Association for Public Policy Analysis & Management

The Community Reinvestment Act (CRA) persuades banks to increase credit supply in low- and moderate- income areas, consistent with safe and sound operations. This paper investigates whether the CRA contributes to increased median income within census tracts during the early 2000’s. Using a regression discontinuity design which utilizes a median income threshold to determine census tract eligibility, I find increased that CRA status does impact median income growth for census tracts close to the threshold for large metropolitan areas. Based on census tract data I found no effect of CRA eligibility status at a national level. However, when I restricted my sample to the top 50 and top 20 Metropolitan Statistical Areas (MSAs), there was a significant and positive effect.

    The CRA lending exams that banks complete do not differentiate between lending to a low- or moderate- income (LMI) buyer or lending to a relatively wealthier borrower who is purchasing a home in a low- or moderate- income census tract. This is an important distinction. Therefore, in order to better gauge what is driving the median income growth in CRA eligible tracts I repeat the analysis using the Home Mortgage Disclosure Act (HMDA) data. A regression discontinuity model is applied to borrower income data in order to test for any discontinuity around the CRA eligibility threshold. The results are consistent in that an effect is observed in larger metropolitan areas. These results support the idea that examiners should differentiate between LMI borrowers and LMI neighborhoods.