Ll ELEV-RY Michégan State University PLACE'IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2/05 amfimts \ >A—m__.___.'..__.....___.._ .m. 1 '-. . , “affix v" - — Plan B Paper __ “.5wa 3. . —_ lender, Kira To UES'K‘IO AN ANALYSIS OF THE COMMUNITY REINVESTMENT ACT AND THE HOME MORTGAGE DISCLOSURE ACT IN IDENTIFYING HOME MORTGAGE DISCRIMINATION AND PROMOTING COMMUNITY DEVELOPMENT LENDING. BY Kira T. Zender A PLAN B PAPER Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF URBAN AND REGIONAL PLANNING Department of Geography Urban and Regional Planning Program May 7, 1994 ---.. , ] ggygume MArr-‘mts: 1T“ ACKNOWLEDGEMENTS I would like to thank Courtney Dufries and Hank Helton of the Federal Reserve Bank of Atlanta, Joan Rogers of the Community Research and Development Group, and Terry Farris of Michigan State University for their assistance in making the paper possible. *rfl- AUTHOR’S NOTE The thoughts and views expressed in this paper are those of the author and do not reflect the views of the Federal Reserve Bank of Atlanta. TABLE OF CONTENTS: INTRODUCTION .........................................p.1 BRIEF HISTORY OF COMMUNITY INVESTMENT LAWS.......p.3 DEFINITION OF DISCRIMINATION.....................p.4 PREMISE OF PAPER.................................p.6 PAPER OUTLINE....................................p.7 HOME MORTGAGE DISCLOSURE ACT..........................p.S PURPOSE........... ..................... . ......... p.8 REQUIREMENTS........... ............... . ......... .p.8 AVAILABILITY OF DATA.............................p.9 HMDA USERS.......................................p.9 EARLY HMDA STUDIES...............................p.10 a t ourn Con t tu ' ................p.11 Detroit Free Press.... ..................... p. 11 Federal Reserve Bank of Boston 1989 Study...p.12 ANALYSIS OF 1989 STUDYOCOODOOCOOOOOOO0.0.0.0....Cp013 Differences in Demand.......................p.13 Lack of Demand/Differences in Communities...p.14 Aggregation of Data ..... ...... .............. p.15 Differences in Income.......................p.16 Discrimination in Housing Markets ........... p.16 CONCLUSIONS FROM EARLY STUDIES...................p.16 FIRREA AND FDICIA AMENDMENTS TO HMDA..................p.17 INTRODUCTION ........ . ................ ............p.17 A New Focus for Determining Discrimination..p.17 Amendments under FDICIAOOOOOOOOOOOOOOOOOOOOOPOIB Amendments Under FIRREA ..................... p.18 FINDINGS FROM EXPANDED HMDA RESULTS ............ ..p.19 General Information............. ....... .....p.20 Significant Findings. 0 O O O O O O O O O O O O O O O O O O O I 0 Op. 20 Similar Findings for 1992.... ..... ..........p.22 IS HMDA DATA SUFFICIENT EVIDENCE OF DISCRIMINATION?. ............... .p.22 BASIC MORTGAGE APPLICATION PROCESS......... ..... .p.23 Verification and Evaluation of Ratios ....... p.24 Obligation Ratios O O O O O O O O 0 O O O O O O 0 O O O O O O O O O O O p O 24 Evaluation of Numbers and Compensating Factors ........................ p.26 SECONDARY MARKETOOOOOOOOOOOO0.0.0.0....0.00.00.00p027 HMDA AND CREDIT REJECTION DIFFERENTIAL TREATMENT OR DISPARATE IMPACT............p.29 INTRODUCTION...OOOOOOOOOOOOOOOOOOOOOOO0....0.....p029 1992 BOSTON FEDERAL RESERVE STUDY................p.29 Study Methodology-Definition of Model.......p.3O Significant Findings........................p.30 Creditworthiness............................p.31 Lender Discretion...........................p.32 Financial Characteristics........... ...... ..p.32 Neighborhood Characteristics/Housing Type...p.32 Loan to Value/Insurance.....................p.33 Secondary Market Flexibility ............. ...p.33 DISPARATE IMPACT OF UNDERWRITING ................. p.34 Disparate Impact ......... . ....... . ...... ....p.34 OTHER STUDIES AND SOLUTIONS..... ..... ....... ..... p.35 New York State Banking Department Study.....p.35 Closing the Gap:........ ....... .............p.35 OTHER FACTORS IN THE HOMEBUYING PROCESS ........ ..p.37 COWUNITY RBIMSTMM ACTCOOCOOOOOOCOOOOOOOOOO0......p039 DEFINITION OF CRA ............. . .................. p.39 BASIS FORCRACOOCOOOOOOOOOO0.00.00.00.00...00.00.0p039 PURPOSE AND SCOPE OF CRA ......................... p.40 ENFORCEMENT OF CRA .............................. .p.40 CRA Examinations ............................ p.41 EXEMPTIONS FROM CRA .............................. p.42 AMENDMENTS TO CRA UNDER FIRREA ................ ...p. 42 EVALUATION SYSTEM ..... .... ...................... p. 43 Additional CRA Requirements.................p. 45 PUblic Access. 0 O O O O O O O O O O O O O O O O ..... O O O O O 0 Op. 45 CONTESTED APPLICATIONS O O O O O O O O O O O O O O O ....... O O O O O p O 4 6 Justice Department Investigations...........p.47 POST FIRREARESPONSE.OOOOOOOOOOOOOOOO...00.......p048 CRA REFORM AND SOLUTIONS FOR CHANGE...................p.49 INTRODUCTION.....................................p.49 CLINTON CRA REFORM DIRECTIVE.....................p.49 CURRENT RESPONSE TO CRA REFORM ................... p.50 RECOMMENDATIONS FOR REFORM.......................p.51 Numerical Criteria..........................p.51 Evaluation of CRA Plans.....................p.52 Special Treatment for Small Institutions....p.53 Incentives... ..... ..........................p.54 Include Credit Services.....................p.54 OTHER CRA REFORM ISSUES..........................p.55 Judgement Factor/Computerized Underwriting..p.55 Market Share Comparison.....................p.55 GENERAL CRITICISMS OF CRA REFORM.................p.56 CONCLUSION. ......... .................. ........... p.56 RECOMMENDATIONS FOR FUTURE ANALYSIS .............. p.56 INTRODUCTION: The Home Mortgage Disclosure Act (HMDA) and the Community Reinvestment Act (CRA) were created by Congress in response to concerns that lenders were not providing fair access to credit to various minority groups and to low income areas and central cities in general. Community groups and activists termed this practice redlining which has been defined as the practice of drawing red lines around disfavored neighborhoods where money would not be lent, regardless of the creditworthiness of individual loan applicants (A Citizen’s Action Guide, 1992, p.3). One of the primary criticisms of redlining was the fact that the lack of lending in certain areas, primarily the inner cities, was contributing to the overall decline of these areas through disinvestment. While residents of low income and