true explanatory variable will be significant in the regression, even though it may play no direct causal role in explaining the behavior in question. Johnston (1963) discusses measurement error and Cain (1986) discusses the implications for detecting discrimination.
Tootel (1996) investigates whether lending patterns in Boston resulted from discriminatory practices based on borrower or neighborhood characteristics (i.e., 'redlining'). He found that lenders were "reluctant to make loans to minorities wherever they apply, and the discrimination is not reflective of a reluctance to extend credit in poor areas that happen to be minority."
Tootell (1996) rules out the possibility that statistical discrimination caused the Boston Fed's results. Lacour-Little (1999) claims that studies relating to these issues, such as Board of Governors of the Federal Reserve System (1993), and Canner and Passmore (1995, 1997), have been inconclusive. Reviews of the most recent literature and audit studies, such as Yinger (1998) and Urban Institute (1999), conclude, however, that whatever motivates market participants, market forces have not yet been sufficient to eradicate mortgage lending discrimination. Moreover, these academic findings are often supported in the statistical evidence presented in court cases that document that despite substantial progress, various forms of discriminatory practices still persist in mortgage and housing markets.
Munnell et al. (1996) is an important study of discrimination in mortgage lending with adequate control variables. It would be valuable to know whether results that hold in Boston also hold in other metropolitan areas or nationwide. Avery, Beeson, and Snidermann (1996) address whether there are similarities in racial differences in lending across the country. They fully use the "new" HMDA data and find racial differences in denial rates across all markets and for all loan types, even after controlling for lender, neighborhood, and applicants' economic characteristics. They note that these differences are not due to property location or neighborhood. Thus, causing them to wonder whether differences in how lenders act toward minorities could be an important explanation.
Overall, Munnell et al. (1996) is an excellent study and has advanced the state of the art considerably, though it is not the last word on this topic. There may be several important methodological issues which have not been addressed by this or any other research on mortgage discrimination, but the Munnell et al. (1996) study establishes a strong presumption that lenders discriminate against blacks and Hispanics in providing mortgage loans.
Endogeneity of Loan-to-Value Ratio and Other Variables
All of the studies on discrimination in loan approval assume that mortgage terms, including loan-to-value (LTV) ratios, are exogenous. Maddala and Trost (1982) and Yezer, Phillips, and Trost (1994) argue that if the assumption does not hold, the estimated coefficients, including the coefficient of the minority status variable, may be biased. Schafer and Ladd (1981), Black and Schweitzer (1985), and Greene, Lovely, and Ondrich (1993) test the assumption and their evidence casts doubt on the assumption. Munnell et al. (1996) respond that after applying the instrumental variable technique to account for the possible endogeneity of the LTV variable and using their data set, the bias was small. They did not, however, give the details of this procedure.
Phillips and Yezer (1996) and Rachlis and Yezer (1993) critique the Boston Fed study because property value, loan amount, loan-to-value ratio, term, down-payment, and use of cosigners, are endogenous in the sense that they are negotiated among the applicant, realtor, and lender throughout the process of determining the acceptability of the final mortgage application.
Turner (1993) discusses another problem which was uncovered in a U.S. Department of Justice investigation of Decatur Federal Savings and Loan in Atlanta in 1992. The Decatur loan officers helped only white applicants to improve their applications (for instance, by guiding applicants to pay off credit card debt). Yinger (1996) discusses similar examples which were uncovered by several pilot studies of lenders' pre-application behavior. The rejection rates based on HMDA data may be a poor measure of true rejection rates because HMDA data do not account for applications that are never filed when loan officers discouraged applications most likely to be rejected. Alternatively, lenders actively market their loans to prospective eligible applicants, which will increase the probability that an application will be accepted. These activities result in the observed rejection rate to be different from the measured rejection rate. Studies of loan approval control for applicant characteristics at the time of the final application, not during the initial inquiry. More formally, the control variables are endogenous, because they are influenced by the lender's coaching. This endogeneity leads to downward bias in the coefficient of the minority status variable. These studies therefore understate discrimination, though the extent of it is unknown.
Default Rate Studies
Several scholars, including Peterson (1981), Van Order, Westin, and Zorn (1993), and Berkovec et al. (1994), have proposed examining loan default rate as an alternative way to test for discrimination in mortgage markets.
Berkovec et al. (1996a, 1996b) claim that if there existed any lending discrimination, then African American applicants would be held to higher standards of creditworthiness than white applicants and would, therefore, default less often than whites. Berkovec et al. (1996a, 1996b) found that African American borrowers default more frequently than do white borrowers, and that their losses are higher than those of whites. The differences were small but statistically significant. Berkovec et al. (1996a, 1996b) concluded that African Americans and whites are treated equally in the default process because economically rational lenders would not discriminate if it were against their economic interest to do so. In a formal discussion of Berkovec et al. (1994), Cappoza (1994) goes so far as to say that the study provides a statistically significant evidence of reverse discrimination. Berkovec et al. (1996a, 1996b) admit the limiting nature of their assumptions and the shortcomings in their data. They also acknowledge that they have no proof that there is no discrimination.
