The Community Reinvestment Act (CRA) is a vital law that has leveraged hundreds of billions of dollars of home loans, small business loans and community development financing for low- and moderate-income (LMI) communities. At the same time, the regulations implementing CRA exams that measure bank performance in serving LMI communities have not been updated in a meaningful fashion in over 25 years. Much has changed in banking and technology since then. An update is needed but not one that is hastily implemented and would weaken the law in a misguided effort to simplify and clarify CRA rules.
Two of the bank agency regulators, the Office of the Comptroller of the Currency (OCC) and the Federal Deposit Insurance Corporation (FDIC), proposed changes to CRA late last year. The public comment period on the proposal recently ended on April 8. NCRC’s latest count is that 1,616 of the comment letters or 85% of the total aligned with NCRC’s position that the agencies must go back to the drawing board and significantly revise their proposal if they wish to maintain and/or improve upon CRA’s effectiveness in increasing lending and investing in LMI communities.
Contrary to the wishes of the great majority of the public comments, the Comptroller of the Currency issued a press statement one day after the comment period thanking those who commented and asserting that he intends to issue a final rule in the first half of this year. In calls with several bank CEOs during the comment period, he sought to clarify “mischaracterizations” of the proposal. The haste and partiality of the Comptroller’s public remarks suggests that he will not proceed carefully nor heed thoughtful voices within the banking industry or among community-based organizations.
Some of the biggest concerns expressed in comments are:
- A decrease in bank’s responsiveness to community needs
- A decrease in total CRA dollars
- Changes to where a bank will be evaluated
- Some disagreement between banks and community groups on what should receive CRA credit, but they agree on two key issues.
The Agencies’ Proposal to Measure CRA Performance Will Decrease Banks’ Responsiveness to Needs
A fundamental part of the CRA reform initiative is refining how CRA exams measure bank performance in lending and investing in communities and providing bank services to them. The most significant part of the OCC and FDIC proposal is a new measure called the CRA evaluation measure that would be a ratio of the dollar amount of CRA activities divided by deposits.
Writing on behalf of their members and the state bank trade associations, the American Bankers Association (ABA) concluded, “we have significant concerns regarding the substance and complexity of the proposed performance measurement framework as well as the substantial costs that would be required to implement it. Accordingly, we request that the agencies refrain from finalizing this aspect of the proposal pending further study.” The Independent Community Bankers Association (ICBA) stated, “The use of nationwide performance benchmarks (in the CRA evaluation measure) may not be sufficiently well-suited to assess the performance of community banks.”
Along with community-based organizations, the bank trade associations are concerned that the CRA evaluation measure would encourage banks to seek the largest and easiest deals and community development financing opportunities in order to boost the numerator of the measure. The Community Bankers Association (CBA) wrote that “a balance sheet-based metric would fail to provide sufficient consideration for lending activities like LMI mortgage and small business loan programs, which are crucial to many communities, but do not result in the large dollar volumes favored under the Measure.” The ICBA amplified this concern in its comment letter saying that the nationwide approach the OCC and FDIC used to generate benchmarks for the CRA evaluation measure cannot adequately address economic conditions or other contextual factors in a smaller bank’s geographical service area. For instance, a number of comment letters cited that a bank’s ratio may not pass the required levels because house prices are relatively low in many inland areas of the country.
Community groups added concerns that an approach that emphasized rigid quantitative formulas and downplayed or discarded the qualitative criteria in CRA exams would render community group comments on bank responsiveness to needs unimportant in the rating decisions of CRA examiners. BLDG Memphis writes, “This single-ratio approach completely disregards whether the community development and financial needs of the community are being served by the bank or its investments. As a result, my organization, that has served my neighborhood for years, and whose experience and expertise is seriously considered as part of the current CRA examination process, will be rendered voiceless.”
