Chart of the day, employment edition

By Felix Salmon
December 2, 2011
employment report counts as a win for the White House.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">


Today’s employment report counts as a win for the White House. Markets care about payrolls; politicians care about unemployment. And so does the country as a whole: the severity with which the BLS website crashes on the first Friday of the month is a direct function of the change in the headline unemployment rate, which, wonderfully, fell to just 8.6% last month.

But while numbers matter in election campaigns, it’s people who vote. And they vote based on their own personal experiences, rather than on macroeconomic statistics. And the fact is that if you’re a person in America, the likelihood that you have a job was unchanged this month: the employment-to-population ratio ticked up just one tenth of one percentage point.

When employed people become unemployed, that’s bad news, and immediately visible in the unemployment rate. When unemployed people leave the labor force entirely, that’s equally bad news, but it’s a tougher measure for the public to connect with, since at that point they’re no longer counted in the unemployment rate. Everybody knows what “unemployment” is; the population which cares about the “employment-to-population ratio”, by contrast, is wholly comprised of wonks.

The plunge in the employment-to-population ratio over the course of the Great Recession is going to be its biggest and most lasting legacy. We’re now back to the levels last seen in the days before most women worked, but we live in a very different world now. In the late 1970s, a woman without a job was much less likely to consider herself unemployed than in the early 2010s. And when she casts her vote in November, the degree to which she’s happy or unhappy with the current administration is going to be much more connected to her actual employment status than it is to whether she’s officially showing up in the unemployment rolls.

Over the next few months, we’ll get a better sense of the signal-to-noise ratio in the 8.6% number. I’m hopeful that we’ve seen the last 9 handle in the headline unemployment data series, and if I’m right, then the optics of the unemployment rate are, at the margin, good for Ds and bad for Rs. But the unemployment rate is not a particularly good gauge of how well the economy is functioning, or how many people have jobs. And I’m very pessimistic that the employment-to-population ratio is going to get back above 60% even over the medium term. It’s certainly not going to get there before the election.

It’s no coincidence that after the employment-to-population ratio plunged, we saw the rise of first the Tea Party and then Occupy Wall Street. Social unrest will not go away so long as more than 40% of the US population is jobless. Some part of that 41.5% is too young to work; some part is too old; some part is in jail; and some minuscule part is simply too rich to care. But at heart, it’s a massive dead weight dragging down the hopes and prospects of hundreds of millions of Americans. And they’re not going to take it quietly.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Felix, you do realize that the employment-to-population ratio for OECD Europe **FOR PERSONS AGE 15-64** didn’t exceed 60% until the late 90s? Right? Was that a “massive dead weight dragging down the hopes and prospects of hundreds of millions” of Europeans? Or are you indulging in simple hyperbole?

Given that your chart (measuring employment as a proportion of the ENTIRE population) has never exceeded 65% in the history of the country, and given that it was below the present levels throughout the 1970s, we’re surely talking about less than 10% of the population that really wants a job and doesn’t have one. This isn’t to say that we don’t have a problem — but it isn’t nearly as large a problem as you suggest. Most of the people who don’t have a job are truly not in the job market (such as my 5 year old son or my 75 year old aunt).

Posted by TFF | Report as abusive

If you want to see employment to population charts for age segments, they are in the link below.

Of the seven age segments, guess which one employment to population segment is higher then it was before the recession? ployment-trends-part-1-the-age-of-worker s/

Posted by david3 | Report as abusive

As Baby Boomers retire, this number is going to keep on getting bigger, and it’ll have nothing to do with economics or politics, just age.

One could say that the lack of demand in the global economy has a lot to do with retiring Baby Boomers incomes falling as they move from Employment to Retirement, from consumption to deleveraging.

Demographically, if the West doesn’t find some young consumers to replace these Baby Boomers, we’re going to have to adapt to a situation not unlike that of Japan. Ironically, the loudest shouts from the right on both sides of the pond are to ban immigration, but accepting these new, young workers could be one of the best ways to ensure prosperity for the rest of us in the decades ahead.

Posted by FifthDecade | Report as abusive

TFF, look at the graph. In the 2008 panic millions of people lost their jobs. It wasn’t a sudden epidemic of 5 year olds or 75 year olds coming into the population. Since then it hasn’t come back up. This means that the people who became unemployed are unable to find work, not that 5 year olds and 75 year olds are refusing to be drawn into the labor market. No one is saying that the employment-population ratio should be 100%. The important thing is the sudden changes, or lack thereof.

The reason the employment/population ratio is useful is that it avoids the problem that the unemployment rate has, namely that it goes down when people get so discouraged that they leave the labor market.

Compared to every other post-WW2 recession, the plunge was deeper. And 4 years in it hasn’t started recovering yet.

Posted by jack512 | Report as abusive

> And they vote based on their own personal experiences, rather than on macroeconomic statistics.

Surprisingly enough, this is false. I have a friend in the NYU politics department who told me that his prior was what you indicate, but that when he saw the evidence there was simply no doubting that beliefs about the status of the general economy have much stronger effects on voting behavior than do assessments of one’s own economic situation.

Posted by dWj | Report as abusive

Also, some part of that population does seem, anecdotally, to be moves back toward single-family households for reasons that aren’t entirely economic. The secular trends in this number I would attribute in large part to benign social changes, but the high-frequency changes no doubt are driven by the economy; I’m reluctant to compare today’s number to that in the late seventies, but the fact that it’s been stuck in the 58-59 range for the past couple years after a large plunge is more meaningful to me.