minority areas were allowed to make deposits in financial institutions, their money was being reinvested and lent out not in their own community but in other more prosperous areas. In their article, "The Invisible Lenders: The Role of Residential Credit in Community Economies," Jean Pogge and David Flax-Hatch describe these issues: The practice of "redlining" was first identified and named in the late 19605 on the West Side of Chicago in the Austin neighborhood. Community residents struggling with school issues discovered that savings and loan associations...had labeled Austin a declining neighborhood and actually drawn a red line around Austin and other neighborhoods on a city map. The lenders had decided the redlined areas were vulnerable to racial change and, therefore, not. a good credit risk. The resulting limitations on the availability of residential credit became a self-fulfilling prophecy, residents could not easily sell or buy homes at normal market prices, prices fell, home improvement loans were not available, homes deteriorated, and finally, many residents sold their houses at a loss and moved out. (p.85) These issues of redlining and disinvestment were brought to light by community activists such as Associated Community Organizations for Reform Now (ACORN) who lobbied successfully for the passage of the Home Mortgage Disclosure Act in 1975 and the Community Reinvestment Act in 1977. Since their inception, these acts have stimulated great change and debate in the housing finance arena. Over time, the once adversarial relationship between community organizations and lending institutions has evolved into numerous cooperative ventures and public-private partnerships between nonprofits, community development corporations, lending institutions and various forms of financing agencies all across the country. Many inroads have been made in the housing finance arena, yet these efforts have not been enough to stem the tide of disinvestment in our urban communities. Instead, issues of disinvestment and discrimination are complex having been created over many decades and caused by numerous factors. Yet, efforts to prevent disinvestment by lenders are similarly complex. As more information is collected through HMDA and community development attempts are made through CRA, more questions arise as to whether or not these regulations and others are the best way to solve the problems of disinvestment and discrimination in lending. In essence, the regulation of financial institutions in this manner is just one portion of the implementation strategy which must be put into place in order to deal with the problems in our cities. BRIEF HISTORY OF COMMUNITY REINVESTMENT LAWS: Over the past two decades, Congress has enacted several fair lending laws in an attempt to address some of the issues associated with redlining and disinvestment. The earliest of these laws was the Fair Housing Act of 1968 which prohibits discrimination in the sale or rental of a dwelling on the basis of race, color, religion, handicap, sex, familial status, or national origin” This act makes it unlawful for any person who engages in the business of making or purchasing residential real estate loans, or in the selling, brokering, or appraising of residential real property, to discriminate on the basis of the factors listed above. A similar act, Equal Credit.Opportunity’Act (ECOA) was passed in 1974 to ensure the availability of credit to all creditworthy applicants without regard. to race, color, religion, national Aorigin, sex, marital status, age, or receipt of public assistance funds(Closing the Gap, 1993, p.26). Both acts were created by Congress in an effort to ensure that credit is not denied to qualified applicants on a prohibited basis or because of the location of the property (HMDA, Federal Reserve of Bank Chicago, p. 2). The Home Mortgage Disclosure Act (HMDA) of 1975 was billed as a "right to know act" which required lenders to provide the number and dollar amount of home loans they originated each year in an effort to make such information available to the public. In 1977, Congress created the Community Reinvestment Act (CRA) which required lenders to establish community development policies and agendas aimed specifically at reinvesting money in the community. Both were later amended substantially by the Financial Institutions Reform and Recovery Act (FIRREA) of 1989 and the Federal Deposit Insurance Corporation Improvement Act (FDICIA) of 1991. DEFINITION OF DISCRIMINATION: The creation and subsequent evolution of the fair lending laws has reflected the debate over how to define discrimination. The dictionary definition of the word discriminate and its derivatives is quite broad. The definitions range from: Discriminate (verb) 1.a To make a clear distinction; distinguish: discriminate among the options available b. To make sensible decisions; judge wisely. 2. To make distinctions on the basis of class or category without regard to individual merit; show preference or' prejudice: ‘was accused, of discriminating against women; discriminated in favor of his cronies. 1. to perceive the distinguishing features of; recognize as distinct: discriminate between right and wrong. 2. To distinguish by noting differences; differentiate: unable to discriminate colors. Discriminating (adjective) 1. a. Able to recognize or draw fine distinctions; perceptive.b. Showing careful judgement or fine taste: a discriminating collector of rare books; a dish for the discriminating palate. 2. Separating into distinct_parts or components; analytical. 