How Many People Should Be Working in America?

Jim Hamilton of UCSD--one of the greatest of the masters of econometric time-series analysis--writes:

Econbrowser: How many people should be working in America?: Quite a few commentators have suggested that the labor force participation rate is a much better indicator of the health of the U.S. labor market than is the unemployment rate. I feel that quite a few commentators have this wrong.... The issue... is the divergent impressions one would draw about the current U.S. labor market depending on which statistic you look at. The unemployment rate, which reflects the number of people who say they're looking for work but are unable to find it, suggests that the U.S. economy is currently in significantly better shape than it's typically been.... The labor force participation rate, which is the number of people who are either working or looking for work as a fraction of the population, suggests that far fewer people are working, relative to the size of the potential labor force, than has typically been the case.... The difference between the two measures has to do with people who aren't working and who further described themselves to the BLS as not actively looking for a job. Large numbers of such people explain both a low unemployment rate and a low labor force participation rate....

It seems very hard to quarrel with the statement that any given month's value for the labor force participation represents the confluence of different factors, some resulting from trends that are in all probability quite benign, and some representing cyclical economic swings. Whenever somebody looks at such a statistic and claims to have inferred what it is saying about purely cyclical forces, they must have used some method for distinguishing between these two kinds of forces.

In my opinion, no such reliable method exists. One idea you might try would be to fit a linear time trend to the data, calling any deviation from that trend line the "cyclical" factor.... [But t]here's no good reason to assume that the factors that are contributing to the trend are in fact evolving in a linear, purely deterministic way....

The method that Bradbury used in order to arrive at her lowest estimate, 1.6 million, of the number of missing jobs... amounts to assuming that the slope of a linear trend fit from 1960-1994 could be extrapolated to 2001-2005 to identify the magnitude that we should normally be expecting for that figure. In the case of mature men, that's maybe not such a bad assumption.... On the other hand, for women aged 35-44, this amounts to assuming that the increase in women's labor force participation rates between 1960 and 1994 should have continued to climb upward, and, since it has not, Bradbury finds 1.1 million "missing" jobs in this group alone....

My point is not that there's a right way and a wrong way to control for trends, but rather that there are some fundamental problems with any way you choose to do it, and any conclusions you draw from these exercises need to be very carefully qualified...

One way to gain more information about what is "trend" and what is "cycle" is to take a look at other time series indicators that we believe have a similar cyclical component. When the labor market is cyclically weak, we believe that the unemployment rate will be higher than trend, that the employment-to-population ratio will be lower than trend, that average hours worked will be relatively low (since firms are likely to cut back both on bodies and on hours when their demand for labor is weak), and that the average duration of unemployment will be relatively high (because more of the fluctuations in quits, firings, and hires that drive the employment side of the business cycle are on the hires side). What do these series look like? Here they are:

The issue at stake is essentially whether the past five years have seen a stable value for the trend unemployment rate and a fall in the underlying trend for the employment-to-population ratio (meaning that our current employment-to-population ratio is actually close to, not far below, the long-run trend) or whether the past five years have seen a stable value for the trend employment-to-population ratio and a fall in the underlying trend value for the unemployment rate. The weekly hours series and the unemployment spell duration series seem to vote with the employment-to-population ratio: three series seem to say that the current cycle component is large, that there has been only a smell recovery from the business cycle trough levels, and that we are still pretty far away from full employment. Only one indicator--the unemployment rate--seems to say that recovery is well advanced and that the cyclical component has substantially shrunk.

This, however, doesn't resolve the mystery: why is the indicator that is the unemployment rate giving a different signal? What has happened to keep workers whom we would have expected--given the behavior of unemployment spell duration, hours, and the employment-to-population ratio--to say that they are unemployed from saying so when the CPS interviewers come to call?

Any Berkeley seniors who know (or are confident they can quickly learn) some time series econometrics and are interested in writing a macro-labor thesis, drop me a note. Some serious quantitative work on this would be genuinely useful...

Comments

[Nuk]

One thing I am interested in knowing is how the data reflects the demographic shift in the workplace. It seems to me that change in family incomes from the one worker income to the two worker income has only intensified in the last twenty odd years. Along that lines, it would be interesting to see the total family income plotted against the unemployment rate or the duration.

[Tim B.]

