CRA: How It Affects Communities and Banks

Since the Community Reinvestment Act (CRA) was written into law in 1977, scores of bankers have brought their institutions into compliance with the regulation and scores of experts have studied the resulting impact, particularly on lending patterns. Some praise the CRA as an effective tool for increasing the amount of credit available to low- and moderate-income people. Others criticize it as an unnecessary regulation.

Less attention has been paid to the CRA's impact on communities and to whether CRA agreements actually bring about change in banks' behavior. Two papers that tackle those topics are summarized here. They were presented this past spring at the Federal Reserve System's 2003 Community Affairs academic research conference in Washington, D.C., and can be found in their entirety at www.chicagofed.org. (Click on "Consumer and Economic Development Research Information Center.")

"The Effects of the Community Reinvestment Act on Local Communities"

The CRA requires federal regulatory agencies to encourage banking institutions to help meet the credit needs of the entire community in which they are chartered to do business. To accomplish this, the agencies review banks' records and take their CRA performance into account when considering their applications for a charter, merger or acquisition.

Whether the CRA has significant, little or no impact on communities is unclear, according to a study by researchers at the Board of Governors of the Federal Reserve System.

"Results are mixed and could potentially provide support for very different views of the effects of the CRA," they said.

"One view is that the CRA does matter and contributes to favorable outcomes for lower-income neighborhoods.

"An alternative view holds that the CRA does not have a significant impact on lower-income neighborhoods."

The outcome of the study on the socio-economic effects of the CRA on local communities was no surprise. This type of research has been scarce precisely because it is difficult to assess what would have taken place had the law not been enacted, the researchers said.

"The CRA does not exist in a vacuum; many changes have taken place over the years that affect the same markets as those targeted by the CRA," they said. Examples are changing regulatory or economic environments.

With that in mind, the researchers developed a framework and tested for changes that may have occurred between 1990 and 2000 in census tracts most likely to show marginal effects: those just below and those just above the income threshold for CRA designation. They examined whether differences between the two were related to CRA mortgage-lending activity.

Supporting the conclusion that the CRA has a positive effect, the analysis found that in CRA-designated census tracts, there were lower vacancy rates, higher homeownership rates and higher growth in owner-occupied units than would have been predicted when compared with changes in the census tracts that were not CRA-eligible. The results, which were statistically significant, appeared to be related to CRA activity. Banks that had outstanding CRA ratings had a higher market share of mortgage lending in the CRA-eligible tracts than in the tracts just above the eligibility threshold.

On the other hand, results for the effect of CRA on crime and median home values showed that lower-income neighborhoods fared worse than would have been predicted. In addition, even though homeownership rates and the growth of owner-occupied units were favorable, the results were not robust. Finally, banks with satisfactory CRA ratings had a lower market share of mortgage lending in CRA-eligible tracts.

The authors point out that one explanation for a lack of definitive results could be limitations in the designs of the tests. Because CRA has historically targeted lower-income census tracts, the tests focused on these neighborhoods. "However, it may be the case, that in practice, compliance with the law focuses more on lower-income borrowers or groupings of neighborhoods that can cut across census tracts both above and below the CRA-eligibility threshold. This may occur because an individual census tract is simply too small to effectively target. If so, then our tests might fail to detect the real impact of the law."

The authors of "The Effects of the Community Reinvestment Act on Local Communities" are Robert B. Avery, Paul S. Calem and Glenn B. Canner.

"What Makes CRA Agreements Work? A Study of Lender Responses to CRA Agreements"

Another study suggests that CRA agreements do, in fact, bring about change in banks' lending behavior. The preliminary results indicate that when lenders enter into a CRA agreement, they increase their targeted lending in a community. Mortgage counseling appears to be particularly important in spurring increased lending, while the existence of review committees seems to depress lending somewhat.

The study, conducted by a professor at the University of Southern California and a professor at the University of Delaware, focused on 51 CRA agreements made between 1993 and 2001. Such agreements are growing in popularity with banks that are eager to improve their CRA ratings. In the last 20 years, more than 300 CRA agreements have been forged between banks and community groups or government entities. Typically, a bank promises to engage in a specified volume of lending to targeted groups or communities, such as lower-income and minority individuals.

The researchers faced a number of challenges when collecting data for the study.

"Identifying and tracking lending institutions...was complicated by the fact that the banking industry underwent considerable consolidation during the 1990s," the study says. "Many of the lenders that entered into the CRA agreements in our data were subsequently purchased by or merged into other institutions and no longer exist."

Even if the researchers could track the original bank through mergers, they could not directly compare loans made toward the end of the study period with the earlier years because the later years included activity by the larger institution.

To compensate for these circumstances, the researchers constructed hypothetical institutions, including the original lender who entered into the CRA agreement and all independent institutions the lender was affiliated with through consolidation between 1993 and 2001.

The data indicated that, during the years the agreement was active, most lenders increased their lending by 65 percent, when measured by number of loans, and 94 percent, when measured by dollar volume. However, the study also showed significant differences among lenders. Some had dramatic increases in lending—more than 25 percent had a two-fold increase in their lending (increases of more than 100 percent). But between 35 percent and 45 percent of the lenders reduced their lending.

A more detailed statistical analysis confirmed the finding that lenders generally increased their targeted lending upon entering a CRA agreement. In addition, the increase in lending occurred gradually over time, with the largest increases being observed in the second and third years after the initiation of an agreement. Beyond this point, there was a great deal of variation in lender experiences, and the authors were unable to conclude that there was a uniform increase in lending.

The most effective agreements included mortgage counseling or other technical assistance for individuals, which would indicate that collaboration between the banks and community organizations is important. However, the study found that when community groups established review committees to oversee the lender's activities, possibly an adversarial move, the level of lending was adversely affected.

Future work will focus on whether lenders satisfied the pledges outlined in the agreements.

A portion of this research was conducted while Bostic was an employee of the Board of Governors of the Federal Reserve System.