Employment is an increasingly lagging indicator, here is more data. Growth and productivity are humming along, but employment as a percentage of population fails to impress. Consider a few reasons why this might be:
1. Productivity growth is high, so employers are disinclined to hire more workers. The workers they already have are producing plenty. This explanation, of course, requires that firms have some market power, so they can’t sell all they want at prevailing market prices. There is then little point in hiring more workers, even if wages were to fall, because the extra output would be hard to sell at a good price. Of course if real demand were rising or otherwise robust, high productivity could cause firms to hire more workers rather than fewer.
This all may lead some of my astute readers to wonder what coordination failure is preventing Say’s Law from holding in this context. Does not supply create its own demand, at least barring a massive increase in the demand to hold money? Read Alex for more on this one.
2. The data may overstate the puzzle in the first place.
3. Maybe we are mismeasuring employment by focusing too much on the payroll survey and not enough on the household survey. But just try mentioning this one, and Brad DeLong will thwack you hard. Do note that Alan Greenspan agrees with Brad. The two measures are often quite different, but I have to go with the payroll survey here.
4. Health care costs per worker are extremely high. True, but this has been the case for some time.
5. Some parts of the Bush tax plan have induced firms to substitute capital expenditures for labor. And in the short run, capital and labor may be more substitutable, and less complementary, than in times past. Information technology can substitute for unskilled labor more than before.
6. Bush introduced a “tax cut for the rich,” rather than undertaking fiscal stimulus of the traditional Keynesian kind; read Brad on this one. I am less sanguine about traditional fiscal policy, plus this doesn’t explain why labor market recovery has been so slow, even if it might have been more rapid under different conditions.
7. Firms are outsourcing jobs overseas. The ever-excellent Daniel Drezner debunks this one, though I am not quite ready to assign it measure zero in effect.
8. For some unknown sociological or economic reason, workers are getting discouraged from seeking work more easily than in times past. They leave the labor force, which keeps the percentage of employed low but the unemployment rate low as well. I don’t rule this out, but I don’t have a contending hypothesis here either.
9. Firms are/were uncertain about terrorism, Iraq, corporate scandals, new information technologies, and sectoral shifts. Rather than locking in with hires, they will hold back and wait. But why would this affect labor more adversely than investment or stock prices?
10. Downward stickiness of wages. OK, this is what classical economics would lead you to expect, but in reality wages have become less sticky over time.
The bottom line: The conventional wisdom will opt for some combination of one and nine. Four, five, seven, and eight may explain further pieces of the puzzle. But I would sooner call the whole thing a continuing mystery. Note that most of these hypotheses imply that the economy can still become quite a bit better yet. Either Bush or Kerry will get credit for this, without deserving the plaudits.
Addendum: Arnold Kling adds to the mix.