Persistence in the Labor Market: The Effect of Early Career Labor Market Conditions on Future Job Prospects

Abstract

This paper analyzes whether people who enter the labor market during economic downturns suffer a setback in their careers that can last indefinitely. I test whether recessions experienced in a person's state of birth between the ages of 16 and 25 have effects that persist at the individual level one to two decades after the fact. I find evidence that such adverse shocks do persist for many years after the fact. The paper contributes to three broad literatures. The first is concerned with persistence in female labor supply. The second attempts to measure scarring effects of unemployment and nonemployment on subsequent labor market outcomes. The third is broadly concerned with the distributional effects of recessions. In each case fundamental identification problems are posed by measures of "external" factors that may not be truly exogenous to the individual worker. The use of labor market conditions from the state of birth at specific ages in this paper permits the identification of persistence effects that are free from such exogeneity concerns. I find that early-career recessions lead to lower employment rates/fewer weeks worked per year for white women, black women, and black men. The measured effects of a 3-4 year recession experienced before age 25 (frequently) persist until at least age forty, sometimes even later in life. The magnitudes range from 1 to 3 percent lower employment rates, and 1.5 to over 5 fewer weeks worked per year. Evidence of a persistence effect for wages is less clear. The wage estimates are probably biased by countercyclical removal of low wage workers from the employment statistics (creating an aggregation bias similar to that documented by Solon, et al., 1994). Regardless, even if the wage effect is taken to be zero on average, a zero wage effect coupled with a negative labor supply effect means a negative impact on earnings for white women, black women, and black men. Because white men show virtually no persistence effect for labor supply, their small estimated negative wage effect suggests a modest adverse shock to earnings as well. So the overall effect on earnings appears to be negative for all four groups of people.

JEL Classification: J23, J24, J64, E32