The Jobs Statistics Trump Should Be Worried About

There is a "mess with jobs," but it's been decades in the making.

The pain is there if you look in the right place.

Photographer: Luke Sharrett/Bloomberg

President Donald Trump says lots of nutty, made-up things that invariably generate lots of appalled reactions from the news media and the wonkocracy. His statements about the unemployment rate have generated lots of appalled reactions, too. But, as I've written before, they're not entirely nutty or made-up. In fact, they address -- albeit in exaggerated Trumpian fashion -- a real measurement problem.

Here, for example, is the president in his much-discussed (and yes, in parts quite nutty) interview last week with Time's Michael Scherer:

I inherited a mess with jobs, despite the statistics, you know, my statistics are even better, but they are not the real statistics because you have millions of people that can’t get a job, OK.

OK! Although I wouldn't put it exactly that way. Here's what Trump should have said if he wanted to do more to court the crucial econowonk demographic:

Yes, the unemployment rate is low -- and going lower! But the job market still isn't in great shape, which the unemployment rate misses because it fails to count the millions of people who have given up looking for work. A better measure to focus on would be the prime-age employment-to-population ratio, or the Hornstein-Kudlyak-Lange Non-Employment Index. Did I pronounce Kudlyak right?

The latter measure was devised in 2014 by Andreas Hornstein of the Federal Reserve Bank of Richmond, Marianna Kudlyak of the Federal Reserve Bank of San Francisco and Fabian Lange of Montreal's McGill University in an attempt to better reflect the role played by labor-force dropouts. These are the people who tell government survey-takers that they neither have a job nor have looked for one in the previous four weeks, and they don't show up at all in the headline unemployment rate. Those who haven't looked for a job in the past four weeks but have in the past 12 months are considered "marginally attached to the labor force" and show up in alternative Bureau of Labor Statistics measures, such as U-6 unemployment, which includes discouraged workers and involuntary part-timers. But all other non-job-holders are omitted from the calculations.

This is not the result of some sneaky plot by recent administrations to make the numbers look better. It's basically the way unemployment has been calculated since the 1940s. 1 The reasoning is that you don't want to count as unemployed people who are extremely unlikely to join the labor market in any case, because that would give employers and economic policy makers a skewed picture of how much slack there is in the labor market. Yes, counting the labor-force dropouts would also invariably make things look worse -- I'm not denying that there's some public-relations thinking at work here, just that there's anything new about it. What is new is how many labor-force dropouts there are, especially male ones, although after rising for decades the female labor-force participation rate has fallen since 2000 as well.

Lots of these dropouts do get jobs eventually. As Hornstein, Kulyak and Lange wrote in 2014:

In the U.S. labor market unemployed individuals that are actively looking for work are more than three times as likely to become employed as those individuals that are not actively looking for work and are considered to be out of the labor force (OLF). Yet, on average, every month twice as many people make the transition from OLF to employment than do from unemployment to employment. These observations on labor market transitions suggest that the standard unemployment rate and its extensions proposed by the Bureau of Labor Statistics are both too coarse and too narrow as measures of resource utilization in the labor market. These measures are too narrow since they exclude a large part of the population that is potentially employable, and they are too coarse since they assume the same labor force attachment for all nonemployed individuals.

So they set out to create a measure that includes the labor-force dropouts, weighted to reflect the propensity of the different subgroups among them (disabled, retired, in school, not in school, discouraged, etc.) to return to the workforce. 2 The latest edition of their non-employment index came out Tuesday, and it showed a rate of 8.2 percent for February. Add in people working part time for economic reasons (aka involuntary part-time workers), and it was 9.3 percent. That compares with a headline unemployment rate of 4.7 percent, which is interesting. Also interesting is how this compares with past non-employment rates -- which we can do because Hornstein, Kulyak and Lange back-filled their index using data going back to 1994.

Another Way to Measure Unemployment

The Hornstein-Kudlyak-Lange Non-Employment Index

Source: Federal Reserve Bank of Richmond

The non-employment index has fallen a lot since the recession, although it hasn't fallen as fast as the unemployment rate. It's still somewhat above its lows of 2007 and 2000, but that's true of the unemployment rate, too. So ... a different picture, but not one that shows all that much more evidence of a "mess with jobs," historically speaking, than the unemployment rate.

But maybe you want even more historical sweep and don't care about all those fancy weightings. That's where the prime-age employment-to-population ratio comes in. It's just the number of employed people, divided by the number of civilians aged 25 through 54 (prisoners and uniformed military aren't part of the calculations):

Prime-Age Employment Still Lags

Employment-to-population ratio, ages 25 through 54

Source: Bureau of Labor Statistics

Note: Y-axis doesn't go to zero.

Here you can see that there are still a lot more people not working than during the economic expansions of the 2000s, 1990s and 1980s. It's a little hard to compare going back further than that, though, because the great wave of women entering the paid workforce swamps almost everything else. So let's just look at men:

Men Not at Work

Male employment-to-population ratio, ages 25 through 54

Source: Bureau of Labor Statistics

Note: Y-axis doesn't go to zero.

In March 1953, only 4 percent of prime-age men didn't have jobs. In February, that figure was 14.7 percent -- which is a lot better than the nearly 20 percent rate at the depths of the last recession but still an awful lot of men without paid work. Men staying in school longer account for some of this. So do stay-at-home dads and very early retirees. But mostly these men have dropped out of the labor force for other, unhappier reasons, as Nicholas Eberstadt recounted in his recent book "Men Without Work: America's Invisible Crisis." 3 I think it's fair to characterize this as "a mess with jobs" -- although it's a mess that's been many decades in the making, and I doubt President Trump really knows what to do about it.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

  1. There was a tightening up of how marginal labor-force attachment is determined in 1994 that had the effect of reducing extended unemployment measures, but it had no impact on the main unemployment rate.

  2. From their 2014 paper: "Our proposed NEI is a weighted average of the population shares of the various subgroups among the unemployed and OLF, where the weight for each subgroup is given by the sample average of its employment transition rate relative to the group with the highest transition rate."

  3. Which I have written about before. Repeatedly.

To contact the author of this story:
Justin Fox at

To contact the editor responsible for this story:
Brooke Sample at

Before it's here, it's on the Bloomberg Terminal.