Similarly, Quercia and Stegman (1992) and Berkovec et al. (1994) found that black borrowers have higher default rates than white borrowers. Becker (1993), Brimelow (1993), Brimelow and Spencer (1993) and Roberts (1993) concluded that there cannot be any discrimination against blacks.
Berkovec et al. (1994) provide three reasons that the above conclusion may not follow.
1. Default rate studies underestimate discrimination because they omit key credit variables, such as the applicant's credit history. Moreover, Ross (1995) and Tootell (1993) showed that because only accepted applications are observed, the default rate approach inevitably understates discrimination.
2. Unobserved credit characteristics could itself be a cause of lender discrimination, called statistical discrimination. Lenders cannot observe all credit characteristics for each applicant, but they can observe default rates for different classes of borrower. If default rate for minorities is higher, lenders can lower their risk by using minority status as a signal for poor unobserved credit characteristics. The result will be a higher rejection rate for minority applicants than for white applicants. In this way, a minority-white disparity in unobserved credit characteristics not only undermines the logic of the default rate approach, but it also gives lenders an incentive to practice statistical discrimination.
3. Mortgage discrimination literature assumes that defaults are under the control of the borrower, not the lender. In this literature the terms default and foreclosure are used as synonymous and models of defaults are estimated using data on foreclosures. But as Quercia and Stegman (1992) put it, "although it is the borrower who stops payments, it is the lender who decides if default has occurred by choosing whether to work with the borrower or to foreclose." Lenders' decisions to more aggressively approach foreclosure with minority borrowers could lead to higher observed default rates for these minorities.
Discrimination in Other Aspects
Lenders select the area that their depository institution would serve. According to the Community Reinvestment Act, depository institutions are obligated to help meet the credit needs of their entire service areas. However, in the 1992 Decatur case in Atlanta, for example, the bank had defined its service area to exclude 75 percent of the African American population in one of Atlanta's major counties.
The loan process can be divided into five steps. The first step is advertising and outreach. Lenders use traditional means to advertise loans, such as newspapers and television, as well as post signs on their windows. This implies that the location of their offices is an important element of their advertising programs. Some lenders may try to reach certain segments of the population in their market area. The second step involves the lender's application procedures. This involves how people are treated when they enter the lender's offices to inquire about a loan and whether the application procedures discourage minority applicants. The third step is loan acceptance. This step was considered at length previously. The fourth step, which occurs simultaneously with the third, is the determination of mortgage loan terms. There is some evidence that the LTV ratio is different for equally qualified minority and white borrowers. Discrimination might also exist in the determination of interest rates or loan maturities or loan types. The fifth step is the loan administration. Lenders may be more likely to decide on foreclosure procedures with minority borrowers in default than with comparable white borrowers.
Discrimination by Other Participants
Mortgage lenders are by no means the only participants in the housing and mortgage markets. The other participants are: real estate brokers, appraisers, private mortgage insurers, home insurers, credit reporting agencies, loan holders, and institutions in the secondary mortgage market (e.g., Federal National Mortgage Association (Fannie Mae) and Federal Home Loan Mortgage Corporation (Freddie Mac)). These participants as well practice racial and ethnic discrimination. Galster (1992), National Conference (1994), Fix and Struyk (1993), Heckman and Siegelman (1993), Cloud and Galster (1993) have performed comparable analysis of the roles of these participants in the housing and mortgage markets.
CRA Impact Studies
In response to the inability of the previous research to assess lender behavior with respect to both income and race, this line of research has attempted to detect the effect of the CRA on credit flows, particularly on flows of mortgage credit, by either (1) Comparing lending patterns; (2) Examining the performance of consolidating institutions; (3) Assessing the impact of CRA Agreements; or (4) Comparing institutions' profitability.
Comparing Lending Patterns
One way to analyze the effect of CRA on mortgage lending is to compare the performance of CRA-regulated institutions to those which are not. Since both groups were influenced by general factors but only regulated institutions were influenced by CRA, the comparison has the potential to highlight the CRA's impact on lending patterns.
If CRA has an effect on the practices of CRA-regulated lenders, then CRA lending performance by these institutions should exceed that by non-regulated lenders, assuming that it is possible to control for the product mix and other factors influencing lending to low- and moderate-income borrowers and neighborhoods. Because of both the methodological challenges of operationalizing appropriate controls and data issues, studies in this line of research have generally drawn tentative conclusions about the impact of CRA.
Evanoff and Segal (1996) compare mortgage lending data over the 1990-95 period. They found that CRA loans were an increasing share of originations made by CRA-regulated institutions.
Gunther et al. (1999) compare CRA-regulated and non-regulated lenders' loans for the purchase of one-to-four family homes. They found that non-regulated institutions increased their loan portfolio share of lower- and moderate-income neighborhood from 11 percent in 1993 to 14 percent in 1997. While, CRA lenders' portfolio share of such loans stayed steady over the same period around 11.5 percent. Gunther et al. (1999) also compare loans to lower- and moderate-income borrowers for the two types of institutions. They found that non-CRA lenders' portfolio share of loans to these borrowers rose from 25 percent in 1993 to 32 percent in 1997, while CRA lenders' portfolio share fell from 26 to 25 percent. The authors use these findings to argue that deregulation and technological advances in the financial services industry were more likely than CRA to have been responsible for the increased access to credit by lower- and moderate-income borrowers and neighborhoods. This argument was challenged by Immergluck (1999), however, who discuss methodological problems that the study faced.
