This is the third in a series of articles reviewing public comments submitted in response to the Federal Reserve’s Advance Notice of Proposed Rulemaking to revise Community Reinvestment Act rules. See the first here and the second here.
Two previous articles compared and contrasted the views of banks and community organizations regarding the Federal Reserve Board’s (board) Advanced Notice of Proposed Rulemaking (ANPR). This article focuses on the views of NCRC member organizations and community-based allies. If community organizations could write the CRA regulations, the CRA would increase in rigor, grade inflation would be curbed and the focus would be on underserved populations including low- and moderate-income populations, tribal communities, people of color, rural communities and people with disabilities.
Overall, NCRC tabulated that the board received 615 comments on the ANPR. Community-based organizations submitted the great majority of the comments (314). Banks and their trade associations sent in 74 comments. Unaffiliated individuals submitted 191 comments and most of these were aligned with NCRC’s position.
Data and test structure influence reinvestment outcomes
A rigorous CRA exam depends on accurate data. A shortcoming with the current exams, particularly the service test that measures access to basic banking services, is a lack of basic data on deposit products and how many accounts are for low- and moderate-income (LMI) customers. CRA examiners tend to report that a bank has special affordable deposit accounts for LMI customers and then award a high rating on the service test without reviewing numbers of accounts or their features.
This does not satisfy the Woodstock Institute’s new President and CEO, Horacio Mendez, who had been a senior official in the CRA department of banks for several years prior. In Woodstock’s comment letter, he affirmed that banks have deposit data by address, enabling analysis of deposits by income category of tracts.
“Any institution who claims that marrying depositor address information with product distribution is a costly regulatory burden is not being forthright – this is done internally (at banks) all the time,” Mendez said. He urged the agencies not to apologize for requesting information from financial institutions that they already have.
The Cities for Financial Empowerment Fund (CFE Fund) reinforced the point about deposit data in a powerful manner. CFE Fund operates the Bank On program that encourages banks to work with public and nonprofit partners to offer deposit accounts with low fees and other affordable features. The CFE Fund collaborated with the Federal Reserve Bank of St. Louis to collect data from banks offering these accounts. In 2018, the effort collected data on 800,000 accounts with 75% of these for new banking customers. It seems clear from this example that the service test can become much more effective with better data collection in holding banks accountable for providing safe and affordable deposit accounts to traditionally underserved populations.
Related to deposit data collection is whether CRA exams should collect and analyze data on consumer lending. Some bank trade associations replied that CRA exams should remain focused on wealth creation activities such as home and small business lending. While that focus is desirable, consumer lending fills an important need considering that Federal Reserve surveys found that many Americans do not have $400 saved for emergencies. At the same time, many consumers resorted to nonbank and unregulated lenders who charged abusive rates for loans that last longer than the life of a car, according to the Greater Rochester Community Reinvestment Coalition. It would be preferable to steer these consumers to banks with more affordable rates; CRA exams that include evaluations with robust data of the amount and affordability of consumer loans could be instrumental in providing real options for underserved populations.
Along these lines, the Association of Neighborhood and Housing Development (ANHD) asserted that CRA exams need suitable data to examine the quality as well as quantity of multifamily housing. ANHD’s comment letter documented several instances of CRA-regulated banks making loans to multifamily property owners that could be paid back only if rents rose to high levels that were no longer affordable for LMI tenants. The owners often resorted to harassment (such as cutting off hot water) that compelled LMI tenants to leave.
“If you asked neighborhood residents, they would likely say the bank (financing abusive landlords/owners) was doing more harm than good in the area. Omitting negative lending practices from an examination does a disservice to the people and neighborhoods CRA was designed to benefit,” ANHD said in their comment letter.
The concern was magnified considering that New York City had private equity investors holding about 100,000 multifamily units. CRA exams must be conducted by astute examiners that consult area residents, require banks to follow best practices like vetting multifamily loan borrowers and use data on debt service coverage ratios to ensure that the loans are affordable and responsible.
