In contrast to nonunion workers, reemployment wages of workers displaced fr
om unionized jobs decline with tenure on the lost job. This finding cannot
easily be explained by firm- or industry-specific human capital accumulatio
n, deferred-pay policies, standard matching models, or a correlation betwee
n tenure and reentry rates into unionized jobs. Possible explanations inclu
de negative selection of senior union workers and a negative causal effect
of unionism on workers' alternative skills. Despite a much flatter predispl
acement tenure-wage profile, displaced union workers' wage losses increase
with tenure at a rate comparable to or higher than those of nonunion worker
s.