Cím: A Grouped-AKM Approach for (Re-)Assessing Bargaining, Sorting and Differential Rent-sharing
Magyar cím: Alkuerő, szelekció és járadék-osztozkodásbéli különbségek (újra-)vizsgálata egy
csoportosított-AKM alapú megközelítésben.
This paper contributes in multiple ways to the branch of literature in labor economics that estimates firm premia flexibly, by allowing firm effects to vary across subsets of the employer’s workforce. First, we provide a novel formulation for such models, which consists of estimating firm-group wage effects on one joint sample, and obtaining underlying firm-specific average effects in a second stage. This approach facilitates a novel, unambiguous decomposition of bargaining versus sorting effects, which could be generalized for differences across more than two groups — as opposed current best practice in the literature. A more detailed assessment of differential rent-sharing is provided. Our model accounts for the possibility of non-random sorting of workers into firms with different firm-specific pass-through rates, a channel that previous studies have neglected. By linking models of differential rent-sharing and firm-group effect decompositions we provide a detailed assessment of the contribution of sorting with respect to both levels and returns to firm-productivity to between group wage differentials. In five specifications, we estimate group-specific firm premia for workers of different gender, education, occupation, age or completed tenure. While these specifications only marginally increase the explanatory power of wage equations, sorting into different premia firms proves to be an important factor in wage differences across gender, age and seniority. Education and occupational wage differentials, on the other hand, are more driven by within-firm differences. Even after accounting for sorting with respect to rent-sharing, we find significantly higher wage returns to firm productivity for males, more educated workers, and for those in better occupations. A detailed decomposition reveals that both sorting with respect to the level of firm productivity and the different returns to productivity changes contribute substantially to the gender gap in firm-group effects, with the latter between-firm channel being somewhat more prominent.