Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Female earnings function parameters are estimated jointly with interruption incidence and fertility decisions. We pay particular attention to the return to experience and the human capital loss parameter when women are allowed to self‐select into various possible combinations of interruption and fertility outcomes. We find that ignoring endogenous fertility decisions may seriously bias the effect of non‐employment spell lengths on female earnings, as well as the return to experience. To a large extent, the negative relationship between female earnings and non‐employment spell length, that is widely documented in the earlier literature, is the result of a misspecification of the earnings functions.