High Unemployment: Cyclical or Structural?

Unemployment is edging down, but the fraction of the unemployed who have been idle more than a year remains near a record high. In the first three months of the year about 30% of the unemployed – more than 3.9 million people – had been jobless for a year or longer. Before the Great Recession the post-war record for long-term unemployment was set in Ronald Reagan’s first term. Even at the worst stage of that slump, however, only about 14% of the unemployed were out of work for more than a year.

The increase in long-term unemployment has raised the specter of a permanent jump in the unemployment rate, one that is linked to a surge in structural unemployment. Most economists distinguish between cyclical unemployment, which rises and falls over the business cycle, and structural unemployment, which persists even when the economy has been expanding for years. Because high unemployment has lasted so long, many observers worry that structural unemployment has now become deeply embedded in the nation’s job market.

If true, what would this mean? It would imply that traditional remedies for high joblessness have become less effective. Congress cannot accelerate the pace of the recovery by spending more or taxing less. The Federal Reserve cannot safely push the unemployment rate down to 4½%, using either standard or nonstandard monetary policies. Minneapolis Fed president Narayana Kocherlakota has suggested structural unemployment is now so high the Federal Reserve will have to worry about inflationary pressures when unemployment dips into the range between 6½% and 8%. This represents a major turnaround compared with most of the past two decades. In the late 1990s and in the years just before the Great Recession, inflationary pressures were not very noticeable when the unemployment rate was well below 6%.

Inflation would be reignited at a higher unemployment rate if the long-term unemployed are not fully available to fill new jobs. Structurally unemployed workers find it harder to land suitable jobs compared with the other unemployed. Their skills make them a poor match for current job openings. A couple of theories can account for the transformation of regular unemployed workers into the structurally unemployed. One emphasizes the deterioration in worker skill or morale that is caused by a long unemployment spell. A second highlights the attitude of employers toward the job applicants who’ve been idle a long time. Employers may regard the long-term unemployed as poor job candidates, even if they have as much schooling and work experience as other job candidates in the unemployment queue.

There is evidence from Europe that the long-term unemployed exert a smaller restraining effect on inflationary pressures compared with workers who have just been laid off. Interestingly, the evidence for this in the United States is much weaker. If America’s long-term unemployed exert a smaller effect than the newly jobless in curbing inflation, the effect is hard to see in the statistics.

At the moment, the critical issue is whether structural unemployment represents a significant fraction of total unemployment. Long-term unemployment is clearly a big part of our job market problem, but is long-term joblessness the same as structural unemployment? At the moment, the best answer is “No.”

Some of the long-term unemployed are certainly poorly equipped to fill job vacancies. But most workers who have been idle a year or longer have the skills needed to fill at least some jobs. Between March 2007 and March 2011 the number of Americans unemployed longer than a year increased by 3 million. More than 28% of the additional long-term unemployed had a college or post-college degree. The reason most of the long-term unemployed have not found jobs is that employers have had few jobs on offer since the unemployed got their layoff slips.

At the beginning of the last decade there were 1.5 unemployed job seekers for each job opening. In the 2001 recession the ratio of job seekers to job vacancies increased to about 2.8 before dropping back to 1.5 in 2007. The Great Recession made it much harder for job seekers to find work. The ratio of unemployed workers to job vacancies soared above 6:1 in late 2009, and it has only declined to 4:1 in the past year. If job seekers are much more numerous than job openings, the unemployed have much lower chances of finding work in a given week. A short spell out of work is much more likely to metastasize into a long spell of unemployment.

In a labor market where job seekers greatly outnumber job openings, we cannot be astonished when long-term unemployment rises. Even if we could magically endow all job seekers with the precisely the skills needed to find work in expanding industries, we would not have reduced the unemployment rate in the Great Recession very much, and for a very simple reason: There were comparatively few job vacancies to fill.

Employers tend to favor new school graduates or the newly laid off over the long-term unemployed. This makes it tougher for workers in long spells of unemployment to find a job. It does not mean the long-term unemployed are structurally unemployed – or unemployable.

To be sure, many employers complain that job seekers lack the skills needed to fill new job openings. We should take these complaints with a grain of salt. It is always cheaper to hire an applicant who has precisely the skills needed to do a given job. But how often does a perfect match occur? Employers can reasonably expect qualified applicants to have the general background and educational attainment needed to do a good job, and they can hope applicants will already have experience in a very similar job. In most cases, however, the specific skills needed for the position will be learned on the job. Employers should expect to pay for a sizeable percentage of on-the-job training costs, and feel lucky if they don’t have to.

Good evidence of skills mismatch and rising structural unemployment should be plainly evident in the wage statistics. If the skills of the nation’s job seekers diverged wildly from the skill needs of growing employers we would expect to see a spike in wages in those occupations where skill shortages are most acute. I have yet to see this kind of evidence in the wage data. Until we find it we should skeptical of claims that today’s excess unemployment is mainly traceable to structural rather than cyclical joblessness.