3. Serving to distinguish; distinctive: a discriminating characteristic.4. Marked by or showing bias; discriminatory Discrimination (noun) 1. The act of discriminating.2. The ability or power to see or make fine distinctions; discernment 3. Treatment or consideration based on class or category rather than individual merit; partiality or prejudice. (The American Heritage Dictionary of the English Language) 3rd edition. p.532) Similarly, the definition of discrimination in lending is equally as broad and has evolved over time. The discovery, identification and redefining of discrimination in lending is reflected in the multitude of studies which attempt to address this topic. Initially, most studies and analysis focused on analyzing lending patterns and credit flows into minority and low income neighborhoods. Then, after HMDA data was expanded under FIRREA in 1989, the emphasis of discrimination in lending changed to analyzing individual loan files for patterns of discrimination in the form of credit rejection and disparate treatment based on race and income. Similarly, recent analyses have focused not on whether discrimination or disparities in lending exist, but why? From the perspective of the lender, the answer to this question has been to evaluate whether certain underwriting criteria have had an adverse impact on minorities. These definitions of discrimination are further defined as: Blatant Discrimination: The explicit use of a protected variable (race, marital status, sex, etc.) in establishing lending guidelines. Disparate Treatment: When two applicants, identical in all aspects except for a particular characteristic (such as gender, race, or property location), receive different treatment based on that characteristic. The lender, knowing the race, sex and property location of applicants, applies different credit standards on the basis of these variables. Even though the policy is not blatant, minorities are nonetheless singled out and unfairly denied credit. Adverse Impact: When seemingly innocuous lender policies result in the unfair denial of credit to protected classes. Adverse impact need not entail conscious mistreatment of minorities. Lender policy must meet two criteria to provide evidence of adverse impact: It must disproportionately deny credit to minorities, and it cannot be rationally related to a legitimate business purpose. (Housing Research News, in American Banker, December 22, 1993, p.9) 5 PREMISE OF PAPER: Since the adoption.of CRA in 1977, neighborhood activists have utilized HMDA and CRA as ways to tap into private funding for housing finance. HMDA data is commonly used to substantiate disinvestment and discrimination claims made under CRA. Over time, both acts have been amended to encourage more disclosure of home mortgage lending data and to improve community reinvestment policies. During this evolutionary process, the validity of HMDA.data in showing the true causes of mortgage discrimination have been debated and the efficiency and effectiveness of CRA have also been questioned. Newspaper series continue to shed light on discrimination and Congress is once again revisiting these issues. Yet, the question remains whether HMDA and CRA can show the entire picture and if lenders are unduly blamed. While it is essential that lenders eliminate, discriminatory' processes, there are. a multitude of additional players and issues which contribute to the problems and issues associated with discrimination in the housing credit market. Discrimination in housing credit markets is most frequently thought of as an issue that could be better understood by analyzing the behavior of lenders - where they market, to whom, and how. But focusing only on lender behavior would preclude discovery of the extent to which other parties - buyers, sellers, insurers, appraisers, and others - contribute to housing discrimination and its effects (Wienk, 1992, p. 236). In light of the impending CRA Reform, what changes should be made to the fair lending laws in order to discourage discrimination in lending. Should efforts be refocused and concentrated on identifying other factors that show discrimination? Should additional techniques be put in place such as testing or can more complex statistical analysis provide the answers we are looking for? Similarly, will changes to these acts weaken the framework which currently supports successful community development lending efforts? Overall, a multitude of policy questions are currently being debated and suggestions are being made to Congress on these issues. PAPER OUTLINE This paper will provide an overview'of the requirements of the HMDA and CRA, followed by a discussion of the typical housing finance process focusing on underwriting factors and additional factors affecting the process. Additionally, it will analyze the most recent HMDA and CRA changes and ‘will conclude with a discussion of future issues being addressed by Congress. HOME MORTGAGE DISCLOSURE ACT PURPOSE: Congress passed HMDA in 1975 as a "right to know" act in response to allegations of redlining made by community groups and other activists. The purpose of the act was to make information regarding home loans available to the public so that individuals could determine if financial institutions were serving the housing credit needs of their communities. HMDA was also seen as a way of- helping regulators identify possible discriminatory lending patterns and as a way to assist public officials in making public sector investments to attract private investment to areas where it is needed (U.S.C. 2802 in Fishbein, 1992, p. 602). REQUIREMENTS: From the outset, HMDA applied to banks, savings and loans associations, and credit unions with assets of at least $10 million, and a branch or main office located in a metropolitan statistical area (MSA)(HMDA,Federal Reserve Bank of Chicago, p.2). It required these lenders to report by census tract, the number and dollar value of home purchase and home improvement loans they originated or purchased in metropolitan areas each year. These loans were then itemized by type: single family conventional loans, home improvement loans, multifamily loans, and loans to nonoccupants (i.e. investors). .Amendments to HMDA.under FIRREA and FDICIA expanded disclosure requirements to mortgage company affiliates of banks, and savings and loan (S & L) holding companies and service corporations, large independent mortgage companies and small independents (Fishbein, 1992, p. 603). AVAILABILITY OF DATA: Earlier amendments made to HMDA in 1980 included provisions requiring the creation of a central depository of all HMDA data in each metropolitan area. These depositories were to be located in public libraries or government offices to ensure that the public would have access to the data. Similarly, HMDA was further amended to require that all regulatory agencies provide aggregate analysis of HMDA data for each metropolitan area. This information was computerized and made available to the public on a yearly basis. HMDA USERS: Computerization and further refinement of data have resulted in a variety of uses and users of HMDA data. Today, HMDA data is used.by community groups, state and local