Comment

Abstract

Previous articleNext article FreeCommentRobert E. HallRobert E. HallStanford University and NBER Search for more articles by this author Stanford University and NBERPDFPDF PLUSFull Text Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreAbout one-third of the working-age US population is neither working nor looking for work in any given month, according to the Current Population Survey. A small group of those who are out of the labor force indicate an interest in working. The authors study that group and attribute the downward trends in unemployment and employment to a downward trend in the fraction of people who are out of the labor force, but interested in working. The paper applies ideas from a model of individual dynamics in the labor market, which emphasizes the roles of transition rates among the three basic categories of labor market status: out of the labor market, unemployed, and employed.The key question in the Current Population Survey that underlies this research is, “Do you want a job now, either full or part time?” The survey asks this question of everybody ages 16 years and older who did not work in the survey week and did not engage in specific job-seeking activities in the past four weeks. The paper diagnoses a secular decline in “desire to work,” especially in the boom of the 1990s. It teaches us a lot about how people behave who are not currently working or looking for work, but are interested in working. The proposition that a period of extraordinarily favorable conditions in the labor market and rapid growth of employment coincided with a decline in interest in working is a surprise. But I’m not convinced that there was a decline in the desire to work in the 1990s; rather, such a decline began around the year 2000.Table 1 gives some basic data about people who were not in the labor force in 2007, a year of normal conditions in the labor market.1 About 34% of the population 16 years and older was out of the labor force, neither employed nor looking for work, and recorded as unemployed. The great majority responded “no” to the question about wanting a job now. Those answering “yes” accounted for 2% of the population. More than half of them, about 1.2% of the population, had not searched for work in the previous year, suggesting that their interest in working had not actually been high enough to take action to get jobs. A small group, accounting for one-quarter of 1% of the population, said they were not available to work now even though they wanted a job now. Overall, in a normal year, 0.6% of the population 16 and older who were out of the labor force wanted jobs, had taken action to find jobs in the prior year, and were currently available to work. By contrast, in that year, 3.0% of the population was not working but had actively looked for work in the past four weeks and were, therefore, counted as unemployed.Table 1. Responses to the Question about Wanting a Job and to Related Questions, 2007 Count (1,000s)Percent of populationWorking age population231,867100Not in labor force78,74433.96Not in labor force, do not want a job now74,04131.93Not in labor force, want a job now4,7032.03Not in labor force, want a job now, did not search for work in previous year2,7481.19Not in labor force, want a job now, searched for work in previous year1,9550.84Not in labor force, want a job now, not available to work now5600.24View Table ImageThe paper is about changes over time in the fraction of the population who were out of the labor force who wanted jobs. Figure 1 shows the fraction over the period from 1994 (when the Current Population Survey was revamped) to 2014. During the vigorous expansion of the 1990s, unemployment fell to a low of 4.0% of the labor force averaged over the year 2000. The fraction of the population not working but desiring work fell dramatically over that period. Unemployment oscillated through 2007 and the fraction desiring work remained roughly constant through 2007. The want-job fraction rose sharply when unemployment skyrocketed in late 2008 and reached its peak in 2010. With the return of unemployment down to close to normal levels, the want-job fraction has fallen only a small amount.Fig. 1. Percent of population age 16+ who said they wanted a job nowView Large ImageDownload PowerPointI find it difficult to reconcile figure 1 with the framework and conclusions of the paper. The interpretation that jumps out of the figure to my eye is that the want-job fraction has the same determinants as does unemployment. In 1994, in this interpretation, the labor market was somewhat soft, with unemployment at 6.1%. Unemployment was high because job-finding rates were abnormally low, according to modern theories of unemployment. The want-job population suffers, in part, from being in a state of concealed unemployment. Job-finding rates in that group are positive but low, and are even lower in a soft market such as in 1994. Higher job-finding rates later in the 1990s resulted in shrinkage of the unemployed population and of the want-job population.Figure 2 limits the want-job population to those who said they were available now. The positive correlation between the percent of the population in that subgroup and the unemployment rate is even more striking with this limitation. The size of this subgroup seems to be determined by essentially the same forces as is the unemployment rate. On the other hand, figure 3 shows that the other subgroup, who were not available to work, did not track unemployment after 2000, though it did decline along with unemployment in the 1990s. Figure 4 limits the population who wanted to work to those who had searched in the previous year. This limitation also isolates a group whose size appears to be determined by the same forces as unemployment.Fig. 2. Percent of population age 16+ who said they wanted a job now and were available to work.View Large ImageDownload PowerPointFig. 