Redlining: the Justice Department cases.

Knowing how the Justice Department gathers and weighs its evidence in a redlining investigation is important for lenders today. The lead attorney in the investigation of Chevy Chase chronicles how the department stacked up its evidence in two much-debated cases.

The justice department's

recent redlining lawsuit against chevy chase federal savings bank and its mortgage subsidiary has been the subject of intense debate and criticism. The department accused the institution of redlining black neighborhoods through sophisticated marketing practices designed to avoid soliciting loans from those areas. The case was settled by a consent decree in which the thrift agreed to commit $11 million to the redlined areas through special loan programs and the opening of branches and offices. As with earlier Justice Department fair lending lawsuits, this case was settled without contested litigation, and thus, Chevy Chase's rebuttal to the department's evidence was not formally articulated in court. Accordingly, this article reflects only evidence relied upon by Justice in bringing its case against Chevy Chase.

Many critics allege the Justice Department brought this case using novel and untested theories. Some have called it a backdoor effort by the Clinton administration to promote a Community Reinvestment Act (CRA) agenda by forcing banks and thrifts to allocate credit equally among white and minority neighborhoods. Others are more pejorative, calling the case "true evil" and a jackbooted assault on the lending industry by government bureaucrats, evoking comparisons to Waco (Washingtonian magazine, July 1995).

Much of this mounting chorus of criticism, which has now reached the halls of Congress, stems from a fundamental misunderstanding of the facts and the law that, in my judgment, clearly justified the department's action. This misunderstanding has led some to unfairly impugn the motives of the Justice Department and its career attorneys and support staff, but more important seems to have created genuine confusion in the industry as to its compliance responsibilities under fair lending laws.

I served as the attorney who headed the Justice Department's investigation of Chevy Chase and many of its prior fair-lending cases. Accordingly, my article is intended to provide some guidance and understanding of the facts and legal principles supporting the department's redlining claims against Chevy Chase. Since I no longer work for the department, the views expressed are my own.

To understand Chevy Chase, it's important to review the Justice Department's seminal 1992 lawsuit against Decatur Federal Savings & Loan Association in Atlanta. In that case, the department developed the investigative techniques and legal principles that would guide its subsequent investigation of Chevy Chase.

Notwithstanding the barrage of criticism, the lessons of Decatur and Chevy Chase appear to be having an impact in many parts of the lending industry. HMDA data for 1994 from across the country show dramatic increases in loans to minority borrowers, particularly blacks and Hispanics. This is a laudable trend that shows there are significant marketing opportunities in underserved minority areas to which lenders are responding. If sustained, this will go a long way toward defusing the public debate over redlining.

Development of the DOJ's legal theories

The Decatur investigation arose out of a series of articles in the Atlanta Journal-Constitution called "The Color of Money," which showed that many of Atlanta's leading lenders made few loans in black neighborhoods. The series also showed that the disparities in loan originations could not be explained by differences in income or a lack of qualified borrowers. The articles further showed that many of these lenders had few branches in minority areas, rarely called on minority Realtors to solicit loan customers and viewed minority neighborhoods as "ceded" to independent mortgage companies that specialized in FHA loans.

Decatur was targeted by Justice because of its large market share and extremely high racial disparities. In early 1991, under the Bush administration, when I and a team of attorneys and support staff were assigned to the Decatur investigation, we followed principles and guidelines that had traditionally been used by the department to prove intentional discrimination under a variety of civil rights laws the attorney general has authority to enforce. These principles emanate from a host of Supreme Court decisions in housing, employment, voting and school desegregation cases during the past 25 years. These decisions establish guidelines for proving purposeful discrimination through circumstantial or indirect proof.

An overview of principles from these cases is important to understanding both Decatur and Chevy Chase. In Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 265-266 (1977), a zoning case under the Fair Housing Act, the court held that proof of discriminatory intent may be established through a "sensitive inquiry into such circumstantial and direct evidence of intent as may be available," that "smoking gun" evidence is not required, nor must there be evidence that race was the sole motive for the defendant's behavior.

