Figure A.4: Relationship between Severity of 1980–1982 Recession in County of Residence and County of Birth.

Notes: Figure plots OLS estimates of the interaction between the 1979–1982 decrease in log real earnings per capita in individuals’ county of birth and indicators for age. The dependent variable is the 1979–1982 decrease in log real earnings per capita in individuals’ county of residence. I estimate this relationship using confidential Census/ACS data for individuals born from 1970–2013 (county of residence observed from 2000–2013) and confidential PSID data for individuals born from 1968–1979 (county of residence observed in 1979). Both regressions include fixed effects for gender, race, and birth year-by-birth state, plus birth year interacted with the 1950–1960, 1960–1970, and 1970–1980 change in log real median family income and log population, and the 1960 level of log population, log population density, percent urban, percent black, percent foreign, percent with a high school degree, and percent of families with income below $3,000. The Census/ACS regression additionally includes fixed effects for age at time of survey by gender and age at time of survey by birth state. The dashed lines are pointwise 95 percent confidence intervals based on standard errors clustered by state. The Census/ACS sample contains 23.8 million individuals born in the continental U.S. with a unique PIK, unique birth county, and non-allocated variables. The PSID sample contains 3,684 individuals born in the continental U.S.

Sources: BEA Regional Economic Accounts, Confidential 2000–2013 Census/ACS data linked to the SSA NUMIDENT file, Confidental PSID data