I examine changes in the incidence of job loss between 1981 and 1995 using
data from the Displaced Workers Surveys (DWS) from 1984-1996. The rate of job loss
followed a cyclical pattern between 1981 and 1991, starting at a high level in the slack
labor market of the early 1980’s, falling during the economic expansion of the middle and
late 1980’s, and rising during the weak labor market at the end of the 1980’s and early
1990’s. A puzzle is that the overall rate of job loss has increased in the 1990’s despite the
sustained expansion. I document these facts and address the possibility that the elevated
rates of job loss in the 1990’s are a statistical artifact resulting from changes in the wording
of key questions in the DWS in 1994 and 1996 exacerbating a problem of misclassification
of some workers as displaced. Using additional data from a debriefing of respondents to the
the February 1996 DWS to adjust rates of job loss to estimate a consistent time series of
job loss rates, I find 1) that the overall rate of job loss has not declined in the 1993-95 time
period, despite the strong labor market and 2) the overall rate of job loss in the 1993-95
period may be as high as it was during the slack labor market of 1989-91 and almost as
high as it was during the very slack labor market of 1981-83.