This white paper also benefited from the vision and many contributions by Gerron Levi, former Senior Director of Government Affairs at NCRC. Contributions were also provided by Charlotte Saltzman, a paralegal at Relman Colfax PLLC.
The Community Reinvestment Act (CRA) was enacted in 1977. It was a response to the nation’s long and painful history of lending discrimination against people of color and the resulting disinvestment and decay in communities of color. Congress’ purpose was to counter decades of official and unofficial redlining and other forms of discrimination by government agencies and the nation’s banks. The law’s author, Senator William Proxmire, and many other supporters over the years have confirmed that this was the impetus for the law.
To counter the effects of discrimination, Congress identified banks’ affirmative duty to meet the credit needs of their entire communities and required federal regulatory agencies to evaluate and rate banks periodically on how well each is meeting that obligation. Limiting the ability to achieve CRA’s central aim, however, regulators have generally interpreted the “entire community” to focus on income. Through regulations and in practice, their chief concern has been how well banks are meeting the needs of low- and moderate-income (LMI) people and neighborhoods. Little direct emphasis has been placed on the needs of people and neighborhoods of color.
It is time for regulators to incorporate an explicit focus on race in core CRA regulations and examination procedures. This should and can be done in a manner that complements, and does not in any way supplant, the longstanding focus on LMI. There are many ways regulators can make CRA more race-conscious that comport with CRA’s language and purpose and with the Equal Protection Clause of the United States Constitution. This paper proposes several and explains why they are constitutional.
The first set of proposals is informational. We call for increasing the collection, calculation and reporting of data about lending to people and neighborhoods of color. This can be done through metrics specific to an individual bank’s performance in its assessment areas (a geographic area defined by the bank as its community) and benchmarks reflecting the demographics of, and the whole market’s collective lending in, the assessment areas. We have designed these measurements to parallel ones proposed last year by the Federal Reserve Board based on LMI.
The benchmarks will help regulators interpret each bank’s performance and enhance their understanding of the local circumstances in which banks operate, called the performance context. To further enhance that understanding, we also propose creating full-time research positions dedicated to studying the racial components of banks’ varying performance contexts, modeled on work done by the Federal Reserve Bank of San Francisco.
These proposals should not be at all controversial from a constitutional perspective because they do not, to any degree at all, allocate government benefits or impose sanctions based on race. They are comparable to a host of race-based information already collected and published by the federal government, the most basic being the census.
The second set of proposals are about how to incorporate a bank’s record of meeting the needs of people and neighborhoods of color in CRA evaluations and ratings. An Equal Protection question arises here because under most circumstances the government may not act on the basis of race. But it may do so in circumstances that satisfy the “strict scrutiny” test applied by courts. Courts might not construe explicitly race-conscious CRA evaluations as triggering strict scrutiny because good CRA ratings are not a limited government benefit — potentially every bank can earn one — but we assume for present purposes that these proposals would be subject to that test.
Strict scrutiny has two basic components. First, the government must have and provide substantial evidence of a compelling interest to justify its consideration of race. Remedying the effects of past and present discrimination is established in the courts as a compelling interest, and there is extensive evidence that discrimination continues to impair access to credit, and other banking services, for people and neighborhoods of color.
Second, the government must act in a manner that is narrowly tailored to achieving its compelling interest. That means several things of particular importance with respect to CRA. Genuine consideration must first be given to race-neutral alternatives and found inadequate. Racial quotas are anathema, but aspirational and flexible targets are proper, particularly when based on local circumstances, and when consideration is given to banks that attempt in good faith to meet race-based goals but ultimately do not. The burden on third parties — primarily non-Hispanic Whites in this context — may not be too great. And race should only be considered with respect to racial groups, bank products and geographical areas where evidence indicates that access is currently curtailed due to past or present discrimination. Each element of our proposal for incorporating race in CRA evaluations satisfies these requirements.
We make proposals for the retail lending, retail services, community development financing and community development services subtests of CRA evaluations. For retail lending, again paralleling the Federal Reserve Board’s pending LMI proposal, banks should have to reach certain statistical thresholds, or make a good faith effort to do so, to obtain a presumption of a satisfactory rating. Performance ranges could be set for the several other possible ratings, too. In the alternative, examiners could be given discretion to treat excellent performance compared to aspirational goals as a subjective “plus factor” when assigning a rating. The “plus factor” approach should be used for the other subtests. Metrics would be involved in some respects, but examiners would use their judgment with the help of guidelines to promote consistency to interpret and decide how much weight to give the numbers.
Evaluations should also consider whether banks are arbitrarily excluding neighborhoods of color from their assessment areas,which is not currently prohibited but should be. If so, this should be factored into ratings. When a bank has multiple assessment areas and only some are selected for a full-scope review in an evaluation (to the extent the distinction between full- and limited-scope reviews is retained), racial demographics should be added to the list of factors considered in deciding which assessment areas to select. Finally, if a bank chooses to be evaluated under a strategic plan approved by its regulator, it should have to conduct outreach to organizations serving neighborhoods of color when developing its plan.
To determine which parts of the country warrant these race-conscious elements in CRA evaluations, the regulatory agencies should conduct and periodically refresh a study to determine where, for what products, and as to what groups discrimination continues to distort the market. This will assure that race is only considered where it needs to be, consistent with the Equal Protection Clause.
Adoption of these proposals will give regulators a deeper understanding of whether banks are meeting the needs of their entire communities by broadening the crucial but too narrow primary focus on LMI. It will motivate banks to expand their focus in the same way. It will bring nearer the day when everyone has fair access to the critical products and services provided by banks. That is the very reason for CRA.
I. Background of the Community Reinvestment Act and Summary of Proposals to More Explicitly Consider Lending to People and Neighborhoods of Color
A. The History, Design and Successes of the Community Reinvestment Act
CRA”, enacted in 1977, places an affirmative obligation on the nation’s banks to meet the credit needs of the entire communities in which they are chartered, including the LMI households and neighborhoods in those communities. Every two to five years, each of the nation’s three bank regulators examine and release a public performance evaluation and CRA rating for each bank it regulates. Evaluations assess how well a bank is meeting its CRA obligations.
Banks are evaluated on how well they provide mortgages, small business and small farm loans, community development debt and equity financing, and financial services to the LMI borrowers and communities in their “assessment areas.” These are geographic areas defined by each bank that, by and large, include neighborhoods around a bank’s branches. In rating banks, examiners also consider their compliance with the nation’s fair lending and consumer protection laws. In addition to the reputational benefits or consequences of a bank’s CRA rating, regulators consider the rating when a bank seeks to expand and open new branches, and merge with or acquire another financial institution. Ratings range from substantial noncompliance to outstanding, and are applied at both component and institutional levels.
Together with other antidiscrimination and disclosure laws enacted during the 1960s and 1970s, CRA was designed to end and remedy years of lending discrimination by the US government and the nation’s private financial institutions which resulted in racial segregation, disinvestment and decline in lower-income neighborhoods and communities of color. Senator William Proxmire, who authored CRA, expressed the law’s intent: to “eliminate the practice of redlining by lending institutions.” Redlining denies credit to would-be borrowers in certain neighborhoods because of the neighborhoods’ racial or ethnic composition, without regard for the creditworthiness of individual credit applicants. The term comes from the red used on maps to denote neighborhoods viewed as being high-risk, including neighborhoods of color.
While de jure and de facto housing and credit discrimination were widespread by governmental and private actors, one notable example that has been extensively documented is the explicitly discriminatory work of a federal agency, the Home Owners’ Loan Corporation (HOLC), created in 1933. The agency developed “Residential Security” maps of major American cities that documented how loan officers, appraisers, and real estate professionals evaluated mortgage lending risk during the era immediately before the surge of suburbanization in the 1950s. Neighborhoods considered high risk, or “hazardous,” were often redlined by lending institutions, denying them access to capital investment which could improve residents’ housing and economic opportunities. Neighborhoods with people of color were systematically deemed higher risk.
CRA and its implementing regulations were enacted and designed to remedy the damage done by these longstanding practices and to ensure that the nation’s banks provide mortgage, small business, and other credit, investment, and financial services to affected borrowers and communities. In many regards, CRA has succeeded in remedying some of the “market failures” associated with underserved communities. In addition to racial discrimination, the law’s drafters recognized that underserved markets would continue to be beset by negative and informational externalities. For example, if lenders fear that other lenders will not lend to areas that are perceived as risky (whether they are or not), they will refrain from lending; the fear is self-fulfilling. The resulting inability to borrow money to buy and improve homes consigns neighborhoods to continuing disinvestment. The lack of home sales increases the perceived risk of neighborhoods because appraisers rely on sales for information on the value of homes when determining appropriate loan amounts. Informational externalities and asymmetries can result in delays, caution about perceived risk and banks charging higher interest rates. Lender expectations of this sort can cause a potentially viable market to suffocate from lack of credit. In the process, borrowers who may otherwise be creditworthy are denied loans and have nowhere else to turn because of the lack of entry by competitive lenders.
CRA has been effective in addressing these problems by facilitating coordination and assuring banks that they will not be the lone participants in thinly traded markets. The statute and its regulations have produced positive information externalities that allow all lenders, whether or not covered by CRA, to better assess and price for risk. By conferring an affirmative and continuing obligation on banks to help meet the credit needs of all the neighborhoods they serve, CRA has not only prompted banks to be more active lenders in LMI areas, but also important participants in multisector efforts to revitalize communities across the country.
Because of CRA, banks have made good strides in LMI markets. They have taken numerous steps, including establishing loan products geared to LMI borrowers; entering loan pooling arrangements; undertaking lending consortiums; and partnering with local groups, community development corporations and community development financial institutions (“CDFIs”) to break down barriers that impede the efficient flow of capital into LMI communities.
The numbers show significant benefits. One accounting and tax advisory firm estimated that the banking sector was the source of 85% of the $10 billion in capital committed to housing tax credit investments in 2012. When bank investors were surveyed about why they were so attracted to housing tax credit investments, they said the principal motivation was their obligations under the CRA investment test. Between 2009 and 2018, banks issued more than $2.2 trillion in home loans, and more than $564 billion in small business loans, to LMI borrowers and communities, and provided at least $2 trillion in community development financing between 1996 and 2017.
CRA and its regulations do not, however, focus on people and neighborhoods of color the way they focus on LMI people and neighborhoods. There are some limited race-conscious provisions, but explicit consideration of race in CRA evaluations and other aspects of CRA is more notable for its omission than inclusion.
B. Successfully Addressing the Continuing Consequences of Discrimination Requires a More Race-Conscious CRA
With a structure focused on income and not race, CRA has not sufficiently addressed the continuing extraordinary financial disparities that are the direct result of persistent and systemic racial bias. For example, according to 2019 survey data, White families had a median wealth of $188,200 and a mean wealth of $983,400; Black families’ median and mean wealth was $24,100 and $142,500, respectively. That is less than 15% of White families. Hispanic families’ median and mean net worth was $36,100 and $165,500. Rates of minority homeownership overall have changed little in the past 25 years, while homeownership rates for African Americans have regressed to levels lower than when the Fair Housing Act was passed in 1968.
The nation’s regulators have acknowledged the outsized racial gaps that CRA has not closed. In 2018, former FDIC chairman and current board member Martin Gruenberg recognized the successes of CRA, but also the significant disparities that remain for communities of color. Both the Republican-appointed former OCC Acting Comptroller Brian Brooks and the Democratic-appointed current Acting Comptroller Michael Hsu have recently cited the continuing challenges facing communities of color.
The Fed, in a 2020 advance notice of proposed rulemaking that is discussed extensively in this paper, makes clear that CRA is a policy tool intended to address this longstanding problem:
CRA invests the Board, the FDIC, and the OCC with broad authority and responsibility for implementing the statute, which provides the agencies with a crucial mechanism for addressing persistent systemic inequity in the financial system for LMI and minority individuals and communities. In particular, the statute and its implementing regulations provide the agencies, regulated banks, and community organizations with the necessary framework to facilitate and support a vital financial ecosystem that supports LMI and minority access to credit and community development.
CRA’s effectiveness at increasing access to credit and to the financial market for LMI individuals indicates that it can be a powerful and effective tool for increasing access among underserved borrowers. But the efficacy has been largely tied to those measures that are explicitly discussed and considered by regulators as part of the CRA examination framework. Lacking an extensive and direct focus on race, CRA’s benefits are often distributed in a manner that is not targeted at ameliorating racial inequity and inequality that exists due to past and, to a lesser degree, continuing, discrimination. To successfully address the legacy of discrimination in the United States’ banking system, CRA’s regulatory framework must be modified to explicitly consider banks’ lending, service and community development activities from the perspective of race.
C. Summary of How CRA Regulations Should Be Made More Race-Conscious
This paper discusses several race-conscious ways that CRA regulations should be modified and explains why they are consistent with the Equal Protection Clause of the United States Constitution. In many cases, we build on the Fed’s 2020 Advance Notice of Proposed Rulemaking (ANPR) by offering race-conscious approaches that are analogous to the Fed’s LMI-focused proposals. These would complement, not replace or in any way diminish, what the Fed proposes.
We summarize the proposals here. In subsequent sections we describe them in greater detail and explain why they are lawful after setting forth the constitutional framework. The final section demonstrates that regulators already possess authority to adopt these race-conscious proposals, i.e., that statutory amendment is not needed.
We first propose significant ways of increasing the collection, calculation and reporting of data about lending to people and neighborhoods of color. The data include metrics about banks’ performance in their assessment areas, as well as community and geographic benchmarks. The benchmarks convey the racial demographics of the assessment areas and of lending activity by all institutions in them (that is, by the market as a whole). They concern retail lending (mortgage, small business, small farm and consumer), branch locations and community development financing (i.e., loans and investments). This information would increase public, examiner and even banks’ own understanding of banks’ success, or lack thereof, in serving people and neighborhoods of color, and the context in which banks operate, termed the “performance context.” These measurements match the Fed’s informational proposals about LMI people and neighborhoods.
To further enhance regulatory understanding of the racial components of banks’ varying performance contexts, we next propose dedicating research positions to studying exactly that. This is intended to overcome inconsistency and lack of depth in efforts to understand performance context attendant to the current dependence on individual examiners whose expertise does not lie in this type of research. This is a problem for racial and non-racial elements of performance context.
Banks may not arbitrarily exclude LMI neighborhoods from their assessment areas. But no such express prohibition applies to neighborhoods of color, only a prohibition against unlawful discrimination in delineating an assessment area. We explain why arbitrary exclusion should be barred for both, so that banks are compelled to pay greater attention to meeting the needs of neighborhoods of color that are fairly considered part of their local communities.
We then propose how race should be incorporated explicitly into CRA evaluations themselves. This would apply when, as to specific racial groups, geographic markets and products, evidence shows that access to credit is less than it would be absent the impact of past and/or present discrimination.
For retail lending, we present two alternatives. The better one is to establish threshold levels of lending to people and neighborhoods of color that banks must reach to obtain a presumption of satisfactory on this part of the evaluation, with a constitutionally important exception for falling short despite a good faith effort. Quantitative performance ranges could also be established to generate recommendations for each rating. The less effective but still beneficial alternative is to give examiners discretion to treat strong performance relative to benchmark-based aspirational goals as a plus factor in assigning a rating, with no negative consequences for failing to meet those goals.
With respect to the evaluation of retail services and community development financing and services, we explain how qualitative consideration should be given to race using the plus factor design. Metrics would be involved in some respects, but examiners would use their judgment with the help of guidelines to promote consistency to interpret and decide how much weight to give the numbers.
Evaluations should also affirmatively review whether neighborhoods of color have been arbitrarily excluded from a bank’s assessment areas, consistent with the proposed addition of that prohibition. And when certain assessment areas are selected for a full-scope review in an evaluation (to the extent the distinction between full- and limited-scope reviews is retained), racial demographics should be added to the list of factors considered in deciding which assessment areas to select.
Finally, we propose that when a bank chooses to be evaluated under a strategic plan approved by its regulator (an option in lieu of the regular evaluation procedure), it must conduct outreach to organizations serving neighborhoods of color.
II. Constitutional Framework for Analyzing Race-Conscious Government Programs
A. Strict Scrutiny: What it Means and When It Applies
When a governmental entity acts on the basis of a racial classification and the action is challenged in court as violating the Constitution’s Equal Protection Clause, the propriety of the government’s action is generally reviewed using a strict scrutiny standard. This can include an isolated government action or a whole program. Under strict scrutiny, the government’s action is appropriate only if the statute or regulation giving rise to it is “narrowly tailored” to meet a “compelling governmental interest:”
When race-based action is necessary to further a compelling government interest, such action does not violate the constitutional guarantee of equal protection so long as the narrow-tailoring requirement is also satisfied.
Strict scrutiny is the most exacting standard of judicial review. The Supreme Court has squarely rejected the proposition that “benign” racial classifications favoring traditionally disadvantaged groups should be subject to a less demanding standard of review.
For some classifications, including those based on gender, courts employ intermediate scrutiny, requiring that the law be “substantially related” to or “substantially further” an “important” government interest.
For classifications not based on suspect grounds, courts employ rational basis review, under which they ask whether the regulation is rationally related to a “legitimate” government interest.
The Supreme Court has made clear that the application of strict scrutiny varies based on context:
“Context matters when reviewing race-based governmental action under the Equal Protection Clause. . . . Not every decision influenced by race is equally objectionable, and strict scrutiny is designed to provide a framework for carefully examining the importance and the sincerity of the reasons advanced by the governmental decisionmaker for the use of race in that particular context.”
Of great importance to consideration of race in CRA context, courts recognize that the government has a compelling interest in remedying the effects of past or present discrimination. This is especially true for discrimination either caused by the government (e.g., through facially discriminatory laws, regulations or practices), or in which the government is a “passive participant” (e.g., distributing funding in a market that suffers from discrimination in a way that does not address or remedy that discrimination). The target must be specific, though, because “claims of general societal discrimination—and even generalized assertions about discrimination in an entire industry—cannot be used to justify race-conscious remedial measures.”
The government must also have a “strong basis in evidence” to support its compelling interest. When the interest is addressing discrimination, this means evidence that discrimination has occurred and requires remedial action. The government must identify evidence of current exclusion from the precise activity or area at issue that is the result of discrimination, demonstrating why a race-conscious remedy is necessary to eliminate that exclusion.
Assuming the government can establish a compelling interest, it must then demonstrate that the action, program or law at issue is narrowly tailored to meet that interest. The inquiry is intended “to ensure that the means chosen ‘fit’ the compelling” interest such that “there is little or no possibility that the motive for the classification was illegitimate racial prejudice or stereotype.” Factors courts consider when determining whether a law is narrowly tailored include whether the law is over- or under-inclusive; the burden created on third parties; the duration of the law (or whether it includes provisions for reevaluation); how “flexible” the program is (including how individualized the consideration given to potential beneficiaries is, provisions allowing for the withholding or granting of benefits for non-race-conscious reasons, and provisions allowing non-minorities to receive the relevant benefit in certain circumstances); and whether the government has considered or previously tried race-neutral approaches and found them inadequate. These factors function as safeguards to assure that race is not used in a manner that exceeds constitutional legitimacy. Being on the wrong side of the analysis on just one of them is likely enough for a court to find a use of race unconstitutional.
B. Principles Taken From Application of Strict Scrutiny to Race-Conscious Programs in Other Contexts
Although strict scrutiny is an exacting standard, it is not inherently or necessarily fatal to the laws or regulations being evaluated. The courts have upheld race-conscious programs in a variety of contexts, including higher education admissions and government contracting, in the past 20 years. We next describe some of these cases to discern principles that support or cabin the potential explicit consideration of race under CRA.