government agencies, news media, consumer groups, banking regulators and financial institutions. Lenders also use HMDA data in a variety of ways that were not initially intended in the original law. For instance, HMDA loan data can be used by lenders to identify markets, target special populations and or specific geographic regions and to evaluate competing lenders’activities (Fishbein, 1992, p. 605). However, the primary use of the expanded HMDA.data is for community groups to support CRA claims against lending institutions. Computerization of the data and improved availability led to increased usage of the data by community groups and other analysts studying lending patterns. Similarly, expanded data requirements under FIRREA increased the number of variables and improved the reporting format to allow for more complex statistical analysis. EARLY HMDA STUDIES: Early studies focused on issues of credit flow and neighborhood income characteristics. Studies were completed using loan data and census-tract data to determine the differences in lending activity between. neighborhoods or between cities and suburbs. The results showed wide disparities between low-income, predominantly minority neighborhoods and more affluent, predominantly white areas (Canner 1982, p.2 in Fishbein, 1992, p. 605). During the late 19805, two newspaper series published in the Atlanta Journal Constitution and Detroit Free Press brought widespread attention to the issues of mortgage lending discrimination. Both of these studies attempted to show substantial loan disparities between black and white neighborhoods while comparing neighborhoods with similar income levels. Overall, these studies concluded that there were substantial discrepancies in lending patterns in black and white neighborhoods. Findings of these newspaper studies were later substantiated by a lending study completed by the Boston Federal Reserve Bank in 1989. 10 Atlanta Journal Constitution: In his 1988 Pulitzer Prize winning "The Color of Money" series, Bill Dedman of the Atlagta Journal Constitutien compared predominantly white and predominantly minority Atlanta neighborhoods with the same income level and found that the white neighborhoods received five times as many home loans from local banks and savings institutions as the black neighborhoods (Fishbein, 1992 p. 605). Atlanta Journal constitution-Second study: A second study was completed by the Atlanta Journal Constitution in January of 1989 based on information obtained under the Freedom of Information Act from the Federal Home Loan Bank (FHLB) of Atlanta. Under this law, the FHLB provided reports from savings and loans for the nation’s largest cities. These reports, which were aggregated by individual metropolitan area, portrayed loan rejection rates based on borrower characteristics. The results were that on the‘whole, black mortgage-loan applicants were rejected roughly twice as often as white applicants in the nation's largest cities(Dedman, 1989, in Fishbein, 1992, p.607). Detroit Free Press: A similar study published by the Detroi; Free Press compared the number of home loans made in the city’s white, middle-income neighborhoods with similarly situated black neighborhoods. The results showed that three times more loans were made in the white 11 neighborhoods than in the black neighborhoods. They attempted to show loan demand by utilizing deed transfer data and information provided through a state disclosure law which contained more information than the HMDA data (Fishbein, 1992, p. 606). 1989 Federal Reserve Bank of Boston study: A lending study conducted by the Federal Reserve Bank of Boston and later published at the request of several members of Congress, supported the findings of the newspaper articles. Like- the Detroit study, the Boston study used deed transfer information in place of HMDA data to estimate demand for mortgages and to analyze the patterns of mortgage lending in the City of Boston. This study attempted to determine whether differences in economic and other nonracial characteristics (primarily neighborhood characteristics) as reported in census data, might account for the disparities. The researchers found that, after controlling for neighborhood factors, predominantly minority neighborhoods in Boston had been granted 24 percent fewer mortgage loans per housing unit than predominantly white areas. The number of mortgage originations relative to the owner- occupied housing stock was 24 percent lower in black neighborhoods than in white neighborhoods, after taking account of economic variables such as income, wealth, and other factors (Bradbury, Case, and Dunham, in Munnell, et. al., 1992, p.5). ' In total, approximately 48,000 property transactions over a five-year period were examined and the results provided further evidence of discrepancies between mortgage origination in white and 12 black neighborhoods. Overall, Bradbury, Case and Dunham concluded that housing and mortgage credit markets were functioning in a way that hurt black neighborhoods: Lower mortgage origination in black neighborhoods cannot be explained away by lower levels of income and wealth, lower rates of housing development, or other neighborhood differences. Even after taking these factors into account, one still finds a substantial discrepancy between mortgage orginations relative to the housing stock in white and black neighborhoods (Bradbury, Case, and Dunham 1989, p.31 in Fishbein, 1992, p.606). ANALYSIS OF 1989 BOSTON FEDERAL RESERVE STUDY: The authors of the study and other critics offered several reasons for the results of the study. Among the most widely noted flaws were the inability to show demand for loans, lack of data about mortgage company activity, and the geographic aggregation of HMDA data which prohibited analysis of individual loan files. Others noted the differences in income and basic economics in minority communities and how this might limit the number of minorities who might apply for loans“ Similarly, discrimination in other parts of the housing market such as the home selection process have also been cited. Differences in Demand: A major criticism of the 1989 study was the difference in demand for home purchase loans between minority and upper income white neighborhoods. Canner and Smith cite a lack of demand for mortgages in minority neighborhoods. Thus, a possible interpretation of the earlier