Abstract
We evaluate Austrian labor market policy focusing on its possible effects upon recurrent unemployment. Without properly considering the selection processes for public training programs, perverse results emerge. Taking the participation decision into account in a bivariate probit setting, Austrian manpower training programs turn out to be a sort of ‘catching up1 strategy: (i) disadvantaged and less motivated job-seekers are given priority in enrollment into training programs and (ii) participation in such courses improves employment stability significantly.