The role of statistics

Statistics are frequently one of the most important indicators of purposeful discrimination where they show long-standing and unexplained racial disparities in the workplace, schools, housing and voting (see, Teamsters v. United States, 431 U.S. 324 [1976] [employment]; Columbus Board of Education v. Penick, 443 U.S. 449 [1979] [schools]; Arlington Heights [housing]; Rogers v. Lodge, 458 U.S. 613 [1982] [voting]). There is no basis in principle for excluding mortgage lending practices from such analysis.

In Teamsters, a Justice Department pattern or practice case, statistics showed severe racial imbalances in a trucking company's work-force, with virtually all of the company's black workers in low-level service-maintenance jobs and almost none in the more lucrative long-distance driver jobs. The court found this statistical imbalance a "telltale sign of purposeful discrimination" because "absent explanation, it is ordinarily to be expected that nondiscriminatory hiring practices will in time result in a work force more or less representative of the racial and ethnic composition of the population of the community from which employees are hired" (see 431 U.S. at 340, n. 20).

In Hazelwood School District v. United States, 433 U.S. 266 (1977), a Justice Department case challenging the hiring practices for public school teachers, the court noted that less than 2 percent of the school district's teachers were black even though at least 5.7 percent of the public school teachers in the district's recruiting market were black. The court found these disparities probative of possible discrimination because they exceeded two to three standard deviations, a test of statistical significance that showed it was highly unlikely they could have occurred by chance. The court went on to note that when "gross statistical disparities can be shown, they alone can in a proper case constitute prima facie proof of a pattern or practice of discrimination" (see 433 U.S. at 307-08). If unexplained by the defendant, the plaintiff may prevail in such instances.

Other circumstantial evidence

Another indicator of intentional discrimination is continued adherence to a policy or practice with well-known and predictable racial results, such as continued use of nonjob-related employment tests that keep minorities out and perpetuate a virtually all-white work force, Washington v. Davis, 426 U.S. 229 (1976), or gerrymandering voting districts or school attendance zones that foreseeably keep minorities out or perpetuate racial segregation (see Gomillion v. Lightfoot, 364 U.S. 339 [1960] [racial gerrymander of municipal voting districts]; Columbus Board of Education v. Penick supra [school attendance zones]).

In Davis, the court held that an invidious discriminatory purpose may be inferred from the totality of relevant facts, including evidence that "the challenged conduct bears more heavily on one race than another" (see 426 U.S. at 242). In Penick, the school board created optional attendance zones for certain neighborhoods that resulted in white students avoiding predominantly black schools. The school board also followed a pattern of siting new schools only in predominantly white or predominantly black areas.

Applying its teachings in Davis, the court held that these and other practices of the school board, while not on their face racially motivated, led to results that were so foreseeable as to constitute deliberate efforts to maintain racial segregation. In Hazelwood, the court noted, in addition to the unexplained statistical disparities, the school district had previously recruited teacher applicants only from predominantly white schools and universities and never visited predominantly black institutions, including a nearby black college. Predictably, this practice resulted in a predominantly white pool of applicants that contributed to few black hirings.

Finally, evidence of a history of racial discrimination in an industry is relevant to assess whether blatantly discriminatory practices were simply replaced by more sophisticated schemes that accomplish the same result or perpetuate past discrimination (see Rogers v. Lodge supra.; Hazelwood School District v. United States, 433 U.S. 266 [1977]. See also Griggs v. Duke Power Co., 401 U.S. 424, 427[1971]).

These principles, fashioned over many years of civil-rights enforcement, stem from the premise that racial bias in modern society can exist in a myriad of subtle and complex forms. The Justice Department's redlining claims against Decatur and Chevy Chase are fully consistent with these principles.

Statistics on market share

In applying these principles to Decatur, Justice found that the thrift had operated for many years in the Atlanta area, indeed since 1927, and that there was a large black population that lived in and around the white areas where it conducted most of its business that received few, if any, loans from the institution.

