The goal of this paper is to examine evidence on the causal effects of early job stability on adult labor market outcomes, by eliminating the bias in this estimated relationship that stems from omitted job match quality or other factors. Specifically, rather than simply estimating least squares regressions of adult wages on youth labor market experiences related to early job stability, along with adult characteristics, youth labor market conditions from the years in which workers entered the labor market are used as instrumental variables for the job stability experienced by workers as youths. The idea is that variation in youth labor market conditions is exogenous to the individual, and therefore generates variation in early job stability that is unrelated to job match quality. This empirical approach, in principle, yields estimates that come closer to measuring the potential effects of policies that would increase early job stability.

Past research that does not fully account for the endogeneity of early job market stability finds little or no evidence that early job stability improves adult labor market outcomes (Gardecki and Neumark, 1998). This suggests that the approach taken in this paper will reach an even more pessimistic conclusion; if unobserved job match quality is a source of positive correlation between adult wages and early job stability, the expectation is that the instrumental variables procedure will lead to even lower estimates of the positive effects of early job stability (or stronger estimates of negative effects). As will become clear, however, the evidence generally points in the opposite direction, with the instrumental variables estimation pointing to substantial benefits from early job stability. After presenting the evidence, the paper considers alternative interpretations of these unanticipated findings. The interpretation that appears to best fit the available facts is that while there are returns to search, there are also positive returns to early job stability. However, heterogeneity in the returns to search is a source of spurious negative correlation between early job stability and adult wages-as those with higher returns to search have lower early job stability but higher adult wages-which is eliminated by the instrumental variables estimation. Thus, the evidence in this paper suggests that exogenous increases in early job stability in youth labor markets in the U.S.-such as might be caused by school-to-work or other programs-would have beneficial effects on the incomes youths eventually earn as adults. no credit check payday loans