The most helpful area of case law for a race-conscious CRA likely lies in government contracting affirmative action cases. For more than 30 years, legal challenges have been raised relating to a variety of government programs intended to increase the proportion of minority-owned businesses that receive (sub)contracts for government-funded construction projects. These programs generally provide some form of preference to minority-owned businesses during the contracting process and have been justified using a variety of more and less successful methods. The central Supreme Court decisions in this area are City of Richmond v. J.A. Croson Co. (Croson) and Adarand Constructors, Inc. v. Peña (Adarand).
Croson involved an ordinance adopted by Richmond, Virginia, requiring 30% of construction contracts to be awarded to minority-owned businesses. There were some facts and findings supporting the program; notably, Congress had determined 20 years prior that racial discrimination was common in construction, and Richmond had found that minority-owned businesses received less than 1% of construction contracts, despite the fact that the city’s population was 50% minority.
The Supreme Court found that was insufficient to justify Richmond’s race-conscious program. The Court observed that the city had not determined the number or proportion of minority-owned businesses in Richmond’s construction market and had no evidence that qualified minority contractors had ever been passed over for contracts. Thus, the city could not say how many contracts minority enterprises would be getting if discrimination exerted no influence, making the 30% figure arbitrary. The problem, the Court found, is that “generalized assertions about discrimination in an entire industry” do not permit properly tailored remedies. That is, a general finding that “there has been past discrimination in an entire industry provides no guidance for a legislative body to determine the precise scope of the injury it seeks to remedy. It has no logical stopping point.”
The Court was also troubled by the inclusion of minority groups against which there was “absolutely no evidence of past discrimination . . . in any aspect of the Richmond construction industry.” This suggested to the Court that “perhaps the city’s purpose was not in fact to remedy past discrimination.”
Seven years later, Adarand, involved a remedial federal program in which prime contractors were given a financial incentive to hire subcontractors certified as small businesses controlled by “socially and economically disadvantaged individuals.” The program included a presumption that minorities were socially and economically disadvantaged. Although the Supreme Court did not decide the merits of Adarand, its discussion in that case has been of critical importance in the years since. Among other elements of the program, the Court specifically discussed the fact that non-minorities were given the opportunity to prove their social disadvantage, and that the presumptions of social and economic disadvantage were rebuttable. This has been an important feature of race-conscious government contracting programs that have survived strict scrutiny in lower courts.
When remanding the case to the Court of Appeals, the Supreme Court indicated that the lower court should “address the question of narrow tailoring” by asking, among other things, “whether there was any consideration of the use of race-neutral means to increase minority business participation in government contracting” and “whether the program was appropriately limited such that it will not last longer than the discriminatory effects it is designed to eliminate.” The Court also focused on the lack of clarity around whether individualized inquiries were necessary to determine economic disadvantage, and how such a showing would be made.
From these cases and lower court cases that followed, a few key considerations emerge.
First, because the compelling interest in contracting cases has to do with remedying the effects of discrimination, the program must be designed and tailored in a manner that is tied to discrimination that is identified with meaningful specificity. This means that broad societal prejudice is insufficient to support an affirmative action program; instead, there must be some evidence of racially motivated discrimination within the economic sector (and geography) in which the program will operate, and this evidence must rise above general assertions of the existence of discrimination in an industry or sector.
The necessary showing can be made with a combination of statistical and anecdotal evidence. The statistical evidence should go beyond simple bottom-line disparities in outcomes for racial groups, such as different rates of approval for mortgage applications. It might do so by, for example, controlling for factors that are relevant to mortgage approval through a regression analysis. Anecdotal evidence might take the form of, for example, interviews with individuals about their personal experiences. If statistical disparities are large enough — the Supreme Court has used the terms “significant” and “gross” — they may provide sufficient evidence of discrimination to demonstrate the government’s requisite compelling interest without the addition of anecdotal evidence. Statistical evidence is the more important of the two.
Second, programs that involve aspirational goals or flexible targets, rather than quotas, meet with more favor from the courts. A program oriented toward achieving full and fair participation (participation that would reflect expected participation absent discrimination and its effects) is permissible—but a program that requires a certain level of minority participation to be achieved, regardless of how feasible that may be and without regard for any good-faith efforts that may have failed, will likely be considered too inflexible to withstand scrutiny.
Third, the less of a burden caused to third parties outside of the race intended to benefit from a program, the more likely the program is to withstand constitutional scrutiny. Flexibility in the program such that an individual is not penalized or prevented from receiving a benefit despite attempts to comply with the program will tend to support the validity of that program (for example, waivers for good faith attempts to hire minority subcontractors, or individualized mechanisms to qualify for or be disqualified from an affirmative action program, which also speak to flexibility). On the other hand, if a purportedly flexible program effectively operates as a categorical grant or denial of a benefit based on a person’s race, the program is unlikely to withstand scrutiny.
Fourth, it is critically important that the governmental body putting forth the race-conscious program has considered the feasibility of race-neutral alternatives to meet its compelling interest, potentially including attempting race-neutral solutions without sufficient success. This overlaps with the requirement (under the compelling interest part of the analysis) of demonstrating the necessity of a race-conscious approach, lending credibility to the government’s assertion that taking race into account is necessary to remediate the effects of racial discrimination.
Higher Education Admissions
Unlike in contracting, the government interest that frequently undergirds affirmative action in the educational context, including admissions to institutions of higher education, is an interest in diversity itself, an end that has been recognized by courts as compelling only in limited circumstances. This difference has significant ramifications for the application of the strict scrutiny standard. Though some are less relevant to the CRA context in which ameliorating the effects of discrimination is the animating goal, certain elements of these cases are nevertheless important.
Similar to the concept of flexibility in contracting cases, flexibility and individual consideration are paramount in the educational context. Programs that include numerical quotas or caps (even if they are based on local demographics) are highly suspect. Courts have shown serious skepticism that the goals of institutional diversity require quotas and have found that quotas designed to reflect local demographics are not sufficiently related to the interests institutional diversity is intended to serve.
On the other hand, programs that focus on individual consideration and take a holistic view both of institutions and applicants to those institutions meet with significantly more favor. In particular, programs that consider race as one of many factors, without recourse to a quota or inflexible point systems (i.e., a method that gives an automatic and pre-determined, often numerical, boost to applications from people of color to make them more competitive), are supported in court decisions. As the Supreme Court explained in upholding the University of Michigan Law School’s race-conscious admissions policy in Grutter v. Bollinger in 2003:
When using race as a “plus” factor in university admissions, a university’s admissions program must remain flexible enough to ensure that each applicant is evaluated as an individual and not in a way that makes an applicant’s race or ethnicity the defining feature of his or her application. The importance of this individualized consideration in the context of a race-conscious admissions program is paramount.
Analogously, it is reasonable to expect that a race-conscious CRA requiring a particular number or percentage of loans be made to minority borrowers, or in minority areas, will not be viewed as sufficiently holistic or flexible; likely neither would a system that automatically boosts a CRA rating in a predetermined manner based on specific race-related lending activities. Instead, considering race in the CRA context is more likely to be permissible if minority lending is considered as one of many factors intended to determine how successfully a financial institution is meeting the needs of its entire community—including borrowers and neighborhoods of color, and in a non-formulaic way.
COVID Relief Programs
Strict scrutiny for programs based on race extends well beyond government contracting and higher education admissions. Although the application of strict scrutiny varies — including to reflect the nature of the government interest at stake — many of the considerations more relevant to CRA carry over across contexts.
In recent months, race-conscious programs arising from COVID relief efforts related to Small Business Administration (SBA) grants and forgiveness of Department of Agriculture (USDA) loans, for example, have been subjected to strict scrutiny by the courts. In both, the program at issue was found not to satisfy strict scrutiny (or to be unlikely to do so), but the context of the cases and details of the decisions help underscore the lessons taken from other contexts and provide additional guidance for when race-conscious programs are permitted.
Vitolo v. Guzman concerned a challenge to COVID relief for minority-owned restaurants, a $29 billion program passed this year as part of the American Rescue Plan Act. For the program’s first 21 days, Congress limited grants to priority applicants — restaurants owned or controlled at least 51% by women, veterans or socially and economically disadvantaged people. Other applicants would not receive grants if the funds ran out in that time. The program includes a rebuttable presumption that certain groups are socially disadvantaged: Black, Hispanic, Native, Asian Pacific, and Subcontinent Asian Americans all receive this presumption. Those not entitled to the presumption can only be deemed socially disadvantaged by demonstrating that they have experienced discrimination by a preponderance of the evidence.
The Sixth Circuit granted a preliminary injunction against the priority period. It held that the government had not adequately shown a compelling interest. In the court’s view, the government relied too generally on “societal discrimination against minority business owners” without “identify[ing] specific instances of past discrimination” It also found inadequate the government’s evidence of intentional discrimination against many of the groups who received the presumption of social disadvantage.
The court was also not satisfied as to narrow tailoring. Of central importance, it held that in this context, the presumption of social disadvantage effectively guaranteed eligibility for minority-owned businesses and ineligibility for non-minority owned businesses (unless women- or veteran-owned). It also cited “obvious race-neutral alternative[s]” in determining that the program was both over-and under-inclusive.
Vitolo is a reminder of the necessity for careful and thorough documentation of discrimination specific to an economic sector or a particular activity to properly support a race-conscious program and to properly tailor the remedial program. It also underscores the importance of ensuring flexibility in the program by preventing any program element from operating as an automatic grant or denial of eligibility.
Wynn v. Vilsack also addressed a program in the American Rescue Plan Act, this time one forgiving federal loans owed by “socially disadvantaged farmers and ranchers.” It emphasizes the importance of considering other remedial efforts that may have altered the playing field and ensuring that the program is flexible and responsive to actual conditions on the ground.
The court recognized the “dark history of past discrimination against minority farmers” by the USDA. But it found that this did adequately support a compelling interest because the evidence did not account for federal attempts to remedy that discrimination and consider whether they had been successful. Although the government argued that prior remedial measures had fallen short, the court determined that the presently existing barriers the government pointed to — for example, weaker financial condition of Black farmers pre-pandemic, disparities in default rates, and weaker credit profiles — were not adequately shown to be the effects of discrimination. Still, the court recognized that the government might be able to make out a compelling interest later in the case when the evidentiary record was further developed, either with evidence that attempts at remediation had in fact been insufficient to remedy past discrimination, or that the government remained a passive participant in discrimination.
The court’s biggest concern was narrow tailoring. The program did not use a presumption of social disadvantage or require economic disadvantage. Rather, “[t]he debt relief provision applies strictly on racial grounds irrespective of any other factor.” The court called this a “rigid, categorical, race-based” program and “the antithesis of flexibility,” and it was fatal. It also objected that the loan forgiveness program “fail[ed] to provide any relief to those who suffered the brunt of the discrimination” that the government had identified, i.e., people who could not get a USDA loan in the first instance or who no longer had one because of discrimination. As such, the program was inflexible and failed to address the discrimination that the government relied on for its compelling interest.
The lessons from Vitolo and Wynn must be taken seriously, but as discussed in this paper, CRA involves a very different type of program and context than these COVID relief measures. That is highly consequential for a strict scrutiny analysis.
C. It Is Likely, Though Not Certain, That Courts Would Apply Strict Scrutiny If Regulators Consider Race When Determining CRA Ratings
If CRA regulations are amended to take into account the race of loan recipients when determining an institution’s CRA rating, it is likely that a court hearing a challenge to those regulations will evaluate the reliance on race using strict scrutiny. There are two significant distinctions, however, that could lead to a different standard of review or, at a minimum, would likely affect how strict scrutiny is applied.
First, unlike many contexts in which the government takes race into account when acting, it is not the race of the would-be recipient of the benefit that figures in the government’s provision of a direct benefit. In the contracting context, the government may decide whether to provide contracts and funds based on minority subcontractors who will be hired on to participate in the project. In the educational context it is the race of the would-be student herself that is the relevant criterion. In the context of CRA, the race of a financial institution’s owners or top managers would be immaterial. A good CRA rating could be attained without regard to the race of the top shareholder, the CEO or anyone else responsible for the institution.
Second, unlike the race-conscious programs at issue in the majority of case law in this area, CRA does not deal in limited resources or limited benefits. There are finite amounts of government contracts and contracting dollars available; schools cannot admit all comers. In this way, traditional affirmative action-type programs involve a zero-sum game: if one business is awarded a contract, it is at the expense of another business that might have received the contract instead. In contrast, good CRA ratings are an unlimited resource: there is no reason every financial institution cannot receive a satisfactory or outstanding rating. Put another way, one financial institution receiving a good CRA rating does not burden another financial institution in the “if you get it, I cannot,” zero-sum sense of burden.
These considerations could support a conclusion that strict scrutiny does not apply to the consideration of race in arriving at CRA ratings. Strict scrutiny is often discussed in connection with programs that differently benefit and burden people based on race and create “racial classifications.” Whether race-conscious ratings would allocate benefits or burdens based on race, or could properly be described as racial classifications, are fair questions to which Equal Protection jurisprudence does not provide clear answers. If strict scrutiny did not apply, CRA would be evaluated under the much more permissive rational basis standard of review.
Nonetheless, in light of broad statements by the Supreme Court and lower federal appellate courts about the need for strict scrutiny analysis any time a government acts on the basis of race (all governmental uses of race are subject to strict scrutiny) and the current makeup of the Supreme Court, the more likely outcome is that making race a component of CRA ratings and requirements would be reviewed under a strict scrutiny standard. Because a bank would have a better chance to receive a good rating based on how it performs in connection with the race of other individuals, we expect courts would look to these expansive statements and find strict scrutiny triggered. For purposes of this article, we presume so.
Rather than holding that rational basis review applies, a court more likely would make use of the distinctions between CRA and other contexts to shape the form strict scrutiny takes. As the Supreme Court has indicated, strict scrutiny varies by context; this variability is reflected in how the doctrine is applied in a given case. The fact that, due to the unlimited nature of the benefit, any burden on third parties is significantly curtailed if not eliminated makes it more likely that CRA would withstand strict-scrutiny review. And the fact that it is not the race of the beneficiary that is relevant in a CRA rating likewise should improve the program’s chances of passing strict scrutiny.
D. Strict Scrutiny Should Not Apply to Regulators’ Collection and Publication of CRA Information About Race
Regulatory collection and publication of data and other information related to minority lending, on the other hand, should not be subject to strict scrutiny. Such collection and reporting is common and its importance is well established. Reports that deal with racial demographics include ones as fundamental as the census itself. Without reporting about race, the government would be significantly hindered in its ability to enforce civil rights and other important statutes and to make sound policy decisions.
The federal Home Mortgage Disclosure Act (HMDA) is the most obvious example of such reporting in the lending context. Mortgage lenders must collect and report to the government application-level information that includes applicant race and census tract demographic information; the government publishes much of it, including the name of the lender. This serves many important purposes. As a recent CFPB report states, “HMDA data are used to assist in determining whether financial institutions are serving the housing credit needs of their local communities; facilitate public entities’ distribution of funds to local communities to attract private investment; and help identify possible discriminatory lending patterns and enforce antidiscrimination statutes.” There are many other examples of the government gathering and publishing data regarding race. Indeed, the Office of Management and Budget publishes guidelines for how to collect and report race-related data.
Gathering and publishing data about race, and publishing reports based on that data, do not amount to or directly affect the dispersal of government benefits. Nor do they entail the use of race in a manner that somehow burdens the rights of a racial group. In other words, they do not constitute the government “acting on the basis of race” in a manner that implicates strict scrutiny. This distinguishes them — by far — from the types of racial classifications by the government that has implicated strict scrutiny review by the courts. It would be good policy to make CRA reporting about race more robust as recommended below, and it should not cause any constitutional concern.
III. Regulators Can Increase Their Understanding of the Role of Race in a Bank’s Performance Context Without Triggering a Heightened Standard of Judicial Review
Banks operate in local communities which differ in many ways that are relevant to a community’s credit needs. In regulatory parlance, this is part of the “performance context.” This section describes important ways regulators can, without triggering strict scrutiny or any other heightened standard of review, systematically improve their understanding, and the public’s understanding, of the racial components of a bank’s performance context. These approaches should not be controversial from a constitutional perspective.
A. Expanding the Fed’s Proposal to Include More Information Regarding Race
The Interagency Q&A Regarding CRA defines performance context as “a broad range of economic, demographic, and institution- and community-specific information . . . .” It includes:
(1) Demographic data on median income levels, distribution of household income, nature of housing stock, housing costs, and other relevant data pertaining to a bank’s assessment area(s);
(2) Any information about lending, investment, and service opportunities in the bank’s assessment area(s) maintained by the bank or obtained from community organizations, state, local, and tribal governments, economic development agencies, or other sources;
(3) The bank’s product offerings and business strategy . . .;
(4) Institutional capacity and constraints, including the size and financial condition of the bank, the economic climate (national, regional, and local), safety and soundness limitations, and any other factors that significantly affect the bank’s ability to provide lending, investments, or services in its assessment area(s);
(5) The bank’s past performance and the performance of similarly situated lenders;
(6) The bank’s public file . . . and any written comments about the bank’s CRA performance . . .; and
(7) Any other information deemed relevant by the [regulator].
i. Retail Lending
For LMI, the Fed proposes to move toward more systematic use of the retail lending performance context in examinations by establishing a series of quantitative “benchmarks.” The benchmarks would tell regulators what portion of the potential credit recipients in an assessment area are LMI and in LMI areas, and what portion of actual credit recipients are LMI and in LMI areas. They generally parallel information currently termed “comparators” and used by examiners, but in a less structured way than the Fed is now proposing.
There are parallel race-related data points for all of these benchmarks, but the Fed does not currently calculate or use them as comparators, or propose to do so as benchmarks. Calculating race-based benchmarks, alongside the LMI ones, would significantly improve regulators’ understanding of banks’ varying performance contexts.
To be more specific, the nature of each LMI benchmark in the Fed’s proposal has three components: community or market; geographic or borrower; and line of business:
Community benchmarks concern the demographics of an assessment area. Market benchmarks concern the demographics of loan recipients aggregated across all lenders.
Geographic benchmarks compare an assessment area’s LMI census tracts to the assessment area as a whole. Borrower benchmarks compare an assessment area’s LMI residents, small businesses, small farms and loan recipients to those of the assessment area as a whole.
There are sets of benchmarks for each of four lines of business: mortgage, small business, small farm and consumer.
This results in sixteen benchmarks, four for each line of business. In mortgage, for example, the community/geographic benchmark would be the percent of owner-occupied homes in an assessment area’s LMI census tracts. The market/geographic benchmark would be the percent of mortgages made in the assessment area’s LMI tracts by all lenders. The community/borrower benchmark would be the percent of families in the assessment area that are LMI. Finally, the market/borrower benchmark would be the percent of mortgages made by all lenders in the assessment area to LMI borrowers.
Identifying counterparts based on race instead of LMI is straightforward. Again using mortgage to illustrate, the community/geographic benchmark would be the percent of owner-occupied homes in an assessment area’s census tracts that are either majority- or disproportionately minority. The market/geographic benchmark would be the percent of mortgages made in the assessment area’s majority- and disproportionately minority tracts by all lenders. The community/borrower benchmark for mortgage would be the percent of families of color in the assessment area. The market/borrower benchmark would be the percent of mortgages made by all lenders in the assessment area to borrowers of color.
The benchmarks would provide reference points for a bank’s own lending. To allow for comparison, the Fed proposes to calculate a geographic distribution metric and a borrower distribution metric for each of a bank’s major product lines. The geographic metric would be the number of the bank’s loans in LMI census tracts divided by the total number of its loans in the assessment area. The borrower metric would be the number of its loans to LMI borrowers divided by the same figure.