Another worrisome aspect of the ratio measure is that it would only count mortgage loans at 25% of their value if the loans were sold to investors or Fannie Mae or Freddie Mac within 90 days. This is supposed to prevent banks gaming the system by buying loans made to LMI households shortly before their CRA exam. However, the Home Mortgage Disclosure Act (HMDA) now has data fields available to CRA examiners that would detect such rapid fire loan purchasing behavior before CRA exams. This discounting proposal is not only unnecessary but also injurious to special affordable housing programs.
The Consumer Bankers Association (CBA) stated, “Banks actively seek lending opportunities in LMI neighborhoods. Banks offer a number of down payment assistance programs, as well as special loan programs designed for LMI home buyers in an effort to originate more loans. Each program has its own requirements, which makes offering them more complex. Further, many such programs, especially in partnership with nonprofits or government agencies, require the loans be sold.” It seems like the agencies designed an effective CRA evaluation measure and discounting procedure only if they want to decrease mortgage lending in LMI communities.
The CRA evaluation measure and its discounting rules also threaten to decrease homeownership for people of color. The National Association of Hispanic Real Estate Professionals (NAHREP) estimated that over the past decade, Hispanics accounted for about 50% of the homeownership growth in America and that many Hispanic home buyers used special affordable home loan products referenced by the CBA comment. NAHREP asserted, “This modification (single metric) has the capacity to incentivize banks to do larger and easier projects, such as bridges or stadiums, instead of smaller, more complex activities, such as extending mortgages to working class people in those communities.” Likewise, the Metropolitan Milwaukee Fair Housing Council is concerned that the ratio’s focus on the big deal will imperil smaller bank grants and other financing in the Take Root Milwaukee, a multi-stakeholder program of education and foreclosure prevention counseling. This program provided counseling to an average of 1,778 homeowners each year and assisted an average of 3,863 potential borrowers seeking education on home purchases.
Mark Willis, former head of community development at JP Morgan Chase for 19 years, concluded, “The NPR’s unwavering focus on just dollars is fundamentally flawed. While it is unclear if the NPR will result in more or less CRA activity measured purely in dollar terms, it is inevitable that, as banks manage what is measured, they will simply focus on the dollars alone and not on the incremental impact they may be having on the wellbeing of LMI communities.”
The Agencies’ Proposal to Measure CRA Performance May Even Decrease CRA Dollars
Stakeholders ranging from bank trade associations and academic think tanks to community-based organizations commented that the OCC and FDIC lacked data to assess whether their proposed benchmarks for the CRA evaluation measure would increase or decrease CRA lending and investing. While the bank trade associations did not attempt to use currently available data to assess this critical question, the Urban Institute agreed with NCRC that data analysis shows that the evaluation measures would likely result in banks relaxing their efforts. The proposed evaluation measure would award banks with a Satisfactory rating if their ratio was 6% and an Outstanding rating if their ratio was 11%.
The Urban Institute’s research revealed that 72% of banks with more than $100 billion in assets would score Outstanding on the proposed evaluation measure and that 94.9% of all banks would. This is in stark contrast to about 10% of banks in recent years scoring Outstanding. The Urban Institute concluded that, “In fact, for the substantial number of institutions whose scores far exceed the outstanding threshold, the proposal could encourage them to cut back on CRA-eligible activities. Moreover, although there is a positive relationship between banks that do well on the existing test and those that do well under the proposed test, some institutions that need to improve based on the existing test would score outstanding on the proposed test, a questionable result.”
NCRC used a different database, the recently released data from a Federal Reserve sample of CRA exams, to conclude that banks would score in a Satisfactory or Outstanding level on their CRA evaluation measures in the great majority of their assessment areas over more than a decade. In addition, NCRC used Call Report data to find that banks’ percentage of retail loans to LMI borrowers and census tracts could decrease while banks still exceeded the proposed benchmarks for the CRA evaluation measure. The sum total of the research suggests that the agencies’ proposal would not adhere to CRA’s legislative mandate that banks continually affirmatively meet community needs. This mandate suggests increasing CRA loans and investments over time, not decreasing them as would likely happen should the agencies implement their proposal.
