Evaluating Business Cycle Models with Labor Market Search

Incorporating labor market search in general equilibrium models has been shown to generate realistic dynamics in employment, job creation, and job destruction and to increase the magnitude and persistence of the impact of productivity shocks on output. This paper studies the extent to which the empirical successes of those models are dependent upon specifications of the matching technology that can exaggerate the level of unemployment, making possible large employment flows in response to business cycle shocks. Even with a more strict definition of unemployment and a less constrained flow of workers into new employment, models with labor market search are still found to generate larger and more persistent propagation of productivity shocks than do traditional real business cycle models. However, the respecification of the job turnover process reduces the size of those effects. Policy experiments find that subsidizing job creation can lead to small reductions in output's response to a negative productivity shock, but it reduces steady state output and consumption and increases unemployment and the volatility of output. Taxing job destruction the opposite effects.