AN ANALYSIS OF SAMPLE ATTRITION IN PANEL-DATA - THE MICHIGAN PANEL STUDY OF INCOME DYNAMIC

By 1989 the Michigan Panel Study on Income Dynamics (PSID) had experie nced approximately 50 percent sample loss from cumulative attrition fr om its initial 1968 membership. We study the effect of this attrition on the unconditional distributions of several socioeconomic variables and on the estimates of several sets of regression coefficients. We pr ovide a statistical framework for conducting tests for attrition bins that draws a sharp distinction between selection on unobservables and on observables and that shows that weighted least squares can generate consistent parameter estimates when selection is based on observables , even when they are endogenous. Our empirical analysis shows that att rition is highly selective and is concentrated among lower socioeconom ic status individuals. We also shows that attrition is concentrated am ong those with more unstable earnings, marriage, and migration histori es. Nevertheless,,we find that these variables explain very little of the attrition in the sample, and that the selection that occurs is mod erated by regression-to-the-mean effects from selection on transitory components that fade over time. Consequently: despite the large amount of attrition, we find no strong evidence that attrition has seriously distorted the representativeness of the PSID through 1989, and consid erable evidence that its cross-sectional representativeness has remain ed roughly intact.