The number of nonagricultural workers “on part-time schedulesfor economic reasons,” shows a strong relationship to businesscycle trends, according to seasonally adjusted data from the CurrentPopulation Survey. The number and proportion of persons involuntarily working part time–sometimes referred to as the “partiallyunemployed”–generally rise during a recession and decline during arecovery period. In a comprehensive examination and analysis of thesedata which appeared in the June 1983 Monthly Labor Review, Robert W.
Bednarzik demonstrated that during cyclical periods, the incidence ofeconomic part-time work moves in the same direction as, but leads,movements in the civilian unemployment rate. Bednarzik explained thatsuch part-time employment typically rises before unemployment begins toincrease during a recession, mainly because employers tend to reducehours of work when possible before laying off employees. Duringrecovery periods, employers usually restore the hours of those onshortened workweeks before rehiring laid-off workers. The main focus ofBednarzik’s analysis, however, was the relationship and variationin cyclical behavior of the two main causes of involuntary part-timework, cutbacks in weekly hours due to slack work and failure to findfull-time work, both of which were seasonally adjusted specifically forhis study. Following up on Bednarzik’s analysis, BLS tested the cyclicalsensitivity and accuracy of the new series and confirmed that these datacaptured more clearly the distinctions between the concepts of personsworking part time involuntarily than did the existing published series,which divided the total number into those who “usually work fulltime” and those who “usually work part time.” Thus, toprovide data users with more relevant series that can isolate the maincauses of part-time work, BLS has replaced the existing usual full- andpart-time series with the new series. Effective with data for January1985, the new series are published in monthly issues of “TheEmployment Situation” news release and Employment and Earnings,and, beginning with this issue, are also published in table 4 in theCurrent Labor Statistics section of the Monthly Labor Review.
Data arepublished for all persons (in agriculture and nonagricultural industriescombined) as well as for persons in nonagricultural industries only.(The former series were limited to workers in nonagriculturalindustries.) Time series based on the new definitions are availableback to 1955 and can be obtained from BLS. The new series clearly show different cyclical behavior, which, inturn, illustrates different underlying labor market problems.
The morecyclical “slack work” series reflects short-run adjustmentsmade by firms to minimize layoffs and subsequent recalls or hirings.Thus, slack work rises sharply during economic downturns, but showsrapid improvement during the early stages of recovery. The”failure to find full-time work” series reflect theexperience, skills, and training of workers; the match of availableworkers to work schedules; and the types and locations of job openings,as well as the general state of the economy. The “failure tofind” series is clearly less cyclical. Indeed, in contrast to the”slack work” component, it typically rises during the earlystages of a recovery, probably because many unemployed workers find andaccept part-time jobs (perhaps after exhausting unemployment insurancebenefits) as a better alternative to remaining fully unemployed withoutcompensation. Recent data illustrate this point. The following tabulation showsthe number of persons (seasonally adjusted) and the percent of totalcivilian employment on part-time schedules for economic reasons duringSeptember of 1982 and 1983 and January 1985: The number of persons involuntarily working part time due to slackwork dropped by 1 million in the first 12 months of recovery from theseries high in September 1982 and by only 265,000 in the subsequent 16months (through January 1985).
During the first 12 months of recovery,the proportion of the total employed comprised by persons on shortworkweeks due to slack work fell from 3.7 to 2.6 percent.
In January1985, the group accounted for 2.3 percent of the employed. This patternof decline was similar to that following the recovery from the 1973-75recession. Thus, it seems clear that this component shows rapidimprovement early in the recovery, as employers rest ore hours of thoseworkers retained but with reduced workweeks before adding new workers,and then improves more slowly as the recovery matures. In contrast, theother major component–persons who can only find part-time jobs–showedno improvement early in the recovery period; indeed, it rose slightly.It did moderate later, but not by the magnitude of the decline in theslack-work component.