The Stubborn Jobless Rate: Puzzling, but Far from Scary

The Dems say the labor market is the worst since the Depression. But the drop in jobs has been milder than in many other recessions

By Robert J. Barro

With the start of the Presidential campaign season, one has to expect a rise in spurious economic claims. Nevertheless, I was surprised by the assertion from several Democratic candidates that the current U.S. labor market is the worst since the Great Depression of the 1930s. This claim has no basis in fact. Measured by unemployment rates or employment growth, the recession of 2001 and its aftermath was not as bad as many post-World War II recessions, notably 1948-49, 1957-58, 1981-82, 1953-54, and 1974-75. The downturn was more comparable with the mild recessions of 1960-61, 1970, and 1990-91.

When gauging a weak labor market, most economists look first at the unemployment rate. This rate, now 6.1%, peaked in the aftermath of the latest recession at 6.4% in June of this year. In contrast, during previous recessions, the jobless rate reached 10.8% in November, 1982; 9% in May, 1975; 7.9% in October, 1949; and 7.5% in July, 1958. The current 6.1% rate is only modestly above the average of 5.6% since January, 1948.

Instead of just looking at how high the jobless rate is, economists often look at how much it has risen during a recession. In the recent slowdown, the unemployment rate increased 2.5 percentage points, from 3.9% at its low point in December, 2000, to 6.4% in June. This rise is substantially smaller than those during the major postwar recessions: 4.4 points in 1948-49, 3.9 in 1974-75, 3.8 in 1957-58, and 3.6 in 1981-82 and 1953-54. More comparable are the increases in the unemployment rate by 2.6 points in the recessions of 1970 and 1990-91 and 2 points in 1960-61.

A more interesting perspective emerges when one looks at employment growth. Let's consider the most commonly used measure, the total non-agricultural employment series from the Bureau of Labor Statistics. Employment peaked at its all-time high of 132.6 million in February, 2001, and fell to 129.8 million in August, 2003, a 2.1% decline. This percentage decline is much smaller than those in the major postwar recessions: 5.2% in 1948-49, 4.4% in 1957-58, 3.4% in 1953-54, 3.1% in 1981-82, and 2.8% in 1974-75. More similar to today's situation was the fall by 2.2% in 1960-61. For 1970 and 1990-91, the decreases were 1.5%.

Although we clearly can reject the claim that the current labor market is the worst since the Great Depression, we still want to know why employment growth has remained negative despite the recent pickup in gross domestic product growth. One common explanation is that employment growth and the jobless rate are lagging indicators. There's some truth to this. When the GDP changes, some of the shift in employment shows up only after one or two quarters. But most of the change usually occurs within the same quarter as the change in GDP. Moreover, even a one- to two-quarter lag would not explain why employment growth has remained sluggish for 2 1/2 years now. Given the historical link between GDP growth and employment, job growth has been lower than expected in the last nine quarters. This extremely unusual pattern suggests that some basic change may have occurred in the relation between employment and GDP growth.

One way to look at the change is that the GDP growth required to get positive employment growth has risen substantially. In the long-term relationship, employment growth tended to be positive whenever GDP growth was mildly positive -- only something less than 1% GDP growth was required. Since 2001, however, GDP growth apparently has had to exceed 2.5% to 3% for employment to begin climbing. That's why employment is still decreasing.

The other way to look at the numbers is that, since GDP growth has been positive and employment growth has been negative, the economy has been able to produce more goods with fewer workers. In other words, productivity measured by GDP per worker has been rising strongly. (An adjustment for hours worked strengthens this scenario.) Since the second quarter of 2001, productivity has grown by 3.1% a year, compared with the long-term average of 1.4%. The key question is how long this high productivity growth will last.

If the high productivity gains continue, the economy will benefit tremendously for many years to come. In the longer term, employment growth corresponds to increases in the labor force, which depends on trends in population and labor-force participation. If the labor force expands at its long-term average of 1.5% to 2% per year, employment growth will be about the same. To determine long-term GDP growth, we have to add the productivity growth rate. Thus, if productivity grows at 3.1%, the GDP growth rate would be around 5%, compared with the historical average of 3.3%. That would be very nice. For Republicans, an important question is whether this pleasantness will materialize by election time in 2004. That is a question I cannot answer.

Robert J. Barro is a professor of economics at Harvard University and a senior fellow of the Hoover Institution (

    Before it's here, it's on the Bloomberg Terminal.