Working Paper: NBER ID: w30919
Authors: Holden A. Diethorn; Gerald R. Marschke
Abstract: We use the labor market for doctorates in the biomedical sciences, where career dislocation is common, as a case study of skill-task mismatch and its consequences. Using longitudinal, worker-level data on biomedical doctorates, we investigate mismatch as an explanation for the negative pecuniary returns to postdoc training. Our data contain unique worker-level job task information that allows us to compare the skills acquired in the years just after graduation to the tasks required in later employment. Our findings reveal a postdoc salary penalty when task mismatch is high, which is frequent, and a salary premium when skills align with tasks. Differences in accumulated task-specific human capital explain the between-sector heterogeneity in the returns to postdoctoral training, including the large and persistent salary penalties from postdoctoral training in industry, and the penalty overall. Task mismatch as a cost of pursuing risky careers in science and in other fields requiring large upfront investments in task-specific human capital has received little attention in the empirical labor literature.
Keywords: task mismatch; salary penalties; biomedical PhD; labor market
JEL Codes: I26; J24; J31; J44
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| postdoctoral training (M53) | likelihood of obtaining research-focused jobs (I23) |
| likelihood of obtaining research-focused jobs (I23) | potential higher salaries (J39) |
| task alignment (Y80) | increased earnings (J31) |
| postdoctoral training (M53) | lower earnings (J31) |
| task mismatch (C78) | lower earnings (J31) |