Optimal short-time work : screening for jobs at risk

Short-time work - a wage subsidy conditional on hour reductions - has become an important tool of labor market policy in many European countries. As the scope of these policies expanded, concerns about side effects due to adverse selection increased. We develop a model of job retention policies in the presence of asymmetric information to study selection into these programs. The social planner wants to prevent excessive job destruction but cannot observe which jobs are truly at risk. We do not restrict the social planner to use hour reductions a priori. Instead, we show that hour reductions of short-time work policies act as a screening mechanism to mitigate the adverse selection problem. This perspective of short-time work as a policy response to an underlying adverse selection problem provides an entirely new rationale for these policies. Our approach can be used to revisit recent empirical findings which rely on employment effects to evaluate existing short-time work schemes. In our model, however, average employment effects across groups are not sufficient to determine whether the policy is efficient. Indeed, we show that an optimal short-time work policy cannot avoid a small degree of adverse selection. This is particularly important in light of recent evidence that firms with small revenue shocks and no discernible employment effects have participated in short-time work programs at large costs to the public. In our model, these costs are information rents which are required to screen for jobs at risk. We calibrate our model to German data before the financial crisis and find that the optimal short-time work policy would have reduced separations by 1.2 - 2.4 percentage points.