Pseudo Panel Data Models With Cohort Interactive Effects Artūras Juodis 10.6084/m9.figshare.4928849.v1 https://tandf.figshare.com/articles/journal_contribution/Pseudo_Panel_Data_Models_With_Cohort_Interactive_Effects/4928849 <p>When genuine panel data samples are not available, repeated cross-sectional surveys can be used to form so-called pseudo panels. In this article, we investigate the properties of linear pseudo panel data estimators with fixed number of cohorts and time observations. We extend standard linear pseudo panel data setup to models with factor residuals by adapting the quasi-differencing approach developed for genuine panels. In a Monte Carlo study, we find that the proposed procedure has good finite sample properties in situations with endogeneity, cohort interactive effects, and near nonidentification. Finally, as an illustration the proposed method is applied to data from Ecuador to study labor supply elasticity. Supplementary materials for this article are available online.</p> 2017-04-27 21:37:56 Cohort interactive effects Labor supply elasticity Pseudo panel data Weak identification