study was that . fewer mortgages were made in black neighborhoods because 13 people in black.neighborhoods did not buy houses as frequently as residents of white neighborhoods and therefore did not apply for as many mortgages (Munnell, et. al.,p.8, 1992). Canner and Smith cite pre-1990 HMDA.data and state that the «demand for home purchase loans is less from lower income groups, than from the upper and middle income groups. They state that home purchase loans in low- or moderate-income neighborhoods constituted a small proportion (approximately 10-12%) of the overall home purchase loans made. Similarly, approximately one-third of the home purchase loans are for properties in upper-income neighborhoods (income at greater than 120% of median family income for the MBA) and the remainder were middle income properties. Lack of Demand\Differences in Communities: Critics of the Atlanta and Boston studies noted that the studies failed to account for the differences in demand between neighborhoods that were compared. Some suggested that these differences in demand may be a result of other factors such as reliance:on government backed loans, use of home improvement loans, and inability to meet the underwriting standards established by lending institutions: The relatively heavy reliance on government-backed loans in Atlanta's minority neighborhoods also may have reflected differences in the ability of applicants in the two groups of neighborhoods to meet the underwriting standards for conventional loans established by creditors, including downpayment amounts and debt-to-income ratios. Information about the amount of assets available for downpayment and levels of debt burden of the Atlanta home buyers was not available (Canner and Smith, 1991, p.865). 14 Aggregation of Data: Canner and Smith claim that by comparing the level of home lending per housing unit in seemingly similar minority and nonminority neighborhoods chosen based on aggregate characteristics such as neighborhood median family income, the study did not account for differences in the economic circumstances of the residents. In essence there may be additional economic factors which are not shown when the data is aggregated. Differences in Income and Basic Economic Situations: Other critics have noted the disparities in¢overall wealth and income between white and minority households. that it is not so much racial discrimination but patterns which reflect fundamental differences in the economic circumstances of population groups (whether already living in or seeking to reside in the different areas) and in market specialization by different types of lending institutions (Canner and Smith, 1991, p.864). In the article entitled "When Even Having Equal Income Is Not Enough," Ronald Zimmerman examines differences in income between whites and minorities: in this country low-and moderate income blacks and some other minorities have significantly less income and 'wealth, on average than their white counterparts....This income polarity also holds true for the middle- and upper-income groups. Within each income group, the incomes of whites tend to cluster near the top of the income range while the incomes of blacks tend to be distributed near the middle to lower end of the income range. Because minority households have fewer cash resources, their ability to purchase homes and increase their wealth is impeded. This suggests that the widespread mortgage loan distribution patterns reflected by the Home Mortgage Disclosure Act data may be largely a consequence of these differences and should not be surprising (p.8). 15 However, Zimmerman goes on to state that studies have shown that even at all income levels, even if income is the same, more mortgage loan applications from black applicants than white applicants are denied. Zimmerman summarizes: a key factor in the answer may be that even at middle- and higher-income levels, a tremendous difference in wealth still exists between blacks and whites. (p.9) Discrimination in the Housing Market: The authors of the Boston study noted that the study did not» account for discrimination in the housing market which in turn might account for lower numbers of minority mortgage applicants. The study, however, could not distinguish between discrimination in the housing market and discrimination in the mortgage market. From the available data, it was not possible to sort out the precise role played by lenders, as opposed to buyers, sellers, developers, realtors, appraisers, insurers, and others (Munnell, et, al., 1992, page 8). CONCLUSIONS FROM THESE EARLY STUDIES: As a result of this study and others, it became increasingly more evident that there was a considerable difference in lending between white and minority areas. Consequently, the question of whether discrimination exists has been redirected to focus on where discrimination occurs and why. HMDA data have long been the primary source of public information about the geographic distribution of home loans originated and purchased by financial institutions. Dozens of studies have examined the distribution of home loans across neighborhoods stratified by residents' income and race...For the most part, one basic lending pattern has stood out: Considerable differences exist in the levels of home lending activity across neighborhoods within the local communities when the neighborhoods are grouped by median family income or racial composition (Canner and Smith, 1991, p.864). 