Decatur's Home Mortgage Disclosure Act (HMDA) data from 1985 to 1990 showed that the overwhelming majority (more than 97 percent) of its mortgage loans were made in majority white census tracts. Computer mapping of its loan originations over the 1985 to 1990 period showed heavy concentrations of loans in white neighborhoods throughout the Atlanta area, but no loans or only a trickle of loans in identifiably black areas.

Justice examined whether these disparities might be explained by socioeconomic differences between white and minority neighborhoods. Majority-black areas typically may include more low-income residents who cannot afford to purchase homes. They also are likely to experience lower population growth, have a higher proportion of rental dwellings, and experience fewer home sales and refinancings than majority-white areas.

In addressing that possible explanation, Justice analyzed Decatur's market share of loans in majority-white census tracts compared with majority-black tracts. This analysis effectively controlled for these explanations because it looked only at loans to presumably creditworthy borrowers. It was appropriate to examine Decatur's market share in the black census tracts because it was a large-volume lender that could reasonably be expected to compete for loans in those areas given their proximity to the white areas in which it had high market share. The market share analysis showed statistically significant disparities - more than four standard deviations under the Hazelwood test - that were unexplained.

Circumstantial evidence - CRA service area delineation

Beyond the statistics, Justice looked at a host of circumstantial evidence that suggested racial considerations influenced Decatur's marketing practices. Decatur arbitrarily excluded most of the black neighborhoods of Atlanta from its marketing territory under the CRA.

In defining its boundaries inside the City of Atlanta, the thrift followed the tracks of a railroad that had historically separated white and black residents of the city. The white areas north of the tracks were included, and the black areas south of the tracks were excluded.

To accomplish this, Decatur selectively had to use banking agency criteria under the CRA in defining its market. It chose the law's "political boundary" option for defining its territory to include virtually all of the predominantly white eastern suburbs and some rural areas using county boundaries, even though at the time (1979) it made few loans and had little branch presence in many of those areas.

In its treatment of Fulton County with its large black urban population in Atlanta, the thrift selected the CRA's "effective lending territory" alternative, which allowed it to include only areas where it made loans and had branches. Since those loans and branches were north of the railroad tracks, it could exclude the black neighborhoods south of the tracks. Almost 75 percent of the black population of Fulton County was excluded from the institution's marketing area with clearly foreseeable results - a mostly white area in which to concentrate its sales and marketing efforts.

Branch and office locations

The Justice Department next examined the history of the 43 branches and eight mortgage offices that Decatur had opened over the years. Only one was ever located in an area that was not predominantly white. That branch (Kirkwood) was a corporate response to the 1968 riots following the assassination of Dr. Martin Luther King. It was an action that the company said was driven by "social concerns" and not profits. Nonetheless, Decatur closed the branch only three years later because company officials said it was losing money.

But Justice found that it was not unusual for some of Decatur's newly opened branches to lose money in their first years of operation. Indeed, several that were opened in white neighborhoods in the 1970s were allowed to remain open after losing considerably more money than the short-lived Kirkwood branch. The only other full-service branch that Decatur Federal had ever closed in its history was a branch (Glenwood) that was opened in the 1950s when the area was predominantly white and closed in the mid-1980s after the area had become predominantly black.

Soliciting mortgage applicants

Justice Department investigators looked closely at the way that Decatur solicited Realtors and builders for new loan business - a centerpiece of its marketing strategy. The thrift relied on a cadre of account executives who used what were described as "preferred call lists" that contained the names of 590 Realtors and builders in the Atlanta area. Only one was a member of the local association of black Realtors, and only four had business addresses in black neighborhoods.

Interviews with Atlanta-area Realtors confirmed that the account executives called almost exclusively on those who operated in white areas. One black Realtor said when she worked at a real estate agency in north Fulton County, she was called constantly by Decatur Federal's account executives. When she transferred to another company in south Fulton County the contact stopped. A former Decatur account executive told Justice investigators that she was specifically instructed by the thrift for unexplained reasons not to solicit loans "south of I-20," which included many of Atlanta's black neighborhoods.

