Monday, March 7, 2011

Working Age Employment/Population Ratios

My preferred employment indicator is the employment/population ratio for working aged men (here age 25-54).  The most recent data (including the point for February released on Friday) is shown above, on a graph since 2000, showing the two most recent recessions, and one recovery between them, along with the sluggish sort-of-recovery since the great recession.  The last data point is pretty good, however, adding to the evidence that the US economy, at least, is starting to recover more strongly.

Here's the long term picture for that series (since 1948) for the context:

Basically, working-age men used to, well, work.  But not so much any more - the ratio has been declining for decades, ratcheting down in each recession, and then failing to fully recover before the next recession.  Presumably the pattern will continue, with this ratio, having dropped dramatically in the great recession, recovering a few percentage points before the next big drop (unless we get a big Middle East oil shock this year in which case it will drop again before it has a chance to recover much at all).

I take this long-term trend to be due a mixture of globalization, technological unemployment as we gradually approach the economic singularity, and displacement of men from the workforce by women.  We are now down to the point where almost 1 in 5 working aged men don't work.

I don't focus on male employment/population because I think women's employment is less important, but rather because it's less easy to interpret due to the major move of women into the workforce in the twentieth century.  Here's the long term picture for the equivalent female ratio:

You can see that from before the 1940s until the 1990s, working age women (also mothering-age women) were increasingly choosing to work.  It rather looks now as though that ratio has peaked and is declining.  Whether this is because women are starting to focus on raising children more again, or due to the same economic and technological forces that are affecting men, I can't say.  I have some anecdotal experience suggesting the former is happening, but probably the latter is important too.

At any rate, here is the shorter-term picture for women:

So far, the nascent employment recovery is less clear than in the male data.  Note the difference in scales here though - men lost about six percentage points of employment during the recession, whereas women only lost three and a half percentage points (hence the "mancession" term).  However, male employment seems to be more volatile on the upside as well as the downside.  If you look again at the two long-term pictures above, you can see that female employment is consistently less volatile through the business cycle than male.  Women are more apt to be employed in service sector jobs in health care, education, etc, and these industries are less cyclical than construction, manufacturing, etc that are more likely to employ men.


jemand said...

Looks like a lot of couples may have decided having one worker in a volatile but higher earning sector (most male jobs) and one worker in a secure but less lucrative position (most female jobs) is a smart way to manage the modern job market.

It may soon turn out this is the *only* way to survive, as wages in both sectors seem to be stagnating and yet food and fuel costs continue to rise...

Also, I wonder how many women *say* they are deciding to stay home for the kids, after failing to find a job for awhile, because it feels better and is a more socially acceptable explanation than just failing to be employable-- which hits one's psyche pretty badly. Unfortunately, men don't usually have this kind of social option for their lack of employment, which may end up straining relationships as these men deal with both intense internal stress and wider social disapproval. On the other hand, I wouldn't be surprised if these narratives soon change, so the disproportionate psychological effects from being unemployed may soon be ending.

Robert said...

Your posts are special, essentially beyond criticism. However, regarding this graph, I will note the following from the Wiki entry on Darrell Huff's 1954 book How to Lie With Statistics.

" ...It also shows how statistical graphs can be used to distort reality, for example by truncating the bottom of a line or bar chart, so that differences seem larger than they are..."

Robert said...

I should add that Huff's complaint was directed toward news sources catering to a less sophisticated audience.

Burk said...

Thanks for a fascinating post. One might offer a different hypothesis from the singularity or other ways to blame workers for not working, which is that jobs are scarcer and worse-paying. Mainstream economists think highly of the non-inflation accelerating unemployment rate (NAIRU), but don't have any good way to know what it is. They just think it exists, and is good for the economy, when in reality, our goal should be employment for everyone who wants to work.

The upshot is that our macroeconomic settings (interest rates, budget deficits, etc.) have all been biased to lower employment than necessary since the 70's, keeping an army of the unemployed in reserve. The effect has been not only to lower wages and wage expectations and lower the share of income going to labor, but it also has forced families to have two people in the work force rather than one, destroyed unions, and shifted power generally to corporations and management.

In markets, power goes to those doing the choosing, so small shifts in the numbers can cause big shifts in power and outcome.