“…Ours is not the first study of the CRA to exploit the income threshold, using RD to estimate the impact across tracts(Avery, Calem and Canner, 2003;Bhutta, 2011;Avery and Brevoort, 2015;Bostic and Lee, 2017;Lee and Bostic, 2020), and some have used difference-in-differences based on changes in CRA eligibility resulting from MSA boundary changes(Ding and Nakamura, 2021;Ringo, 2017;Ding, Lee and Bostic, 2020), although none of these use firm-level data or examine racial inequality. Our identification of the CRA effect exploits a much large number of changes associated with the redefinition of CRA eligibility nationwide, as described further below.…”