16 FIRREA AND FDICIA AMENDMENTS TO HMDA INTRODUCTION : As a result of the debate sparked by the Boston study and reoccurring media coverage, Congress revisited the issue of discrimination in lending. Amendments made to HMDA under the Financial Institutions Reform and Recovery Act (FIRREA) of 1989 and again in 1991 under the Federal Deposit Insurance Corporation Improvement. Act (FDICIA) have significantly‘ changed. reporting requirements and strengthened the role of HMDA in detecting discrimination in lending. Studies using expanded HMDA data have shown discrepancies in lending rates between whites and minorities. However, many still debate whether HMDA.data is sufficient proof of discrimination. A New Focus for Determining Discrimination: Under FIRREA, data disclosure for the HMDA must include information on three‘additional variables: race, gender and income. With these additional factors, the emphasis in lending discrimination switched from analysis of mortgage credit flow into similar neighborhoods divided by race, to analysis of individual loan files and cases of credit rejection which evidenced discrimination. Given access to individual loan files, the concern has changed from redlining, that is differential treatment by lenders based on location of property, to‘discrimination defined as 17 differential treatment of applicants based on, race or other personal traits rather than economic characteristics (Munnell, et. al., 1992, p.9). Amendments Under FDICIA: Changes under the FDICIA included expanding HMDA reporting requirements to all small mortgage companies (thereby making almost all firms in the full-time business of mortgage lending subject to HMDA reporting). The FDICIA also mandated that lenders must provide loan applicants with a copy of the appraisal report of the property to be purchased so that applicants can contest underappraisals by unfamiliar real estate agents. Amendments under FDICIA also called for the Justice Department to investigate discrimination cases and to seek actual and punitive damages. Several such cases have recently been conducted (See Shawmut case under Community Reinvestment Act). Amendments Under FIRREA: Significant changes to HMDA include expanded data reporting requirements, a new reporting format, improved tabulation of data and accessibility, and changes in reporting coverage. These changes are explained in detail below: 1) Egpanded Data: - Expanded data to include race, gender, and income level of borrowers and loan applicants. 0 Mandated data collection for ALL loan applications even if credit is not granted. 18 2) New Reporting Format: 0 Changed loan reporting format to Loan Application Registers (LARs). 0 Required lenders to file LARs with their federal regulator. 0 LARs give option to cite reasons for loan denial by category. 3) Imgroved Tabulation of Data and Accessibility: 0 Required the Federal Financial Institutions Examination Council (FFIEC) to tabulate data into reports. - Expanded tabulation format from one annual report totalling loan activity in each metropolitan area to as many as 30 tables for each lender depending on information provided by lender and the metropolitan area. 0 Designated HUD as collector of mortgage company data. Cha d Re ort Covera : 0 Extended coverage to 400 independent mortgage companies 0 Required institutions to report loans sold within the same calender year of purchase or origination and also required classification of loans by type of secondary market entity. 0 Required the FFIEC to disclose certain raw data contained on the LARS for individual lenders. 0 Exempted small depositories with assets of $30 million or less. FINDINGS FROM EXPANDED HMDA RESULTS: The first expanded HMDA data results were made public in October of 1991. Studies of the 1990 HMDA data by the Federal Reserve Board showed large disparities. Many of these disparities were addressed by a study completed by Glenn B. Canner and Dolores S. Smith of the Federal Reserve Board staff using preliminary HMDA data. Canner and Smith reported on the following categories: Volume of Application Use of Various Home Purchase Loan Products Overall Approval Rates Approval Rates of Minorities Conventional Home Purchase Loans Government-Backed Home Purchase Loans Home Improvement Loans Relation of Approval Rates to Neighborhood Income and Composition 19 Approval of Home Purchase Loan Applications Neighborhood Income Neighborhood Racial Composition Neighborhood Income and Racial Composition Approval of Home Improvement Loan Applications Canner and Smith, 1991, p.867-873) A..... General Information: In total, Canner and Smith stated that approximately 5.26 million home loan applications were reported under HMDA for 1990. These applications were broken down by housing size of one to four families or multifamily which.was defined as five or more families. The majority of applications were for homes which were one to four families. They were as follows: 3.09 million for home purchase, 1.02 million for home refinancing, 1.10 million for home improvement, The remaining balance, (approximately .05 million) was reported for multifamily dwellings, however this figure‘was not broken down into the categories listed above. canner and Smith also noted that approximately 74% of the home purchase loan applications were for conventional mortgage loans and the remainder were for government- backed forms of credit such as FHA, VA and FmHA loans (Canner and Smith, 1991, p. 867). significant Findings: Significant findings from the preliminary HMDA data included evidence that minorities had a strong reliance on government backed loans, minorities had higher denial rates, and denial rates 20 increased as the proportion of minorities increased. The results of the study are discussed in more detail below. Strong Reliance on Government Backed Loans by Minorities: 0 Government-backed loans are much more likely to be used by households with relatively low incomes than by households with high incomes. 0 Black applicants and (to a smaller extent Hispanic applicants) are more likely than either white or Asian applicants to seek government-backed home purchase loans 0 Even after controlling for applicant income the data still indicates that blacks, and to a lesser extent Hispanics, are more likely than whites to use FHA and VA loans. 0 Overall Approval Rates for conventional home purchase loans and government-backed loans were 72.3 and 71.1 respectively. e D nial Rates fo A 1 nts° 0 Nationally, about 14.4 percent of white applicants for conventional home purchase loans were denied credit in 1990 as compared to 33.9 percent for black applicants and 21.4 percent for Hispanics. o Denial rates for Government Backed Home Purchase Loans were: 26.3 percent for blacks, 18.4 percent for Hispanics, and 12.8 percent for Asians, compared with 12.1 percent for whites. 0 Overall, denial rates for Home Improvement Loans are higher than for home purchase loans. 0 Denial rates for Home Improvement Loans were 36.9 percent for black, 32.5 percent for Hispanic, and 24.6 percent for Asian American, and 17 percent for white applicants. Increasing Denial Rates as the Proportion of Minority Residenfig Increase: 0 Loan denial rates decline as the income of the residents of an area increases. 0 Loan denial rates increase as the proportion of minority residents increases. 