Advertising and commission system

Decatur did not use black-owned or minority-directed radio stations or newspapers to solicit mortgage business. Targeted advertising is often necessary to dispel the image in the black community that an institution's loan business is oriented to the white community, particularly if there has been a legacy of past discrimination.

Minority-directed advertising was a chief recommendation of an Atlanta mayoral commission formed after the "Color of Money" series. The commission found that failure to conduct such advertising contributed to the predominantly white flow of credit applicants to many Atlanta-area lenders.

When presented with that recommendation, Decatur rejected it, considering its advertising in the Atlanta Journal-Constitution as sufficient. Decatur's commission system, based on a percentage of the loan amount, mirrored what is typical for the industry in that it rewarded account executives for bringing in high-end loans. However, by doing so, it also operated as a foreseeable deterrent to developing business in minority neighborhoods where housing values are disproportionately lower than in most white neighborhoods.

FHA/VA lending

Decatur Federal's many ads in the Atlanta Journal-Constitution never mentioned the availability of FHA or VA loans - loans it was aware held great demand in black neighborhoods. And despite its being a HUD-endorsed FHA lender, the thrift had only a few applications for those loans - mostly from whites.

Decatur's employment profile

Justice Department investigators examined Decatur's employment profile in the key jobs that affected its loan underwriting and origination practices. The employment of few, if any, minorities in these positions may hamper a lender's ability to develop loan business in minority areas, particularly those that operate in urban areas such as Atlanta that have large minority populations and available minority workers for these jobs. Decatur employed no blacks as underwriters or staff appraisers (indeed, only one black was employed as an outside fee appraiser); only one of its 18 account executives was black, and only three of its 12 loan processors were black. While the employment of few minorities was not itself a violation of the Fair Housing Act or ECOA, these positions frequently involve close or direct contact with Realtors and mortgage loan applicants. Having mostly white employees in these positions was viewed by Justice as further circumstantial evidence that Decatur's lending practices were geared to the white community.

Historical evidence

Finally, as part of its investigation, Justice examined the history of overt racial discrimination in lending in Atlanta and other parts of the country. In the 1940s and 1950s many Atlanta lenders refused to do business with black customers, forcing them to rely on several black-owned thrifts and a life insurance company for credit.

Even into the 1960s and 1970s many lending institutions in Atlanta and other parts of the country were reluctant to make loans in minority areas because of overtly discriminatory appraisal theories that viewed the presence of blacks and other minority groups as "undesirable" influences on property values. As recently as 1974, many lenders in cities throughout the country openly acknowledged taking race into account in their marketing and lending strategies.

In 1974, HUD released a survey of 582 savings and loan associations in 50 cities that showed that 18 percent of the respondents admitted they would not make loans in neighborhoods with high concentrations of minority residents. Another 17 percent said they "considered" the presence of racial and ethnic minorities a negative factor in conducting loan business in those neighborhoods. The overtly racial appraisal theories remained in many standard industry texts until 1977, when the Justice Department, exercising its pattern or practice authority, successfully sued the appraisal industry to get them removed.

The purpose of this historical market review was not to hold lenders such as Decatur personally responsible for past practices they did not singlehandedly shape, or prove present day fair lending violations. There was no preconceived bias or guilt by association intended by this inquiry. As in many pattern or practice investigations undertaken by the Justice Department, evidence of past discrimination in an industry is an inescapable context for assessing whether recent practices perpetuate the effects of prior discriminatory policies, particularly of lenders that were in operation during times when overt discrimination existed in some quarters of the industry.

The total evidence gathered led Justice to conclude that Decatur Federal had for many years pursued a sales and marketing strategy for home loans that intentionally avoided black neighborhoods because of race. This evidence was sufficient to accuse the institution of redlining, although in settling the case Decatur denied all charges of discrimination.