Brad, can you add the unemployment series to the employment/population series? This would give us a "people interested in working" series - I'd be curious what the trendline is on that...

Also, the rise in two-income families probably has a great deal of impact on the average length of unemployment. In general, if you just lost 100% of your income, you'll take any job to get a percentage of it back; on the otherhand, if you lost 50% of your income, you can probably afford to wait it out for a bit and look for a good job and you're less likely to move to a job because the other income-earner presumably works near your original location.

Anecdotally, my parents faced the same dillema the last time my father was laid-off - my father got a job in NJ very quickly but decided not to take it because that would mean my mother would have to leave her job and they would be back to one income for a period of time anyways, so they waited it out for 5-6 months and until father got another job more locally.

Tim

[ken melvin]

Couple or three things bothering me about the graphs:

How is it that the homeless who appeared in the eighties didn't go away when unemployment went down? Seems similar the phenomenon of the sixties and seventies where people who lost their jobs to automation in the rust belt and the mechanical cotton picker in the south were still around but somehow disappeared from the data roles.

Does the average work week data include part time workers? To my observation the hours worked per week by full time employees have gone up.

Where's a graph of the ratio of salaries or wages paid to GDP or national income?

[James Wimberley]

The way Brad puts the question at the end of his post discourages ameateurs from responding, but here goes anyway. Wages are low in the USA today, compared to historical experience, and especially for un- and semi-skilled jobs at the margin. Perhaps there's a perverse supply curve - more people have to take jobs (even several jobs) to make ends meet. But if you can get by without - for example, a housewife with an employed partner - low wages make it unattractive to bother. People aren't stupid; the housewife might be willing to work at higher wages, but the way things are there's no point in looking.
This split between those who have to work and those who don't would show up in marked differences in the survey response by family status.

[spencer]

One series from census where you might find an answer to part of the noncyclical component is the series on wife working. From the 1950s till the 1990s this series showed steady growth as it rose from just over 20% of families in the 1950s to almost 50% of families in the 1990s. But for the last 5 plus years this series has flatten out and quit growing.

This may reflect several factors, but part of the explanation may be that with weak wage growth and strong increases in child care costs it is no longer
so advantagous for the wife to bring in a second income.
By the time you deduct all the costs of the wife working--child care, clothing and other expenses, higher taxes, etc. --- the net addition from the second income is not that significant. Consequently, we are seeing a peaking of wife working. This may be the most important non-cyclical explanation of the lower participation rate.

[hrned]

Sure, this bears serious investigation to find out what's really going on. My hypothesis: many white collar baby boomers displaced in last 5 years are still without normal jobs, but are living on savings or investing in real estate or working as "consultants" when they can.

[hrned]

Sure, this bears serious investigation to find out what's really going on. My hypothesis: many white collar baby boomers displaced in last 5 years are still without normal jobs, but are living on savings or investing in real estate or working as "consultants" when they can.

[Barry]

Psnecer, that's a good explanation, except that there's an obvious cyclical component for that - in a good labor market, it's easier to find work which pays well, is not too bad, etc. In a bad labor market, it's harder.

[save_the_rustbelt]

"Sure, this bears serious investigation to find out what's really going on. My hypothesis: many white collar baby boomers displaced in last 5 years are still without normal jobs, but are living on savings or investing in real estate or working as "consultants" when they can."

Anecdotally I can tell you this is very true. Drop by any Home Depot and talk to the 50-somethings selling paint.

I'm not up to the detailed statistical analysis many of you are, but being in the trenches - this labor market is very weak in many areas.

Any good work on "underemployment" out there?

Definition of a Consultant: an unemployed white guy who owns three white shirts and a decent briefcase

Ok, so I'm not a comedian.

[A Harless]

Brad,

I think you're understating the case by looking just at BLS data. The most striking indicator is the Conference Board Help Wanted Index. It fell from a cyclical high of 93 in Feb. 1999 to a (30-year) low of 36 in May 2003, and it stands at 38 as of June 2005. In part this reflects the increased use of online advertising as opposed to newspapers, but that trend started long before the peak, and it would have to be an implausibly rapid shift to alter the conclusion that businesses are not looking very hard for workers. (I've done some work on adjusting the HWI using market share data and the Monster Employment Index, which tracks online advertising, and the result indicates that overall recruitment efforts over the last 2 years have barely increased relative to the size of the existing stock of jobs.)

