BOSTON — If you scrutinize the U.S. labor market numbers from the last two years of economic recovery, you're left with what seems to be a paradox. Since the recession's low point, in November 2001, the number of employed people 16 and older has risen, but the number of jobs on the formal payrolls of employers remains below recessionary levels.
The different numbers have conveniently provided politicians with a choice, depending on which points they wanted to make, but they've perplexed economists.
Some analysts have concluded that this gap must represent the different ways in which statistics are kept by the two main sources of national data on employment developments.
But here's a more likely explanation. The number of people on formal payrolls remains low because new jobs tend to be ones that don't show up on payrolls. Employment gains are among the self-employed and contract workers, or in the informal "gray" and "black" labor markets. People are doing temporary day work or contracting that's kept off the books. These don't tend to be highly paid jobs or jobs with benefits like health insurance, and they are often performed by immigrants, especially undocumented immigrants.
Each month the U.S. Bureau of Labor Statistics releases two sets of employment data. One set of estimates is based on a survey of households across the nation and reflects what people say about their employment status; the other is based on a survey of employers, public and private, who report how many people are on their payrolls. If you listen to what the employees say, civilian employment in the nation has increased by about 2.3 million workers in the last couple of years. But if you look at what the employers report, the U.S. lost nearly 800,000 jobs between November 2001 and November 2003.
It's no surprise that the two estimates don't match perfectly. When we looked at employment data from the first two years of recovery in the previous five recessions, we found that gaps were normal. But not on the scale we are now seeing. In those earlier recessions, the difference in the two sets of numbers averaged just 292,000. This time the difference is nearly 3.1 million, or more than 10 times the size of the gap between the two sets of numbers in other recessions we examined.
The tendency among those who use these statistics is to pick one survey over the other depending on methodological -- or political -- preferences and simply dismiss the findings of the other survey. But it's important to try to understand the gap, because what it really tells us is that the nature of work is changing.