Gunther (2000) and Schill and Wachter (1994) also compare mortgage lending between the CRA-regulated and non-regulated institutions. Neither study was able to confirm that the CRA alters mortgage credit allocation within low- and moderate-income or minority neighborhoods. Dahl, Evanoff, and Spivey (2000) point out that a drawback of these studies is that mortgage companies may be influenced by potential regulations even though they are not currently subject to the CRA.
Ford Foundation (2000) study used the Harvard Joint Center for Housing Studies' enhanced HMDA database on home purchase and refinance loans for the period 1993-2000 to carefully control for loan product mix. It applied multivariate models to isolate the effect of CRA while controlling for the influence of other factors such as economic and housing conditions, borrowers' demographic traits, and lender characteristics. The analysis reported the following three key findings:
1. CRA lenders have changed their behavior. CRA lenders originate a higher proportion of CRA-eligible loans than they would if CRA did not exist, and they seem to reject fewer CRA-eligible loan applications than they would if CRA did not exist.
2. CRA lenders appear to have captured a higher share of the CRA-eligible lending market than they would have if CRA were not in place.
3. The study reported that CRA's quantitative impact was small but significant. These findings are consistent with the observation that CRA-regulated institutions continue to lead others in extending prime conventional loans to lower- and moderate-income borrowers and neighborhoods.
Litan, Retsinas, Belsky, Fauth, Kennedy, and Leonard (2001) principal findings support the results of Ford Foundation (2000). Moreover, based on their interviews with experts in four metropolitan areas--Boston, Detroit, Denver and Houston--they reported that lenders, civic leaders, and public officials alike believe that CRA has made a material difference in the behavior of lending institutions and in community credit flows.
Apgar and Duda (2003) analyze the effect of the CRA on regulated lenders by comparing their home purchase lending record with that of other lenders. Their analysis suggests that CRA-regulated institutions continue to lead the market in the provision of prime, conventional residential mortgage loans to low- and moderate-income borrowers and neighborhoods, particularly in terms of their greater outreach to minority. Avery, Calem, and Canner (2003) and Belsky, Schill, and Yezer (2001) find similar results.
Performance of Consolidating Institutions
Another way to assess the impact of CRA is to analyze the mortgage activity of institutions that have been especially active in mergers and acquisitions. The logic behind such an approach is that since regulators evaluate a financial institution's application for merger and acquisition in light of the institution's past performance and adherence with respect to the CRA rules, then consolidating institutions should be especially attentive to CRA-related activities in order to pave the way for regulatory approval.
Avery et al. (1999) found that the proportion of CRA home mortgage loans by consolidating organizations typically increased in the counties in which they had branch offices. Moreover, CRA loans as a share of total home purchase loans increased more among consolidating institutions than among institutions that did not engage in merger activity in the same counties. The authors also found, however, that consolidating institutions lost market share to independent mortgage and finance companies and credit unions. The authors summarize their findings as consistent "with the view that CRA has been effective in encouraging bank organizations, particularly those involved in consolidations, to serve LMI and minority borrowers and neighborhoods."
Dahl, Evanoff, and Spivey (2000) analyze changes in low- and moderate-income mortgage lending for 170 banks that experienced a CRA rating downgrade between 1991 and 1995. They hypothesize that these banks, relative to banks which did not experience a rating downgrade, would increase their quantities of low- and moderate-income lending. This expected response is based on potentially adverse consequences of public criticism and/or possible regulatory denials of applications for future mergers or other organizational changes. However, their empirical results fail to support this hypothesis. The findings are consistent with the contention that during this period regulators stressed adjustments in the lending process of banks (e.g., documentation of lending programs and efforts directed at targeted markets) more than lending performance.
Impact of CRA Agreements
A third way to detect the impact of CRA is to compare mortgage lending in areas covered by CRA agreements with those that are not, and to compare lending by lenders that have signed CRA agreements with those that have not.
Schwartz (1998) looks at the effect of CRA agreement on lender behavior. He compares mortgage and home improvement lending in 1994 by banks with and without CRA agreements. His results indicated that the existence of an agreement positively impacted the bank lending to low- and moderate-income and minority households and neighborhoods, with the most dramatic difference being on lending to black households. Schwartz (1998) also found that those institutions with agreements had higher approval rates for low-income and minority borrowers than institutions that had not entered into such agreements. He did not look, however, at lending behavior before and after signing agreements, and he did not control for other factors that may have generated the patterns he observed. For example, it is possible that the decision to sign an agreement is endogenous. That is, lenders with greater ability to meet CRA obligations sign agreements, in effect taking credit for actions they would have undertaken anyway.
Schlay (1999) found that lending increased to low- and moderate-income and minority borrowers and neighborhoods in all cities examined, suggesting that the extent of CRA organizing in a particular city is not necessarily predictive of the increase in lending to low- and moderate-income borrowers and neighborhoods. She notes that 'regulation from below,' in the form of CRA-organizing, may have created CRA enforcement at the federal level that affected all markets.
