Jobless Rate Is Not the New Normal

By CHRISTINA D. ROMER

Published: April 9, 2011

IT takes a while to match people with jobs, so even in normal times the unemployment rate is never zero. Before the recent recession, in the view of many economists, the lowest sustainable rate of unemployment in the United States — a level known as the normal or natural rate of unemployment — was around 5 percent.

David G. Klein

Multimedia

 Weekend Business

The turmoil of the last few years, however, has shaken up the economy. Is it possible that it has affected the natural rate of unemployment — increasing it to 8 or even 9 percent? Such a climb would imply that the prospects for a rebound in output and employment have been greatly reduced — and that high unemployment would be our new normal.

This is implicitly the view of some Federal Reserve policy makers, who say that there is nothing more the central bank can do to lower unemployment. And it’s the view of those who say “structural” factors are the main cause of our current high unemployment, which stood at 8.8 percent in March.

Fortunately, there is a more compelling explanation. Strong evidence suggests that the natural rate of unemployment actually hasn’t risen very much. Instead, the elevated unemployment rate appears to reflect mainly cyclical factors, particularly a lingering shortfall in consumer spending and business investment.

Consider the effects of changes in industrial composition. The housing bust and financial crisis led to a decline in construction and finance employment that is likely to be long-lasting. Does that imply a substantial rise in normal unemployment? Almost surely not. Compared with the pre-crisis days, about 1.3 million more construction and finance workers are out of work. Even if they all remained permanently unemployed — which is obviously unrealistic — this change would add less than a percentage point to the normal unemployment rate.

More fundamentally, the economy can usually cope with changes in industrial composition. During normal times, industries decline and grow, and displaced workers move to new sectors. For example, manufacturing jobs declined steadily as a share of total employment in the 1990s, yet the normal unemployment rate remained very low.

The real problem today is that jobs are scarce in just about all sectors. An important study, published in 2010 and recently updated, showed that workers who have lost jobs in construction and finance have been leaving the ranks of the unemployed at almost the same rate as those laid off in less troubled industries. The problem isn’t about particular sectors; it’s about a general lack of hiring.

What about declining geographic mobility? Today, about 11 million families are underwater on their mortgages, which means they owe more than their homes are currently worth. This could make it harder for them to sell their homes and move to jobs in other regions.

But the argument that such “house lock” is a source of high unemployment runs into two empirical walls. First, jobs are not plentiful anywhere. In the most recent data, the unemployment rate in every state was above its level before the recession. So our unemployment problem wouldn’t go away if only people could move more easily.

Second, if house lock were an important factor, we would expect to see greater declines in labor mobility in states with more underwater mortgages, and among homeowners compared with renters. A study scheduled for publication in The Journal of Economic Perspectives finds no support for either of these hypotheses.

Narayana Kocherlakota, the president of the Federal Reserve Bank of Minneapolis, made waves last August when he pointed to a recent climb in posted job vacancies as evidence that current unemployment is mainly structural. Normally, there wouldn’t be this many vacancies until unemployment were closer to normal. Therefore, he argued, a severe mismatch between unemployed workers and the skill requirements or location of available jobs had raised the normal unemployment rate by as much as three percentage points.

But this analysis misses the fact that early in recoveries, vacancies typically rise relative to unemployment. Also, as discussed by Peter A. Diamond in his recent Nobel prize lecture, businesses that are relatively more plentiful today — for example, larger companies and those outside of construction — tend to post their vacancies more consistently. Thus, the changing composition of companies helps explain the unusual rise in vacancies.

When experts weigh the evidence, they come down strongly on the side that normal unemployment has not risen greatly. Once a year, the Survey of Professional Forecastersasks respondents for their estimate of the natural unemployment rate. In the third quarter of 2010, the median estimate was 5.78 percent, almost exactly one percentage point higher than in the third quarter of 2007, just before the recession started. (The highest estimate was 6.8 percent.) And the Congressional Budget Office uses 5.2 percent as its estimate of the natural rate.

All of this suggests that most of our high unemployment is still the consequence of low demand. Consumers remain hesitant to spend because unemployment and debt are high. Companies are unwilling or unable to invest because customers are few and credit is still tight.

This diagnosis suggests that the appropriate remedy is to stimulate demand. In February,I suggested a number of steps the Federal Reserve could take. Some additional fiscal measures would also be useful. More public investment (as the president has advocated), additional aid to state and local governments, and a cut in payroll taxes for employers would all help. Given the severity of the long-run budget problem, short-run fiscal stimulus should only be undertaken as part of a comprehensive package of gradual spending cuts and tax increases. That would give the economy the jolt it needs, while providing reassurance that the United States will remain solvent over the long haul.

REGARDLESS of the cause of extended high unemployment, it is a disaster for families, the economy and government budgets. Thus, if I am wrong, and more unemployment is structural than the current evidence suggests, this is no excuse for washing our hands of the problem. Only the nature of the needed policy response would change. Instead of focusing on increasing demand, we would need policies to help workers and jobs find one another, measures to move workers to where the jobs are (or vice versa), training programs and better education.

And even though today’s unemployment appears mainly cyclical, it could turn structural. The longer that unemployment remains high for cyclical reasons, the more likely that job prospects for unemployed workers will be permanently damaged. In a number of European countries in the 1980s, for example, prolonged recession appears to have caused normal unemployment to rise sharply. Getting cyclical unemployment down quickly is the surest way to prevent that from happening in the United States. 

Christina D. Romer is an economics professor at the University of California, Berkeley, and was the chairwoman of President Obama’s Council of Economic Advisers.

A version of this article appeared in print on April 10, 2011, on page BU5 of the New York edition.