Posted by dWj | Report as abusive

“The important thing is the sudden changes, or lack thereof.”

Sure, I agree. It is a serious issue. But jack, LOOK AT THE GRAPH! The dip from the peak to the valley is under 7%. It is irresponsible for Felix to toss off percentages in the 40s in an attempt to exaggerate the magnitude of the problem.

Posted by TFF | Report as abusive

“Of the seven age segments, guess which one employment to population segment is higher then it was before the recession?”

Economic necessity, in many cases, though it is also part of a broader trend. And they are holding on to jobs that the younger generations desperately need.

Posted by TFF | Report as abusive

I have to echo TFF here that this chart, by itself, doesn’t tell you much. Clearly there is an issue when the recession hit, but there was also a drop in employment. But employment to population ratios include all sorts of other trends, such as people going back to school, retiring, having maternity leave, etc. If Felix wants to wonk out, he should actually wonk out. Or he could make his point much more simply by showing the change in different categories of unemployment.

Posted by MKCurious | Report as abusive

Demographics. Baby Boomers were mostly born between 1946 and 1950. Add 60-65 and you get 2006 to 2015. The start of that rush to retirement came one year before the beginning of the crash and the midpoint is 2011.

That seems too much of a coincidence to me. Even if the demographics aren’t causal, the correlation between these events must be pretty high, and the effects magnifying.

Posted by FifthDecade | Report as abusive

When Salmon says that social unrest will not go away so long as more than 40% of the US population is jobless, we should assume he means in contrast to when the economy was at full employment, such as in 2000, when the corresponding number was less than 36%, or 2007, when it was less than 37%.

Posted by jack512 | Report as abusive

If that is his point, jack, then why doesn’t he phrase it that way? Felix could have said, “We need jobs for another 5% of the population to reach full employment and eliminate unrest.” That would have been quite clear. Instead he lumped that 5% in with the 35%+ of the population that doesn’t WANT a job.

Besides, I’m skeptical of the connection being made. The percentage of the population employed has changed dramatically over time, and there is no good reason to believe the “natural” figure is above 60%. Europe has lived for decades with an employment rate below that, and they don’t have any more bomb-throwing revolutionaries than we do. As the Baby Boomers age, we will likely see our own natural employment rate fall as well.

I do agree with your earlier point that the sharp DROP in employment is more significant than the number employed.

Posted by TFF | Report as abusive

The most imporntant issue in the post is that the population of workers to non-workers has droped meaningfully.

Fewer people providing for more people virtually insures a lower level of prosperity.

Best hopes for more people finding jobs they like working at well into their golden years!

Posted by y2kurtus | Report as abusive

The employment-population ratio for prime age workers, which is not affected by demography, shows the same non-recovery of employment:  /02/meh-and-i-say-that-with-feeling/

Posted by jack512 | Report as abusive

Jack, that isn’t the graph that Felix was citing. You cannot defend an invalid argument by saying, “Well, the conclusion is correct.”

y2kurtus has a better interpretation of the data when he writes, “The most important issue in the post is that the population of workers to non-workers has dropped meaningfully. Fewer people providing for more people virtually insures a lower level of prosperity.” Yet another thing that Felix COULD have said instead of trumpeting the horrors of an economy in which fewer than 60% are employed.

I will note that Krugman’s graph shows a drop from 80% to 75% of the “prime age” population holding a job. A 5% decline, as I suggested earlier. Trumpeting the 40% figure is hyperbole, and especially pointless when you realize that it is STILL a higher employment ratio than Europe has ever had. It is the fact of the change that is significant, not the absolute number.

Demographics encompasses more than age distribution. In a broader sense it also includes household structure. This article suggests that more women are leaving the workforce (perhaps to focus more on family and household) nomy/men_women_jobs/index.htm

I can imagine plenty of other interpretations of that data, and note that job losses for men are still double those for women over the last decade, but it does illustrate that this is a more complex issue than can possibly be described by employment/population.

Posted by TFF | Report as abusive

Quick question — do these figures include ALL employment? Or just full-time employment? Do they count the paper route and the 10-hour-week job I held at a supermarket in high school?

Reason I ask is that the numbers seem high if counting just full-time employment. If I work from 25 to 65 and die at 80, then I would be employed for 50% of my life. I realize that is just one life-path out of many, and that this is a single time-slice of the population rather than a longitudinal study, but I’m still not sure how you would ever get a figure as high as 65%.

If including part-time employment, then it is hard to interpret the data sensibly. Part-time employment and full-time employment simply aren’t equivalent.

Posted by TFF | Report as abusive

@TFF: It includes all employment, but the denominator is only working-age adults. (Something like 18-65, I think.)

Posted by saralonde | Report as abusive

TFF raises a good point – as many full time jobs have been replaced with part time jobs, and many people have had to double up and do more than one part time job to make enough to live on, are these workers counted twice?

Posted by FifthDecade | Report as abusive

saralonde, as best I can tell they are counting civilian non-institutional population age 16+.

FifthDecade, if you read their “technical notes” they explain that there are two separate reports — in the household survey, each individual is counted once. In the establishment survey, they will be double-counted.

Posted by TFF | Report as abusive