Bostic (2001) looks for the effect of signing an agreement on overall lending in the counties in which the participating lender operated branches. Bostic (2001) finds that the number of newly-initiated CRA agreements in a county is significantly associated with 3-year changes in conventional mortgage lending, particularly in lending to low- and moderate-income and minority borrowers and neighborhoods. He also found, however, that these effects do not persist over time, disappearing almost completely after 3 years. Bostic (2001) concludes that the effectiveness of CRA agreements in increasing lending activity is ultimately determined by the persistence of sophistication of community groups in monitoring compliance with CRA agreements.
The logic of this line of research is to determine if the CRA has caused significant differences between the profitability of CRA-related lending and other lending for a given product type.
For proponents of the efficient markets hypothesis, the CRA is a threat to lender profitability. For others, the CRA has the potential to increase profitability. This is because the two groups believe respectively that: (1) the CRA, by forcing financial institutions to make unprofitable loans, hurts profitability, or (2) the CRA, by directing lenders to potentially overlook markets, helps profitability. Therefore, by examining the relative profitability of financial institutions, this line of research may also be used to judge the appropriateness of the two competing views.
Research on the profitability of CRA-related lending has followed two approaches. First, some researchers have focused on overall institutional profitability. Examples in this line of research are: Board of Governors of the Federal Reserve Board (1993), Canner and Passmore (1997), Malmquist, Phillips-Patrick, and Rossi (1997). These studies have generally found that banking institutions that are relatively more active in CRA-related lending activities have levels of profitability that are not significantly different from those of other institutions.
The second approach has been to survey banking institutions directly to obtain information about the profitability of their CRA-related and other lending activities. The results of this approach have been mixed. Harrison (1999) and Meeker and Myers (1996) suggest that CRA-related mortgage lending is profitable for most institutions, but not as profitable as other traditional mortgage lending. Board of Governors of the Federal Reserve Board (2000), Canner, Laderman, Lehnert, and Passmore (2002) and Canner and Passmore (1997) suggest that for most banks the profitability of LMI home purchase lending has become comparable to that of other home purchase lending.
Johnson and Sarkar (1996) examine the effect of passage of the CRA on the stock market valuations of banks and savings and loan associations. They conclude that the expected costs of the CRA exceed the expected benefits among small institutions only.
Taken together, the results of existing studies are mixed, largely because each had methodological limitations that limit their broad generalization. Nevertheless, the studies consistently find suggestion of a CRA effect, but one that their authors cannot definitively document, meaning that the debate on CRA's usefulness continues to simmer.
This paper reviews the empirical evidence on the community reinvestment act (CRA) in light of three major different perspectives: (1) The lending market is efficient, (2) The lending market is inefficient due to illegal "discrimination," and (3) The lending market is socially inefficient as caused by "externalities." The empirical evidence is reviewed for the evaluation of the relative merit of these perspectives. The empirical evidence was divided into two categories: (1) The evidence on discrimination and (2) The evidence on CRA impacts.
As Caskey (1994), Listokin and Casey (1980), and Schill and Wachter (1993) have also noted, the existing empirical research reveals that lender behavior has many facets and that it occurs in a complex environment. Several excellent studies have addressed some important aspects of lender discrimination, but the complexity of the phenomenon has led them to a number of methodological obstacles.
The debate preceding the enactment of CRA has continued to this date. The evidence on efficient markets in bank lending seems inconclusive. The controversy will continue for the foreseeable future. An observer with a strong prior belief in the ability of market forces to take advantage of profitable opportunities can easily remain unconvinced by the evidence on the effects of the CRA. On the other hand, an observer with a strong prior belief in the prevalence of market failures in lending will find striking confirmation in such evidence. Between these two extremes lie a range of reasonable assessments.
CRA-regulated lenders refer to federally regulated banks and thrifts as well as their mortgage company and finance company affiliates. CRA-eligible loans refer to loans made to lower-income households and/or households living in lower-income areas. Assessment areas refers to areas where they maintain deposit taking operations.
Lower-income borrowers are defined as having incomes less than 80 percent of metropolitan area median income, and lower-income communities are census tracts with 1990 median family income that was less than 80 percent of their metropolitan area median.
Low- and moderate-income (LMI) borrowers, and borrowers who live in LMI areas, and loans to such persons, are referred to as "CRA-eligible borrowers," "CRA-eligible loans," or "CRA-eligible lending", respectively.
The term minority is used to mean African American or Hispanic and the term white is used to mean non-Hispanic white, even though many Hispanic Americans have European or white racial backgrounds. Black is used as a synonym for African American.
Ahlbrant, Jr., R.S. (1977). Exploratory Research on the Redlining Phenomenon. AREUEA Journal, 5, 473-481.
Apgar, W.C. & Duda, M. (2003). The Twenty-Fifth Anniversary of the Community Reinvestment Act: Past Accomplishments and Future Regulatory Challenges. Economic Policy Review, Federal Reserve Bank of New York, Policy Review, June, 169-191.
