Association for Public Policy Analysis & Management

We estimate the long-run effect of the Home Owners Loan Corporation (HOLC) maps on children born just prior to the drawing of the maps in the late 1930s. Aaronson et. al. (2019) show the maps led to a decline in homeownership rates, home values, and rents, and an increase in racial segregation over the next seven decades, likely through disinvestment caused by higher borrowing costs in lower-graded neighborhoods.

This project takes advantage of the Census Longitudinal Infrastructure Project (CLIP), described in Alexander et. al (2015). CLIP links the address-level 1940 Decennial Census with contemporary datasets, including the 2000 Decennial Census, the Survey of Income and Program Participation (SIPP) and the Current Population Survey (CPS). That allows us to observe later life outcomes including human capital, earnings, and some measures of wealth and health.

We employ two identification strategies to estimate the causal impact of the HOLC maps on these later life outcomes. First, we compare children growing up within city blocks but separated by an HOLC boundary that resulted in differences in access to household credit. We then compare these differences to a placebo set of children that were separated by either a) boundaries that received the same HOLC grade and therefore should have had no difference in credit access or b) hypothetical boundaries that mimic pre-existing demographic characteristics along our treatment boundaries. A second independent identification strategy takes advantage of the HOLC’s decision to only grade cities with a population of 40,000 or more at the time the maps were developed. We compare children in cities that were just above 40,000 in population to children in cities just below.

We find robust evidence that the maps had a statistically and economically significant causal effect on the educational attainment of children, including the propensity to graduate high school, attend college, and graduate from college. We also find evidence that the maps had causal effects of lowered income, increased poverty, and increased disability rates later in life for children in the lowest graded neighborhoods.

Finally, we plan to use a link to the SSA Numident file, which records deaths, to investigate the effect of the HOLC maps on longevity.

References

Aaronson, Daniel, Daniel Hartley and Bhashkar Mazumder, 2019, “The Effects of the 1930s HOLC “Redlining” Maps,” Working paper, Federal Reserve Bank of Chicago.

Alexander, J. Trent, Todd Gardner, Catherine G. Massey and Amy O’Hara, 2015, “Creating a Longitudinal Data Infrastructure at the Census Bureau,” Working paper, U.S. Census Bureau.