Gender, Selection into Employment, and the Wage Impact of Immigration

Monopsony, Efficiency, and the Regularization of Undocumented Immigrants

“…The cross-section estimates of 𝛼 1 and 𝛼 2 in columns 1-2 indicate that the average wage of low-educated French men in the Paris region fell after the policy change. As noted by Borjas and Edo (2021), the employment response likely changed the sample composition of earners, producing a selection bias in the estimates of 𝛼 1 and 𝛼 2 . If there is positive selection into employment and if regularization indeed increased the wage of low-skill workers, the mean wage of the French men who entered the market after 1982 would be below the mean wage of the pre-existing workers.…”

“…One possible way of correcting for selection bias (called the "identification at infinity" method in the literature) is to isolate a subsample of workers for whom selection into employment is unlikely to matter (Chamberlain, 1986;Heckman, 1990;Mulligan and Rubinstein, 2008;and Blau, Kahn, Boboshko and Comey, 2021). The employment probability of French men increases substantially with the presence of children (Borjas and Edo, 2021). We use the "identification at infinity" approach by reestimating the model in the subsample of French men who had at least one child below age 18.…”