Again, there are clear analogues based on race: the number of loans in majority- and disproportionately minority census tracts divided by the total number of loans, and the number of loans to minority borrowers divided by the total number of loans. Adding these metrics would provide important information about a bank’s success, or lack thereof, in extending credit to people and neighborhoods of color.
ii. Retail Services (Branch Locations)
For large banks, the Fed also proposes branch distribution benchmarks for retail services. Three are community benchmarks: the percentage of census tracts, households and total businesses in an assessment area by tract income level (low, moderate, middle, upper and unknown). The fourth is a market benchmark: the percentage of all the bank branches in the assessment area by tract income level.
Like the retail lending benchmarks, these branch distribution benchmarks are easily adjusted to address race. Instead of dividing census tracts by five income levels, they could be divided by tract minority percentage (e.g., 0-20%, 20-40%, etc.). One community benchmark would be the percentage of census tracts in the assessment area that are 0-20% people of color, etc.; one would be the percentage portion of households in tracts that are 0-20% people of color, etc.; one would be the percent of businesses that are in tracts that are 0-20% people of color, etc. And the market benchmark would be the percentage of the assessment area’s bank branches in census tracts that are 0-20% people of color, etc. Using hypothetical numbers, and incorporating a hypothetical bank’s own branch distribution as the Fed proposes to do with respect to LMI, all of the branch distribution benchmarks might be combined in one chart as follows:
iii. Community Development Financing
The Fed also proposes community development financing metrics and benchmarks for large retail banks. The metric would be the dollar amount of a bank’s community development loans and investments in the assessment area divided by its deposits in the assessment area.
There would be a local and a national benchmark. The local benchmark would be the total of all large retail banks’ community development loans and investments in the assessment area divided by their collective deposits in the assessment area. There would be two variations of a parallel national benchmark, depending on whether the assessment area is in a metropolitan or nonmetropolitan area. It would be all large retail banks’ community development loans and investments in all metropolitan (or nonmetropolitan) areas in the country divided by their deposits in those areas.
We propose a complementary variation of the community development financing benchmarks and metric focused on neighborhoods with a substantial population of color. The numerator for the benchmark calculations would be community development loans and investments by large retail banks in majority- and disproportionately minority census tracts of the assessment area (or metropolitan/nonmetropolitan parts of the country). The denominator would remain all deposits, whether in the assessment area or metropolitan/nonmetropolitan parts of the country. The community development financing metric would be the bank’s community development loans, and investments in the same tracts in the assessment area, divided by all of its deposits in the assessment area.
The metrics and benchmarks would provide a more precise understanding of how effectively community development funds are reaching neighborhoods of color, not just assessment areas as a whole. The ones proposed by the Fed are important to see whether local funds are being diverted to entirely distinct regions, but the high-level picture they paint would be incomplete without others that zoom in on the neighborhoods with a heightened need for the funds.
iv. Dashboards for Banks
The Fed proposes to provide an online portal for banks with dashboards showing some of the LMI information for each of a bank’s assessment areas and product lines. For each of a bank’s major lines of business, the proposed dashboard would incorporate the calculation of “thresholds” (based on the benchmarks) used to determine whether the bank receives a presumption of a satisfactory rating. How parallel race-based benchmarks and metrics could similarly and properly be incorporated in examinations is discussed supra in Section V.D. Here we only address the more limited step of presenting the racial information by dashboard in a way that is easily understood, and useful for banks and regulators alike.
For a given line of business, the dashboard could show: (1) the borrower distribution metric; (2) the two borrower benchmarks (community and market); (3) the geographic distribution metric; and (4) the two geographic benchmarks (also community and market). As with the LMI proposal, updating the figures based on data for the prior quarter or year would make them reflect not only the local community, but also “changes in the local business cycle.”
In this hypothetical example, the dashboard would quickly and clearly convey that the bank’s mortgage lending in neighborhoods of color does not stand out as good or bad, but that it is having difficulty reaching borrowers of color. This information would help guide the bank in assessing its strategies for serving its entire community, and examiners in reviewing the bank’s responsiveness to the entire community.
The Fed proposes to publish the community development financing benchmarks on publicly available dashboards “to provide the most transparency and clarity to banks to allow them, and the public, to track their performance.” That should include the additional benchmarks proposed above focused on funding in neighborhoods of color. Each large retail bank’s own dashboard should also include its own metrics in an easily understood format, so it can tell how well it is performing relative to its peers.
v. Making Data Available to the Public
In addition to ascertaining the data points concerning race and making at least some of them available on banks’ dashboards, regulators should publish the data as part of the public portion of each written CRA performance evaluation. This would allow all stakeholders and interested members of the community to better understand how effectively the examined institution, and the market as a whole, are serving the needs of individuals and neighborhoods of color (which is necessary to understand how well they are serving the credit needs of their whole community). This information would be an invaluable complement to the longstanding public disclosure of HMDA data about mortgage lending, and the forthcoming public disclosure of small business lending data regarding race under section 1071 of the Dodd-Frank Act.
The Massachusetts Division of Banks already includes similar data in its public CRA evaluations. For example, it has included granular mortgage application information broken out by race and ethnicity, alongside the same data aggregated from all lenders in the assessment area and each group’s percentage of the population (e.g., that the examined bank received 0.9% of its applications from African Americans, that the market as a whole received 1.4% of its applications from African Americans, and that African Americans comprised 2.7% of the population).
B. Dedicating Full Time Research Positions to Studying How Race Affects Performance Context
As the Fed’s ANPR explains, the performance context is not simply the data points discussed above. In addition to more types of quantitative data, it includes qualitative information such as community investment needs and opportunities, and CRA strategies that have proved effective locally. This type of information can be obtained from, for example, interviews with community stakeholders.
Typically, developing an understanding of and writing up the performance context for a CRA exam falls to the assigned examiner. As researchers from the San Francisco Fed explain, this creates problems. First, bank examiners are not researchers, and so there is a skills mismatch between examiners and the task of properly researching and identifying the performance context. Second, because the same or overlapping performance contexts are assessed by different people, they are “likely accessing different data sources, which may create a lack of consistency across exams.” This decentralized structure inevitably diminishes the ability of regulators to gain a deep, comprehensive understanding of the performance context and, consequently, of bank performance.
The San Francisco Fed responded with an approach that can readily be adapted to race. It created a formal partnership across departments, assigning one researcher to work full-time “developing the performance contexts for all CRA exams conducted within the Federal Reserve’s 12th district.” The researcher works with outreach managers across the district to leverage their community contacts. Centralizing development of the performance context this way has “created consistency across exams in terms of data sources and analysis, making it easier to do bank-to-bank or year-to-year comparisons,” and allowing examiners to be more up to date about changes that “affect banks’ CRA activities and opportunities.”
The Fed should adopt a similar approach with respect to race so that its CRA exams are informed by a greater understanding of how race affects community and individual banking needs. Specifically, it should dedicate a full-time research position in each of its twelve districts. Better yet, it should make this a coordinated effort with other federal financial regulators to assure that important knowledge is shared by all and that examinations are consistent. The researchers should focus on, and have expertise regarding, lending to communities and borrowers of color. Such communities should include those that are majority-minority and disproportionately minority for the same reasons discussed earlier in connection with benchmarks and metrics. The researchers would work closely with examiners and any other researchers who, like at the San Francisco Fed, are dedicated to analyzing performance context. The public portion of written CRA evaluations should be augmented to include these more thorough and comprehensive performance context understandings, in addition to information already included. This would extend the benefit from regulators and banks to all stakeholders and the public at large.
A priority for these new research positions should be establishing an information collection project for race like one previously maintained by the San Francisco Fed’s Community Indicators Project, which focused on LMI households and communities. The San Francisco Fed semi-annually surveyed “representatives from banks, nonprofits and community based organizations, foundations, local governments, and the private sector” “to collect insights from community leaders about the conditions and trends affecting low-income households and communities.” The project also incorporated data about employment, education, personal investments, rents and housing. The results were published for the public.
A parallel indicators-type project should be set up to better understand conditions and trends affecting households and neighborhoods of color. It should aim to identify areas in communities of color where banks are not adequately serving lending, banking and community development needs, to the extent those needs are not being met, and approaches that have been successful in meeting them. A project like this is essential so that policymakers, CRA examiners, community stakeholders and the public can thoroughly assess how well banks are meeting CRA’s goal of assuring that the needs of the “entire community” are well-served.
IV. A Compelling Interest Supports the Explicit Consideration of Race in CRA Requirements and Evaluations
As discussed above, the government has a compelling interest that supports taking race-conscious steps to ameliorate the continuing effects of past or present discrimination. This includes assuring that it does not distribute benefits in a manner that perpetuates those effects. These propositions are clear from Supreme Court precedent, and undoubtedly apply to discrimination in lending.
As also noted above, that is not the end of a compelling interest inquiry under strict scrutiny. The validity of the interest must be shown, not just asserted. That is, there must be a “strong basis in evidence” of the need for race-conscious remedial action to address the discrimination. This means robust and focused findings that identify current exclusion, grounded in discrimination, from the precise activity at issue. The evidence must allow a reviewing court to understand why the particular remedy enacted is needed to cure the exclusion. As the Supreme Court has explained:
Proper findings in this regard are necessary to define both the scope of the injury and the extent of the remedy necessary to cure its effects. Such findings also serve to assure all citizens that the deviation from the norm of equal treatment of all racial and ethnic groups is a temporary matter, a measure taken in the service of the goal of equality itself.
This standard is focused on the need for a government response and does not require “definitive proof of discrimination.”
For a national program, the compelling interest may be demonstrated at the national level. Whether, and in what form, the program is justified at the local level is then addressed as part of the narrow-tailoring analysis.
As we now show, there is a strong basis in evidence of the need for race-conscious remedial action in connection with the types of lending covered by CRA. The evidence so demonstrating is much too extensive to describe or cite here. Appendix A to this paper provides references to NCRC publications in this area, which in turn cite extensive work by a number of groups and experts who have contributed to this research and literature over the years. These publications supplement and complement the citations within the body of this paper, which further support the strong basis in evidence.
A. There is Extensive Evidence of Public and Private Discrimination Affecting All Types of Credit Covered by CRA
As discussed above, CRA was created to address the impact of discrimination on US credit and financial markets. Both private and public actors engaged in widespread redlining, building discrimination into the foundation of modern credit markets.
Through much of the twentieth century, the federal government and private lenders denied communities of color access to prime mortgage products, significantly curtailing homeownership opportunities for Americans of color, and allowing discriminatory practices that both stripped wealth from and prevented the accumulation of wealth by residents of neighborhoods of color. In the 1930s, the federal government created the HOLC and the FHA. The HOLC assisted homeowners who were in default on their mortgages and in foreclosure. The FHA provided direct assistance to individuals to finance home purchases and federal insurance to back private mortgage lending.
HOLC and FHA programs expanded access to affordable mortgages, assisting millions of families to obtain the financing needed to own their own homes. But people of color were systematically excluded from these programs from the start. This meant they were also largely excluded from private credit markets, because private lenders depended heavily on government programs.
HOLC surveyed neighborhoods based on lending risk and created Residential Security Maps that graded neighborhoods from A (best) to D (hazardous). Minority presence in a neighborhood systematically led to lower grades; the presence of Black people or other “inharmonious” racial or social groups could lead to a D grade. D-grade neighborhoods were depicted on HOLC maps in red (i.e., redlined), to indicate that lending in those areas was risky and discouraged. These maps were used by the FHA.
In 1939, the FHA outlined its core principles in a piece called “The Structure and Growth of Residential Neighborhoods in American Cities,” which “emphasized the importance of maintaining racial segregation.” The FHA also “gathered extensive data on the racial composition of neighborhoods and instructed financial institutions not to lend to households in integrated or predominantly African American areas.”
Historically redlined neighborhoods continue to face significant hurdles. Wealth, health and many other disparities remain pronounced. The redlining and de jure discrimination of the twentieth century is still felt acutely today, including in credit and financial markets.
Beyond the redlining and official discrimination that began in the 1930s and continued for decades, extensive evidence makes plain that there has been discrimination and disparity in the housing market and in mortgage, refinance and home improvement credit markets. There has been systemic and extensive discrimination in farm credit, including by the United States Department of Agriculture. And there has been discrimination in small business lending. The government has engaged in discrimination explicitly, and the government also has distributed benefits in a manner that is affected by (and sometimes perpetuates) discrimination by private entities.
This discrimination did not end with demise of explicit redlining, or with the passage of the Fair Housing Act in 1968, the Equal Credit Opportunity Act in 1974 or CRA in 1977. Although discrimination may have changed in form from the overt racial policies of much of the twentieth century, discrimination and disparities that are caused or perpetuated on discriminatory lines have not been relegated to the past. Statistical and other evidence remains abundant. Indeed, only last month the US Department of Justice and the OCC announced actions against a bank for redlining Black and Hispanic neighborhoods in Houston.
B. Extensive Evidence Shows that Discrimination and Its Effects Limit Access to Quality Credit for People and Communities of Color
Many have documented the continuing effects of discrimination on the financial markets and on access to credit. Redlining, other forms of discrimination, and their legacy are felt in myriad ways, from disparities in homeownership rates that are worse now than even prior to the passage of the Fair Housing Act, to disparities in wealth and discouragement in the lending process that prevent minority-owned small businesses from having full and fair access to credit. In 2021, all 12 districts of the Federal Reserve System hosted a multipart program called “Racism and the Economy: Understanding the implications of structural racism in America’s economy and advancing actions to improve economic outcomes for all.” Contributors documented how the continuing effects of discrimination influence housing, entrepreneurship and business lending, education, employment and numerous other areas of society and the economy. All of these in turn influence the finances of Americans of color and their access to credit.
To take homeownership as an example, numerous studies demonstrate that systemic racism has affected and continues to affect homeownership rates in markets across the United States. These disparities lead to and compound disparities in wealth, which in turn affect minority business owners’ and entrepreneurs’ ability to access credit and capital. It is regrettably no surprise that the most recent HMDA data, from 2020, continued to demonstrate wide disparities between Whites and people of color in lending for housing.
With respect to small business lending, Black and Hispanic Americans continue to rely more heavily on personal and family savings as a source of financing than White Americans, despite having only a fraction of the wealth of White Americans. Black-owned businesses are turned down for loans twice as frequently as White-owned firms. Recent studies indicate that small businesses in communities of color were not able to access federal COVID-19 relief on par with access by businesses in White areas, including due to possible discrimination.
In short, historical discrimination entrenched inequalities in access to credit and in financial markets. As noted above, Appendix A provides NCRC publications that help to document this fact, supplementing the references in this paper. These inequalities have not been fully addressed and eliminated. The lingering and substantial effects, compounded in some cases by current instances of discrimination, prevent people of color from accessing credit fully and fairly. This provides the required strong basis in evidence to conclude that the government can, and should, adopt race-conscious remedies, including by amending CRA regulations to take race into account in a much more central way than it currently does.
C. Regulators Should Conduct Further Research to Supplement this Evidence, Just as the US Department of Justice Did in Support of Race-Conscious Contracting Programs
The evidence described above of the government’s compelling interest in addressing credit discrimination is strong, but a comprehensive analysis by regulators would make it even more so. This would increase the likelihood that race-conscious CRA components, if subjected to strict scrutiny, will pass muster.
Our proposal that regulators undertake this project is rooted in work the US Department of Justice (DOJ) did to support racial preferences in government procurement in response to the Supreme Court’s 1995 decision in Adarand. Adarand made clear that strict scrutiny applies to government affirmative action programs, creating uncertainty about the constitutionality of existing programs. DOJ surveyed the field to shore up evidence that the preference program was based on a compelling interest as Adarand required.
This effort resulted in a 1996 report summarizing:
the long legislative record that underpins the acts of Congress that authorize affirmative action measures in procurement . . .
Congressional hearings and reports that bear on the problems that discrimination poses for minority opportunity in our society, but that are not strictly related to specific legislation authorizing affirmative action in government procurement;
 recent studies from around the country that document the effects of racial discrimination on the procurement opportunities of minority-owned businesses at the state and local level; and
 works by social scientists, economists, and other academic researchers on the manner in which the various forms of discrimination act together to restrict business opportunities for members of racial and ethnic minority groups.
DOJ’s report has played an important role in judicial decisions holding that race-conscious procurement programs are properly supported by a compelling interest. The government has also updated and expanded the supporting evidentiary record in the years since in connection with various constitutional challenges.
Regulators should conduct a similar analysis focused on inferior access to credit stemming from discrimination. This would significantly aid in the defense of race-conscious CRA elements that we expect would be subject to strict scrutiny, which we describe in the following section.
V. Race Can and Should Be Incorporated Explicitly In CRA Requirements and Evaluations In a Narrowly Tailored Way
A. Race-Neutral Efforts Have Been Tried Without Sufficient Success
One narrow tailoring factor serves, at least as a practical matter, as a prerequisite to any race-conscious effort when strict scrutiny is triggered. This factor generally prohibits the government from adopting race-conscious methods to remedy the continuing effects of discrimination without first concluding that race-neutral methods are inadequate. This determination may come from actual experience with race-neutral approaches, but that is not required; “serious, good faith consideration of workable race-neutral alternatives that will achieve” the desired result is enough.
There is some ambiguity in Supreme Court precedent regarding how comprehensive the consideration or prior unsuccessful experience must be. In 2003, the Court explained that “[n]arrow tailoring does not require exhaustion of every conceivable race-neutral alternative.” Ten years later, though, it quoted that exact statement but added that the “reviewing court must ultimately be satisfied that no workable race-neutral alternatives” would achieve the government’s interest “about as well and at tolerable administrative expense.” Notwithstanding some uncertainty about just where the line is, it is clear that the government must have a convincing explanation — well beyond mere assertion — of why a race-conscious approach is truly needed.
There is a very strong basis for concluding that race-neutral alternatives are insufficient to overcome the impact of discrimination in lending. The history of the Community Reinvestment Act itself is exhibit number one. Next month CRA will be 44 years old. The core components of the statute and its implementing regulations have been race-neutral throughout, with primary attention given to census tracts identified based on income and some focus on other factors closely related to economic well-being. While illegal racial discrimination adversely affects a bank’s performance rating, that is a far cry from incorporating evaluation of how well a bank is meeting the needs of people and neighborhoods of color. For example, one bank might largely ignore neighborhoods of color because of racial prejudice, while another ignores the same neighborhoods because it has always focused on others. In each case, the needs of the neighborhoods of color are poorly served, but likely only the former would be viewed as illegal discrimination, and therefore properly negatively considered by an examiner.
There have been additional race-neutral programs for decades, yet discrimination and the effects of discrimination in lending persist. Longstanding race-neutral programs of the Federal Housing Administration, Department of Agriculture Rural Development and Veterans Administration facilitate mortgage lending primarily to people with low- and moderate-incomes. People who take advantage might otherwise not be able to obtain a loan or one with reasonable terms.
The Equal Credit Opportunity Act’s special purpose credit program provisions are another race neutral attempt to overcome reduced access to credit rooted in discrimination. Under a SPCP, private organizations — but not the government — may offer preferential access to credit to a group defined in part by race (or by other characteristics that ordinarily may not be considered). SPCPs are an important tool, and we commend institutions that have adopted them, but they have been taken up much too sparingly by the private sector despite being allowed since 1976.
In addition to programs addressing credit specifically, there have long been many race-neutral government programs focused on economic disadvantage and opportunity, which are crucial underpinnings of access to credit (or lack thereof). For example, the federal government supplements incomes through the Temporary Assistance for Needy Families, Housing Choice Vouchers and Supplemental Nutrition Assistance programs. It supports access to higher education through Pell Grants, student loans and loan forgiveness. It subsidizes health insurance premiums under the Affordable Care Act. Other programs provide job training. All of these race-neutral programs and many more aim to put people on, and help them maintain, the sound financial footing that reputable lenders want to see when evaluating applications for credit.
