A previous version of this working paper was originally published in January 2012.

The model with a constant separation rate fails to produce realistic volatility and productivity responsiveness of the separation rate and worker flows. The specification with endogenous separation succeeds along these dimensions. Allowing for on-the-job search enables the model to replicate the Beveridge curve. All specifications, however, fail to generate sufficient volatility of the job finding rate. While adopting the Hagedorn-Manovskii calibration remedies this problem, the volume of job-to-job transitions in the on-the-job search specification becomes essentially zero.