Test structure must be designed to promote a mix of lending and investments
As with data collection, reforms to the structure of CRA exam subtests must carefully promote a mix of activities responsive to needs rather than favoring one type of activity over others. This issue is prominent in the proposal to create a separate community development (CD) subtest. Community development financing has wider neighborhood benefits than home and small business lending which is also vital but assists individual homeowners and small businesses. A useful contrast to keep in mind is a CD construction loan for a 100-unit rental development as opposed to a retail loan for a homebuyer. Neighborhoods need both retail lending and CD financing in order to revitalize and thrive.
Currently, CD lending is part of the lending subtest while CD investments are considered in a separate subtest. Over the years, proponents of a proposed CD test that combines CD lending and investment suggest that such a test would promote the most efficient type of CD financing rather than causing banks to ensure that they have a minimum of either CD lending or investing in order to pass the lending or investment subtests. However, a new problem emerges from the proposed CD test; CD loans are usually offered in greater amounts than CD investments, perhaps because of the ease of the financing. Some stakeholders are concerned that a CD test may tilt the financing even more to CD lending and starve communities of needed CD investments.
We support the proposal to establish a separate community development test, but oppose the suggestion that the CD lending and CD investments (subtotals) would be combined. Equity investments and contributions are vital to communities while providing lower returns to banks, and must therefore continue to be valued and evaluated separately. The board also proposes to encourage patient CD lending which could further favor CD lending as compared to CD investing. Both lending and investment are critical to affordable housing and economic development such that they should be examined separately. We think the rules should prioritize annual lending and investments. Impact scoring could be used to reward patient and portfolio CD activity, as well as impactful CD efforts.
The board proposed a single ratio as a quantitative measure that combined CD loans and investments and compared those to deposits. There is no reason why this one measure cannot be supplemented by separate ratios for CD loans divided by deposits and CD investments divided by deposits. Each of the three measures can be weighed in such a manner as to encourage a mix of activities responsive to local needs. In addition, as Greenlining and CRC suggested, the CD loans and investments should receive qualitative scores (impact scores in the board’s parlance) that reflect the degree to which the financing is innovative and responsive to needs. Community organizations were perceptively aware of the complexities of local needs and provided insightful input into how to design tests to ensure that exams do not become merely bean counting exercises that maximize the volume of CD financing but do not consider its quality or contribution to community vitality.
CRA must be focused on underserved populations
The board asked in question two of its 99 questions how to address racial inequities. Racial disparities in retail lending, bank branching and services have been stubborn and persistent over the decades. Community organizations offered several instances of this in their comment letters. For example, ANHD stated that “22% of New York City is Black and 29% Latino, yet fewer than 8% of all home purchase loans in 2017 went to non-Hispanic Black borrowers and 8% to Hispanic borrowers of any race. It reached just 9% in each category in 2019. Consistently, Black and Latinix borrowers are denied loans at a greater rate…and have been steered to higher-cost products.”
To remedy these inequities, NCRC and our member organizations urged the board to add performance measures on CRA exams measuring lending and services to people and communities of color just like the current measures for LMI borrowers and communities. For example, a new metric could be the percent of loans in majority-minority census tracts. In addition, NCRC had proposed a version of this for a category of tracts considered underserved as measured by low levels of lending per capita.
If the board does not consider bank performance to people and communities of color in a manner analogous to LMI borrowers and communities, CRC listed a number of compelling alternatives. It stated:
The rules should at least provide a mechanism so that superior bank reinvestment in neighborhoods of color and to borrowers of color can enhance a CRA rating and poor service can result in a lower rating. This can be accomplished through impact scoring across all products and services, or through consideration of these issues in evaluating a bank’s performance context. Such consideration should take into account any and all disparities in marketing, denials, originations, pricing…default rates, etc. The board should enhance participation by communities and people of color in the strategic plan process and in conducting community contact (outreach).