That the federal government’s varied, longstanding and costly race-neutral efforts have been insufficient to overcome the deep and abiding impact of discrimination on access to credit is plain from the evidence described above in section IV. The conclusion that race-conscious remedies are needed is exceedingly well founded.
B. These Proposals Would Not Burden White Neighborhoods or People
While any burden on others should be identified and assessed with respect to each specific way that race-consciousness might be incorporated into CRA, as we do below, there is an important overarching point about burden: nothing proposed here would alter the fundamental touchstone of CRA evaluations, i.e., assessing whether a bank is “meeting the credit needs of its entire community.” The proposals here are for additional vantage points from which to assess whether a bank is doing so, not to alter what a bank is required to do. As discussed above in connection with explaining why strict scrutiny might be an inappropriately high standard for judicial review in this context, there is no call for diminished treatment of non-minority neighborhoods or people.
Even if a few banks might respond to a race-conscious CRA by decreasing to an extent their focus on non-minority neighborhoods or people, the cumulative effect, considering all financial institutions’ behavior, would be marginal at most, and the effect on any given White borrower would likely be minimal or nonexistent. It is unlikely that banks would respond to greater scrutiny of how well they are serving neighborhoods and people of color by forsaking other parts of their communities, much less by doing so in any collectively significant way. Narrow tailoring’s concern with the burden on third parties is not implicated in any meaningful way by the proposals discussed below, and so does not suggest any difficulty satisfying strict scrutiny. In other words, a race-conscious CRA would likely expand access to credit to parts of the community now being ignored, without meaningfully restricting access to those who already benefit from banks’ outreach and lending.
C. Banks Should Be Prohibited From Arbitrarily Excluding Neighborhoods of Color From Their Assessment Areas
When delineating their assessment areas, banks “[m]ay not arbitrarily exclude low- or moderate-income geographies, taking into account the bank’s size and financial condition.” But there is no rule against arbitrarily excluding neighborhoods of color. With respect to race, the sole prohibition is much narrower; only illegal discrimination is barred. As with LMI neighborhoods, banks should be precluded from arbitrarily excluding neighborhoods of color from their assessment areas. This would compel them to pay greater attention to serving such neighborhoods and would satisfy a strict scrutiny analysis.
Delineating assessment areas in a manner that elides neighborhoods of color is not uncommon. In extreme cases, doing so has been cited as evidence of redlining in lawsuits brought by the federal government. In suing BancorpSouth Bank, for example, the Department of Justice and the Consumer Financial Protection Bureau stated that:
“[The Bank’s] assessment area excluded all 91 high-minority neighborhoods and 137 of 142 majority-minority neighborhoods (96.5 percent) in the Memphis MSA, including minority neighborhoods near existing branches that easily could have been drawn into the assessment area”;
“For counties with few or no majority-minority neighborhoods, the Bank selected all or most of the county to be in its assessment area”; and
“[T]he Bank’s delineation excluded approximately 85 percent of the 523,000 African Americans living in the Memphis MSA from its CRA assessment area.”
Other cases cite similar evidence:
“Eagle’s CRA assessment area excluded 50 of the 61 majority-black census tracts within the City of St. Louis, and all of the 27 majority-black census tracts in St. Louis County.”
“Excluding Indianapolis-Marion County and its 50 majority-Black census tracts from the Bank’s Community Reinvestment Act (‘CRA’) assessment area, while including overwhelmingly white counties.”
“This assessment area in New York State excludes and forms a semi-circle around the four counties with the highest proportion of majority-Black-and-Hispanic neighborhoods in the State” and “Hudson City’s assessment area in the Camden MSA excludes all 337 majority-Black-and-Hispanic census tracts.”
“KleinBank drew its main CRA assessment area to be horsehoe-shaped; it includes the majority-white suburbs, and carves out the urban areas of Minneapolis and St. Paul that have higher proportions of minority populations.”
In all of these cases, assessment area delineation combined with extensive other evidence — primarily analyses of HMDA data, branch locations and marketing, and sometimes even explicitly discriminatory policies and practices — were used to expose the defendants’ conduct as unlawfully discriminatory, as alleged by the government.
But avoiding neighborhoods of color in a deliberate and demonstrably discriminatory way is less common than it used to be. The larger problem is banks that do not adequately serve these neighborhoods for more prosaic reasons, e.g., their focus is elsewhere. Neighborhoods of color may escape these banks’ assessment areas for reasons that are more about neglect than ill intent. As currently constituted, CRA regulations and evaluations do not provide a sufficient counterweight to assure that these neighborhoods are firmly planted on banks’ radar, as they must be if the country is to overcome the impact of discrimination on banking. Regulatory compliance currently does not create adequate incentive for banks to pay greater attention, but the incentive would be greater if the arbitrary exclusion of neighborhoods of color were prohibited. Because banks would face consequences for neglect that deprives these neighborhoods of the access they need, neglect would become less common. Banks would have to pay attention.
The case of Dallas County in Texas is illustrative. Interstate 30 is a significant racial dividing line. North of the highway is disproportionately white; south is disproportionately Black and Hispanic. A Dallas television news station found that more than 100 banks serve Dallas County, but twenty percent exclude parts of the county south of I-30 from their assessment area. The government has not sued any banks for discrimination based on their Dallas County assessment areas and there has not otherwise been any finding of discrimination, meaning regulators are not viewing this racially significant exclusion as in conflict with CRA regulations.
Adding a prohibition against the arbitrary exclusion of neighborhoods of color to CRA raises none of the concerns that are the focus of strict scrutiny:
- It involves no quotas and is not otherwise inflexible. No particular number of neighborhoods of color would have to be considered to determine whether they are arbitrarily excluded; this would vary with local residential patterns. Implementing regulations and/or examiner guidance can be constructed to allow meaningful discretion to consider a variety of factors that are relevant to delineating an assessment area. This is in accord with how a “bank’s size and financial condition” are already relevant to the parallel prohibition regarding LMI neighborhoods.
- It would not be over- or under-inclusive. Rather, it would primarily be a tool for addressing one aspect of the broader inquiry into whether a bank is serving its entire community. Because of the continuing focus on LMI neighborhoods, it also would not be the only tool explicitly called for. And as a freestanding prohibition against only arbitrary exclusion, it would not force the inclusion or exclusion of any neighborhoods that are not reasonably included or excluded.
- It would not burden third parties. Avoiding arbitrary exclusion of neighborhoods of color dictates nothing about the inclusion/exclusion of other neighborhoods because the size of an assessment area is up to the bank.
Prohibiting the arbitrary exclusion of neighborhoods of color from assessment areas would further the central purpose of CRA by bringing equal banking opportunities instead of neglect to many people and would be in accord with constitutional requirements.
D. Incorporating Race in CRA Evaluations
Many factors, especially how well LMI neighborhoods and people are served, are considered when regulators evaluate and assign a CRA rating to a bank. Race should be considered, too, by evaluating how well a bank is serving neighborhoods and people of color. We describe here how this can be done in a manner that is practical and constitutional, dovetails with the Fed’s current ANPR, and would help ameliorate the continuing impact of past and present racial discrimination on access to credit.
We first describe three ways in which consideration of race should be incorporated generally, then address its incorporation in the four specific subtests proposed by the Fed, and then its incorporation in strategic plans that banks may ask to be evaluated under. The four subtests are: the retail lending subtest and retail services subtest that comprise the retail test, and the community development financing subtest and the community development services subtest that comprise the community development test.
1. Identifying Groups, Markets and Products Where Race Should Be Incorporated in Evaluations
To satisfy narrow tailoring requirements, race should not automatically be included in all aspects of all evaluations. Rather, race should factor only insofar as access to credit is currently negatively impacted by the continuing effects of past or present discrimination with respect to (1) a specific racial group, (2) a specific product and (3) the specific assessment area evaluated.
For example, if research demonstrates that discrimination continues to impact access to mortgage credit for African Americans in a bank’s Chicago assessment area, then how well the bank is meeting the mortgage credit needs of African Americans should be part of the lending portion of the bank’s evaluation for that assessment area. But for a contrary example, if there is no significant evidence of discrimination affecting access to mortgage credit for Asian Americans in a Jackson, Mississippi, assessment area then a Jackson bank’s success in meeting Asian Americans’ credit needs would not be part of the lending part of its evaluation. If discrimination is affecting a racial group’s access to any product in an assessment area, then that race would be considered in the service and community development parts of an evaluation.
To facilitate this conditional consideration of race, the relevant regulators should undertake a joint study to determine which racial groups face reduced access to credit due to discrimination in different parts of the country, and which products are affected by that discrimination. It should primarily be a statistical study, supplemented by anecdotal evidence where it is readily available and reasonably recent. The study should be conducted at a fine enough level to permit application to most assessment areas; this should be more granular than the state level because assessment areas are likely to be drawn more locally than statewide by mirroring the borders of a metropolitan statistical area (MSA) or a metropolitan division (MD). Impaired access to credit stemming from discrimination in rural counties, particularly those with large populations of people of color, should also be considered by the study.
The study should be updated on an annual or other reasonable periodic basis. This will assure that race is no longer used in evaluations after the impact of discrimination has been overcome.
An examiner would apply the study’s findings to an examinee’s assessment area(s) to determine whether and in what fashion race should be incorporated into an examination. The criteria for that determination would be public so that banks know well in advance. The following are different ways that, where the study supports doing so, race should be considered.
2. Selection of Assessment Areas for Full-Scope Review
But should the distinction remain, then if the study findings demonstrate that race should be incorporated in the evaluation of one or more elements of a bank’s activities, race should be used in deciding which assessment areas receive a full-scope examination. As the name indicates, a full-scope review is the most thorough and rigorous. In a full-scope review, “examiners complete all of the procedures for an assessment area.” As described by the FDIC, “[t]he primary difference between full scope and limited scope assessment area reviews is the inclusion of qualitative factors and presentation of full performance context in the performance evaluation for full scope assessment areas.” This is significant. Johns Hopkins studied small business lending in Baltimore,a city that is over 60% Black,and found that full-scope reviews encompassing the city were not conducted of the institutions with the largest depository holdings in the city.
For “large” and “intermediate small” institutions with multiple assessment areas, at least one from each state and one from each multistate MSA or MD must receive a full-scope review. At least one must receive a full-scope review for small institutions with multiple assessment areas.
Pursuant to interagency procedures, examiners consider a range of factors in selecting assessment areas for full-scope review. One factor explicitly incorporates LMI:
The level of the institution’s lending, investment, and service activity in the different assessment areas, including in low- and moderate-income areas, designated disaster areas, or distressed or underserved nonmetropolitan middle-income geographies designated by the Agencies based on (a) rates of poverty, unemployment, and population loss or (b) population size, density and dispersion.
Other factors concern lending, investment and service opportunities in an area; the other institutions present and the importance of the institution being examined to an area; public and community group comment about the institution’s CRA performance; population size of an area; any anomalies in CRA or HMDA data for an area; how long it has been since an area’s last full-scope examination; the institution’s past CRA performance; issues identified in CRA examinations of other institutions; and the examiner’s knowledge of an area or similar areas.
None of these focus on race. This omission should be addressed by amending the factor quoted above to read:
. . . including in low- and moderate-income areas, majority- or disproportionately minority areas, designated disaster areas . . . .
This would assure that an important element of the “entire community” is not overlooked in determining which assessment areas are evaluated most carefully.
Adding race in this manner would comport with narrow tailoring because, analogous to the Supreme Court’s holding that consideration of race is a proper “plus factor” among many others in a university’s holistic admissions process, lending to relevant racial groups would be one consideration among many, and would not affect a bank’s rating in an inflexible quantitative and formulaic way. Race would not be “the defining feature” of assessment areas; each would receive “individualized consideration” based on many features. And because there is no maximum number of assessment areas that may receive a full-scope examination, no areas need be excluded. But even if consideration of race tips an examiner away from a full-scope review of a low-minority area and that is viewed as a burden, this should not be a problem under Supreme Court precedent: fair consideration would still have been given to the other area based on other factors.
3. Review of Assessment Areas for Arbitrary Exclusion of Neighborhoods of Color
CRA examinations consider whether a bank arbitrarily excludes LMI census tracts from its assessment areas. Currently, there is no parallel inquiry regarding majority- or disproportionately minority tracts. This reflects the current regulations on delineating assessment areas discussed above. With respect to race, delineation is only reviewed for illegal discrimination. This has led to the exclusion of neighborhoods of color from assessment areas, creating a barrier to additional lending in those neighborhoods.
In conjunction with amending regulations to preclude the arbitrary exclusion of neighborhoods of color from assessment areas, examiners should affirmatively consider whether a bank has done so. The focus for any particular assessment area would be groups as to which the joint regulatory study identifies a continuing impact of discrimination. If the study only implicated discrimination against African Americans, the question would be whether majority- or disproportionately Black neighborhoods have been arbitrarily excluded.
Coupling the underlying prohibition with this enforcement mechanism would give banks a stronger incentive to self-police for the unintended, but nonetheless damaging, omission of neighborhoods of color. It would be constitutional for the same reasons as amending the regulatory prohibition.
4. Incorporating Race in the Retail Lending Subtest
Two alternatives are proposed here for incorporating race in the lending subtest portion of CRA examinations. Both meet the narrow tailoring requirements of strict scrutiny, though in part they draw support from different strains of precedent.
The first alternative parallels the Fed’s proposal for LMI lending. An institution would need to meet or exceed a threshold level of lending to people and neighborhoods of color, or make a good faith effort to do so, to be eligible for a presumption of satisfactory. Regulators could also set performance range presumptions for the possible conclusions, from substantial noncompliance to outstanding.
The second alternative would set aspirational goals for lending to people and neighborhoods of color. If an institution performs well relative to its goals, the examiner could use that as a plus factor in assigning a rating. This alternative would likely be less effective in promoting minority lending.
Though each approach relies on numerical targets, case law regarding government contracting programs demonstrates that this is consistent with narrow tailoring if done correctly. The Supreme Court has clearly stated that “some attention to numbers” is not by itself equivalent to a quota and does not render consideration of race insufficiently flexible.
Also to accord with narrow tailoring requirements, the racial groups considered in setting the targets would be those identified through the joint regulatory study as facing diminished access in the assessment area because of the continuing effects of discrimination.
i. Alternative 1: Thresholds and Performance Ranges
The United States Department of Transportation (DOT) funds states to hire contractors for highway and other transportation projects. States must set goals for the percent of these funds that will go to “disadvantaged business enterprises,” or “DBEs.” These are small businesses that are minority-owned and run by people who are socially and economically disadvantaged. People of color presumptively qualify subject to financial limits.
States must follow DOT regulations to establish their goals. The goals must reflect, for the local market, “the level of DBE participation [they] would expect absent the effects of discrimination.” The regulations require states to determine the relative availability of DBEs in the market, and whether evidence indicates that an adjustment to that figure should be made to arrive at a goal, and describe proper ways of making adjustments. A state’s goal must be submitted to DOT for approval. Because these regulations assure that goals are “reasonable . . . given particular local conditions,” they have repeatedly been found consistent with narrow tailoring.
DOT may not penalize a state for failing to meet its goal unless the state did not make a good faith effort. Courts have cited this as an important factor in assuring that the program is adequately flexible to comport with narrow tailoring. Program regulations explain what constitutes a good faith effort. Significant sanctions may be imposed absent success or a good faith effort, including the suspension or termination of federal program funding.
This provides a model for incorporating lending goals into CRA evaluations. The metrics proposed above for minority lending can provide a locally based foundation for setting goals in a manner that is narrowly tailored and parallels the Fed’s proposal for setting threshold levels (i.e., goals) for LMI lending.
For a given line of business in a given assessment area, the Fed proposes to compare a bank’s geographic and borrower distribution metrics to local thresholds. The thresholds would be derived from the community and market benchmarks. These are the same LMI-based metrics and benchmarks discussed above, which are based on local circumstances.
Specifically, for each type of distribution metric (geographic and borrower), the community and market benchmarks would be multiplied by a fixed percentage (under 100%), and the lower result would be the threshold for that type of distribution. If the community/borrower benchmark in the assessment area is 44% and the market/borrower benchmark is 35%, and the fixed percentage is 65% for community and 70% for market, then the borrower distribution threshold is 24.5%. If, for each line of business, a bank’s two distribution metrics both meet or exceed their thresholds, then the bank would presumptively receive a satisfactory rating on the retail lending subtest for that assessment area. The presumption could only be rebutted “in cases of consumer compliance violations involving discrimination and other illegal credit practices.”
This methodology can readily be adapted to the race-based counterpart metrics discussed above. Like the local benchmarks they would be based on, updating the thresholds periodically would assure that they adjust with changing circumstances.
To comport with narrow tailoring, one important difference would be that a bank’s good faith effort to meet its thresholds would be sufficient to obtain the presumption of satisfactory. Bank regulators could identify indicia of good faith, just as DOT does for DBE contracting. This would avoid rigid numerical requirements and provide for consideration of qualitative factors. Indicia could include, inter alia, affirmative marketing, adoption of special purpose credit programs and product design tailored to the needs of communities of color.
A bank would need to receive a satisfactory rating on both the LMI and race elements of this expanded retail lending subtest to receive an overall satisfactory rating on the subtest, meaning that race would not be the defining feature of the evaluation. The dashboard might show the bank’s performance compared to the thresholds for both LMI and race like this:
mortgage line of business unless the examiner determines that it made a good faith effort to reach its borrower distribution metric based on race. This makes sense, because the figures suggest the bank may disproportionately be serving non-minority customers even when lending in neighborhoods of color, thereby failing to meet credit needs across its entire community.
The next part of the Fed’s proposal for focusing on LMI — establishing performance ranges — can also easily be adapted to consideration of race in a narrowly tailored manner.
The Fed proposes to create a performance range for each of the possible ratings. As with the thresholds for a presumption of satisfactory, the ranges would be based on the metrics and benchmarks. A formula (accounting for the bank’s business model and credit opportunities and needs) would be used to combine the different lines of business and measurements to reach an overall conclusion for the retail lending subtest in an assessment area.
Importantly, this would only be a recommended conclusion. The examiner would also consider aspects of the bank’s performance context not factored into the quantitative measures, and other qualitative factors, to determine whether to depart from the recommendation. This would be based on judgment, not formula. A bank might be found “satisfactory” despite a “needs to improve” recommendation if capacity and constraint issues would undermine the safety and soundness of additional lending, to use a Fed example. A list of relevant factors (which ought to be non-exhaustive) would be provided for examiners to consider.
The same performance range procedure can be used with respect to race. Allowing for consideration of good faith efforts in addition to the other non-formulaic factors would ensure that a bank’s performance in connection with race would be assessed with adequate flexibility. The last step — combining the conclusions for LMI and race for a final rating on the subtest — also can (and should be) based on examiner judgment, perhaps utilizing a simple matrix to provide a nonbinding suggested outcome.
Stepping back from the mechanics, evaluating how effectively a bank is serving the needs of both LMI and minority communities in its assessment area(s) would give examiners a more comprehensive and nuanced view into how well the bank is serving the entirety of its local communities. This would better serve CRA’s purpose because that is the statutory touchstone, not LMI communities alone.
Doing so through presumptions and performance ranges as outlined would ensure transparency and predictability for banks and the public, while avoiding the rigidity with respect to race that the Constitution does not permit. If at the end of the day, a bank’s retail lending subtest rating is pulled down because of unjustified poor performance with respect to race, it would parallel the constitutionally legitimate penalizing of a state for an unjustified failure to meet DBE contracting goals. But the point of this construct, of course, is to make that outcome less likely, by encouraging banks to pay sufficient attention to people and neighborhoods of color. Doing so like this would satisfy strict scrutiny.
ii. Alternative 2: Exceeding Aspirational Goals as a Plus Factor in Evaluations
A lesser but still valuable improvement would be for regulators to establish aspirational goals that banks should strive to meet for lending to people and neighborhoods of color. These goals would not trigger any consequences for a bank if they are not met. The Fed is already considering a version of this that would allow “substantive and meaningful engagement” with minority depository institutions to potentially elevate an overall rating from satisfactory to outstanding.
