The alleged failure of the economy to create jobs is an illusion that stems from a survey of employment conducted each month by the U.S. Bureau of Labor Statistics (BLS), commonly known as the payroll survey. The payroll survey has problems measuring the modern workforce and contains a unique methodological problem: It systematically overcounts the jobs of many workers when they change employers. But worker turnover has declined significantly since the 1990s, artificially deflating payrolls.

Since the recession ended in November 2001, payroll jobs are down by over 700,000, in contrast to the additional 1.9 million Americans who say they are working, according to the latest Labor Department household survey. A close examination of the two surveys suggests that payrolls are missing something. Congress should not move to protect jobs or meddle in labor markets if, as the facts show, those markets are functioning well.

• The payroll survey double-counts many workers who change jobs and is now artificially deflated because job turnover is down. Decelerating turnover in 2002-2003 explains up to 1 million jobs "lost" in the payroll survey since 2001.
• The payroll survey does not count the surge in self-employment. The household survey has recorded a surge of 650,000 self-employed workers. This number may be even higher if modern workers in limited liability companies or in consulting positions with traditional firms are not identifying themselves as self-employed.
• Revisions of the payroll survey have frequently been in the millions. Overestimates of 1 million payroll jobs were common in the early 1990s and also in 2002. Underestimates of 1 million jobs were common during the 1993-1996 period.
• The disparity between the two BLS employment surveys is cyclical. The disparity widens during recessions and narrows during periods of rapid growth in gross domestic product (GDP). Such variation strongly suggests a statistical bias in one of the surveys.
• The BLS household survey indicates record high employment. The disparity of 3 million jobs (in employment growth) between the household and payroll surveys since the recovery began is unprecedented.

Pessimists have tried to defend the payroll survey because it is bigger and less volatile than household survey data. The danger in the "bigger is better" rationale is mistaking quantity for quality. The big sample in the payroll survey is filled with double-counts and miscounts. The supposedly high monthly variability in the household survey is no worse than the large revisions in payroll data, except that revised payroll variability is historically invisible.

The Modern Workforce
One could think of the payroll survey as counting all the "brown-eyed" workers at traditional firms and "blue-eyed" workers at start-ups. It does not count "green-eyed" individuals who are self-employed, consult 20 hours a week, or simply home-school their children.

Further, workers who leave the IBM payroll for full-time consulting roles with IBM are still likely to consider themselves IBM employees. Likewise, partners at a limited liability company (LLC, a new company form) often consider themselves traditional employees, while the government classifies them as self-employed. The number of LLCs has swelled from near zero in 1993 to over 700,000. Clearly, a "hazel-eyed" workforce--self-employees who consider themselves payroll "brown-eyes"--is emerging. Yet these consultants and partners, like every farmer and sole proprietor in America, are considered jobless by the payroll survey.

Decelerating Turnover Since 2001
The payroll survey double-counts any individual who changes jobs during the pay period in which the worker is on two payrolls. "If a person leaves one job and starts another during a relatively short time span, they could appear on both employers' payrolls," said Labor Department economists in late 2003. Constant turnover means that the payroll survey systematically overestimates the level of jobs.

Because worker turnover has declined since 1999, the measure of payroll jobs has been deflated by up to 1.77 million jobs. In the two years since the recession ended, gross job flows are down by 1.2 percent, deflating the payroll survey by 1 million jobs. Thus, a net growth in payroll jobs in the real economy looks like a job loss in currently published data.

The bottom line is that either the current payroll data will be revised significantly or they will not. If they are revised, the household survey will be essentially vindicated. The more intriguing possibility is that there are structural problems within the payroll survey that have only just now surfaced in the wake of the odd recovery of the "new" economy. The loss of payroll jobs may be permanent as the workforce shifts to new forms.

Conclusion
Policymakers and analysts should treat payroll data with caution when making comparisons to employment levels in 2001 and earlier years. A better measure of employment is the household survey, and analysts can now point with confidence to the employment of an additional 1.9 million workers since the recession ended.

Tim Kane, Ph.D., is Research Fellow in Macroeconomics in the Center for Data Analysis at The Heritage Foundation.