Ardalan, K. (2005a). Community Reinvestment Act and Efficient Markets Debate: Overview. Under Review by an Academic Journal.
Ardalan, K. (2005b). Community Reinvestment Act and Efficient Markets Debate: Review of Theory. Under Review by an Academic Journal.
Art, R.C. (1987). Social Responsibility in Bank Credit Decisions: The Community Reinvestment Act one Decade Later. Pacific Law Journal, 18, 1071-1139.
Avery, R.B., P.E. Beeson & M.S. Sniderman (1996). Accounting for Racial Differences in Housing Credit Markets. Chapter 3 in J. Goering & R. Wienk (Eds.), Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC, Urban Institute Press.
Avery, R.B., R.W. Bostic, P.S. Calem, & G.B. Canner (1999). Trends in Home Purchase Lending: Consolidation and the Community Reinvestment Act. Federal Reserve Bulletin, 85, 81-102.
Avery, R.B. & T.M. Buynak (1981). Mortgage Redlining: Some New Evidence. Federal Reserve Bank of Cleveland Economic Review, Summer, 18-32.
Avery, R.B., P. Calem & G. Canner (2003). The Effects of the Community Reinvestment Act on Local Communities. Board of Governors of the Federal Reserve System.
Barth, J.B., J.J. Cordes & A.M.J. Yezer (1979). "Financial Institution Regulations, Redlining and Mortgage Markets" Proceedings of a Conference on the Regulation of Financial Institutions, Federal Reserve Bank of Boston Conference Series No. 21, October, 101-143.
Becker, G.S. (1993). The Evidence against Banks Doesn't Prove Bias. Business Week, 19, 18.
Benston, G.J. (1979). Redlining Research: A Review and Critical Analysis Discussion. Proceedings of a Conference on the Regulation of Financial Institutions, Federal Reserve Bank of Boston Conference Series No. 21, October, 144-195.
Benston, G.J. (1981). Mortgage Redlining Research: A Review and Critical Analysis. Journal of Bank Research, 12, 8-23.
Berkovec, J.A., G.B. Canner, S.A. Gabriel & T.H. Hannan (1994). Race, Redlining, and Residential Mortgage Loan Performance. Journal of Real Estate Finance and Economics, 9, 263-294.
Berkovec, J.A., G.B. Canner, S.A. Gabriel & T.H. Hannan (1996a). Mortgage Discrimination and FHA Loan Performance. Chapter 7 in J. Goering & R. Wienk (Eds.), Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC: The Urban Institute Press.
Berkovec, J.A., G.B. Canner, T.H. Hannan & S.A. Gabriel. Race, Redlining, and Residential Mortgage Defaults: Evidence from the FHA-Insured Single-Family Loan Program. Chapter 6 in J. Goering & R. Wienk (Eds.), Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC: The Urban Institute Press.
Black, H.A. & R.L. Schweitzer (1985). A Canonical Analysis of Mortgage Lending Terms: Testing for Lending Discrimination at a Commercial Bank. Urban Studies, 22, 13-19.
Black, H.A., R.L. Schweitzer & L. Mandell (1978). Discrimination in Mortgage Lending. American Economic Review, 68, 186-191.
Board of Governors of the Federal Reserve System (1993). Report to the Congress on Community Development Lending by Depository Institutions. Washington, DC: Federal Reserve Board, October.
Board of Governors of the Federal Reserve System (2000). The Performance and Profitability of CRA-Related Lending. Washington, DC: Federal Reserve Board, July.
Bostic, R.W. & B. Robinson (2001). Do CRA Agreements Influence Lending Patterns? Working Paper for the Board of Governors of the Federal Reserve System.
Bradbury, K.L., K.E. Case & C.R. Dunham (1989). Geographic Patterns of Mortgage Lending in Boston, 1982-1987. New England Economic Review, September/October, 3-30.
Brimelow, P. (1993). Racism at Work? National Review, 12, 42.
Brimelow, P. & L. Spencer (1993). The Hidden Clue. Forbes, 4, 48.
Browne, L.E. (1993). Boston Fed Study Shows Race Bias. Letter to the Editor, Wall Street Journal, September, 21, A23.
Browne, L.E. & G.M.B. Tootell (1995). Mortgage Lending in Boston--A Response to the Critics. New England Journal of Economic Review, September/October, 53-78.
Caine, G.G. (1986). The Economic Analysis of Labor Market Discrimination: A Survey. In E.O. Ashenfelter & R. Layard (Eds.), Handbook of Labor Economics, Vol. 1. (pp. 693-785). New York: Elsevier Science Publishers.
Calem, P. & M. Stutzer (1994). The Simple Analytics of Observed Discrimination in Credit Markets. Manuscript, Federal Reserve Bank of Philadelphia, June 7.
Canner, G.B. (1982). Redlining: Research and Federal Legislative Response. Staff Study, 121, Washington, DC: Board of Governors of the Federal Reserve System.
Canner, G.B. & S.A. Gabriel (1992). Market Segmentation and Lender Specialization in the Primary and Secondary Mortgage Markets. Housing Policy Debate, 3(2), 241-329.