Aspirational goals have repeatedly been found to satisfy narrow tailoring as long as they are grounded in relevant local conditions and are not arbitrary. In the DOT program discussed above, DOT has a national goal that at least 10% of contract dollars will go to DBEs. It is explicitly aspirational. Most courts have paid limited attention to this national goal because it is not treated as any state or contractor’s own goal.
Any race-based aspirational goals under CRA should be a percentage of the benchmarks. This is important because even though the goals would not have consequences, there must be evidence that they are reasonable. But unlike the percentages used to set thresholds, here they should be no less than 100% of the relevant benchmark. That is because a bank’s goal should be to serve people and neighborhoods of color at least as well as the local market as a whole is serving them, given that goals would only be adopted in places where access to credit continues to be suppressed by the effects of past or present discrimination.
Going one step further,but still short of the threshold and performance level proposal, regulators could allow examiners to treat doing well compared to goals as a plus factor. Exceeding goals, especially by a significant margin, might lead an examiner to raise the bank’s rating on the subtest from what it would otherwise be in the absence of any consideration of race (beyond consideration of illegal discrimination, that is).
This is analogous to how, under Grutter, admissions offices may constitutionally consider race as a plus factor in pursuit of a diverse student body, thereby offering admission to someone who might otherwise be declined. As in admissions, using race like this would supplement, not displace, consideration of other aspects of a bank’s performance. That consideration would remain “highly individualized [and] holistic,” with “serious consideration [given] to all the ways a [bank] might contribute to” meeting the credit needs of its whole community. Performance with respect to race would not become the “defining feature” of the evaluation.
With this approach, it is important to provide examiners discretion to determine subjectively when to raise a rating. That is because the Supreme Court has clearly rejected assigning a particular number of bonus points to an application for admission based on race on the ground that it undermines properly flexible and individualized consideration. But guidelines could, and should, be adopted to guide this discretion.
Using the achievement of race-based goals as a plus factor would have a salutary effect on some banks and would satisfy narrow tailoring. It would not, however, be as effective or provide as much useful information to the public and regulators about bank performance as incorporating consideration of race into thresholds and performance ranges.
5. Incorporating Race in the Retail Services Subtest
The Fed proposes to update the retail services subtest in several ways. They concern clarification and emphasis, not a wholesale new approach. The proposal is based on judgment, including when considering the meaning and importance of metrics. The Fed itself says that the subtest “would use a predominately qualitative approach, while incorporating new quantitative measures.” The Fed seeks to leverage data, but not make examiners beholden to it.
Adding consideration of how well banking services are provided to people and neighborhoods of color, using the subjective plus factor approach, would be narrowly tailored and fit well with the Fed’s proposal.
The Fed would divide the subtest into two components, delivery systems and deposit products. The former is about getting banking services to people, the latter about having adequate deposit products.
The delivery system evaluation would have four elements: branch distribution (i.e., location), record of opening and closing branches, branch-related services (e.g., extended hours, non-English capacity, free or low-cost check cashing) and non-branch delivery systems (e.g., ATMs, mobile, and internet). Using the branch distribution benchmarks and metric discussed above, the branch distribution analysis would assess branch accessibility for LMI people and neighborhoods in particular. Additional consideration might be given for maintaining branches in banking deserts. The record of opening and closing branches would likewise be reviewed with particular focus on how those changes affect LMI access. The services element would consider how effectively a bank’s expanded offerings improve access to or decrease costs of financial services, again especially for LMI people and neighborhoods. The non-branch delivery systems review shares the same LMI focus.
The branch distribution and openings/closings information would be assessed qualitatively, even though they would include consideration of quantitative information. This recognizes that “[a]n examiner could evaluate these data in different ways based on performance context.” Thus, unlike with respect to the lending subtest, there would be no thresholds, and evaluation of “[t]he record of opening and closing branches would continue to rely on examiner judgment.”
The Fed has not yet offered much specificity for assessing branch-related services and non-branch delivery systems, but it appears that these would also be considered subjectively via examiner judgment. The Fed is seeking input on how more data could be used in these areas to allow for more consistency, but they do not seem to lend themselves to an exclusively formulaic evaluation.
The deposit products component of the retail services subtest would “evaluat[e] a bank’s deposit products, including checking and savings accounts, focusing on those tailored to meet the needs of LMI individuals.” The Fed’s goal is to “elevate and strengthen the evaluation of deposit products” by assessing their availability, usage and impact. The types of products it is focused on include, among others, accounts with low-cost transactions, individual development accounts and accounts with low or no opening deposit requirements. There is little specificity so far in the Fed’s proposal for deposit products, but it appears to contemplate an ultimately subjective evaluation that would incorporate consideration of data to some degree. Presumably criteria would be developed to help examiners reach a decision, but those decisions would be discretionary based on all the relevant factors.
The multifaceted and subjective character of the Fed’s proposal dovetails well with narrow tailoring considerations. Those were the very characteristics relied on by the Supreme Court in approving the admissions plan in Grutter.
Without reducing the attention given to serving LMI (and overall community) needs, all the factors to be considered with respect to LMI can also be considered with respect to race. For example, whether branches are located in high-minority neighborhoods and whether new branches have opened in those neighborhoods can be taken into account. Such inquiries complement, and in no way diminish, the parallel inquiries based on LMI. When the answers to the race-conscious questions demonstrate that a bank is doing a good job of serving people and neighborhoods of color, examiners should be permitted to use that as a plus factor in selecting a rating on the subtest. This is very much in line with Grutter, except that it is even more clearly permissible here because unlike admission to a school a good result on the race-conscious aspects of an evaluation is available to all.
6. Incorporating Race in the Community Development Financing Subtest
The Fed proposes a new community development financing subtest for large retail banks, as well as wholesale and limited purpose banks. It would combine consideration of both community development loans and investments.
The Fed’s focus here is on the assessment area as a whole, not LMI individuals or neighborhoods specifically. That is because there are many kinds of activities that can have community development as a primary purpose. The scope encompasses activities that benefit LMI, disaster and underserved or distressed nonmetropolitan middle-income areas, and activities that provide affordable housing or community services for LMI people, or finance certain small businesses or farms.
Once again, the proposal — with one important but easily accomplished adjustment — translates well and in a constitutional manner to consideration of race. To a degree, in fact, the Fed’s proposal already incorporates some of the race-conscious ideas presented here.
The Fed proposes to use the community development financing metric measured against thresholds derived from the local and national community development financing benchmarks. If a bank’s metric surpassed a threshold, it would gain a presumption of a satisfactory rating on this subtest, as with the retail lending subtest. Here the Fed hesitates, however, recognizing that data limitations might counsel using the thresholds only as guidelines and not as the basis for a presumption, at least until more data are available.
Judgment based on qualitative considerations would factor in significantly either way. Under the presumption approach, judgment would be used by an examiner to decide between a satisfactory and an outstanding rating if a bank qualified for the presumption. If a bank did not qualify or the presumption approach were not adopted, judgment would be used in conjunction with the quantitative measurements to decide between any of the possible ratings.
The qualitative considerations would include performance context, “emphasiz[ing] a bank’s capacity and constraints.” They would also include “especially impactful activities, such as a significant capital investment in a [minority depository institution]” and “smaller dollar activities” with an outsized impact.
Helping to guide examiner judgment would be impact scores from one (lowest) to three (best) that examiners would assign to each community development financing activity. The impact scores would be based on data and any other qualitative information a bank chooses to provide regarding an activity. The Fed is considering “specifying that economic development activity focused on . . . minority-owned small businesses would be considered responsive and impactful in development of a Community Development test conclusion or rating.” As is both evident and proper, this proposal already heads in a race-conscious direction.
The same quantitative measurements, translated to race in the manner explained above, should be used to create aspirational goals specific to assessment areas for community development financing that benefits neighborhoods of color. The local benchmark, not the national ones, should be used in setting goals to ensure that the goals are properly tied to local circumstances, as narrow tailoring requires. Aspirational community development financing goals are proper here for the same reasons explained with respect to retail lending.
Examiners should have discretion to consider success measured against those goals, as well as qualitative factors such as financing that are particularly impactful for neighborhoods of color, as plus factors in assigning a rating on the subtest. Again, as with retail lending, this would not displace consideration of other factors that are not race-conscious and it would not make financing benefiting neighborhoods of color the defining feature of the subtest. Rather, an examiner’s attention would remain on the whole of a bank’s community development financing performance including the many kinds that currently can qualify, with race added as only a part of the mix in a non-formulaic manner. That is a narrowly tailored approach.
7. Incorporating Race in the Community Development Services Subtest
The community development services subtest proposed by the Fed “would be primarily qualitative.” It does not involve any presumptions or thresholds. Relevant services “include activities such as service on boards of directors for community development organizations or on loan committees for CDFIs, financial literacy activities targeting LMI individuals, and technical assistance for small businesses.” The Fed is exploring incorporation of some quantitative data (“such as the ratio of community development services hours to the number of bank employees”), but it appears it would still be subject to judgmental review by examiners.
Race can and should be incorporated here by expanding what counts as community service so that relevant services targeting people and neighborhoods of color are counted. This parallels how similarly targeted activities should be counted in the financing subtest. Examiners should have discretion to consider services targeted in this way as a plus factor on the services subtest. For the same reasons identified above, this would be narrowly tailored.
E. Banks Submitting Strategic Plans Should Have to Demonstrate That Plan Development Included Outreach to Organizations Serving Neighborhoods of Color
CRA regulations afford a bank the option of being evaluated under a strategic plan that it develops and submits to its regulator for approval. The plan must include measurable goals for a satisfactory rating and address lending, community investment, and the provision of banking services.
In developing a strategic plan, a bank must informally seek suggestions from members of the public in the covered assessment area(s). It must then formally solicit public comment once a plan has been developed before submission for approval. The public’s involvement in developing a bank’s plan, and the public’s response, factor in whether the regulator approves the plan.
The public input requirement should be modified to require outreach to groups that serve neighborhoods of color when a plan covers an area with many people of color. This could include nonprofit organizations, media, houses of worship, government agencies and informal networks, among others. For example, it might include local Black and Hispanic chambers of commerce, NAACP and Urban League chapters, housing counseling agencies, Minority Business Development Agency Centers, reinvestment organizations and community development corporations.
This proposal is in harmony with the central aim of CRA because it would assure consideration of input from an underserved segment of the “entire community.” Like assuring that neighborhoods of color are not arbitrarily excluded from assessment areas, it would serve to help put, or keep, banks on the course laid out by CRA decades ago without changing that course.
It would also easily satisfy strict scrutiny:
- The proposal involves no quotas and is not otherwise inflexible. No specific groups or number of groups would need to be solicited for input. Rather, local circumstances would suggest who a bank might want to contact. Banks would not be required to adopt recommendations, only to seek and consider them. Consistent with current procedures, regulators reviewing a plan would have ample discretion to determine whether participation by groups serving neighborhoods of color was adequate, and how the adequacy or inadequacy of their participation factors into the final decision of whether to approve a plan.
- It would not be over- or under-inclusive. Public input from groups serving neighborhoods of color would not be in lieu of input from any other segment of the community. The core requirement for input from an assessment area’s public writ large would remain just as it is. Accordingly, there would be no exclusion or reduction of input from other groups that regulators have long since determined should be heard from. Nor would there be over-inclusion because the requirement would be limited to areas with many people of color where the the impact of discrimination on lending persists.
- It would not burden third parties. Again, this would not replace input from other groups. Banks seeking approval of a strategic plan would need to seek out and listen to their voices in exactly the same way they must today.
Requiring outreach to groups serving neighborhoods of color in connection with CRA strategic plans would promote inclusion without excluding anyone and would satisfy strict scrutiny.
VI. The Text of CRA Is Not an Obstacle Because It Permits Consideration of Race
Even if constitutional, regulators may only incorporate race-conscious criteria if permitted by CRA. Though mostly quiet on the subject, for several reasons we believe the statute does allow it.
First, the statute contains no facial prohibition against taking account of race.
Second, the statute is phrased in a way that fairly encompasses awareness of race. Its aim is “to encourage [regulated financial institutions] to help meet the credit needs of the local communities in which they are chartered,” and regulators are to “assess the institution’s record of meeting the credit needs of the entire community, including low- and moderate-income neighborhoods.” Residents and neighborhoods of color plainly are part of these communities. Moreover, the emphasis on LMI neighborhoods is not exclusive; rather, it demonstrates that regulators may properly look at constituent parts of a community to aid in evaluating performance in the “entire community.”
Third, the statue already requires race-consciousness in two discrete ways. First:
In assessing and taking into account . . . the record of a nonminority-owned and nonwomen-owned financial institution, the appropriate Federal financial supervisory agency may consider as a factor capital investment, loan participation, and other ventures undertaken by the institution in cooperation with minority– and women-owned financial institutions and low-income credit unions provided that these activities help meet the credit needs of local communities in which such institutions and credit unions are chartered.
In the case of any depository institution which donates, sells on favorable terms (as determined by the appropriate Federal financial supervisory agency), or makes available on a rent-free basis any branch of such institution which is located in any predominantly minority neighborhood to any minority depository institution or women’s depository institution, the amount of the contribution or the amount of the loss incurred in connection with such activity may be a factor in determining whether the depository institution is meeting the credit needs of the institution’s community for purposes of this chapter.
Both instances are significantly narrower than the proposals made herein, but they undermine any broad claim that Congress intended regulators to be blind to race when evaluating how well institutions are meeting community credit needs. And nothing on their face suggests that these are the only two considerations of race Congress intended to permit.
Though of less significance, CRA’s legislative history demonstrates that overcoming redlining on the basis of race was part of Congress’ purpose in passing the law. Race-consciousness by regulators would be consistent with this purpose. And regulators have, at least occasionally, incorporated race into their CRA decision making. This indicates they have not found it clear that doing so is inconsistent with the statute. Indeed, the Fed’s current proposal begins with a section on the statutory purpose and history, the opening sentence of which states that the “CRA . . . provides the [regulatory] agencies with a crucial mechanism for addressing persistent systemic inequity in the financial system for LMI and minority individuals and communities.”
While we conclude that the statute itself permits regulators to apply new race-conscious criteria under CRA, certainly it does not require regulators to do so or give them detailed guidance should they opt to do so. This might give rise to a constitutional challenge under the nondelegation doctrine. That doctrine recognizes “that Congress generally cannot delegate its legislative power to another Branch.”
The bar for satisfying the nondelegation bar is low, however, and has been cleared here. “The constitutional question is whether Congress has supplied an intelligible principle to guide the delegee’s use of discretion. So the answer requires construing the challenged statute to figure out what task it delegates and what instructions it provides.” Rejecting a nondelegation challenge to the SBA’s minority contracting preference program, one court stated:
This Court notes that the Supreme Court has “found the requisite ‘intelligible principle’ lacking in only two statutes, one of which provided literally no guidance for the exercise of discretion, and the other of which conferred authority to regulate the entire economy on the basis of no more precise a standard than stimulating the economy by assuring ‘fair competition.’”
One element of the challenge in Rothe was that Congress had not, purportedly, given the SBA adequate guidance about who should benefit. That assertion was rejected because the statute defined “socially disadvantaged individuals” and gave examples (Black, Hispanic, Native and Asian Pacific Americans; Indian tribes; and Native Hawaiian Organizations). Here CRA defines “minority” by reference to another statute, which in turn defines it as meaning “any Black American, Native American, Hispanic American, or Asian American.”
In addition to statutory language, it is proper to consider the purpose and factual background of a statute in determining whether an agency has been given sufficient guidance. The purpose of ensuring that financial institutions meet the needs of everyone, enacted against the backdrop of a long history of disserving specific parts of their communities, would provide substantial guidance to regulators deciding how and to what degree to incorporate race explicitly into CRA rules and evaluations.
The necessary “intelligible principle” likely is present in CRA such that the regulatory adoption of race-conscious criteria would not violate the nondelegation doctrine.
It is time to bring a broad and explicit focus on race to Community Reinvestment Act regulations and evaluations. The evidence confirms that more than forty years after the law was passed, it is in fact well past time. The legacy of government-sponsored and private discrimination continues to warp credit markets profoundly, preventing people and neighborhoods of color from having fair access to credit and other banking services.
There are many ways, consistent with the Equal Protection Clause of the United States Constitution, CRA can be improved by making it more race-conscious. Data should systematically be collected and published showing how well banks are meeting the needs of people and neighborhoods of color. This data and parallel qualitative information should be considered when assigning ratings on evaluations. And, as set forth above, there are more.
The addition of race will supplement, not displace, the many other important factors addressed by CRA. It will give regulators a deeper understanding of whether banks are meeting the needs of their entire communities by broadening the crucial, but too narrow primary focus on LMI. It will motivate banks to broaden their focus in the same way. It will bring nearer the day when everyone has fair access to the critical products and services provided by banks. That is the very reason for CRA.
The NCRC reports identified below represent just a portion of the expansive body of literature on disparities and discrimination affecting homeownership, lending and finance. Many more studies, articles and reports are cited throughout these publications and should be examined and taken into account when considering the topics discussed in this paper. These NCRC publications and the sources cited in them should not be taken as a complete list of the important work in this field demonstrating the compelling governmental interest in adopting race-conscious regulations under the Community Reinvestment Act.
Jason Richardson, Joshua Devine, Jad Edlebi, 2020 HMDA Preliminary Analysis, (Aug. 19, 2021), https://ncrc.org/2020-hmda-preliminary-analysis/.
- Initial analysis of 2020 HMDA data showing home buying trends throughout the year; the study finds, for example, that white and Asian homeowners increased their share of refinance lending in 2020.
Dedrick Asante-Muhammad et al., Black Entrepreneurship’s Lethal Pre-Existing Condition: The Racial Wealth Divide During The COVID Crisis (Apr. 6, 2021), https://ncrc.org/black-entrepreneurships-lethal-pre-existing-condition-the-racial-wealth-divide-during-the-covid-crisis/.
- Report detailing the racial wealth divide in the United States before the COVID-19 pandemic.
Dedrick Asante-Muhammad, Jamie Buell, Joshua Devine, 60% Black Homeownership: A Radical Goal For Black Wealth Development (Mar. 2, 2021), https://www.ncrc.org/60-black-homeownership-a-radical-goal-for-black-wealth-development/#elementor-action%3Aaction%3Dpopup%3Aopen%26settings%3DeyJpZCI6IjkwMjg2IiwidG9nZ2xlIjpmYWxzZX0%3D.
- Report detailing the historic and continuing racial gap in homeownership, particularly for Black Americans. It underscores the need to increase homeownership among Black Americans and proposes a 60% Black homeownership rate goal.
Anneliese Lederer & Sara Oros, Are Modifications for Small Businesses a Possibility In The COVID-19 Pandemic? (Feb. 2, 2021), https://ncrc.org/are-loan-modifications-for-small-businesses-a-possibility-in-the-covid-19-pandemic/.
- Analysis of survey results of small business owners across the United States to evaluate their financial health during the pandemic. The study finds significant racial disparities in financial stability, as well as the assistance provided by lenders to these businesses.
Jad Edlebi, Research Brief: Bank Branch Closure Update (2017 – 2020) (Dec. 14, 2020), https://ncrc.org/research-brief-bank-branch-closure-update-2017-2020/.