0 For the most part, whether the neighborhood is low or moderate income, middle income, or upper income, the proportion of home purchase loan applicants denied credit increases as the percentage of minority residents increases. 21 similar Findings for 1992 HMDA Data: The HMDA data for 1992 was released in the fall of 1993. It showed similar patterns of disparity between the races. This data shows that 36 percent of black applicants for mortgage loans were turned down, while only 16 percent of white applicants had their loan applications denied (England, January, 1994, p.41). IS EXPANDED HMDA DATA SUFFICIENT EVIDENCE OF DISCRIMINATION?“ While these statements show strong disparities in lending patterns by race and income, many dispute whether HMDA data is sufficient to prove mortgage credit discrimination. Many argue that the expanded HMDA data does not include important factors which effect mortgage credit decisions, specifically factors of creditworthiness and collateral: The data have important limitations, however, and care'must be taken in drawing conclusions from observed lending patterns. Foremost among these limitations is a lack of information about factors that are important in determining the creditworthiness of applicants and the adequacy of the collateral offered as security for their loans. Without taking into account such information, one cannot determine whether individual applicants or applicants grouped by a common characteristic (such as race or gender, have been treated fairly (Canner and Smith, 1991, p. 859). Canner and Smith cite a list of omitted variables which are important to the mortgage decision but not directly shown by HMDA data. These factors are all integral parts of the decision to grant a mortgage. They include creditworthiness, property value, debt history, collateral, consumer's income, outstanding debts, 22 equity, amount of downpayment, employment experience, debt repayment history, and property appraisal (Canner and Smith, 1991, p.875). Thus, in order to determine if discrimination in lending is occurring in the form of credit rejection it is important to evaluate all the factors which are considered in the lending process. Galster suggests statistical analysis using all variables to compensate for these omissions: The recently released 1990 HMDA data will be inadequate for this task, because crucial control variables such as credit and employment history, indebtedness, and assets and characteristics of the property were not gathered (Galster 1991b) ...What is needed is a sophisticated, multivariate estimate of demand.and supply functions for mortgages based.on a large, robust, current database of individual households and (accepted and rejected) applicants for mortgages. Creation of such a database would necessitate access to lenders’loan files, presumably mandated by federal or state regulatory agencies.(Galster, 1992, p.650) A study completed by the Federal Reserve Bank of Boston entitled, "Mortgage Lending in Boston: Interpreting HMDA Data," (Munnell et al.) attempted to incorporate all the variables associated with the mortgage lending process into a statistical model to test to see if race 'was a significant factor. In completing the study the Federal Reserve Bank of Boston augmented HMDA data collected from lenders’ loan files with 38 additional variables. BASIC MORTGAGE APPLICATION PROCESS The following is a description of the basic mortgage application process as it was described by Munnell, et. al. The three main steps of the basic mortgage application process are a 23 quick review of the application for viability, verification of the information and appraisal of the property, and evaluation of the obligation ratios and consideration of any compensating factors which might influence the decision (Munnell, et a1, 1991, p.10). 'Once a lender is selected, the applicant completes a standard loan application form which is reviewed by an intake person or a loan officer to ascertain whether the loan is viable. .At this time the loan application can be denied if the information does not appear to be viable. However, there is some concern that applicantS~ are informally prescreened and turned away before any application is ever completed. Some have suggested the use of paired testing to identify discrimination at the prescreening stage (Fishbein, 1992, p. 621),(Galster, 1992, 651). Verification and Evaluation of Ratios: Once the application is deemed viable, the lender investigates and verifies the financial ability and inclination of the applicant to repay the loan, and determines if there are sufficient liquid funds for a down payment and closing costs. These factors are verified by checking employment, credit history, and.bank.deposits. Obligation Ratios: After credit history and employment are established, the lender must evaluate several key ratios which summarize the applicant's wealth, income, debts, and assets. Together these ratios are used to evaluate whether the applicant has the ability to support the loan. These ratios are: 24 Housing Expense/Income: which measures proposed monthly housing expenses relative to income. Total Debt Payments/Income: which measures the total debt payment obligations relative to income. In general, lenders estimate a household should spend approximately one-fourth of its income (28%) on housing and only about one third of its income (36%) on total indebtedness. Similarly, these ratios are guidelines established by Fannie Mae and Freddie Mac which indicate whether the mortgage can be sold in the government insured secondary market.