Chevy Chase case is similar

The Justice Department looked at virtually the same categories of evidence in developing its redlining claim against Chevy Chase. The only difference was that Decatur was brought under the Bush administration while Chevy Chase was a Clinton administration initiative. The similarities are:

* Like Decatur, Chevy Chase operated in a metropolitan area (Washington, D.C.) that has a large black population concentrated in discrete, well-defined geographic areas of the District and Prince George's County, Maryland. Those areas and neighborhoods were near or surrounded by majority white neighborhoods and communities where Chevy Chase conducted significant mortgage business.

As one of the largest originators of home mortgages in the Washington area, it was reasonable to expect that it would include the area's black neighborhoods in its marketing plans and originate significant numbers of loans there.

* As in Atlanta, the history of residential lending in the Washington D.C. area was important to understanding how the market was served or not served by local lending institutions. In 1977, the Washington, D.C., Commission on Residential Mortgage Investment issued a report chronicling the history of disinvestment of the city's black neighborhoods. The report found that lenders refused to make loans in "changing neighborhoods" so that "[b]y the 1950s, Washington, D.C. had become a 'tale of two cities' with two distinct housing markets and two separate sources of lending." The District's black neighborhoods were for many years serviced primarily by black-owned financial institutions, such as the Industrial Bank and the United Bank of Washington, and small independent mortgage companies that specialized in FHA and VA loans.

In the aftermath of the riots in 1968 following the assassination of Dr. Martin Luther King, redlining in the Washington, D.C., area became even more widespread and prompted the D.C. City Council to enact a tough anti-redlining ordinance in 1973. During this time many lenders, including Chevy Chase's mortgage company, moved their offices outside the District to suburbs in Maryland and Northern Virginia. Some significantly curtailed their lending and branch presences in D.C. neighborhoods, except for the predominantly white northwest section of the District.

* Chevy Chase's HMDA data from 1985 to 1992 showed it made the vast majority of its loans (rarely less than 97 percent) in majority-white census tracts. Computer mapping of Chevy Chase's loan originations from 1976 to 1993 showed consistent race-based patterns similar to Decatur's. Large clusters of loans appeared year after year in white areas, such as Fairfax County, Virginia; Montgomery County, Maryland; and Northwest Washington, D.C., patterns that always stopped abruptly at the edge of most majority black neighborhoods in the District and Prince George's County. Similar patterns were shown in mapping the institution's loan applications.

These lending patterns were particularly striking to Justice investigators because, as in Decatur, the white areas where Chevy Chase had a large share of mortgage applications and loans were close to or straddled the black neighborhoods that it failed to serve. This 18-year pattern, as revealed by computer mapping, continued up to the commencement of the department's investigation in 1993.

* As in Decatur, the Justice Department conducted a market-share analysis to determine whether these lending patterns might be explained by a lower demand for mortgage loans in majority black areas because of factors such as less new home building activity and fewer qualified borrowers. This analysis showed statistically significant disparities in Chevy Chase's market share of loans in black census tracts compared with white census tracts.

For example, in 1992 Chevy Chase originated 2.1 percent of all purchase and refinance loans in majority-white census tracts but only 0.3 percent of those loans in majority-black tracts. This disparity in market share constituted 4.8 units of standard deviation, which meant that the likelihood these disparities could have occurred by chance variation in loan originations unrelated to the racial composition of the census tracts was less than one in one-half million.

The Justice Department's market-share analysis covered a seven-year period (1987-1993) and showed disparities that were as high as Decatur's - more than four standard deviations in some years. Chevy Chase's overall lending patterns, including home improvement loans and residential construction and commercial loans, showed little lending in majority-black areas.

* Like Decatur, Chevy Chase had direct endorsement authority from HUD to originate FHA loans, yet made few such loans in majority-black neighborhoods. As in Atlanta, FHA loans in the Washington, D.C., area account for a significantly higher percentage of mortgage loans in black neighborhoods than white neighborhoods. From 1985 to 1992 Chevy Chase originated 2,312 FHA loans in the Washington metropolitan area, only 69 (2.9 percent) of which were in majority-black census tracts.