- Supplemental analysis of bank branch closures throughout the United States. Among other elements, the study examines closures in medium-to-high minority areas and low-to-moderate minority areas, and finds that closures hit minority communities especially hard in a number of geographies. The study finds that closures disproportionately occur in places where few branches exist.
Agatha So & Jason Richardson, Hispanic Mortgage Lending: 2019 HMDA Analysis (Nov. 17, 2020), https://ncrc.org/hispanic-mortgage-lending-2019-analysis/.
- Analysis of 2019 HMDA data for Hispanic loan applicants finding disparate treatment and underservice in the mortgage market.
Anneliese Lederer & Sara Oros, Lending Discrimination During COVID-19: Black and Hispanic Women-Owned Businesses (Nov. 10, 2020), https://www.ncrc.org/lending-discrimination-during-covid-19-black-and-hispanic-women-owned-businesses/.
- Analysis of fair lending testing showing that financial institutions continued to discriminate against Black females and Hispanic male borrowers in Paycheck Protection Program loan distribution.
Jason Richardson et al., Redlining and Neighborhood Health (Sep. 10, 2020), https://www.ncrc.org/holc-health/.
- Report examining the health outcomes of historically redlined neighborhoods showing that historic redlining is correlated with present demographics. It finds a statistically significant relationship between redlining and population health, including those pre-existing conditions that can heighten risk of morbidity in COVID-19 patients and general life expectancy.
Seema Agnani & Jason Richardson, Mortgage Lending In The Asian American And Pacific Islander Community (Aug. 6, 2020), https://ncrc.org/mortgage-lending-in-the-asian-american-and-pacific-islander-community/.
- Analysis of 2018 HMDA data specifically related to lending in the Asian American and Pacific Islander Community. It finds disparities and variations within the community; Asian Indian and Chinese homebuyers, for example, were less likely than Filipino borrowers to use non-conventional loans.
Bruce Mitchell, Starved!: Significant Mortgage Lending Disparities Still Persist (Sep. 26, 2019), https://www.ncrc.org/starved-significant-mortgage-lending-disparities-still-persist/.
- Analysis of 38,049 census tracts finding that historically redlined neighborhoods continue to receive significantly fewer home mortgage and business loans.
Amber Lee, Bruce Mitchell, Anneliese Lederer, Disinvestment, Discouragement and Inequity in Small Business Lending (Sep. 4, 2019), https://ncrc.org/disinvestment/.
- Study of publicly available data on business lending and NCRC-conducted testing to analyze disparities in bank lending for small businesses based on race.
Bruce Mitchell, HOLC “Redlining” Maps: The Persistent Structure Of Segregation And Economic Inequality, (Mar. 20, 2018), https://www.ncrc.org/holc/.
- Study of historical redlining by HOLC that compares historical HOLC risk rating with recent social and economic conditions, as well as current levels of segregation and economic inequality. It reveals that economic and racial segregation created by historical redlining persists.
NCRC, Home Mortgage Lending In St. Louis, Milwaukee, Minneapolis, And Surrounding Areas (July 18, 2016), https://ncrc.org/home-mortgage-lending-in-st-louis-milwaukee-minneapolis-and-surrounding-areas/.
- Analysis of mortgage lending in St. Louis, Missouri finding significant disparities in mortgage credit distribution and access to banking resources based on income level and racial demographics of a neighborhood.
NCRC, Access To Capital And Credit In Appalachia And The Impact Of The Financial Crisis And Recession On Commercial Lending And Finance In The Region (Nov. 26, 2013), https://www.ncrc.org/access-to-capital-and-credit-in-appalachia-and-the-impact-of-the-financial-crisis-and-recession-on-commercial-lending-and-finance-in-the-region/.
- Study of recent trends in availability of capital and credit for small businesses in the Appalachian Region.
Tamara Jayasundera et al., Foreclosure in the Nation’s Capital: How Unfair and Reckless Lending Undermines Homeownership (Apr. 29, 2010), https://www.ncrc.org/foreclosure-in-the-nations-capital-how-unfair-and-reckless-lending-undermines-homeownership/.
- Study of subprime loans, non-traditional lending, and foreclosure in the Washington, D.C. Metropolitan Statistical Area which confirms racial disparities in loan type and foreclosure, even when borrowers had similar credit scores, incomes, loan-to-value ratios, and neighborhood characteristics. It finds that borrowers of color are more likely to receive subprime loans and were disproportionately affected by the foreclosure crisis.
The National Council of Negro Women & NCRC, Assessing the Double Burden: Examining Racial and Gender Disparities in Mortgage Lending (June 2009), https://ncrc.org/wp-content/uploads/2009/07/ncrc%20nosheild%20june%2009.pdf.
- Analysis of HMDA data finding significant racial/ethnic disparities in high-cost lending practices. Black and Hispanic borrowers were significantly more likely to receive high-cost loans than white borrowers within their income band.
 For simplicity, we use “race” to connote both race and ethnicity.
 See 12 U.S.C. § 2901 et. seq.
 The Office of the Comptroller of the Currency (OCC), the Federal Reserve Board (Fed), and the Federal Depository Insurance Corporation (FDIC). The Consumer Financial Protection Bureau (CFPB) also conducts fair lending reviews on banks with over $10 billion in assets, which are considered part of a bank’s CRA examination. Together, these agencies (and the National Credit Union Administration) comprise the Federal Financial Institutions Examination Council (FFIEC) and all CRA performance evaluations and ratings are posted on the FFIEC website. See https://www.ffiec.gov/cra/default.htm.
 See 12 U.S.C. § 2903(a)(1).
See 12 U.S.C. § 2903(a) (“In connection with its examination of a financial institution, the appropriate Federal financial supervisory agency shall—(1) assess the institution’s record of meeting the credit needs of its entire community, including low- and moderate-income neighborhoods, consistent with the safe and sound operation of such institution; and (2) take such record into account in its evaluation of an application for a deposit facility by such institution.”); see also 12 U.S.C. § 2902(3) (defining “application for a deposit facility”).
 12 U.S.C. § 2906(b)(2).
 Federal Reserve Chairman Ben Bernanke noted in 2007, for example, that public and congressional concerns about the deteriorating condition of America’s cities, particularly lower-income and minority neighborhoods, led to the enactment of CRA. Chairman Ben S. Bernanke, Board of Governors of the Federal Reserve System, The Community Reinvestment Act: Its Evolution and New Challenges (2007), https://www.federalreserve.gov/newsevents/speech/Bernanke20070330a.htm.
 123 Cong. Rec. 17604 (1977).
 See Candace Jackson, What is Redlining? (New York Times, Aug. 17, 2021), https://www.nytimes.com/2021/08/17/realestate/what-is-redlining.html.
Many books and papers have detailed the pervasiveness of racial discrimination in federal, state and local policies as well as by private financial institutions. See, e.g., Richard Rothstein, The Color of Law: A Forgotten History of How Our Government Segregated America (2017).
 Robert E. Litan, Nicolas P. Retsinas, Eric S. Belsky, Susan White Haag, The Community Reinvestment Act After Financial Modernization: A Baseline Report, (US Treasury Dept. 2000) at 46, (Economic Rationales); see also David C. Ling and Susan M. Wachter, Information Externalities in Home Mortgage Underwriting, 44 Journal of Urban Economics 317-32 (1998) (provides evidence of information externalities in neighborhood decline); Jack M. Guttentag & Susan M. Wachter, Redlining and Public Policy (Edwin Elton & Martin J. Gruber, New York University Graduate School of Business, 1980) at 1-50.
 See Guttentag et al., supra note 12, at 1-50.
 CohnReznick, The Community Reinvestment Act and its Effect on Housing Tax Credit Pricing (A CohnReznick Report 2013) at 10, https://ahic.org/images/downloads/Research_and_Education/the_community_reinvestment_act_and_its_effect_on_housing_tax.pdf.
 FFIEC, CRA National Aggregate Reports, https://www.ffiec.gov/craadweb/national.aspx (NCRC calculations of data from the FFIEC webpage); NCRC, Proposed Changes to CRA Puts Billions In Lending At Risk Each Year (Feb. 2020), https://ncrc.org/proposed-changes-to-cra-puts-billions-in-lending-at-risk-each-year/.
 See infra at § VI.
 Neil Bhutta et al., Changes in U.S. Family Finances from 2016 to 2019: Evidence from the Survey of Consumer Finances, 106 Fed. Rsrv. Bull. 5 (2020), https://www.federalreserve.gov/publications/files/scf20.pdf.
 Urban Institute, Data Talk: Black Homeownership Gap: Research Trends and Why the Growing Gap Matters (2019), https://www.urban.org/events/black-homeownership-gap-research-trends-and-why-growing-gap-matters.
 Martin J. Gruenberg, The Community Reinvestment Act: Its Origins, Evolution, and Future (2018), https://www.fdic.gov/news/speeches/2018/spoct2918.html (Numerous studies have underscored the critically important impact CRA has had in expanding access to credit in low- and moderate-income urban and rural communities since it was enacted. While large challenges remain, such as the significant disparities in access to credit in minority neighborhoods, the progress has been real.”) (internal citations omitted).
 Remarks by Former Acting Comptroller Brian P. Brooks at the Access to Capital Forum (July 17, 2020), https://www.occ.gov/news-issuances/speeches/2020/pub-speech-2020-96.pdf, (“While we can applaud the trillions of dollars that CRA has encouraged banks to spend in the communities they serve since 1977, we must be honest that we have not moved the needle much on important economic equality indicators…Minority homeownership has not changed much in 25 years, and African American homeownership has regressed to 1968 levels. The wealth gap is as large as it has ever been.…Banks have closed thousands of branches in poor areas, starved for financial services. The status quo has not worked.”) (internal citations omitted).
 Statement of Michael J. Hsu, Acting Comptroller of the Currency, before the US Senate Committee on Banking, Housing, and Urban Affairs (Aug. 3, 2021), https://www.occ.gov/news-issuances/congressional-testimony/2021/ct-occ-2021-79-written.pdf (noting “In the average US metropolitan area, homes in neighborhoods where the share of the population is 50 percent Black are valued at roughly half the price as homes in neighborhoods with no Black residents, suggesting that the most important source of generation wealth building has been denied this segment of the population.”).
 Federal Reserve System, Community Reinvestment Act, Advance Notice of Proposed Rulemaking (2020 ANPR), 85 Fed. Reg. 66412 (Oct. 19, 2020).
 Brooks Access to Capital Forum, supra note 21 (“While we can applaud the trillions of dollars that CRA has encouraged banks to spend in the communities they serve since 1977, we must be honest that we have not moved the needle much on important economic equality indicators.”).
 See Statement of Michael J. Hsu, Acting Comptroller of the Currency, before the US House Financial Services Committee (May 19, 2021), https://financialservices.house.gov/uploadedfiles/hhrg-117-ba00-wstate-hsum-20210519.pdf (“Historically, many low-income individuals have been treated by banks as either credits to be avoided or credits to be exploited. The OCC’s twin missions of ensuring equal access to financial services and fair treatment speak to both of these challenges…To address this problem, the OCC must work to strengthen regulations implementing the Community Reinvestment Act (CRA).”).
 The Fed has not yet taken the next step of publishing a notice of proposed rulemaking.
 Grutter v. Bollinger, 539 US 306, 326-27 (2003); see also, e.g., Adarand Constructors, Inc. v. Pena, 515 US 200, 227, 235 (1995).
 Id. at 227.
 See, e.g., Craig v. Boren, 429 US 190, 197 (1976).
 See, e.g., City of New Orleans v. Dukes, 427 US 297, 303 (1976) (per curiam).
 Grutter, 539 US at 327.
 See, e.g., City of Richmond v. J.A. Croson Co., 488 US 469, 496-497 (1989) (discussing the distinction between remedying “wrongs worked by specific instances of racial discrimination” and “remedying the effects of societal discrimination”) (internal marks omitted); Midwest Fence Corp. v U.S. Dep’t of Transp., 840 F.3d 932, 935 (7th Cir. 2016); Associated Gen. Contractors of Am., San Diego Chapter Inc. v. California, 713 F.3d 1187, 1196-1198 (9th Cir. 2013) (discussing lingering effects of discrimination); Ensley Branch, N.A.A.C.P. v. Siebels, 31 F.3d 1548, 1565 (11th Cir. 1994) (interest in remedying past or present discrimination is “widely accepted as compelling”); cf. Grutter, 539 US at 328 (“we have never held that the only governmental use of race that can survive strict scrutiny is remedying past discrimination.”).
 See Associated Gen. Contractors of Am., 713 F.3d at 1197 (government actor could meet strict scrutiny standards “if, looking at the evidence in its entirety, the data . . . suggest[ ] that public dollars are being poured into ‘a system of racial exclusion practiced by elements of the local construction industry.’”)(quoting Croson, 488 US at 492); Concrete Works of Colo., Inc. v. Denver, 321 F.3d 950, 958 (10th Cir. 2010) (Denver “clearly may take measures to remedy its own discrimination or even to prevent itself from acting as a ‘passive participant in a system of racial exclusion practiced by elements of the local construction industry.’”) (quoting Croson, 488 US at 492); W. States Paving Co. v. Wash. State Dep’t of Transp., 407 F.3d 983, 991 (9th Cir. 2005); see also Adarand Constructors, Inc. v. Slater, 228 F.3d 1147, 1165 (10th Cir. 2000) (Slater) (“the federal government has a compelling interest in not perpetuating the effects of racial discrimination in its own distribution of federal funds”); DynaLantic Corp. v. U.S. Dep’t of Def., 885 F. Supp. 2d 237, 252 (D.D.C 2012) (“the government may implement race-conscious programs not only for the purpose of correcting its own discrimination, but also ‘to prevent itself from acting as a ‘passive participant’ in private discrimination in the relevant industries or markets’”) (internal citations omitted).
 W. States Paving Co., 407 F.3d at 1002.
 See, e.g., Midwest Fence Corp., 840 F. 3d at 945 (quoting Wygant v. Jackson Bd. of Educ., 476 US 267, 277 (1986)).
 See Croson, 488 US at 510 (“Proper findings in this regard are necessary to define both the scope of the injury and the extent of the remedy necessary to cure its effects.”).
 Grutter, 539 US at 333 (cleaned up).
 See id. at 333-42; HB Rowe Co., Inc. v. Tippett, 615 F.3d 233, 252 (4th Cir. 2010); Midwest Fence Corp., 840 F.3d at 942-46; Sherbrooke Turf, Inc. v. Minn. Dep’t of Trans., 345 F.3d 964, 971-73 (8th Cir. 2003); W. States Paving Co., 407 F.3d at 993-95; Slater, 228 F.3d at 1178-87.
 In Adarand, 515 US at 237, Justice O’Connor explicitly pushed back against the idea that strict scrutiny is “strict in theory, but fatal in fact.”
 488 US 469 (1988).
 515 US 200 (1995).
 Croson, 488 US at 499.
 Id. at 510.
 Id. at 498.
 Id. (internal citations omitted); accord W. States Paving Co., 407 F.3d at 1002.
 Croson, 488 US at 506 (emphasis in original).
 Id. To emphasize its point that including people beyond African Americans was over-inclusive given the (lack of) evidence, the Court observed that “[i]t may well be that Richmond has never had an Aleut or Eskimo citizen.” Id.
 Adarand, 515 US at 204, 208. The opinion evidences some lack of clarity as to whether members of minority groups were entitled to a presumption of economic disadvantage, but that did not appear to be a key driver of the decision. See id. at 207-08 (the subcontracting program “appears to require an individualized . . . showing of economic disadvantage. . . . A different set of regulations, however, says that members of minority groups wishing to participate in the . . . subcontracting program are entitled to a race-based presumption of social and economic disadvantage. . . . . We are left with some uncertainty as to whether participation in the . . . subcontracting program requires an individualized showing of economic damage.”). Other programs based on presumed social and economic disadvantage clearly include an economic limit. See, e.g., 13 C.F.R. § 124.104(c)(2) (net worth limit in Small Business Administration program).
 Adarand, 515 US at 207.
 Id. at 208. Although the Court did not specifically discuss whether these factors would help the program survive strict scrutiny, instead remanding the case, these factors likely would have met with approval by the Court in light of additional precedent in the area.
 See, e.g., HB Rowe Co., Inc., 615 F.3d at 252 (4th Cir. 2010); Midwest Fence Corp., 840 F.3d at 942-46; Sherbrooke Turf, Inc. v. Minn. Dep’t of Trans., 345 F.3d 964, 971-73 (8th Cir. 2003); W. States Paving Co., 407 F.3d at 993-95; Slater, 228 F.3d at 1178-87.
 Id. at 237-38 (internal citations and quotation marks omitted).
 Id. at 238.
 See, e.g., W. States Paving Co., 407 F.3d at 1002.
 See Midwest Fence Corp., 840 F.3d at 950-51.
 See id. at 949 n.1, 951. These might be individuals actually seeking financial services or individuals engaged in testing through mystery shopping. See generally, e.g., NCRC, Racial and Gender Mystery Shopping for Entrepreneurial Loans, https://ncrc.org/wp-content/uploads/2020/02/NCRC-Mytery-Shopping-Race-and-Gender-v8.pdf.
 Croson, 488 US at 509; see also, e.g., Slater, 228 F.3d at 1173-74 (quoting Croson’s explanation of the relevance of “significant statistical disparities.”)
 See Associated General Contractors of Am., 713 F.3d at 1197-98 (anecdotal accounts of discrimination “confirmed” findings of statistical disparity study); Mountain W. Holding Co. v. Montana, 691 F. App’x 326, 331 (9th Cir. 2017) (“Without a statistical basis, the State cannot rely on anecdotal evidence alone.”)
 See, e.g., W. States Paving Co., 407 F.3d at 994 (“A quota system is the hallmark of an inflexible affirmative action program.”)
 See Vitolo v. Guzman, 999 F.3d 353, 363 (6th Cir. 2021).
 See Grutter, 539 US at 335; Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 US 701, 723-724 (2007); id. at 726 (“The plans are tied to each district’s specific racial demographics, rather than to any pedagogic concept of the level of diversity needed to obtain the asserted educational benefits.”)
 See, e.g., Gratz v. Bollinger, 539 US 244, 255-56 (2003) (describing university admissions procedure in which applicants could score up to a maximum of 150 points, in which an applicant received 20 points “based upon his or her membership in an underrepresented racial or ethnic minority group.”).; id. at 270 (“We find that the University’s policy, which automatically distributes 20 points, or one-fifth of the points needed to guarantee admission, to every single ‘underrepresented minority’ applicant solely because of race, is not narrowly tailored to achieve the interest in educational diversity that respondents claim justifies their program.”)
 See Parents Involved, 551 US at 722-23 (discussing the “highly individualized, holistic review” upheld in Grutter).
 539 US at 336-37.
 See, e.g., Bethune-Hill v. Va. State Bd. of Elections, 580 U.S. ___, 137 S. Ct. 788 (2017) (discussing applicability of and requirements for triggering strict scrutiny in context of racial gerrymandering). Cases involving alleged infringement of other fundamental rights—for example, content-based restrictions on speech implicating rights under the First Amendment—also may be evaluated under strict scrutiny.
 999 F.3d 353 (6th Cir. 2021).
 Pub. L. No. 117-2 § 5003(c)(3)(A).
 13 C.F.R. § 124.103(b)(1).
 Id. § 124.103(c)(1).
 Vitolo, 999 F.3d at 361.
 See id. at 361-62. The court also discussed a lack of evidence that the government had participated in the discrimination to be addressed through the program, rejecting the government’s argument that “delays” in providing funds to minority businesses and the fact that “prior relief programs had failed to reach” minority businesses did not establish discrimination. Id. at 362.
