Identifying sorting: in theory

We argue that using wage data alone, it is virtually impossible to identify whether Assortative Matching between worker and firm types is positive or negative. In standard competitive matching models the wages are determined by the marginal contribution of a worker, and the marginal contribution might be higher or lower for low productivity firms depending on the production function. For every production function that induces positive sorting we can find a production function that induces negative sorting but generates identical wages. This arises even when we allow for non-competitive mismatch, for example due to search frictions. Even though we cannot identify the sign of the sorting, we can identify the strength, i.e., the magnitude of the cross-partial, and the associated welfare loss. While we show analytically that standard fixed effects regressions are not suitable to recover the strength of sorting, we propose an alternative procedure that measures the strength of sorting in the presence of search frictions independent of the sign of the sorting.