The Community Reinvestment Act (CRA), first passed by Congress in 1977, was created to respond to America’s history of redlining, in which lenders refused to lend to Black and Latino households and other households of color. Although the CRA does not explicitly consider race or ethnicity at the national level, it does focus on ensuring that banks respond to the credit needs of the communities in which they do business.

But these rules apply only to banks and thrifts, which are regulated by the Federal Reserve, the Comptroller of the Currency, and the Federal Deposit Insurance Corporation. The CRA does not apply to nonbanks, including credit unions and independent mortgage banks (IMBs), despite these institutions accounting for 60 percent of all mortgage originations, including 75 percent of all government and government agency mortgages. Whereas most banks and thrifts have a defined geographic footprint, most IMBs operate over wide geographic areas, using a combination of retail branches, call centers, and wholesale lending through mortgage brokers and correspondent institutions.

With IMBs constituting such a large proportion of mortgage lending, a flurry of activity at the state level has occurred over the past few years to adopt versions of the CRA to cover nonbank mortgage lenders. Massachusetts has long had CRA rules examine both banks and nonbank lenders, and recently, Illinois and New York have adopted—but not yet implemented—versions of the CRA that cover nonbank mortgage lenders. California, Maryland, and Pennsylvania have also considered adopting similar rules, but these state-level efforts are currently inactive. The federal regulators are updating CRA rules at the national level.

Interestingly, the California bill, introduced in 2022, did explicitly propose expanding the CRA rules to consider race and ethnicity. If this change were made, how would banks and IMBs compare? Our recent research demonstrates that nationally and across the six states that have adopted or are considering adopting CRA rules for nonbanks, IMBs perform a greater share of mortgage lending to borrowers and neighborhoods of color than their bank counterparts.

IMBs lend more to borrowers of color than banks do, and both exceed the current homeowner share

To assess whether current mortgage lending meets the credit needs of borrowers of color, we use the share of homeowners of color as a benchmark and compare it with the composition of bank lending and nonbank lending by race and ethnicity. This comparison tells us whether borrowers and neighborhoods of color receive a share of loans in line with current homeownership rates.

But the current share of homeowners is just a lower-bound estimate of the actual credit needs, as homeownership increases with age, and the younger population has a larger share of people of color than does the older population. Homeowners of color constitute 30.3 percent of homeowners ages 25 to 44 but only 18.8 percent of homeowners ages 65 to 74.

We found that the total share of mortgage lending to borrowers of color is 33.2 percent nationally, exceeding the 24.4 percent national share of homeowners of color. IMB lending to borrowers of color is significantly higher than the bank lending share (36.7 percent compared with 26.8 percent), and both exceed the current homeowner share.

As we look at the six states considering adopting CRA versions to cover nonbanks, we see a pattern that is broadly consistent with the national-level results. In every one of the six states, banks and IMBs outperform the homeowner benchmark. In five of the six states, IMBs do more lending to borrowers of color than banks do, with the shares comparable in New York.

IMBs lend more to neighborhoods of color than banks do, and both fall short of the current homeowner share

We also looked at mortgage lending to neighborhoods where at least 70 percent of the neighborhood consists of households of color. Nationally, the share of total mortgage lending to these neighborhoods (8.0 percent) falls short of the share of homeowners in these neighborhoods (10.8 percent). This pattern holds across five of the six states, with Massachusetts the only exception, as there are few such neighborhoods in the state.

By separating bank and IMB mortgage lending, we find that IMB lending exceeds the share of bank lending in neighborhoods of color. Nationally, the share of IMB lending to these neighborhoods is 9.3 percent, well exceeding the 5.7 percent for bank lending—a 3.6 percentage-point difference. This pattern, in which bank lending lags IMB lending, holds across all the six states.

Policymakers must weigh the cost concerns and data benefits of adding CRA rules

With IMBs originating more mortgages each year than banks that are subject to the CRA, some states have included IMBs in their CRA regimes, while others are considering it. Our work shows that IMBs, nationally and across all the states considering CRA changes, do more lending to borrowers and communities of color than banks do.

The national CRA covers banks but not IMBs. State-level CRA requirements could theoretically help institutions better serve the credit needs of communities of color in these states, but they do impose significant costs on the covered institutions and the regulatory agencies charged with implementation. As policymakers contemplate changes, they may want to consider the following:

• What is the best way to increase lending to borrowers of color and borrowers with low incomes? Given the costs of setting up a CRA examination framework, hiring examiners, and preparing supervisory reports, state funds may be better used to enhance programs that target markets serving borrowers of color and borrowers with low incomes. The expansion of special purpose credit programs by state housing finance agencies, for example, could offer a more efficient use of funds. With banks already covered by federal statute and nonbanks doing more lending to borrowers of color and borrowers with low incomes, would offering more targeted programs encourage more lending to underserved groups than state-level CRA requirements?
• Can data be used to better prioritize a CRA focus? For states that have already enacted the CRA for banks and IMBs, individual bank–level data could help prioritize CRA attention and institutions that lag the statewide benchmark.

As demographics change, ensuring that financial institutions can meet communities’ specific credit needs remains a priority for policymakers. Understanding where communities turn for mortgage lending and weighing trade-offs based on costs will only become more critical for targeting assistance.