Canner, G.B., S.A. Gabriel & J.M. Woolley (1991). Race, Default Risk and Mortgage Lending: A Study of the FHA and Conventional Loan Markets. Southern Economic Journal, 58, July, 249-262.
Canner, G., E. Laderman, A. Lehnert & W. Passmore (2002). Does the Community Reinvestment Act (CRA) Cause Banks to Provide a Subsidy to Some Mortgage Borrowers? Finance and Economic Discussion Series, 2002-19, Board of Governors of the Federal Reserve System.
Canner, G.B. & W. Passmore (1995). Implementing CRA: What is the Target? Proceedings of the 31st Annual Conference on Bank Structure and Competition, May, 171-191.
Canner, G. & W. Passmore (1997). The Community Reinvestment Act and the Profitability of Mortgage-Oriented Banks. Federal Reserve Board, March.
Canner, G.B., W. Passmore, W. Cook & L. Kirch, Lisa (1995). Home Purchase Lending in Low-Income Neighborhoods and to Low-Income Borrowers. Federal Reserve Bulletin, 81(2).
Canner, G.B. & D.S. Smith (1991). Home Mortgage Disclosure Act: Expanded Data on Residential Lending. Federal Reserve Bulletin, 77, November, 859-884.
Canner, G.B. & D.S. Smith (1992). Expanded HMDA Data on Residential Lending: One Year Later. Federal Reserve Bulletin, 78, November, 801-824.
Cappoza, D.R. (1994). Comments on 'Race, Redlining, and Residential Mortgage Loan Performance'. Journal of Real Estate Finance and Economics, 9, November, 295-298.
Carr, J.H. & I.F. Megbolugbe (1993). The Federal Reserve Bank of Boston Study of Mortgage Lending Revisited. Journal of Housing Research, 4(2), 277-313.
Caskey, J. (1994). Bank Representation in Low-Income and Minority Urban Communities. Urban Affairs Quarterly, 29, June, 617-638.
Cityscape (1999). Commemorating the 30th Anniversary of the Fair Housing Act.
Dahl, D., D.D. Evanoff & M.F. Spivey (2000). Does the Community Reinvestment Act Influence Lending? An Analysis of Changes in Bank Low-Income Mortgage Activity. Federal Reserve Bank of Chicago, Working Paper No. WP 200006, May.
Dedman, B. (1988). The Color of Money: Home Mortgage Practices Discriminate against Blacks. Atlanta Journal--Constitution, May, 1-16.
Evanoff, D.D. & Segal, L.M. (1996). CRA and Fair Lending Regulations: Resulting Trends in Mortgage Lending. Federal Reserve Bank of Chicago Economic Perspective, 20(6), 19-46.
Ford Founation (2002). The 25th Anniversary of the Community Reinvestment Act: Access to Capital in an Evolving Financial Services System. Joint Center for Housing Studies, Harvard University, March.
Galster, G.C. (1991). Housing Discrimination and the Urban Poverty of African Americans. Journal of Housing Research, 2, 1-36.
Galster, G.C. (1992). Research on Discrimination in Housing and Mortgage Markets: Assessment and Future Directions. Housing Policy Debate, 3(2), 639-683.
General Accounting Office (1995). Community Reinvestment Act: Challenges Remain to Successfully Implement CRA. Report to Congressional Requesters, November.
Glennon, D. & M. Stengel (1994). An Evaluation of the Federal Reserve Bank of Boston's Study of Racial Discrimination in Mortgage Lending. Economic and Policy Analysis Working Paper 94-2, Comptroller of the Currency, Washington, DC.
Goering, J. (1986). Introduction. In J. Goering (Ed.), Housing Desegregation and Federal Policy, (pp. 1-8). Chapel Hill, North Carolina: University of North Carolina Press.
Goering, J. & R. Wienk (Eds.) (1996), Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC: The Urban Institute Press.
Goering, J. & R. Wienk (1996). An Overview. Chapter 1 in J. Goering & R. Wienk (eds.), Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC: The Urban Institute Press.
Greene, V.L., M.E. Lovely & J.I. Ondrich (1993). Do Community-Based Long-Term Care Services Reduce Nursing Home Use. Journal of Human Resources, 28, Spring, 297-317.
Greeston, P., C.D. MacRae & C. Pedone (1975). The Effects of FHA Activity in Older, Urban, Declining Areas: A Review of Existing, Related Analyses. Urban Institute Report No. 220-1, Washington, DC: Urban Institute.
Gunther, J. (2000). Should CRA Stand for 'Community Redundancy Act. Regulation, 23(3), 56-60.
Gunther, J.W., K. Klemme, K. & K.J. Robinson (1999). Redlining or Red Herring? Southwest Economy, May/June, 8-13.
Hammer, S. & G. Associates (1979). The Role of the Real Estate Sector in Neighborhood Change. Final report prepared for Office of Policy Development and Research, Pub. No. HUD-PDR-380, Washington, DC: U.S. Department of Housing and Urban Development.
Harrison, D. (1999). Bank Specific Characteristics as a determinant of Community Reinvestment Act (CRA) Performance. Working Paper, University of Vermont.
Holmes, A. & P. Horvitz (1994). Mortgage Redlining: Race, Risk, and Demand. Journal of Finance, 49, March, 81-99.
