Incarceration, employment, and earnings: Dynamics and differences

Abstract

We study the dynamics of incarceration, employment, and earnings. Our hidden Markov model distinguishes between first-time and repeat incarceration, between persistent and transitory nonemployment and earnings risks, and accounts for nonresponse bias. We estimate the model via maximum likelihood using the National Longitudinal Survey of Youth 1979, accounting for the large differences in incarceration rates by race, education level, and gender. First-time incarceration is associated with 33% (50%) lower expected lifetime earnings and 6 (10) fewer years of employment for Black (White) men with a high school degree. Among less-educated men, differences in incarceration and nonemployment can explain around half the Black-White lifetime earnings gap.