A decline in pension coverage during the 1980s for males in the privat
e sector is confirmed and is contrasted with the federal, state, and l
ocal sectors in which no such decline is found. As a consequence, the
governmental advantage in the probability of coverage grew over the 19
80s. This remains true even in the most appropriate testing methodolog
y in which worker earnings are endogenous and in which otherwise const
ant characteristics are moved between years and sectors. Government wa
ge differentials are estimated as a byproduct of this methodology and
show a tendency toward comparability over the 1980s.