Horne, D.K. (1994). Evaluating the Role of Race in Mortgage Lending. FDIC Banking Review, 28, Spring, 297-317.
Hutchinson, P.M., J.R. Ostas & J.D. Reed (1977). A Survey and Comparison of Redlining Influences in Urban Mortgage Lending Markets. AREUEA Journal, 5, 463-472.
Immergluck, D. (1999). Unfinished Business: Increases in African-American Home Buying and Continuing Residential Segregation in the Chicago Region. Chicago: Woodstock Institute.
Jaffey, D. (1976). Credit for Financing Housing Investment: Risk Factors and Capital Markets. In Housing in the Seventies
Working Papers (491-511), National Housing Policy Review, Washington, DC: U.S. Government Printing Office.
Johnson, S.A. & S.K. Sarkar (1996). The Valuation Effects of the 1977 Community Reinvestment Act and Its Enforcement. Journal of Banking and Finance, 20.
Johnston, J. (1963). Econometric Methods. New York: McGraw-Hill.
King, A.T. (1981). Discrimination in Mortgage Lending: A Study of Three Cities. New York University, Solomon Brothers Center for the Study of Financial Institutions Monograph Series in Finance and Economics.
Lacker, J.M. (1995). Neighborhoods and Banking. Economic Quarterly, Federal Reserve Bank of Richmond, 81(2), 13-38.
LaCour-Little, M. (1999). Discrimination in Mortgage Lending. Journal of Real Estate Literature, 7(1), 15-49.
Ladd, H.F. (1998). Evidence on Discrimination in Mortgage Lending. Journal of Economic Perspectives, 12(2), 41-62.
Laderman, L. (2004). Has the CRA Increased Lending for Low-Income Home Purchases? Federal Reserve Bank of San Francisco Economic Letter, No. 2004.
Lake, R. (1986). Postscript: Unresolved Themes in the Evolution of Fair Housing. In J. Goering (Ed.), Housing Desegregation and Federal Policy (pp. 313-326). Chapel Hill, NC: University of North Carolina Press.
Leven, C., J. Little, H. Nourse & R. Read (1976). Neighborhood Change: Lessons in the Dynamics of Urban Decay. New York: Praeger.
Liebowitz, S. (1993). A Study that Deserves No Credit. Wall Street Journal, September 1, A14.
Listokin, D. & S. Casey (1980). Mortgage Lending and Race: Conceptual and Analytic Perspectives on the Urban Financing Problem. New Brunswick, New Jersey: Rutgers University Press.
Litan, R.E., N.P. Retsinas, E.S. Belsky, G. Fauth, M. Kennedy & P. Leonard (2001). The Community Reinvestment Act after Financial Modernization: A Final Report. Department of Treasury.
Litan, R.E., N.P. Retsinas, E.S. Belsky & S. White Haag, Susan (2000). The Community Reinvestment Act After Financial Modernization: A Baseline Report. Department of Treasury.
Maddala, G.S. & R. P. Trost (1982). On Measuring Discrimination in Loan Markets. Housing Finance Review, 1, 245-268.
Malmquist, D., F. Phillips-Patrick, C. Rossi (1997). The Economics of Low-Income Mortgage Lending. Journal of Financial Services Research, 11, 169-188.
Meeker, L. & F. Myers (1996). Community Reinvestment Act Lending: Is It Profitable? Federal Reserve Board of Kansas City Financial Industry Perspectives, 13-35.
Morrow-Jones, H. (1986). Neighborhood Change and the Federal Housing Administration: Some Theoretical and Empirical Issues. Urban Studies, 23, October, 419-428.
Munnell, A.H., L.E. Browne, J. McEneaney, & G.M.B. Tootell (1996). Mortgage Lending in Boston: Interpreting HMDA Data. Federal Reserve Bank of Boston Working Paper 92-07, October, 1992, American Economic Review.
National Center for Health Statistics (1994). Health, Unites States, 1993. Hyattsville, MD: Public Health Service.
Perle, E.D., K. Lynch & J. Horner (1993). Model Specification and Local Mortgage Market Behavior. Journal of Housing Research, 4(2), 225-243.
Peterson, R.L. (1981). An Investigation of Sex Discrimination in Commercial Banks' Direct Consumer Lending. Bell Journal of Economics, 12, Autumn, 547-561.
Phillips, R.F. & A.M. Yezer (1996). Self Selection and Tests for Bias and Risk in Mortgage Lending: Can You Price the Mortgage if You Don't Know the Risk. Journal of Real Estate Research, 11(1), 87-102.
Quercia, R.G. & M.A. Stegman (1992). Residential Mortgage Default: A Review of the Literature. Journal of Housing Research, 3(2), 341-379.
Rachlis, M.B. & A.M. Yezer (1993). Serious Flaws in Statistical Tests for Discrimination in Mortgage Lending. Journal of Housing Research, 4(2), 315-336.
Roberts, P.C. (1993). Banks in the Line of Fire. Journal of Housing Research, 3(2), 341-379.
Ross, S. (1995). Mortgage Lending Discrimination and Racial Differences in Loan Performance. Working Paper 95-131, Department of Economics, University of Connecticut, Storrs.
