Answering this question is extremely difficult, probably requiring carefully designed experiments. The compendium of research on school-to-work programs discussed in the NCRVE report by Stem, et al. (1994) suggests that researchers are a long way from a definitive answer to this question. A more limited goal is to ask whether youths who appear to be in unstable or dead-end jobs early in their careers suffer adverse labor market consequences as adults. It seems that, minimally, a case for attempting to replace current-perhaps chaotic-methods of job shopping with programs that induce earlier job stability requires evidence that those youths who experience unstable jobs or “floundering about” in their early years in the labor market suffer longer-term consequences.

Assessing such evidence poses econometric problems, however. As a leading example, job matching models suggest that the statistical association between adult wages and early job stability may not provide a good estimate of the increase in adult wages that might ensue from increased early job stability, because good matches are relatively more likely to have resulted in early job stability (Jovanovic, 1979; Mortensen, 1978). That is, these models predict non-random selection of individuals (based on their job matches) into early job market stability, in which case this statistical association will overstate the returns to increasing early job market stability for those who would otherwise have low stability, unless the induced increases in early job stability are accompanied by better matches. debt