3. Percent of population age 16+ who said they wanted a job now but were not availableView Large ImageDownload PowerPointFig. 4. Percent of population age 16+ who said they wanted a job now and had searched in the previous year.View Large ImageDownload PowerPointThe slowing down of job finding in weak labor markets slows down flows for the want-job population, just as it slows down job finding for all types of job seekers. Flinn and Heckman (1983) advanced the idea that people should be designated as unemployed if their job-finding rates are above a reasonable threshold. Results in the meeting-time version of the paper, but not in the final version, showed that the transition rate from want-job status to employment is 15% per month, compared to a general rate from unemployment to employment of 27%. But the transition rate from want-job status to unemployment is 47 percent per month, so the job-finding rate over longer spans of months may be closer to the level among the unemployed. Thus by the Flinn-Heckman logic, at least some of the people in the want-job category should probably be included in the count of the unemployed. More generally, I conclude that this paper and other recent research suggests the benefit of reconsidering the definition of unemployment along Flinn-Heckman lines. The basic idea would be to fit a job-finding probability model using a wider set of answers to the CPS questions—not just recent jobseeking activities, but answers to questions about wanting a job and planning to start a job search. This definition of unemployment would set a threshold value for the estimated probability and classify as unemployed all the nonworking respondents whose fitted job-finding probability met the threshold.The paper seeks to use the answers to the want-a-job question to measure a concept the authors call the desire to work. To me, the desire to work is the difference between the payoff of working and the payoff of nonwork activities. People who are neither looking for work nor working are those with a negative desire to work. A change in the economic environment that raises the general desire to work will result in an increase in the labor force participation rate, which is the fraction of the population with a positive desire to work. By this logic, the best measure of the desire to work is the participation rate. Figure 5 compares the participation rate to the want-job fraction of the population. The participation rate rose a little during the boom of the 1990s, at the same time that the large decline in the want-job fraction declined sharply. Starting in 2000, participation began to fall. Participation began to fall more rapidly around the time of the Great Recession, and the fall continued at about the same rate as the recovery began in the labor market in 2011. Nothing in the figures supports the idea that participation and the want-job fraction share common determinants. Rather, the want-job fraction responds to labor market tightness and thus moves in concert with unemployment. There is little agreement about the determinants of the movements of participation. Demographics have only a modest contribution.Fig. 5. The labor force participation rate and the fraction of nonparticipants wanting jobsView Large ImageDownload PowerPointThe paper studies monthly transitions in the Current Population Survey. This approach may give a distorted impression of labor market dynamics because of classification errors in the survey and because of the high incidence of very short jobs (durations of less than one month). Krueger, Cramer, and Cho (2014) were the pioneers in studying job-finding success over multimonth spans, which helps overcome both problems. Hall and Schulhofer-Wohl (2015) pursue the idea for job-finding rates. The claim of the paper (p. 459) that transitions are so fast that the labor market is always close to its steady state is an artifact of the first-order assumption. There is a growing realization that it actually takes a long time to return to the steady state after a shock—witness the persistent overhang of long-duration unemployment.The paper makes a valuable contribution by calling attention to the neglected want-job questions in the Current Population Survey. Much more research should be done to integrate the answers to those questions into models of labor-market dynamics. However, I’m not sold on the attempt of the paper to link the answers to the want-job question to the concept of the desire to work.EndnotesFor acknowledgments, sources of research support, and disclosure of the author’s material financial relationships, if any, please see http://www.nber.org/chapters/c13603.ack.1 . A spreadsheet with all of the data and calculations in this discussion is available from my website.ReferencesFlinn, Christopher J., and James J. Heckman. 1983. “Are Unemployment and Out of the Labor Force Behaviorally Distinct Labor Force States?” Journal of Labor Economics 1 (1): 28–42.First citation in articleLinkGoogle ScholarHall, Robert E., and Sam Schulhofer-Wohl. 2015. “Measuring Job-Finding Rates and Matching Efficiency with Heterogeneous Jobseekers” NBER Working Paper no. 20939, Cambridge, MA.First citation in articleGoogle ScholarKrueger, Alan B., Judd Cramer, and David Cho. 2014. “Are the Long-Term Unemployed on the Margins of the Labor Market?” Brookings Papers on Economic Activity, Spring. http://www.brookings.edu/about/projects/bpea/papers/2014/are-longterm-unemployed-margins-labor-market.First citation in articleGoogle Scholar Previous articleNext article DetailsFiguresReferencesCited by NBER Macroeconomics Annual Volume 302015 Sponsored by the National Bureau of Economic Research (NBER) Article DOIhttps://doi.org/10.1086/685971 Views: 205Total views on this site © 2016 by the National Bureau of Economic Research. All rights reserved.PDF download Crossref reports no articles citing this article.