* Chevy Chase had an even larger network of branches and mortgage offices than Decatur - 77 branches and 18 mortgage offices throughout the Washington metropolitan area, with considerable cross-selling between the branches and mortgage offices. Only four of Chevy Chase's branches were in majority-black census tracts at the time of the Justice investigation, and two of those branches were not opened by Chevy Chase but were acquired through the purchase of another institution. A third branch was opened at a time when the area was majority-white and the fourth was located near Annapolis, Maryland, and outside the Washington, D.C., area. Only one of Chevy Chase's 18 mortgage offices was in a majority-black census tract, and it was not opened until May 1993, shortly before the Justice investigation.

* The evidence reviewed by Justice Department investigators suggested that Chevy Chase delineated its CRA market to exclude black neighborhoods. Using the political boundary method, Chevy Chase initially included all of the District of Columbia in its service area when in 1986 it first came under CRA by joining the federal deposit insurance system.

As shown by computer mapping, since at least 1976 it made substantial numbers of mortgage loans in the District, but virtually all of those loans were in the predominantly white upper northwest area of the city. After formally identifying all of the District as part of its service area, Chevy Chase continued to originate the vast majority of its mortgage loans in upper northwest D.C. with little increase in lending in the city's black neighborhoods.

In 1989, Chevy Chase removed the District entirely from its service territory for reasons that were never explained in the minutes kept by the thrift of CRA meetings. When Justice investigators asked for an explanation, they were told that the "decision" was made in 1987 because Chevy Chase at that time had no branches in the District. This explanation seemed suspect to Justice because CRA regulations specifically permitted banks to include in their service territories areas where they did not have branches. Moreover, at the time of the thrifts 1989 resolution formally removing the District from its CRA market, Chevy Chase had recently opened one office in the District and was planning to open another, both in white neighborhoods of the District.

The thrift acknowledged during the DOJ investigation that there was no basis for its decision to remove the District in 1989. The District remained excluded from its service area until 1992 when its regulator, the Office of Thrift Supervision, criticized the exclusion during a CRA/fair lending compliance review. In response to that criticism, Chevy Chase chose to include only the predominantly white northwest part of the District rather than all of the city. This resulted in the exclusion of more than 80 percent of the District's black population from its market.

* Like Decatur, Chevy Chase relied principally on a virtually all-white staff of employees (called loan originators) to solicit mortgage applications. Only two (2.2 percent) of its 91 loan originators were black. Justice found they targeted their sales calls almost exclusively on Realtors in white neighborhoods and rarely, if ever, called on Realtors active in black neighborhoods.

* Justice obtained corroborating anecdotal information of redlining from interviews with former employees. A former Chevy Chase employee with direct knowledge of the thrift's marketing practices in D.C. told Justice investigators that he was instructed in 1980 not to solicit mortgage applications on properties in identifiably black areas, instructions that were reiterated as recently as 1991. The explanation given was that the institution "was not familiar" with properties in those areas.

* Finally, like Decatur, Chevy Chase never solicited mortgage applications through newspapers, radio stations or other media oriented to the black community. It relied primarily on the Washington Post for newspaper advertisements, yet, like Decatur, never used those ads to solicit FHA or VA loans. Its commission system also provided incentives for seeking loans on higher-priced homes, which further inhibited marketing in black neighborhoods.

So why the furor over Chevy Chase?

Some critics note that in Decatur Justice uncovered evidence of discrimination against actual black applicants for mortgage loans but no such evidence was found in Chevy Chase. They then leap to the unwarranted conclusion that there can be no redlining without proof that specific individuals were rejected for loans because they came from black areas or sought homes in black neighborhoods. These critics also point to some lower court decisions, such as Cartwright v. American Savings & Loan Ass'n, 880 F. 2d 912 (7th Cir. 1989) and Thomas v. First Federal Savings Bank of Indiana, 653 E Supp. 1330 (N.D. Ind. 1987) that required such proof where the plaintiffs claimed they were rejected for loans because of alleged redlining. The focus of these cases was on the lenders' reasons for rejecting the plaintiffs' individual loan applications rather than on broad, multi-faceted lending activities that are frequently the focus of Justice Department pattern or practice investigations.

