Why Aren't More Americans Working?
According to today's employment report, 59.8 percent of Americans ages 16 and older had jobs in February. That's the highest employment-to-population ratio in years, and the rate of increase is clearly on the rise.
Look back some more years, though, and the story is different. The recent gains are real, but by the standards of the past few decades, a 59.8 percent employment-to-population ratio isn't impressive.
Let this be another lesson in how the presentation of information shapes our understanding of it. The second chart paints a gloomy picture -- the picture that Donald Trump may be referring to when he says the true unemployment rate is 40 percent or higher. A 59.8 percent employment-to-population ratio means that 40.2 percent of American civilians 16 and over don't have jobs. That percentage includes high-school students, 100-year-olds and lots of other people who don't want or need jobs, so the true unemployment rate clearly isn't 40 percent. Still, in April 2000 the employment-to-population ratio peaked at 64.7 percent. Now it's significantly lower. What's going on?
The answer that I keep gravitating to is that despite the 4.9 percent unemployment rate, the job market is still pretty weak, and probably malfunctioning in some way. This isn't the only possible answer. In 2014, for example, two economists at the Federal Reserve Bank of New York divided people responding to the Census Bureau's Current Population Survey (from which the unemployment rate and the charts in this article are derived) into 280 cohorts defined by "birth, sex, race/ethnicity, and educational attainment." They determined that most of the decline in the employment-to-population ratio since 2000 could be explained by the changing makeup of the population.
But demographics aren't destiny. The employment-to-population ratios by age group, for example, have changed a lot since 1990. First, the women:
And now the men:
The rise in the percentage of 55-and-older Americans with jobs is striking, if not exactly surprising as people stay healthier longer and retirement benefits get skimpier. The biggest declines in employment-to-population ratios have been among 20- to 24-year-olds. That's partly due to people staying in school longer, which isn't necessarily a bad thing, but also seems to reflect a tougher job market for the young.
When you slice things by educational attainment , another interesting trend stands out:
On the whole, more education means you're more likely to be working. But while the employment-to-population ratio has declined among everybody with a high-school degree and up, those without high school diplomas are more likely to be working now than they were in the 1990s. I think immigration is the explanation. Immigrants are much more likely not to have completed high school than native-born Americans , and they're also more likely to have a job than native-born Americans.
When I last speculated as to why the U.S. employment-to-population ratio was so stubbornly low, a few readers berated me in the comments and in e-mails for not mentioning increased immigration as a possible cause. They're right that a big influx of working-age people from abroad might at least temporarily drive down the employment-to-population ratio. But the fact that immigration has probably been driving up the employment-to-population ratio of at least one segment of the population is an indication that its effects can be complex and surprising.
If you're looking for something to blame for the declining employment-to-population ratio, robots and workers abroad seem to be likelier culprits. Employment in manufacturing, which has been strongly affected by automation and overseas competition, fell from 17.3 million in February 2000 to 12.3 million last month. That's a lot of lost jobs for an economy to replace. Then there are the other possible causes that I trotted out last time:
Maybe the U.S. system of unemployment insurance and job retraining and placement is busted. Maybe the perverse incentives built into the Social Security Disability Insurance program are keeping people who could work out of the labor force. Maybe the U.S. educational system is doing an especially poor job of preparing people for work. Maybe increasing geographic divergence in employment in the U.S. is leaving job seekers stranded far from jobs. Maybe poor child-care options are keeping American women at home. Maybe U.S. corporations, under pressure from capital markets, are spending so much money on share buybacks that they’re underinvesting in labor.
There is at least one positive aspect to today's low employment-to-population ratio, though. It could mean that the U.S. economy can add lots and lots more jobs without much threat of inflation. Wages actually fell slightly in February, even as job gains surprised on the upside. That could be because there are still millions of Americans who would like to be working but aren't.
That's seasonally adjusted; unadjusted it was 59.4 percent.
Along those same lines, the y-axes on the charts in this column don't go to zero. I drew them that way because I didn't want to waste a lot of white space. Some people think that's misleading. I understand their argument, but still think they should get over it.
Active-duty military and people living in institutions (prisons, nursing homes, mental hospitals) aren't included.
These data series only go back to 1992, which is why this chart starts then instead of in 1990 like most of the others.
They're also more likely to have a Ph.D.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the author of this story:
Justin Fox at email@example.com
To contact the editor responsible for this story:
Zara Kessler at firstname.lastname@example.org