Newswise — Amid recent news of falling unemployment and the beginnings of payroll employment growth that would allow labor market to emerge from the so called "jobless recovery," few labor market analysts and members of the media have closely examined the ever-widening gap between the employment growth estimates of the U.S. Department of Labor's two most important monthly surveys, the Current Population Survey (CPS) and the Current Employment Statistics survey (CES), which are used to provide an accurate employment picture for the nation each month. According to a new report issued today from Northeastern University's Center for Labor Market Studies, the growing gulf between these two government surveys' estimates of employment growth since the recession calls for the attention of the nation's policymakers and economic analysts. More people are definitely finding employment, the researchers say, but it's how they're working and being compensated that makes this period of recovery so historically unique.
The CPS (a household survey of the employment status of persons 16 and older) and the CES surveys (a survey of employers' payroll employment levels) are conducted by the U.S. Department of Labor each month. While both the CPS and CES surveys offer estimates of the number of employed, the CES payroll survey's monthly estimates of the number of wage and salary jobs is most frequently relied upon by economists, the media, and financial analysts to identify recent job growth while the CPS findings are more frequently used to track overall unemployment developments. But while the two surveys generated mirror estimates of job loss during the 2001 recession, they've produced extraordinarily different estimates of the nation's employment picture since the economic recovery that began in November 2001.
From the trough of the recession in November 2001 through November 2003, the CPS survey estimates the number of employed persons has risen by nearly 2.4 million while the CES survey indicates that the number of wage and salary jobs in November 2003 is still some 726,000 below its November 2001 level. This chasm " to the tune of some 3.11 million jobs " is historically unique, the researchers say, pointing out that what so many have termed a "jobless recovery" is, in fact, something else entirely.
While economist Andrew Sum acknowledges that there is often a gap between the two measures of employment growth in the first two years after recovery from national recessions, he and his colleagues Paul Harrington and Ishwar Khatiwada examined the last five recessions (from 1969-70 to 1990-91) and noted that never before has the gulf between the two surveys' estimates of employment growth been so wide. Between November 2001 and November 2003, they point out that, the difference between the surveys' estimates of employment growth was nearly ten times as high as the average difference between these two surveys' employment growth estimates in the five preceding recessions (3.1 million verses 300,000).
"The absolute size of the differences between the employment growth estimates of the two surveys for the November 2001 to November 2003 period are historically unprecedented," Sum said. "At no other point in this nation's last five recovery periods have so many people been employed as independent contract workers, as temporarily self-employed, or paid under the table. Since the bottom of the recession, many people have managed to find work " sometimes multiple jobs " but they often don't show up on employers' payroll because the nature of the work arrangement is different."
"The CES is depicting a much more substantial set of sustained job losses than the CPS survey," Harrington said. "The normal pattern of recovery is strong growth in the number of wage and salary jobs. With the CES employment statistics for the past two years, we see the opposite taking place. The nation's experience during the first 24 months of recovery from the recession of 2001 is, thus, clearly an anomaly."
In their report, the researchers point out some fundamental differences between the two surveys' procedures for estimating employment. The CES counts jobs rather than people (i.e. if the same person were to hold two jobs, he would be counted twice). CES also includes 14 and 15 year-olds on the payrolls while the CPS only counts employed persons 16 and older. The CPS survey, however, provides a much broader measure of employment, including farm workers, the self-employed, household workers, contract workers, unpaid family workers, private household workers as well as those working for pay off the books, including legal and illegal immigrants not taken into account by the CES survey.
The paper argues that the differences between the employment growth estimates of the CPS and CES surveys are not statistical artifacts but instead reflect a number of fundamental adjustments by workers and firms in the absence of any substantive payroll employment growth during the current recovery, including the use of contract labor arrangements, temporary shifts into self-employment, working off the books, and the continued strong growth of the immigrant labor force, many of whom are undocumented.
The report examines conceptual differences in employment measures between the two surveys and finds that, while growth in self-employment and farm labor can help explain part of the gap (about one-sixth) between the CPS and CES growth estimates, the bulk of the difference appears to be attributable to the increased use by firms of independent contractors who will be counted by the CPS but not by the CES and to the growth of employment in the informal economy, including the hiring of many undocumented immigrants over the past three years. Since 2000, foreign immigration has contributed some 2.3 million new workers in the U.S. Finally, the researchers say, a part of the gap between the CPS and CES estimates is due to a population adjustment to the CPS survey in January 2003 that added approximately 600,000 workers to the CPS employment totals. Nearly two-thirds of the workers added to the January 2003 CPS employment totals would likely have been employed prior to the November 2001 recession.
Sum and Harrington call for the CPS " the less frequently cited of the two measures " to be taken more seriously by policymakers, economists and the media.
"There are many more persons with jobs in the U.S. since November 2001 than indicated by the payroll survey," Sum said. "This also helps explain why the nation's official unemployment rate is much lower than would have been the case if employment had continued to decline to the extent as indicated by the CES payroll survey. The media's frequent reference to a 'jobless recovery' needs to be substantially revised. The higher levels of employment revealed by the CPS also indicate that labor productivity growth in the past two years likely has been somewhat exaggerated."
Northeastern University, located in the heart of Boston, Massachusetts, is a world leader in cooperative education and recognized for its expert faculty and first-rate academic and research facilities. Through co-op, Northeastern undergraduates alternate semesters of full-time study with semesters of paid work in fields relevant to their professional interests and major, giving them nearly two years of professional experience upon graduation. The majority of Northeastern graduates receive a job offer from a co-op employer. Cited for excellence two years running by U.S. News & World Report, Northeastern was named a top college in the Northeast by the Princeton Review 2003/04. In addition, Northeastern's career services was awarded top honors by Kaplan Newsweek's "Unofficial Insiders Guide to the 320 Most Interesting Colleges and Universities," 2003 edition. For more information, please visit http://www.northeastern.edu.
RSS