Moreover, it would be unusual for a lender today to admit in its records that it rejected a mortgage application because of the racial makeup of the area in which the property is located, and this "smoking gun" evidence is not required to prove purposeful discrimination. As shown by Decatur and Chevy Chase, redlining can be just as effectively pursued through subtle marketing practices that avoid minority areas as by explicitly instructing underwriters to reject applications from those areas. Moreover, were a lender to convey those instructions today, they would become "red flags" for banking agency examiners and perhaps trigger complaints from the rejected applicants to HUD and Justice that would further risk revealing the lender's hidden motives.

Finally, if a lender receives few applications from minorities precisely because of racially discriminatory marketing, it may be extremely difficult to detect from the underwriting files any pattern of rejections based on property location because of the small number of files to review. Imposing such a requirement would only encourage lenders intent on redlining not to attract many applicants from minority areas, leaving only preapplication testing as a possible means of proving redlining.

The Justice Department is not alone in recognizing that discriminatory marketing practices can result in redlining. OTS fair lending regulations have embraced this concept for many years. The OTS regulations state:

"Discrimination in lending is not limited to loan decisions and underwriting standards; a savings association does not meet its obligations to the community or implement its equal lending responsibility if its marketing practices and business relationships with developers and real estate brokers improperly restrict its clientele to segments of the community" (12 C.F.R. [571.24(d)]).

As thrift institutions, these regulations were binding on Decatur and Chevy Chase.

Critics also misread the Justice Department's findings in Decatur. It's true that using a statistical model, Justice found a pattern of underwriting discrimination against black mortgage applicants at Decatur but a similar model failed to show significant discrimination in Chevy Chase. However, in Decatur, Justice found that the applicant's race, rather then the location of the property, caused these rejections, and this discrimination stemmed from subtle bias by the bank's underwriting staff about the credit qualifications of blacks. This claim was never linked to the redlining claim, which, as seen from the discussion above, was clearly focused on Decatur's marketing practices.

The effects test

Some critics allege that Justice accused Chevy Chase of failing to allocate credit proportionately among white and minority areas under a disparate impact theory. This is a loaded and unfair criticism.

First, disparate impact theory does not require credit allocation. The standard only mandates that lenders refrain from practices that operate as "artificial, arbitrary, and unnecessary barriers" to equal opportunities for minorities and other protected classes (see Griggs v. Duke Power Co. 401 U.S. 424, 431 [1971]). Under this legal standard, business practices that disproportionately and adversely affect housing or credit opportunities for protected classes, or which perpetuate racial segregation, are unlawful unless required by business necessity and there is no less discriminatory alternative. Absence of a discriminatory motive is not a defense under this standard.

The disparate impact test is recognized by virtually all of the lower courts in the United States in fair housing and fair lending cases, although the Supreme Court has yet to rule definitively on its application under the Fair Housing Act. The test also is adopted specifically by the federal banking and enforcement agencies in their Interagency Policy Statement on Discrimination in Lending, 59 Fed. Reg. 18266.

The disparate impact test appears applicable to redlining cases, and it is briefly mentioned in the Justice Department's complaint against Chevy Chase. However, critics of Chevy Chase who focus on the department's disparate impact claim blithely ignore the evidence Justice found showing intentional discrimination, namely that race influenced where Chevy Chase chose to conduct its business operations and originate mortgage loans.

What lies ahead?

It's unclear whether the Justice Department intends to open any new redlining investigations on the scale of Chevy Chase in the near future. It has not disavowed the redlining theory developed in Decatur and Chevy Chase, and it adopted a similar theory in its recent insurance redlining lawsuit against the American Family Insurance Company in Milwaukee, Wisconsin. Yet it is unclear whether the federal banking agencies have adopted the Justice Department approach in looking for possible redlining in their fair lending compliance examinations.