Another interesting piece of information from the Conference Board is the response to questions about job availability in their Consumer Confidence survey. In Mar. 2000, 53.3% reported jobs "plentiful" and only 10.6% "hard to get". The trough was in September 2003, with 9.9% "plentiful" and 35.1% "hard to get". Today the two are about even at around 23%, somewhere in the middle, but closer to the bottom than the top.

I'll also offer an anecdote. A friend of mine -- a software engineer with over a decade of experience -- joined the Peace Corps last year following an unsuccessful job search after losing his job during the tech meltdown. While I doubt that the Peace Corps per se has much to do with the low unemployment rate, I suspect that there are a lot of people like him who really want jobs but who have made commitments that keep them out of the domestic labor force. In all likelihood, he'll be back next year when his Peace Corps enlistment expires, and I expect we'll see a lot of others filtering back over the next few years as well (assuming that the recovery continues).

[Chris R]

This is the million-dollar question that's vexed RBC types. As Prof. Hamilton points out, our assumptions about the nature of the trend matter quite a bit--random walk with drift, trend-stationarity, covariance-stationarity, whatever. In the case of hours worked (or of TFP for that matter) we don't have much clear guidance from the data. This is an ongoing research issue that has led to many an argument between Valerie Ramey and Larry Christiano.

The way that I read the charts in the post, unemployment and employment/population levels look very 1995-1997 (as does U-6 if you go to the BLS's website). Should we detrend these things? I don't think so (see unemployment levels in the 1950s and 1960s) but reasonable people can differ. Darned low-frequency movements. Participation rates (not shown here) do seem to show some softness and might have some trends of their own. So far, we have two indicators indicating some strength and one indicating moderate weakness in the labor market.

Hours/workweek is an interesting one. I just don't know what to make of it; it has had a long term downward trend since the nineteenth century as people don't have to spend all their time working in order to survive. Is it a unit root process with some drift? Is it trend-stationary? Is it even measured properly? That's a tough call. I just don't know what to make of this.

Unemployment duration looks pretty unambiguously bad. We're at early 1980s or early 1990s levels here. Why don't the 1970s and early 80s look extraordinarly bad on this chart, as a poster pointed out? Measurement? Still, I'd say that this indicates weakness.

So we have two labor market indicators showing strength, one showing ambiguity, and two showing weakness. If I had to do some Bayesian averaging, I'd say that the labor market is much improved over 2001-2003 but still not at its peak.

Getting more precise would be an awfully big project for an undergraduate since there are so many unsolved issues, and the econometrics can get hairy. I can recommend a textbook for reference....

[A Harless]

If you make no attempt at all to detrend the series, you're left (if you believe in a Phillips curve or something like it) with a puzzle as to why inflation was still so low at the end of the millennium. With UE rate and Emp/Pop both at 25-year extremes, the economy should have been boiling over. (UE duration is interesting: it is a lagging indicator, so its information not timely, but it has had a much more stable relationship with inflation than the other indicators.)

[Chris R]

Agreed. At any rate, Brad, I like this brand of post. It elevates the discussion of these topics.

[Maurice Lanselle]

Considering duration and UE rate together to get a turnover rate (akin to the "plentiful/rare" mentioned above), one notes that the current 4.5% x (50wks/20wks) corresponds to less than 12% turnover, whereas four years ago the rate was about 4% x (50wks/12.5wks) = 16%. Opportunities are not as plentiful?

There is the obvious issue of obsolescence, mis-matching, and retraining: in tough or uncertain times, employers often seek "proven" skills, which makes changes "career changes" difficult or impossible to negotiate. With the apparently low number of hours worked, employers can be patient in the hiring search and take up the slack with those employed (plus temps), can't they?

Some definitions of the measures (or pointers to such defs) would help. Is unemployment a percent of population or of (employed + unemployed), as is generally the case? If the former, then the sum seems to be constant at ca. 68 for the past twenty years. Also, using log scale is often helpful when eye-balling time series.

Are "hours worked" per person or per job? How are people with multiple jobs counted? Multiple part-time jobs? Are "employed" people with one or more jobs, individuals paid for labor, or equivalent full-time jobs (as calculated by HR software)? Add to the "consultants" the structural impact of temping, outsourcing, etc and the measurments get really iffy.