The Agencies Botched their Proposal Regarding Where Banks Should be Evaluated
Similar to the proposed CRA evaluation measure, the agencies botched their proposal to expand the geographical reach of CRA evaluations by proposing to use data that is not currently collected by banks and that would not reflect where banks conduct their activity. Assessment areas (AAs) are geographical areas on CRA exams that correspond to areas in which banks have their branches. While many banks still primarily lend through their branches, some newer banks offer lending and other services mostly through the internet. Others lend through brokers or use loan officers that do not necessarily operate in branches.
Expanding AAs to encompass non-branch lending has been a hot topic for several years but the agencies did not learn from this debate and discussion. The agencies proposed to designate AAs outside of branch networks based on where high volumes of deposits are concentrated. However, the bank trade associations stated in their letters that they currently do not collect deposit data that includes the address of their customers outside of their branches.
Moreover, the ICBA agrees with NCRC in that it would rather use a readily available data resource, lending data, to designate additional assessment areas. The ICBA also suggests that lending data indicates a relationship with borrowers and communities that would better enable a bank to serve an additional assessment area. The trade association states, “Community banks are rooted in their local communities and they understand their local needs. They have relationships with local businesses and families and can accurately assess where loans and investments will have the biggest positive impact.”
Where community groups part with banks is where to designate more AAs. The National Association of Affordable Home Lenders (NAAHL), the ABA and the CBA have slightly different versions of allowing banks to serve LMI borrowers and communities nationally provided they are adequately serving these populations in their AAs. The CBA coined a new phrase “Reinvestment Redistribution” in which bank activity would somehow find underserved communities outside of branch networks.
Community groups would rather reduce bank discretion because without guardrails, banks may be tempted to find LMI communities in metropolitan areas or states that are easier to serve. Reinvestment Partners stated, “Regulators should assign CRA obligations to digital financial institutions (FIs) in areas that, all things being equal, have fewer bank branches on a per capita basis. We believe that the assignment to a digital FI of an assessment area would not necessitate that a private company alters how it manages the operation of its business. Digital banks already use digital advertising to target consumers. They would apply the same technique to our proposed method.”
NCRC proposes a similar method for assigning additional areas for banks to serve. The agencies would designate counties with relatively few loans per capita as underserved and would allow banks to serve them provided banks have met needs in areas in which they have branches. In addition, lending data should be used to identify additional AAs outside of branch networks in which banks have significant market share of loans. NCRC found that non-traditional banks without branches in an area do not make as high a percentage of loans to LMI borrowers as banks with branches. To avoid non-traditional banks failing their CRA exams, examiners could compare them to their peer non-traditional banks as a primary measure but should also not award them high marks on the retail lending part of their exams if they are far behind traditional banks.
Banks and Community Groups Disagree on What Should Count on CRA exams but Share Concerns about Agency Proposals
Community-based organizations, many of which serve the most disadvantaged LMI communities, want CRA to retain a focus on LMI communities that have experienced redlining and disinvestment — a focus shared by Congress when it passed CRA. Banks tend to have a more elastic interpretation of CRA’s mandate to serve community needs as a requirement to serve needs that are more general. Impartial regulators would strive to strike a balance between these views, but it is unclear if this current crop of agency heads are up to the task.
An example of this difference is how financial education is considered on CRA exams. NAAHL and the ABA would like financial education provided to any income group to count on CRA exams, citing difficulties in documenting incomes of clients and the need for financial education in schools. Community groups, on the other hand, want financial education to be targeted to LMI populations that are disproportionately under- or unbanked. In addition, community groups have been tracking the incomes of their clients for several years as required not only by CRA but by other federal programs.
