Publications [#164654] of Christopher D. Timmins

Working Papers

1. T. MaCurdy and C. Timmins, Bounding the influence of attrition on intertemporal wage variation in the NLSY (2000)
   (last updated on 2009/10/22)

   Abstract:
   This paper analyzes wage dynamics in the National Longitudinal Survey of Youth, controlling for the effects of censoring caused by non-random attrition. Nonrandom attrition, caused by individuals failing to appear for interviews or choosing not to work, is common in longitudinal surveys like they NLSY and can bias statistical analyses. Techniques to control for the effects of non-random censoring on the dynamics of mean wages require a great deal of knowledge of (or assumptions about) the censoring process. We adapt the non-parametric bounding techniques of Manski for use with a newly proposed Smoothed GMM quantile estimator to overcome this problem by studying the dynamics of wage percentiles. Results suggest that non-random attrition does not pose serious problems for the analysis of men{\~A}{\^a}{\~A}\leavevmode\hbox{\rm\rlap/c}{\^a}$\neg${\^a}\leavevmode\hbox{\rm\rlap/c}s wages, but that the combination of multiple sources of attrition leads to significant potential biases in the study of women{\~A}{\^a}{\~A}\leavevmode\hbox{\rm\rlap/c}{\^a}$\neg${\^a}\leavevmode\hbox{\rm\rlap/c}s wages.

   Keywords:
   ARMA models • NLSY • Wage growth relationships