Posted on Tuesday, September 14, 2010
AMERICANS are used to thinking of their job market as lithe and supple. Employment snaps back quickly after recessions. Workers routinely shuttle between industries and cities to wherever jobs are abundant. But in the past decade, the labour market has resembled an ageing athlete. Each new injury is more painful and takes longer to heal. More than a year into the current economic recovery the unemployment rate remains stuck close to 10%, raising concerns about the kind of sclerosis that continental Europe suffered in the 1980s.
The slow rehabilitation is in part because the economy suffered a trauma, not a scrape. The fall in GDP during the last recession was easily the largest of the post-war period, and output remains well below its potential. Few had expected a rapid return to full employment, but even modest expectations for jobs growth have not been met. Employment has actually fallen since the end of the recession; and unemployment would be even higher than it is were it not for discouraged would-be jobseekers quitting the workforce. Some economists now fret that other barriers besides weak demand stand between workers and jobs, and that high unemployment is partly “structural” in nature.
The case begins with some kinks in recent data. Rising GDP has not led to the fall in unemployment predicted by Okun’s Law, established in the 1960s by Arthur Okun, an economist. The figures have also departed from the Beveridge curve (named for a British economist, William Beveridge) which relates job vacancies and the jobless rate. Unemployment has failed to fall in a way consistent with the increase in job openings.
Such deviations are perhaps too short-lived to be conclusive. But they jar because there are other reasons to believe that structural obstacles to jobs growth have risen. For instance, jobless benefits have been extended to 99 weeks in some states with high unemployment, compared with the usual limit of 26 weeks. Such payments provide crucial support to the long-term jobless and help to prop up aggregate demand. But they also push up the unemployment rate by discouraging workers from looking for jobs as assiduously as they might otherwise do.
It is easy to overstate the effects of this. A study from the Federal Reserve Bank of San Francisco* found that those eligible for extended benefits were unemployed for 1.6 weeks longer than those who could not claim support. This effect is equivalent to a rise of just 0.4 percentage points in the unemployment rate (though other studies suggest the effect may be somewhat larger). By the standards of continental Europe, jobless benefits in America are quite mean relative to wages.
A bigger worry is that jobseekers no longer have the skills demanded by employers. Half of the 8m jobs lost went in construction and manufacturing, and those departing these industries may struggle to adapt to jobs in more vibrant areas such as education and health services. The cost of this skills mismatch is compounded by America’s housing bust. Many owe more on mortgages than their homes are worth. Households often opt to stay put rather than default, leaving them trapped in places with high unemployment and unable to move to where jobs are plentiful. The rise of the two-income household has also made workers less mobile than they were: it is harder to move in search of jobs if there are two careers to consider.
How important are these factors? Very, says Narayana Kocherlakota, president of the Federal Reserve Bank of Minneapolis. He recently caused a stir by arguing that “most” of America’s unemployment is thanks to such mismatches, and hence not easily alleviated with looser monetary policy. Most American policymakers believe that structural joblessness has risen little, if at all.
One of the few concrete estimates comes from the IMF. A recent report compared the skill levels of the unemployed with indicators of the skills required by employers, to create state-level indices of mismatches. It used local mortgage-default and foreclosure figures to estimate geographical immobility. The results suggest that each of these factors acts to magnify the impact of the other. The authors conclude that, because of these rigidities, the unemployment rate consistent with stable inflation—roughly speaking, the structural rate—rose from around 5% in 2007 to between 6% and 6.75% by 2009.
A weak jobs recovery might push such estimates up further. The long-term unemployed—those who have been out of work for more than 26 weeks—now account for almost half of the jobless. A worry is that this cohort may become unemployable as their skills atrophy and they become increasingly detached from the informal networks that would lead them to new jobs.
Cutting out the middle men
The downturn may have accelerated a secular jobs trend that had been masked by the low- to middle-skilled work that the housing boom created. New technology has caused a hollowing-out and polarisation of the workforce. Employment growth for middle-skilled workers has steadily declined because of automation, but the productivity of high-skilled workers has boomed.
Sorting the cyclical from the structural is tricky; and excessively gloomy conclusions can be drawn in the midst of a downturn. The authors of the IMF study note that the structural barriers to employment growth may be transitory and could fade as the economy recovers. The rate at which workers leave unemployment is uniformly weak across sectors, suggesting that skills mismatch is not yet as important as weak demand in delaying hiring. The absence of price and wage pressures points to plenty of slack in the labour market. Yet it would be unwise for policymakers to hope that stronger demand will mop up all the jobless. America’s jobs market is not quite as flexible as it once was.
- the Economist