Labor market fluctuations in the small and in the large

Shimer's calibrated version of the Mortensen–Pissarides model generates unemployment fluctuates much smaller than the data. Hagedorn and Manovskii present an alternative calibration that yields fluctuations consistent with the data, but this has been challenged by Costain and Reiter, who say it generates unrealistically big differences in unemployment from the differences in policy we see across countries. We argue this concern might be unwarranted, because one cannot assume that elasticities relevant for small changes work for large changes. Models with fixed factors in market or household production can generate large effects from small changes and reasonable effects from large changes. This is reminiscent of attempts to improve the labor market in the Kydland–Prescott model, especially ones incorporating household production, like Benhabib, Rogerson, and Wright.