Schaffer, R. & H.F. Ladd (1980). Equal Credit Opportunity Accessibility to Mortgage Funds by Women and Minorities. Final report prepared for Office of Policy Development and Research, Pub. No. HUD-PDR-551-1. Washington, DC: Department of Housing and Urban Development.
Schaffer, R. & H.F. Ladd (1981). Discrimination in Mortgage Lending. Cambridge, Mass.: MIT Press.
Schill, M.H. & S.M. Wachter (1993). A Tale of Two Cities: Racial and Ethnic Geographic Disparities in Home Mortgage Lending in Boston and Philadelphia. Journal of Housing Research, 4(2), 245-275.
Schill, M.H. & S.M. Wachter (1994). Borrower and Neighborhood Racial and Income Characteristics and Financial Institution Mortgage Application Screening. Journal of Real Estate Finance and Economics, 9, December, 223-239.
Schlay, A.B. (1988). Not in That Neighborhood: The Effects of Population and Housing on the Distribution of Mortgage Finance within the Chicago SMSA. Social Science Research, 17, 137-163.
Schlay, A.B. (1989). Financing Community: Methods for Assessing Residential Credit Disparities, Market Barriers, and Institutional Reinvestment Performance in the Metropolis. Journal of Urban Affairs, 11(3), 201-223.
Schlay, A.B. (1999). Influencing Agents of Urban Structure: Evaluating the Effects of Community Reinvestment Organizing on Bank Residential Lending Practices. Urban Affairs Review, 35(2), 247-278.
Schwartz, A. (1998). Bank Lending to Minority an Low-Income Households and Neighborhoods: Do Community Reinvestment Agreements Make a Difference? Journal of Urban Affairs, 20(3), 269-301.
Shiller, R. (2002). From Efficient Market Theory to Behavioral Finance. Cowles Foundation Discussion Paper No. 1385, October.
Stegman, M. (1972). Housing Investment in the Inner City. Cambridge, Mass.: MIT Press.
Tootell, G. (1993). Defaults, Denials, and Discrimination in Mortgage Lending. New England Economic Review, September/October, 45-51.
Tootell, G. (1996). Turning a Critical Eye on the Critics. In J. Goering & R. Wienk, Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC: The Urban Institute Press.
Urban Institute (1999). A Study of the GSE's Single Family Underwriting Guidelines, Report for the Office of Policy Development and Research. U.S. Department of Housing and Urban Development.
van Order, R., A.M. Westin & P. Zorn (1993). Effects of the Racial Composition of Neighborhoods on Default, and Implications for Racial Discrimination in Mortgage Markets. Paper presented at meetings of American Economics Association, January 8, Anaheim, California.
Warner, A. & J. Ingram (1982). A Test of Discrimination in a Mortgage Market. Journal of Bank Research, 13, 116-124.
Weicher, J. (1979). Government Urban Policy and the Lender. Urban Institute Report 252-0000-1-7. Washington, DC: Urban Institute.
White Haag, S. (2000). Community Reinvestment and Cities: A Literature Review of CRA's Impact and Future. The Brookings Institution, Center on Urban and Metropolitan Policy, March.
Wiginton, J. (1980). A Note on the Comparison of Logit and Discriminant Models of Consumer Credit Behavior. Journal of Financial and Quantitative Analysis, 15, 757-770.
Yezer, A.M.J., R.F. Phillips & R.P. Trost, Robert P. (1994). Bias in Estimates of Discrimination and Default in Mortgage Lending: The Effects of Simultaneity and Self-Selection. Journal of Real Estate Finance and Economics, 9(3), November, 197-216.
Yinger, J. (1996). Discrimination in Mortgage Lending: A Literature Review. Chapter 2 in J. Goering & R. Wienk (Eds.), Mortgage Lending, Racial Discrimination, and Federal Policy. Washington, DC: The Urban Institute Press.
Yinger, J. (1998). Evidence on Discrimination in Consumer Markets. Journal of Economic Perspectives, 12(23), 23-40.
Yinger, J., G. Galster, B. Smith & F. Eggers (1979). Status of Research into Racial Discrimination and Segregation in American Market. U.S. Department of Housing and Urban Development Occasional Papers, 6, 55-175.
Zandi, M. (1993). Boston Fed's Bias Study Was Deeply Flawed. American Banker, 19, 13.
Kavous Ardalan, Marist College
Table 1: Control Variables Used by Munnell et al. (1996) Probability of Default Housing expense/income Total debt payments/income Net wealth Consumer credit history Mortgage credit history Public record history Probability of unemployment Self-employed Loan/approval value--low Loan/approval value--medium Loan/approval value--high Costs of Default Denied private mortgage insurance Census tract dummies Loan Characteristics Two- to four-family house Lender ID Personal Characteristics Race Source: Munnell et al. (1996, Table 2)
|Printer friendly Cite/link Email Feedback|
|Publication:||Academy of Banking Studies Journal|
|Date:||Jan 1, 2006|
|Previous Article:||Comparing credit unions with commercial banks: implications for public policy.|
|Next Article:||Letter from the editor.|