 Id. at 363 (“The non-designated races start with a much higher hurdle. They must bring forward evidence that they suffered episodes of discrimination, which have ‘negatively impacted’ their ‘advancement in the business world,’ and which caused them to suffer ‘chronic and substantial social disadvantage.’ . . . Put this high hurdle against the rapid depletion of the money and the 21-day window, and the hurdle becomes a wall.”)
 Id. (for example, giving “priority consideration to all small business owners who have not yet received coronavirus relief funds).
 __ F. Supp. 3d __, 2021 WL 2580678, at *1 (M.D. Fla. June 23, 2021).
 Id. at *5.
 See id.
 Id. at *5 (“the actual evidentiary support for the inadequacy of past remedial measures is limited and largely conclusory”); see id. at *6.
 See id. at 11.
 Id. at *7.
 Id. at *10. The court also held that there was little evidence that the government gave serious consideration to, or tried, race-neutral alternatives. Id. at *11.
 See, e.g., Adarand, 515 US at 227.
 Grutter, 539 US at 326-27; see also, e.g., Adarand, 515 US at 224 (“any person, of whatever race, has the right to demand that any government actor subject to the Constitution justify any racial classification subjecting that person to unequal treatment under the strictest judicial scrutiny.”); Grutter, 539 US at 326 (“We apply strict scrutiny to all racial classifications to smoke out illegitimate uses of race by assuring that [the] government is pursuing a goal important enough to warrant use of a highly suspect tool.”)(cleaned up).
 See 28 U.S.C. § 2801 et seq.; 12 C.F.R. § 1003.1 et seq.
 CFPB, Data Point: 2019 Mortgage Market Activity and Trends, A First Look at the 2019 HMDA Data (June 2020), https://files.consumerfinance.gov/f/documents/cfpb_2019-mortgage-market-activity-trends_report.pdf, at 4.
 See, e.g., USDA National Agriculture Statistics Service, Census of Agriculture, https://www.nass.usda.gov/Publications/AgCensus/2012/Online_Resources/Race,_Ethnicity_and_Gender_Profiles/; CDC National Center for Health Statistics, National Health and Nutrition and Examination Survey, https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?cycle=2017-2020; US Department of Education, Office of Civil Rights, Annual Report to the Secretary, the President, and the Congress, Fiscal Year 2019 (2020), https://www2.ed.gov/about/reports/annual/ocr/report-to-president-and-secretary-of-education-2019.pdf.
 See 81. Fed. Reg. 67398 (Sep. 30, 2016) (Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity).
 81 Fed. Reg. 48506, at 48535 (§ __.21(b)-1) (July 25, 2016) (issued by the OCC, the Fed, and the FDIC) (Interagency Q&A)).
 12 C.F.R. § 345.21(b) (FDIC); 12 C.F.R. § 228.21(b) (Fed); 12 C.F.R. Part. 25, App. C, § 25.21(b) (OCC).
 2020 ANPR, 85 Fed. Reg. 66410 at 66414; see also id. (“The Board also considers applicable performance context information to inform its analysis and conclusions when conducting CRA examinations.”)
 See 12 U.S.C. § 2901(a)(3) (“regulated financial institutions have continuing and affirmative obligation to help meet the credit needs of the local communities in which they are chartered”) (emphasis added); id. § 2901(b) (“It is the purpose of this chapter to require each appropriate Federal financial supervisory agency to use its authority when examining financial institutions, to encourage such institutions to help meet the credit needs of the local communities in which they are chartered consistent with the safe and sound operation of such institutions.”) (emphasis added).
 2020 ANPR, 85 Fed. Reg. at 66422-23.
 See, e.g., id. at 66420, 66422.
 Including both kinds of tracts recognizes that there are large differences in performance context across the country with respect to the number of minority residents. It would assure that attention to minority lending is not diminished where the minority population is too small or integrated to create majority-minority tracts. There are many ways to define tracts that are disproportionately minority. For example, comparing the percent of African Americans in each census tract with the percent of African Americans in the core-based statistical area as a whole, similar to the way in which neighborhood income is calculated. Alternatively, an analysis of spatial clustering of specific racial groups could be used to detect small concentrations of ethnic or racial minorities that correspond with lower levels of investment. In any case, it is important to look at neighborhood race in a local context where ‘majority-minority’ may not accurately portray local conditions or structural barriers to investment.
 2020 ANPR, 85 Fed. Reg. at 66421-22.
 For these bank-specific metrics as well as the benchmarks, the Fed proposes using the number of loans instead of the dollar amount of the loans. See id. In both the LMI and race contexts, this better captures how well lenders are serving their “entire communities,” 12 U.S.C. § 2903(a)(1), because it avoids overweighting larger loans likelier to go to non-LMI and non-minority borrowers.
 2020 ANPR, 85 Fed. Reg. at 66429-31.
 Id. These benchmarks might only be calculated if a bank has minimum number of branches in the assessment area. That minimum has not been proposed yet. See id. at 66431.
 This chart is constructed to mirror the LMI one in the Fed’s proposal. See id. at 66431.
 2020 ANPR, 85 Fed. Reg. at 66438-39.
 Id. at 66439-40.
 Id. at 66424-25
 Displaying both benchmarks departs somewhat from the Fed’s proposal. See id. at 66423-25. Under the proposal, “[f]or each distribution metric, the lower of the community threshold or market threshold would be selected as the binding threshold,” id. at 66424, and only that one published on the dashboard, id. at 66425.
 Id. at 66420.
 Id. at 66440. Since the statement is specific to benchmarks and omits mention of metrics, it appears not to contemplate public disclosure of institution-specific information about community development financing. As addressed next, we believe public disclosure should include institution-specific information.
 See 12 U.S.C. § 2906(a)(2).
 NCRC, Massachusetts CRA for Mortgage Companies: A Good Starting Point for Federal Policy (2021) at 11, https://ncrc.org/massachusetts-cra-for-mortgage-companies-a-good-starting-point-for-federal-policy/.
 See, e.g., Mass. Division of Banks & FDIC, CRA Performance Evaluation—BayCoastBank (Sept. 5, 2017), https://www.mass.gov/doc/baycoast-bank-cra-pe/download, at 22.
 Laura Choi & William Dowling, Federal Reserve Bank of San Francisco, Understanding Community Development Needs Through CRA Performance Context (Dec. 2014), https://www.frbsf.org/community-development/files/Understanding-Community-Development-Needs-through-the-CRA-Performance-Context.pdf. at 5-6.
 Id. at 6. The 12th district includes Alaska, Arizona, California, Hawaii, Idaho, Nevada, Oregon, Utah, and Washington—plus American Samoa, Guam, and the Commonwealth of the Northern Mariana Islands. See Federal Reserve Bank of San Francisco, About Us, https://www.frbsf.org/our-district/about/.
 Id. at 6.
 Federal Reserve Bank of San Francisco, The 12th District Vantage Point Community Indicators Project (Feb. 2014), at 26, https://www.frbsf.org/community-development/publications/vantage-point/2014/march/2013-community-indicators-survey/.
 Croson, 488 US at 510.
 Midwest Fence Corp., 840 F.3d at 945.
 See, e.g., W. States Paving Co., 407 .F.3d at 997.
 See, e.g., id. at 996 (“to be narrowly tailored, a national program must be limited to those parts of the country where its race-based measures are demonstrably needed”) (citation and internal quotation marks omitted; emphasis in original).
 Mitchell, HOLC “Redlining” Maps, supra note 11.
 See John Kimble, Insuring Inequality: The Role of the Federal Housing Administration in the Urban Ghettoization of African Americans, 32 Law & Social Inquiry 399, 402 (2007), available at https://www.jstor.org/stable/20108708 (“The passage of the National Housing Act in 1934 inaugurated a vigorously interventionist approach to the [financial crisis of the Depression] in which the federal government could orchestrate private market activity without acting as a mortgage lender. The act created the Federal Housing Administration and empowered it to coordinate several mortgage market reforms, chief among which was to insure lenders against any loss on loans made for purchasing homes.”); Ellen Seidman & Andrew Jakabovics, Learning from the Past (Center for American Progress, Sep. 19, 2008), https://www.americanprogress.org/issues/economy/reports/2008/09/19/4902/learning-from-the-past/; John Griffith, The Federal Housing Administration Saved the Housing Market (Center for American Progress, Oct. 11, 2012), https://www.americanprogress.org/issues/economy/reports/2012/10/11/40824/the-federal-housing-administration-saved-the-housing-market/.
 See, e.g., Seidman & Jakabovics, Learning from the Past, supra note 120 (between 1933 and 1936, HOLC issued approximately 1 million refinance loans with an average value of slightly more than $3,000 per property; approximately 20% ultimately were foreclosed upon); Alan S. Blinder, From the New Deal, a Way Out of Mess (New York Times, Feb 24, 2008), https://www.nytimes.com/2008/02/24/business/24view.html; John Griffith, The Federal Housing Administration Saved the Housing Market, supra note 120; NFHA, The Case for Fair Housing https://nationalfairhousing.org/wp-content/uploads/2017/07/TRENDS-REPORT-2017-FINAL.pdf (2017) at 17 (“One of the great contributions of the HOLC was the creation of the low down payment, long-term, fixed-rate, fully-amortizing mortgage. Although the 30 year, fixed-rate loan is the most common type of mortgage in the US housing market today, before the HOLC mortgages had very short terms (five to ten years) and were non-amortizing, so that at the end of the five- to ten-year period, borrowers needed to take out a new loan to pay off the remaining principal balance. The HOLC’s innovative mortgage product eliminated much of the volatility of the mortgage market and made mortgages (and therefore homeownership) less risky for borrowers.”)
 See, e.g., Becky Little, How a New Deal Housing Program Enforced Segregation (Oct. 20 2020),https://www.history.com/news/housing-segregation-new-deal-program (“One of the multiple programs a newly-elected Franklin D. Roosevelt established to stimulate the economy offered home-buying aid for Americans—but only white Americans. The Federal Housing Administration, operated through the New Deal’s National Housing Act of 1934, promoted homeownership by providing federal backing of loans—guaranteeing mortgages. But from its inception, the FHA limited assistance to prospective white buyers.”); Terry Gross, A ‘Forgotten History’ Of How The U.S. Government Segregated America (May 3, 2017), https://www.npr.org/2017/05/03/526655831/a-forgotten-history-of-how-the-u-s-government-segregated-america; but see Amy E. Hillier, Who Received Loans? Home Owners’ Loan Corporation Lending and Discrimination in Philadelphia in the 1930’s, 2 Journal of Planning History 3 (2003) (HOLC did make loans to African Americans, Jews, and immigrants in Philadelphia; HOLC supported racial segregation in the process of reselling properties acquired through foreclosure).
 See NFHA, The Case for Fair Housing, supranote 121, at 18 (“HOLC undertook a survey to assess risk in cities all across the country–virtually all of those with populations of 40,000 persons or greater. Through this undertaking, the HOLC standardized and formalized the appraisal process. Its methodology was not new; rather, it codified the typical private sector practices of the time, which reflected a view of neighborhood dynamics and their impact on property values that was blatantly discriminatory against Black people, immigrants from certain countries, and some religious groups.”)
 See University of Richmond, Mapping Inequality, https://dsl.richmond.edu/panorama/redlining/#loc=5/39.745/-97.217&text=intro; NFHA, The Case for Fair Housing supra,note 121, at 18 (“On the so-called Residential Security maps on which its survey results were recorded, these were color-coded green, blue, yellow, and red, respectively. Neighborhoods could be coded red, or ‘hazardous,’ for a number of reasons, one of which was the presence of Blacks or other ‘inharmonious’ racial or social groups.”)
 See Mitchell, HOLC “Redlining” Maps, supra note 11. HOLC maps document which areas were considered lower risk, and therefore preferred for loans, and higher-risk areas where lending was discouraged.); University of Richmond, Mapping Inequality, supra note 124 (“Conservative, responsible lenders, in HOLC judgment, would ‘refuse to make loans in [hazardous] areas or only on a conservative basis.’” (internal marks omitted)).
 See Kimble, Insuring Inequality, supra note 120, at 405 (FHA’s “City Survey Files” contained HOLC maps).
 Id. at 404.
 Id. at 405; see also id. at 406 (discussing FHA Real Property Inventories).
 Brenda Richardson, Redlining’s Legacy Of Inequality: Low Homeownership Rates, Less Equity For Black Households (June 11, 2020) https://www.forbes.com/sites/brendarichardson/2020/06/11/redlinings-legacy-of-inequality-low-homeownership-rates-less-equity-for-black-households/?sh=216a8c442a7c; Mitchell, HOLC “Redlining” Maps, supra note 11; Tracy Jan, Redlining was banned 50 years ago. It’s still hurting minorities today. (Mar. 28. 2018), https://www.washingtonpost.com/news/wonk/wp/2018/03/28/redlining-was-banned-50-years-ago-its-still-hurting-minorities-today/ (covering NCRC HOLC report).
 See, e.g., Jason Richardson et al., Redlining and Neighborhood Health, https://ncrc.org/holc-health/; Maria Godoy, In U.S. Cities, The Health Effects Of Past Housing Discrimination Are Plain To See (Nov. 19, 2020), https://www.npr.org/sections/health-shots/2020/11/19/911909187/in-u-s-cities-the-health-effects-of-past-housing-discrimination-are-plain-to-see (“In city after city across the US, from Milwaukee to Miami, researchers have found a disturbing pattern: People who live in neighborhoods that were once subjected to a discriminatory lending practice called redlining are today more likely to experience shorter life spans – sometimes, as much as 20 or 30 years shorter than other neighborhoods in the same city.” (in part covering NCRC study)); Nancy Krieger et al., Cancer Stage at Diagnosis, Historical Redlining, and Current Neighborhood Characteristics: Breast, Cervical, Lung, and Colorectal Cancers, Massachusetts, 2001–2015, 189 American Journal of Epidemiology 1605-75, https://academic.oup.com/aje/article/189/10/1065/5812653, (living in a formerly HOLC redlined area imposed an elevated risk for late-stage cancer diagnosis, even for residents of census tracts with present-day economic and racial privilege);
 Hiroko Tabuchi & Nadja Popovich, People of Color Breathe More Hazardous Air. The Sources Are Everywhere. (Apr. 28, 2021), https://www.nytimes.com/2021/04/28/climate/air-pollution-minorities.html; Brad Plumber & Nadja Popovich, How Decades of Racist Housing Policy Left Neighborhoods Sweltering (Aug. 24, 2020), https://www.nytimes.com/interactive/2020/08/24/climate/racism-redlining-cities-global-warming.html; Ariel Drehobl, Lauren Ross, Roxana Ayala, How High Are Household Energy Burdens?, https://www.aceee.org/sites/default/files/pdfs/u2006.pdf (disparate access to energy-efficient appliances, upgrades).
 See Margery Austin Turner & Felicity Skidmore, Mortgage Lending Discrimination: A Review of Existing Evidence (June 1999), https://www.urban.org/sites/default/files/publication/66151/309090-Mortgage-Lending-Discrimination.PDF (1999 review/meta-analysis); Emmanuel Martinez & Lauren Kirchner, The Secret Bias Hidden in Mortgage-Approval Algorithms (The Markup, Aug. 25, 2021), https://themarkup.org/denied/2021/08/25/the-secret-bias-hidden-in-mortgage-approval-algorithms.
 See, e.g., Michael E. Hodge, Mark C. Dawkins, Jaxk H. Reeves, A Case Study of Mortgage Refinancing Discrimination: African American Intergenerational Wealth (Jan. 19 2007) https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1475-682X.2007.00176.x; Robert Bartlett et al., Consumer-Lending Discrimination in the FinTech Era (Nov. 2019), https://faculty.haas.berkeley.edu/morse/research/papers/discrim.pdf; Diana Olick, A troubling tale of a Black man trying to refinance his mortgage (Aug. 19, 2020), https://www.cnbc.com/2020/08/19/lenders-deny-mortgages-for-blacks-at-a-rate-80percent-higher-than-whites.html.
 Alix Carlin & Eileen Divringi, How Improvement Lending in the Third Reserve District: Patterns by Income, Race, and Gender (Dec. 2018), https://www.philadelphiafed.org/community-development/housing-and-neighborhoods/home-improvement-lending-in-the-third-federal-reserve-district-patterns-by-income-race-and-gender.
 See, e.g., EWG, Timeline: Black farmers and the USDA, 1920 to present, https://www.ewg.org/research/black-farmer-usda-timeline/; National Rosenberg & Bryce Wilson Stucki, How USDA distorted data to conceal decades of discrimination against Black farmers (2019) https://thecounter.org/usda-black-farmers-discrimination-tom-vilsack-reparations-civil-rights/; Ximena Bustillo, ‘Rampant issues’: Black farmers are still left out at USDA (July 4, 2021) https://www.politico.com/news/2021/07/05/black-farmers-left-out-usda-497876; Lloyd Wright, Discrimination has become a fact of life for Black farmers–that must end (Oct. 9, 2020), https://fortune.com/2020/10/09/black-farmers-usda-racism-pigford/.
 See, e.g., Amber Lee, Bruce Mitchell, Anneliese Lederer, Disinvestment, Discouragement and Inequity in Small Business Lending, https://ncrc.org/disinvestment/; David G. Blanchflower, Phillip B. Levine, & David J. Zimmerman, Discrimination in the Small-Business Credit Market, 85 The Review of Economics and Statistics 930 (Nov. 2003), https://www.jstor.org/stable/3211816; Federal Reserve Bank of Atlanta, Small Business Credit Survey: 2018 Report on Minority-Owned Firms (December 2019), https://www.fedsmallbusiness.org/medialibrary/fedsmallbusiness/files/2019/20191211-ced-minority-owned-firms-report.pdf; Robert D. Fairlie & Alicia M. Robb, Disparities in Capital Access between Minority and Non-Minority-Owned Businesses: The Troubling Reality of Capital Limitations Faced by MBEs (January 2010), https://www.mbda.gov/sites/default/files/migrated/files-attachments/DisparitiesinCapitalAccessReport.pdf.
 See, e.g., Hilary Hurd Anyaso, Racial discrimination in mortgage market persistent over last four decades (Jan. 23 2020), https://news.northwestern.edu/stories/2020/01/racial-discrimination-in-mortgage-market-persistent-over-last-four-decades/.
 See, e.g., Daniel Aaronson, Daniel Hartley, Bhash Mazumder, The Effects of the 1930s HOLC “Redlining” Maps (Aug. 2020) https://www.chicagofed.org/publications/working-papers/2017/wp2017-12; Bruce Mitchell, Starved!: Significant Mortgage Lending Disparities Still Persist (Sep. 26, 2019) https://www.ncrc.org/starved-significant-mortgage-lending-disparities-still-persist/.
 Department of Justice, Justice Dep’t and OCC Announce Actions to Resolve Lending Discrimination Claims Against Cadence Bank (Aug. 30, 2021), https://www.justice.gov/opa/pr/justice-department-and-office-comptroller-currency-announce-actions-resolve-lending.
 Urban Institute, Data Talk: Black Homeownership Gap: Research Trends and Why the Growing Gap Matters supra note 19.
 See, e.g., Lee, Mitchell, & Lederer, Disinvestment, Discouragement and Inequity in Small Business Lending, supra note 136, at 9-10.
 Federal Reserve Bank of Minneapolis, Racism and the Economy (2021), https://www.minneapolisfed.org/policy/racism-and-the-economy.
 Alyssa Augustine, Fed’s “Racism and the Economy” series explores housing inequity (Mar. 5, 2021), https://www.minneapolisfed.org/article/2021/feds-racism-and-the-economy-series-explores-housing-inequity.
