Unemployment in America: Still a U, not a V

The Economist

RETAIL sales, manufacturing activity and stock prices all show the American economy shook off its mid-2010 lethargy and picked up a head of steam as the year closed. But to anyone who thought a V-shaped recovery was taking hold, the December employment report is a sobering reality check. Non-farm payrolls rose by 103,000, or 0.1%. Private payrolls, a better gauge of the economy’s underlying momentum, rose 113,000; declining state and local government employment held back the total. That was rather deflating after a private tally on Wednesday predicted a gain of as much as 300,000.

The Labor Department revised up the totals for October and November by a combined 70,000, so the total report is mildly encouraging. Private employment has grown an average of 126,000 per month since July, comfortably above the roughly 75,000 to 100,000 needed to keep up with growth in the working-age population. (The government numbers have been heavily distorted by hiring and firing for the federal census.) But growth of more than twice that pace is typical of recoveries. This morning, Ben Bernanke, the Federal Reserve chairman, told Congress it could take four or five years "for the job market to normalize fully."

There was a bolt of unexpected good news in the unemployment rate, which plunged to 9.4% from 9.8%. It’s common for payrolls and the unemployment rate to behave differently because they’re drawn from different surveys (the first from a large sample of employers, the second from a much smaller sample of households). The unemployment rate dropped because household employment rose 297,000 while the number of unemployed dropped 556,000, meaning the overall labour force shrank. Such big changes to household employment and unemployment are routine and should be discounted. The unemployment rate, though, is quite stable, and a 0.4% drop is unusual, the sharpest since April, 1998.

It does not suggest the underlying economy has suddenly picked up momentum. If anything, other indicators from the household survey are frustratingly weak. An improving job market usually draws a flood of job hunters into the labour force, raising the participation rate—the share of the working-age population either working or looking for work. Instead, the participation rate fell to 64.3% in December, the lowest since 1984. True, the average work week ticked up to 33.6 hours, but evidence of latent demand for labour remains scant.

So why did the unemployment rate drop so much? Probably because it was inexplicably high to start with. A traditional macroeconomic rule of thumb known as Okun’s Law predicts that, given the performance of gross domestic product in the last few years, the unemployment rate should only be around 8.5%, reckons Alan Krueger of Princeton University. Why it reached 10% remains a mystery. In any case, December’s drop may be part of a long overdue return to a more normal level. Sometimes, the most powerful explanation for economic behaviour is reversion to the mean.