(see below) In the past, lenders traditionally expected buyers to make a downpayment amount of at least 20 percent of the purchase price of a house. However, buyers can pay as little as 5 percent down provided they purchase private mortgage insurance, which protects the lender in case the borrower fails to repay the loan. Other such special programs are available through the secondary market such.as the Fannie Mae 3/2 Option and Freddie.Mac's Affordable Gold program (England, January, 1994, p. 44). If the application is still viable, an appraisal of the property is completed in order to calculate the loan-to-value ratio. The loan-to-value ratio compares the amount and terms of the loan to the appraised value of the property. The standard ratio is 80%, however, if private mortgage insurance is used the ratio can be higher (typically, 80% is the ratio used by the secondary market purchasers). Lenders must also evaluate the type and terms of the loan requested as well as personal characteristics such as 25 age which may effect the ability to continue working or dependents which may require more money to support. Evaluation of numbers and Compensating Factors: Less then 20 percent of borrowers have perfect applications and lenders must weigh compensating factors in order to approve applications (Munnell, et. al., 1992, p.12). For example a high debt to income ratio can be compensated for with a large down payment,a good record of carrying high housing expenses, a strong propensity to save and a high level of liquid assets, or an excellent potential for future earnings based on education and training. Similarly, credit history problems can sometimes be compensated for with the following: favorable letters from creditors, extenuating circumstances such as an,adverse judgment in a civil suit or prior life circumstances which.have changed for the better. It is important that potential flaws in a loan application be brought to the attention of the applicant so that they may have the opportunity to correct the problems which prevent them from securing a loan. (Some note that white applicants have a larger propensity to be "coached" as to how to improve their applications) . Another alternative is to provide credit counseling and homebuyer seminars to assist prospective applicants. 26 SECONDARY MARKET: The secondary market plays a large role in home financing. Fishbein notes that lenders covered by HMDA sold approximately 2.3 million loans to secondary market entities in 1990 (p.619). The secondary financiers purchase loans from the primary market and provide government backed insurance on these loans. The secondary market financiers are: Federal National Mortgage Association-(FNMA) referred to as Fannie Mae, which is a federally chartered private corporation providing a secondary market for residential mortgages. Federal Home Loan Mortgage Corporation (FHLMC) referred to as Freddie Mac, which is a quasi-governmental agency that purchases mortgages from insured depository institutions and HUD-approved mortgage bankers. Government National Mortgage Association (GNMA) referred to as Ginnie Mae, is a government corporation which provides a secondary market for housing mortgages and special assistance to housing mortgages financed under special HUD mortgage insurance programs. These agencies have established certain guidelines based on underwriting criteria under which they will purchase and insure a loan. Fannie Mae and Freddie Mac guidelines are 28% for housing expense/income, 36% for total debt payments/income and 80% loan to value ratios for purchasing home loans from the primary market. However, these are just guidelines which can be changed based on compensating factors (see Munnell, et. al., 1992). However, the primary lender must take these standards and guidelines into account when attempting to sell off loans to the secondary market. Thus, some critics argue that this process may be unfairly impacting the decision to approve the loan: 27 Some argue that the need to conform to secondary market underwriting guidelines...reduces the willingness of local lenders to be flexible in idiosyncratic cases - cases most often presented by minority applicants. Yet others argue that, by spreading risk and augmenting liquidity of lenders, secondary markets are a boon to putatively riskier segments of the population (Galster, 1992, p. 657). 28 HMDA AND CREDIT REJECTION Differential Treatment or Disparate Impact? INTRODUCTION: In essence, the basis of the Boston Study was to evaluate what factors went into the home purchase process and to attempt to identify through the use of statistical modeling where credit is rejected and if there is differential treatment based on race. From this study, it was determined that minorities often receive disparate treatment» In some cases this disparate treatment may be a result of arbitrary underwriting standards which unintentionally have a disparate or adverse impact on minorities. 1992 BOSTON FEDERAL RESERVE STUDY: The study was based on the 1990 expanded HMDA DATA which showed that minorities in the Boston.Metropolitan Statistical Area were 2.7 times more likely to be denied mortgage loans as whites. In order to account for the omission of key variables such as credit histories, loan-to-value ratios, and other factors, the Boston Fed augmented the study with some 38 additional variables collected from individual loan files selected to cover the financial and employment variables considered in the home mortgage lending decision. Including this information reduced the disparity between minority and white denials from 2.7 to 1 ratio cited above to a ratio of 1.6 to 1 respectively (Munnell et. al., 1992, page 2). 29 Study Methodology-Definition of the Model: Information on 38 variables was requested for 1,200 conventional mortgage loan applications made by blacks and Hispanics in 1990 and from a random sample of 3,300 applications made by whites from lenders in the Boston Metropolitan Statistical Area. Additional data about neighborhood characteristics was gathered from census data and a statistical model was created to test whether race was a significant factor in the lending decision once financial, employment and neighborhood characteristics were taken into account. These variables were summarized in the following model: ~ P(D) f(F,R,L,T,C) where: P(D) Probability of a lender denying a mortgage application. F= Applicant’s ability to carry the loan R= Risks of default L= Potential loss associated with default and foreclosure T= Terms of the loan C= If the lenders judgment is influenced by the race or other personal characteristics of the applicant which might affect the likelihood of denial (Munnell,et. al., 1992, p. 13). Significant Findings of 1992 Study: The primary finding of the Boston study is that overall a significant gap in lending between whites and minorities is still evident, when other factors such as financial, employment, and neighborhood characteristics, are taken into account. The study finds that in Boston, black and Hispanic applicants with the same economic and property characteristics as white applicants would 30 experience a