Cross-sectional versus panel estimates of union wage effects

The main objective of this paper is to compare and reconcile the cross-sectional and the panel estimates of union wage effects. It uses Panel Survey of Income Dynamics (PSID) data for a period of 6 years (1979-1984). Unlike what Freeman (1984) predicted (i.e., panel union wage estimates would be smaller than pure cross-sectional estimates due to the cumulative nature of the errors in reporting union status in the former), both the cross-sectional and panel estimates of union wage effects are very similar ranging from 11 to 13%. unions' seem to have flattening effect on the age-earnings profiles. Not surprisingly, the rate of return to schooling of non-union members is larger than that of the union members.