摘要：This paper appraises the literature on the macroeconomic effects of job reallocations. We overview different methodological approaches to the problem of observational equivalence of aggregate and sectoral shocks and draw two main conclusions. First, the non-directional nature of reallocation shocks holds the key to the fundamental identification problem. The second conclusion is that sectoral reallocation of labor has been responsible for no less that 1/4 and no more than 2/3 of the variance of aggregate unemployment in postwar data. This wide range indicates that the importance of labor reallocation may change over time, being larger at particular historical junctures.