Better targeting of underserved areas would also help address racial and income disparities. A hot topic in CRA reform is how to consider CD financing when a bank engages in CD financing outside of its assessment areas or geographical areas that are evaluated and rated on exams. Some stakeholders suggested that banks should receive credit for CD financing anywhere it is offered. Community groups advocated for a more targeted approach. The Housing Assistance Council (HAC) discussed prioritizing persistently poor counties, those with poverty rates of 20% or more for the last three decades. There were 395 of these counties and 90% of majority-African American rural counties were persistently poor as were 69% of majority-Native American rural counties. If a bank seeks credit outside of its assessment areas, CD financing in these counties and other underserved counties (NCRC offered a formulation based on loans per capita) could be weighed more heavily than CD financing elsewhere.
Likewise, community organizations debunked the board’s proposal for nationwide assessment areas for so-called internet banks, maintaining that this would allow these banks to focus on areas where it is easier to serve LMI borrowers and overlook areas with greater needs but that are harder to serve. The Woodstock Institute stated that instead of nationwide assessment areas, internet banks could prioritize underserved areas as their assessment areas, similar to NCRC’s proposal in a recent white paper.
CDFIs and MDIs are not a substitute for more focus on underserved populations in CRA exams
The board asked how to consider bank support and financing of community development financial institutions (CDFIs) or minority depository institutions (MDIs) as a means of promoting access to capital and credit in underserved areas. Many CDFIs and MDIs are dedicated to serving traditionally underserved areas. However, there should be a caution to not over rely on them. As the Woodstock Institute asserted:
If banks are provided a regulatory “out” by pushing high-touch low-wealth clients to CDFIs, then they will do so. At the same time that they receive CRA consideration for this relationship, they are working to counter the goal of “mainstreaming” LMI and minority customers and reinforcing segregation in the financial industry.
In other words, CRA exams should not provide incentives for banks to write a few large checks to CDFIs and MDIs and scale down their retail lending or direct services to underserved areas. Banks have scale that CDFIs and MDIs are unlikely to reach anytime soon. Therefore, CRA reform will need to avoid creating incentives for banks to scale down their retail activities. Test design, careful construction of quantitative and qualitative evaluation measures and maintaining examiner discretion will be key to avoiding bank abandonment of direct lending and service to LMI and other underserved communities. For example, the lending and service subtests should probably have the highest weights so as to avoid banks loading up on CDFI and MDI support in their CD subtests. Within the CD subtests, impact scoring can penalize huge imbalances between bank direct CD financing and support for CDFIs and MDIs.
The HOPE credit union and policy institute maintained in its letter that MDIs with proven track records should be favored in CRA reform. It stated, “The mechanism by which to do this would be for the CRA to incorporate the “minority lending institution definition included in the Consolidated Appropriations Act of 2021. In addition to being an MDI, minority lending institutions are ones where a majority of both the number and dollar volume of arm’s-length, on-balance sheet financial products…are directed at minorities or majority-minority census tracts or equivalents.” Perhaps impact scoring could award CD loans and investments in MDIs higher scores if the MDIs match the definition offered by HOPE.
CRA grade inflation must be combated
One important bottom line of CRA reform is that the current rate of CRA grade inflation must be combated if we hope to channel significantly more loans, investments and services in LMI neighborhoods, communities of color and other underserved communities. Currently, close to 10% of banks earn Outstanding ratings and 90% receive a Satisfactory ratings. Banks’ efforts will eventually stagnate if this several-year ratings distribution does not change to reflect more real distinctions in performance. To this end, NCRC has argued for the retention of five ratings in the subtests and a point scale to accompany the overall four ratings (which are mandated by statute). For example, even if a bank scores Satisfactory overall, if a point scale indicates it is in the lower part of Satisfactory ratings range, it will have more motivation to improve its future performance.
The reforms should include rigor along the lines suggested above. Somehow, race should be more carefully considered so organizations like St. Louis Equal Housing and Community Reinvestment Alliance that have succeeded in achieving reforms through HUD complaints based on racial disparities no longer have to be the cops of last resort fixing errors that bank CRA exams overlooked with Satisfactory ratings. Ratings should no longer be a “pat on the head” as the Metropolitan Milwaukee Fair Housing Council stated but instead should have real and motivational meaning.
Josh Silver is a senior policy advisor at NCRC.