 Victor W. Hwange, The “Right to Start” Can Expand Entrepreneurial Opportunity in Community of Colors and Beyond https://www.minneapolisfed.org/~/media/assets/events/2021/racism-and-the-economy-focus-on-entrepreneurship/hwang-proposal.pdf?la=en; Jay Lindsay, Overcoming history: Entrepreneurs of color forced to confront constraints from the past, (June 7. 2021), https://www.minneapolisfed.org/article/2021/overcoming-history-entrepreneurs-of-color-forced-to-confront-constraints-from-the-past.
 Alyssa Augustine, Fed’s “Racism and the Economy” series explores racial inequity in the education system (Jan. 15, 2021), https://www.minneapolisfed.org/article/2021/feds-racism-and-the-economy-series-explores-racial-inequity-in-the-education-system.
 Federal Reserve Bank of Minneapolis, Fed’s “Racism and the Economy” series explores racial inequities in labor markets (Dec. 2, 2020) https://www.minneapolisfed.org/article/2020/feds-racism-and-the-economy-series-explores-racial-inequities-in-labor-markets.
 See, e.g., Ben Horowitz et al., Systemic racism haunts homeownership rates in Minnesota (Feb. 25, 2021) https://www.minneapolisfed.org/article/2021/systemic-racism-haunts-homeownership-rates-in-minnesota.
 See, e.g., Fairlie & Robb, Disparities in Capital Access between Minority and Non-Minority-Owned Businesses, supra note 136; Algernon Austin, The Color of Entrepreneurship: Why The Racial Gap Among Firms Costs The U.S. Billions (Apr. 20, 2016), http://globalpolicysolutions.org/report/color-entrepreneurship-racial-gap-among-firms-costs-u-s-billions/.
 See Alicia Robb, Financing Patterns and Credit Market Experiences: A Comparison by Race and Ethnicity for U.S. Employer Firms (Feb. 2018), https://www.sba.gov/sites/default/files/Financing_Patterns_and_Credit_Market_Experiences_report.pdf at 3-4.
 See, Gene Marks, Black-owned firms are twice as likely to be rejected for loans. Is this discrimination? (Jan. 16, 2020), https://www.theguardian.com/business/2020/jan/16/black-owned-firms-are-twice-as-likely-to-be-rejected-for-loans-is-this-discrimination.
 See, e.g., Sifan Lieu & Joseph Parilla, New data shows small business in communities of color has unequal access to federal COVID-19 relief (Sep. 17, 2020), https://www.brookings.edu/research/new-data-shows-small-businesses-in-communities-of-color-had-unequal-access-to-federal-covid-19-relief/; Anneliese Lederer & Sara Oros. Lending Discrimination During COVID-19: Black and Hispanic Women-Owned Businesses, https://www.ncrc.org/lending-discrimination-during-covid-19-black-and-hispanic-women-owned-businesses/.
 Adarand, 515 US at 227.
 The Compelling Interest for Affirmative Action in Federal Procurement: A Preliminary Survey, 61 Fed. Reg. 26050, 26050-51 (May 23, 1996).
 See, e.g., Slater, 228 F.3d at 1176 (“we conclude that the evidence cited by the government and its amici, particularly that contained in The Compelling Interest, 61 Fed. Reg. 26,050, more than satisfies the government’s burden of production regarding the compelling interest for a race-conscious remedy.”).
 Evidence developed after adoption of a race-conscious program is admissible in defense of the program. See, e.g., Slater, 228 F.3d at 1166; Wynn, 2021 WL 2580678, at *4.
 Grutter, 539 US at 339.
 Fisher v Univ. of Austin, 570 US 297, 312 (2013) (internal quotation marks and citations omitted).
 See, e.g., 12 U.S.C. § 2903(a)(1); 12 C.F.R. § 228.11(b)(1).
 See, e.g., 12 C.F.R. § 228.12(g)(4)(iii)(A) (rates of unemployment and population loss).
 See, e.g., 12 C.F.R. § 228.28(c); OCC, Federal Reserve Board, and FDIC, Large Institution CRA Examination Procedures (April 2014) at 2, available at https://www.ffiec.gov/cra/pdf/cra_exlarge.pdfat 18 (§§ 13-15) (Interagency Large Institution Exam Procedures).
 See 15 U.S.C. § 1691(c)(3); 12 C.F.R. § 1002.8(b)(2).
 See Pub. L. 94-239, § 701(c)(3), 90 Stat. 251, 251 (1976).
 See Administration for Children and Families, Temporary Assistance for Needy Families (TANF), https://www.acf.hhs.gov/ofa/programs/temporary-assistance-needy-families-tanf; HUD, About the Housing Choice Vouchers Program, https://www.hud.gov/program_offices/public_indian_housing/programs/hcv/about; USDA, Supplemental Nutrition Assistance Program (SNAP), https://www.fns.usda.gov/snap/supplemental-nutrition-assistance-program.
 See, e.g., US Department of Education, Office of Federal Student Aid, Types of Financial Aid, https://studentaid.gov/understand-aid/types;US Department of Education, Office of Federal Student Aid, Federal Pell Grants, https://studentaid.gov/understand-aid/types/grants/pell; US Department of Education, Office of Federal Student Aid, Income-Driven Repayment (IDR) Plan Request, https://studentaid.gov/articles/student-loan-forgiveness/; US Department of Education, Student Loans, Forgiveness, https://www2.ed.gov/fund/grants-college.html?src=pn.
 12 U.S.C. § 2903(a)(1).
 See 12 C.F.R. § 228.41(e)(3).
 Id. § 228.41(e)(2). Facially neutral policies that cause exclusion may constitute illegal discrimination under the disparate impact test if they do not serve a legitimate business need or if that need can be met through a less discriminatory alternative. See, e.g., Texas Dept. of Hous. & Comm. Affairs v. Inclusive Communities Project, Inc., 576 US 519 (2015) (finding disparate impact claims cognizable under the Fair Housing Act). This is narrower than arbitrary exclusion.
 U.S. and CFPB v. BancorpSouth Bank, Case No. 1:16-cv-118-GHD-DAS, at ¶¶ 28-30 (N.D. Miss. Filed June 29, 2016), available at https://files.consumerfinance.gov/f/documents/201606_cfpb_bancorpsouth-joint-complaint.pdf.
 United States v. Eagle Bank & Tr. Co., Case No. 4:15-cv-01492, at ¶16 (E.D. Mo. Filed Sept. 29, 2015), available at https://www.justice.gov/crt/case-document/complaint-united-states-v-eagle-bank-and-trust-company-missouri-ed-mo.
 United States v. First Merchants Bank, Case No. 1:19-cv-02365-JPH-MPB, at ¶11(a) (S.D. Ind. Filed June 13, 2019), available at https://www.justice.gov/crt/case-document/file/1174041/download.
 CFPB and U.S. v. Hudson City Savings Bank, F.S.B., Case No. 2:15-cv-07056-CCC-JBC (D.N.J. filed Sept. 24, 2015), at ¶¶ 38, 45, available at https://files.consumerfinance.gov/f/201509_cfpb_hudson-city-joint-complaint.pdf.
 United States v. KleinBank, Case No. 0:17-cv-00136, at ¶ 18 (D. Minn. Filed Jan. 13, 2017), available at https://www.justice.gov/crt/case-document/file/927076/download.
 Disregard due to nefarious but well-hidden reasons would likewise be diminished. Discriminatory banks that are subtle enough to avoid being identified as such would recognize that any racially-motivated assessment area manipulation will likely be identified as a problem, even if not as a violation of anti-discrimination laws.
 WFAA, ‘They Underestimate What We Can Do’: WFAA Finds Banks Exclude Blacks, Hispanics in Southern Dallas From Access to Loans (Nov. 22, 2020), https://www.wfaa.com/article/news/local/investigates/banking-below-30-southern-dallas-cut-off-by-freeway-also-left-off-banking-maps/287-10557dd3-bbf4-44a2-b786-44c6347a6e48.
 See 12 C.F.R. § 228.41(e)(3).
 For example, such a standard would not require, or even encourage, a bank to include a 20% minority area in its assessment area in the Washington, D.C. area, which is 45% non-Hispanic white.
 The Asian population is less than one percent in the city of Jackson. See US Census Bureau, QuickFacts, Jackson City, Mississippi, https://www.census.gov/quickfacts/jacksoncitymississippi. Discrimination against Asians would, or course, remain relevant to the evaluation.
 Anecdotal evidence might be gathered from, for example, nonprofit organizations dedicated to addressing fair lending and other civil rights issues. As part of the study, regulators could systematically request relevant anecdotal information from such organizations.
 See 12 C.F.R. § 228.41(b)-(c).
 See 2020 ANPR, 85 Fed. Reg. at 66454.
 See id. at 66411, 66455. NCRC supports eliminating it because smaller metropolitan areas and rural areas are disproportionately designated as limited-scope assessment areas, which limits the weight or influence they have on CRA examinations. This can lead banks to neglect credit needs in these areas since they effectively do not count as much.
 OCC, Federal Reserve Board, and FDIC, Large Institution CRA Examination Procedures, supra note 163; OCC, Federal Reserve Board, FDIC, and OTS, Intermediate Small Institution CRA Examination Procedures (July 2007) at 1, available at https://www.ffiec.gov/cra/pdf/isbank.pdf. (Interagency Intermediate Small Institution Exam Procedures). Classifications are based on asset size. See, e.g., Explanation of the Community Reinvestment Act Asset-Size Threshold Change, https://www.ffiec.gov/cra/pdf/2021_Asset_Size_Threshold.pdf.
 FDIC, Consumer Compliance Examination Manual: CRA Scoping Guidance (May 2017), https://www.fdic.gov/resources/supervision-and-examinations/consumer-compliance-examination-manual/index.html, at XI-10.4.
 See US Census Bureau, QuickFacts, Baltimore city, Maryland, https://www.census.gov/quickfacts/fact/table/baltimorecitymaryland,US/PST045219.
 Interagency Large Institution Exam Procedures at 2, Interagency Intermediate Small Institution Exam Procedures at 1.
 OCC, Federal Reserve Board, FDIC, and OTS, Small Institution CRA Examination Procedures (July 2007) at 1, available at https://www.ffiec.gov/cra/pdf/cra_exsmall.pdf (Interagency Small Institution Exam Procedures).
 Interagency Large Institution Exam Procedures at 1 (¶ 2b) (emphasis added; footnotes omitted); see also Interagency Intermediate Small Institution Exam Procedures at 1 (¶ 1b) (nearly identical language); Interagency Small Institution Exam Procedures at 1 (¶ 1b) (nearly identical language).
 See, e.g., Interagency Large Institution Exam Procedures at 2-3.
 Grutter, 539 US at 336-37.
 Id. at 341.
 See Interagency Large Institution Exam Procedures at 4 (“Review the institution’s stated assessment area(s) to ensure that it . . . [d]oes not reflect illegal disicrimination; and  [d]oes not arbitrarily exclude any low- or moderate-income area(s) taking into account the institution’s size, branching structure, and financial condition”); Interagency Intermediate Small Institution Exam Procedures at 3-4 (same); Interagency Small Institution Exam Procedures at 2-3) (same); see also, e.g., OCC, CRA Performance Evaluation of Third Federal Savings and Loan Ass’n of Cleveland (Feb. 24, 2020), at 8, 9, 22, 24, 25 (assessment areas reviewed do “not arbitrarily exclude any low- or moderate-income CTs”), available at https://www.occ.gov/static/cra/craeval/Feb21/704544.pdf.
 Grutter, 539 US at 336 (alteration omitted).
 See 49 C.F.R. § 26.45(a)(1) (unless an exception applies, “you must set an overall goal for DBE participation in your DOT-assisted contracts.”). This can apply to entities other than states, but for simplicity we refer only to states.
 49 C.F.R. § 26.5.
 Id. §§ 26.41(a)-(b), 26.67(a).
 Id. § 26.45(b).
 Id. § 26.45(c)-(e), (g)-(h).
 Id. § 26.45(f).
 Slater, 228 F.3d at 1182.
 See, e.g., id. at 1182-83; W. States Paving Co., 407 F.3d at 995.
 49 C.F.R. § 26.47(a).
 See, e.g., W. States Paving Co., 407 F.3d at 994 (noting that this “contrasts sharply with the rigid quotas invalided [by the Supreme Court] in Croson”); Sherbrooke Turf, Inc., 345 F.3d at 972; Slater, 228 F.3d at 1182. States, in turn, may not penalize contractors that make a good faith effort to hire DBEs as subcontractors. 49 C.F.R. § 26.53; id. Part 26, App. A. This, too, has been an important consideration in judicial determinations that the program is sufficiently flexible. See, e.g., W. States Paving Co., 407 F.3d at 994.
 Id. § 26.47(c).
 Id. § 26.101(a).
 These are the fixed percentages proposed by the Fed. 2020 ANPR, 85 Fed. Reg. at 66425. Whether these are too low is not addressed in this paper.
 0.65 times 44 is 28.6; 0.7 times 35 is 24.5; 24.5 is lower than 28.6.
 2020 ANPR, 85 Fed. Reg. at 66419-24. The Fed has not yet proposed what the fixed percentage should be. See id. at 66423.
 Id. at 66425. Currently, no level of lending is specified as indicative of satisfactory performance. Id. at 66420.
 This refers to marketing that targets members of a protected class (e.g., potential Black borrowers) to increase diversity in the application pool without diminishing efforts to attract others. Courts and regulators approve of affirmative marketing. See, e.g., South-Suburban Hous. Ctr. v. Greater S. Suburban Bd. Of Realtors, 935 F.2d 868 (7th Cir. 1991); Interagency Policy Statement on Discrimination in Lending (1994), 59 Fed. Reg. 18266, 18270 (April 15, 1994).
 How the performance ranges for satisfactory would relate to the threshold for a presumption of satisfactory is not clear, and the Fed sought input on this question. Id. at 66428.
 Id. at 66427-28.
 Id. at 66428.
 The Fed states in the ANPR that “many stakeholders . . . raised concerns about a lack of transparency and predictability required to achieve a particular rating.” 2020 ANPR, 85 Fed. Reg. at 66420.
 Id. at 66459.
 See, e.g., Slater, 228 F.3d at 1181 (“[t]his aspirational goal is reasonably construed as narrowly tailored.”)
 49 C.F.R. § 26.41(a); Id. § 26.41(b).
 Id. § 26.41(b).
 See, e.g., W. States Paving Co., 407 F.3d at 989, 994-95.
 See 49 C.F.R. § 26.41(c) (“The national 10 percent goal does not authorize or require recipients to set overall or contract goals at the 10 percent level, or any other particular level, or to take any special administrative steps if their goals are above or below 10 percent.”)
 See Slater, 228 F.3d at 1182 (criticizing Federal Highway Administration’s 12-15% goal; “we can find no explanation in the record before us.”)
 Grutter, 539 US at 337.
 See Gratz, 539 US at 271-72.
 Indeed, Justice Kennedy’s dissent criticized the admissions program for lacking such guidelines. See Grutter, 539 US at 392.
 Id. at 66428.
 2020 ANPR, 85 Fed. Reg. at 66429.
 Id. at 66430, 66432.
 Id. at 66430-31.
 Id. at 66431-32.
 Id. at 66432.
 Id. at 66430-31.
 Id. at 66431.
 Id. at 66430, 66431.
 Id. at 66432.
 Id. at 66429.
 Id. at 66433.
 Id. at 66432. “An [Individual] Development Account] is a special bank account that helps you save for your education, the purchase of a first home or to start a business.” Social Security Admin., Spotlight on Individual Development Accounts, available at https://www.ssa.gov/ssi/spotlights/spot-individual-development.htm.
 2020 ANPR, 85 Fed. Reg. at 66437.
 Id. at 66438.
 In some respects the focus is broader than the assessment area because some community development activities affect a broader area. Id. at 66437. In such instances when community development financing performance is evaluated at a state, multistate, or institution level, race can and should be incorporated in a manner that parallels what we propose here.
 See Interagency Q&A, 81 Fed. Reg. at 48530 (§ __.12(h)-8); see also 2020 ANPR, 85 Fed. Reg. at 66414 & n.42.
 Id. at 66440-41; see also supra § III.A (discussing metrics and benchmarks). The Fed has not yet proposed whether to base the threshold on the local or national benchmark, or a combination of the two. 2020 ANPR, 85 Fed. Reg. at 66441 (“More precise and comprehensive data would aid in analyzing these and other options.”)
 Id. n.115.
 Id. at 66441; see also id. at 66448-49 (proposing additional ways to encourage support for minority depository institutions).
 Id. at 66442.
 Id. at 66446
 See supra § III.A.
 2020 ANPR, 85 Fed. Reg. at 66442.
 Id. at 66443.
 See 12 C.F.R. § 228.27(a). This would continue under the Fed’s proposal. See 2020 ANPR, 85 Fed. Reg. at 66411 (“any bank would have the option to be evaluated pursuant to an approved strategic plan.”)
 See id. § 228.27(f)(1), (3). A bank may specify that if it fails to substantially meet those goals, it will be evaluated in the ordinary manner. See id. § 228.27(f)(4).
 See id. § 228.27(d)(1).
 See id. § 228.27(d)(2).
 See id. § 228.27(g)(2).
 In fact, though not a subject of this paper, a parallel new requirement to seek input from groups that serve LMI communities would be beneficial and welcome.
 12 U.S.C. § 2901(b)
 Id. § 2903(a)(1)
 Id. § 2903(b) (added by amendment in 1992) (emphasis added).
 Id. § 2907(a) (added by amendment in 1991) (emphasis added).
 See, e.g., supra § I.A; NCRC, The purpose and design of the Community Reinvestment Act (CRA): An examination of the 1977 hearings and passage of CRA, available at https://ncrc.org/the-purpose-and-design-of-the-community-reinvestment-act-cra-an-examination-of-the-1977-hearings-and-passage-of-the-cra/. Consistent with the legislative history, DOJ stated in a redlining complaint in 2016 that “CRA is an important safeguard for minority neighborhoods that have been traditionally underserved by creditors.” United States v. BancorpSouth Bank, Case No. 1:16cv118-GHD-DAS (N.D. Miss. filed June 29, 2016) at ¶ 24.
 See, e.g., Federal Reserve Bank of Richmond, CRA Performance Evaluation – Signet Bank (Jan. 15, 1996) at 18, available at https://www.federalreserve.gov/dcca/cra/1996/460024.pdf.
 2020 ANPR, 85 Fed. Reg. at 66412; see also id. (“Congress enacted CRA in 1977 primarily to address economic challenges in predominantly minority urban neighborhoods that had suffered from decades of disinvestment and other inequities.”)
 Mistretta v. United States, 488 US 361, 372 (1989).
 Gundy v. United States, 139 S. Ct. 2116, 2123 (2019) (plurality).
 Rothe Dev., Inc. v. Dep’t of Defense, 107 F. Supp. 3d 183, 212 (D.D.C. 2015) (quoting Whitman v. Am. Trucking Ass’n, 531 US 457, 474 (2001)) (citations omitted).
 Id. at 212.
 12 U.S.C. § 2907(b)(3); note to 12 U.S.C. § 1811 (1989) (Pub. L. 12 101–73 § 217(5)(J)).
 See, e.g., TOMAC, Taxpayers of Mich. Against Casinos v. Norton, 433 F.3d 852, 866 (D.C. Cir. 2011) (citing Am. Power & Light Co. v. SEC, 329 US 90, 104 (1946)).
About Relman Colfax PLLC
Relman Colfax PLLC is a national civil rights law firm dedicated to protecting civil rights and enforcing our nation’s civil rights laws. Glenn Schlactus is a partner at Relman Colfax PLLC and has concentrated his practice on enforcement of and compliance with civil rights laws. He has been with the firm since 2006. Sacha Markano-Stark joined Relman Colfax PLLC as an attorney in 2021. More: www.relmanlaw.com.