Abstract: Does public employment pay differently over a career, and for whom? I answer using French administrative panel data (DADS-EDP, 2012--2019) and a dynamic model of employment and earnings with five labor-market states and latent transition and income classes, estimated by a sequential Expectation-Maximization algorithm. Three findings emerge. First, a lifetime public premium exists: lifetime earnings are higher in the public sector up to the 46th percentile of the earnings distribution, and up to the 30th percentile when selection on latent types is removed. Second, the public sector exhibits smaller differences in lifetime earnings for men and women by about 13\%, through better retention, more equitable hours, and better-paid part-time work. Third, the premium is not uniform: highly educated workers face a lifetime penalty in the public sector, reflecting compressed public wage grids. The results reconcile hourly-wage penalty estimates with lifetime premium estimates as different estimands of the sectoral gap, implying that there is no single measure of the public pay gap.

Abstract: We introduce occupation as a third dimension of wage decompositions, alongside workers and firms, and find that it is central to how wages are dispersed and how workers and firms sort. Occupation belongs to neither side alone: workers choose occupations, and firms choose which occupations to employ. Using French administrative data for 2015--2019, we estimate a model of wages and mobility with latent worker and firm types and observed occupations. Under every decomposition we use, occupation accounts for at least as much wage variance as firm type and far more than worker type, and the dominant sorting channel runs between firms and occupations. Occupation accounts for a quarter to a third of the AKM worker effect and for two-fifths of the firm share in a discrete-type decomposition. Our model estimates little portable worker heterogeneity in the data and still reproduces the standard AKM decomposition. Applying the event-study design of Card, Heining and Kline (2013) to occupation changes within firms, we find that workers gain the premium of a better-paid occupation on promotion and largely keep it after moving down: they lose 19 percent of the gain, against 88 percent for firm premia. Occupation premia are acquired and retained, while firm premia stay with the employer.

Abstract: This paper investigates the structural reliability of the AKM variance decomposition (developed by Abowd et al. [1999]) within sparse bipartite worker-firm networks. I demonstrate that the incidental-parameter bias inherent in plug-in estimators is driven by the spectral properties of the graph Laplacian, where a low average node degree ($\bar{d}$) serves as a key diagnostic for estimation fragility. Crucially, the analysis reveals that distortions in sorting covariance and worker variance shares are not merely artifacts of limited mobility; rather, they arise from the interaction of idiosyncratic noise with the weighting structure of the bipartite graph.Through Monte Carlo simulations, performance of some bias correction methods are evaluated. The results indicate that while these corrections reduce bias, they remain susceptible to noise-amplification when the network's spectral density is low, a phenomenon that systematically attenuates evidence of positive assortative sorting. By identifying $\bar{d}$ as a key statistic for the severity of spectral collapse, this research provides a theory for characterizing the structural sensitivity of variance component estimators to the underlying worker-firm network.

Abstract: This study examines the impact of privatization on female full-time employment, with implications for UN SDG 5 on gender equality. Using forecasted treatment effects and bootstrapping, data from 1997-2007 are analyzed to estimate the Average Treatment Effect on female full-time employment. Deterministic trends forecast counterfactual outcomes, while bootstrapping estimates the ATE's confidence intervals. Findings reveal a significant decline in female full-time employment post-privatization. This case study highlights the importance of considering gender costs in privatization policies and underscores the need for robust methods in policy analysis. Further research should explore long-term implications and mechanisms affecting gender equality.