On February 25, 1995, the assistant attorney of the Justice Department's Civil Rights Division, Deval Patrick, sent a letter to the Mortgage Bankers Association of America (MBA)and other industry trade groups, responding to a number of fair lending questions. One of those questions asked if "niche" lenders that based their marketing decisions on economic characteristics of potential borrowers, which produces lending patterns concentrated in white areas, might be subject to prosecution. In response, the assistant attorney general seemed to say that if the lender "adopt[s] its 'niche' under circumstances from which a discriminatory purpose could fairly be inferred" it might be the target of a redlining investigation.

One issue not squarely addressed in Patrick's letter is whether Justice, at least for now, will limit its redlining investigations to lenders like Decatur and Chevy Chase that are obligated by the CRA to serve their entire communities. Clearly, that obligation makes it more difficult for banks and thrifts to justify not lending in minority neighborhoods within their service areas.

On the other hand, much of the redlining evidence in Decatur and Chevy Chase was not dependent on CRA obligations. Computer mapping and market-share analysis of a large mortgage banker not covered by CRA may show a de facto market that is equally suggestive of racial redlining. Other evidence, such as offices located only in white areas and sales calls targeted exclusively or primarily to Realtors in white areas may be directly tied to the lending and market-share disparities.

This does not mean that every lender that falls short on any of these indices, such as advertising and commission systems, is guilty of intentional racial redlining. While they may be indicators of a problem, they are not in themselves determinative of unlawful discrimination.

The Justice Department's investigations of Decatur and Chevy Chase involved painstaking analyses of the lenders' records and activities over many years and numerous interviews with Realtors and former employees who provided strong anecdotal evidence of redlining. It also remains to be seen how the disparate impact test might apply to these cases.

Private class-action litigation over alleged redlining may increase in the wake of Decatur and Chevy Chase, and community groups are likely to continue to use CRA and other means to force lenders to focus their attention on marketing loans in minority areas. Preapplication testing may also become an important vehicle for detecting redlining in investigations undertaken by HUD, Justice and possibly the banking agencies.

Things to keep in mind

As a precaution, all lenders, whether covered by CRA or not, should carefully examine their marketing strategies, keeping in mind the types of evidence Justice examined in its redlining investigations of Decatur and Chevy Chase. HUD's fair lending "best practices" agreement with the MBA and the department's individual agreements with mortgage bankers provide useful guides for developing marketing programs geared to minority areas. Other publications, such as the Federal Deposit Insurance Corporation's "Side by Side: A Guide to Fair Lending" and the Federal Reserve Bank of Boston's "Closing the Gap: A Guide to Equal Opportunity Lending," emphasize nondiscriminatory marketing practices and provide helpful advice.

However, there is no substitute for careful monitoring and critical self-analysis of marketing practices by any lender concerned about redlining. Market share studies in white and minority areas and computer mapping are essential to any such review, as are reviews of sales calls by loan originators to ensure they are not avoiding direct solicitations of Realtors in minority areas. For example, some loan originators may rely only on fax transmission to Realtors in minority areas to solicit business, but they engage in numerous personal visits to agents in white areas, accompanied sometimes by video presentations to the entire staff of the realty office.

Failure to exercise the same aggressiveness in soliciting business from agents in minority areas can only hinder the lender's ability to increase its market share in those areas. Incentives to solicit business in minority areas should be explored, such as creative salary and commission systems that encourage loan originators to prospect for loans in low- and moderate-income minority neighborhoods.

Employing minorities in loan originator positions is also important. They are more likely to be familiar with agents in minority neighborhoods. They are likely to be more familiar with how best to use targeted marketing and advertising to reach minority customers. They also may be well equipped to assist minority borrowers with credit blemishes through the underwriting process.

Richard J. Ritter is a fair lending consultant based in Washington, D.C. He was formerly a special litigation counsel in the Civil Rights Division of the U.S. Department of Justice assigned to fair lending enforcement.