Another area of disagreement is allowing banks to serve larger businesses and receive credit on CRA exams. Banks are generally in favor of increasing the revenue size that defines eligible small businesses from $1 to $2 million while community groups oppose this. Expanding Black Business Credit Initiative (EBBC), comprised of seven Community Development Financial Institutions (CDFIs) that are Black-led or focused on financing Black-owned business, stated that this change would divert lending from businesses owned by people of color. EBBC maintained, “In terms of revenue (as measured by annual receipts) White owned businesses, have over $640,000 in annual receipts, dwarfing that of businesses of color ($160,000) and Black-owned businesses ($73,000).” Moreover, “the percentage of bank commercial loans under $1 million is already shrinking, from the mid-30s in the 1990s to now less than 20% of total outstanding commercial loans.” These comments imply that any further relaxation of the revenue size of small businesses would further decrease lending to the businesses most in need of credit.
Encouragingly, the bank trade associations and community groups agree on two important aspects of the proposed rule. Both stakeholders oppose the agencies’ proposal to stop considering lending in LMI census tracts. Similar to the themes in community group letters, NAAHL states, “In fact, the seminal modern analysis of urban poverty, The Truly Disadvantaged, by Harvard University’s William Julius Wilson, identified the concentration of poverty and the loss of middle-income families as a fundamental source of inner-city distress.”
No longer considering lending in LMI tracts would exacerbate concentrations of poverty. As CBA put it, “But the very nexus of the CRA regulation is based upon the concept of encouraging banks to lend in communities from which deposits are collected and eliminating geographic redlining. The exclusion from CRA consideration of mortgage lending in LMI tracts seems counter to those ideals. Any new CRA framework should not establish an explicit mandate to keep LMI communities LMI. Such a new framework is completely counter to a core historic objective of CRA, that is, economic integration which brings greater opportunity to LMI families and areas.”
The agencies had proposed to delete consideration of lending in LMI tracts because they were concerned that such consideration would promote gentrification. However, NCRC and others have proposed better methods to deal with this issue including not allowing loans to non-LMI borrowers to count in LMI areas experiencing rapid home value increases.
Finally, the agencies had proposed to delete from the regulatory definition of community development, the criterion of economic development that included the provision of jobs for LMI workers. NAAHL’s statement summarized the position of several community and bank stakeholders on this issue:
“We urge the Agencies to reconsider this dismissive decision. The Brookings Institution has found that 53 million Americans between the ages of 18 to 64—accounting for 44% of all workers—qualify as ‘low-wage’…While it may be unfortunate that low-wage work is so widespread in the U.S., these jobs are critically important to LMI people and communities. Moreover, there is considerable room to include LMI jobs that pay substantially higher wages, based on 80 percent of the U.S. median household income of $61,937 in 2018. In our members’ experience, the current CRA policy has indeed proven workable and should be retained.”
Conclusion: Stop, Go Back and Engage Banks and Community Groups in More Dialogue
The major stakeholders, banks and community groups, agree that more than half of the agencies’ proposal goes in the wrong direction. Namely, the proposed evaluation measures and how to add geographical areas on CRA exams would have unintended and negative consequences. In the third area of reform, what counts on CRA exams, the two groups of stakeholders agree on certain matters but disagree on others. Thus, the most constructive course of action for the OCC and FDIC would be to table their proposal and work with the Federal Reserve Board in crafting a proposal that is based on careful analysis of data as well as more responsive to the thoughtful input of the stakeholders.
In addition, some well-constructed discussions among stakeholders regarding differences in what counts may help to narrow areas of disagreement and identify workable compromises. On other hand, if the agencies pursue a hardheaded and hasty conclusion to this rulemaking, they risk enacting a rule that would be a great disservice to LMI communities as they emerge from the COVID crisis. It would be a highly controversial and unpopular rule likely to be overturned by fair-minded regulators at a future date. This is no legacy for which any group of agency heads want to be remembered.
Josh Silver is NCRC’s Senior Advisor of Policy and Government Affairs
Photo courtesy of ANHD’s Melanie Breault.