The Changing Nature of Unemployment

Unemployment in the United States rose for 15 years (1968-83) and fell over the next 17 years (1983-2000), with cyclical swings in between. In the 1990s, the unemployment rate fell below 4 percent for the first time since 1969. While overall unemployment levels of the late 1990s were nearly as low as those of the late 1960s, the employment picture for less-skilled males was decidedly worse. In fact, the often praised boom of the 1990s represented little progress for this group of American workers.

In a previous study, University of Chicago Graduate School of Business professors Kevin M. Murphy and Robert H. Topel, along with Chinhui Juhn of the University of Houston examined long-term changes in joblessness among American men between 1967 and 1989. Their initial study showed that a steep, steady decline in demand for low-skilled workers had created high rates of unemployment, labor-force withdrawal, and long spells of joblessness for this group. These workers represented a wide range of occupations and skill types, but were mainly low-wage workers.

This decline in the demand for low-skilled workers can be traced to numerous industrial changes, including the decline of manufacturing as a major source of employment. In the past three decades, areas such as business services, finance, insurance, and real estate have grown considerably, all of which heavily favor higher-skilled workers. Technological changes, changing patterns of international trade, and competition from abroad also have reduced the earning power of low-skilled workers while benefiting those with greater skills and education.

Murphy, Topel, and Juhn's recent study, "Current Unemployment, Historically Contemplated," revisits these issues while taking account of the economic boom of the 1990s. Using data from the Bureau of the Census and the Bureau of Labor Statistics, the authors analyze the evolution of joblessness in the United States from 1967 to 2000 among American men ages 19 to 49. Did the expansion of the 1990s really return the U.S. labor market to conditions of the late 1960s, as the unemployment statistics would seem to indicate?

As it turns out, the shift toward long-term joblessness implies that particular rates of unemployment have a much different meaning today than in past decades.

"There has been a growing fraction of less-skilled men who would have been counted among the unemployed thirty years ago who have simply withdrawn from the labor force," says Topel.

In the 1990s, even as unemployment was falling, time spent out of the labor force was rising. In spite of what has been reported in the media, the authors point out that the long-term changes in the labor market have greatly favored skilled workers. These high-skilled workers have enjoyed the perks of the economic boom times, with rising wages and stable employment opportunities. On the other side of the spectrum, the changing labor market has had a far greater impact on less-skilled, less-educated workers below the median of the wage distribution, who have seen their earning power diminish over the past thirty years.

The study shows that falling unemployment rates in the 1990s greatly exaggerated the improvement in labor market conditions for male workers as a whole, and for less-skilled males in particular.

Beyond the Unemployment Rate

The unemployment rate may be the most widely understood measure of economic health, but there are many nuances to consider. The authors distinguish between three different rates: unemployment (out of work but looking for work during the past four weeks); joblessness (out of work, whether or not looking for work); and nonparticipation (out of work and not looking for work).

Murphy, Topel, and Juhn show that rising nonparticipation rates offset the low rate of unemployment in the 1990s. Therefore, even with a historically low unemployment rate, the economic outlook for low-skilled workers has become increasingly grim. This rise in nonparticipation appears to be due to an expansion of the government disability program in the early 1990s, as well as continued low levels of real wages for less-skilled men.

Job opportunities continue to exist, but those jobs have become less attractive to the large fraction of the labor force represented by less-skilled men. Since the 1970s, there has been a decline in the relative and absolute wages for less-skilled workers relative to wages for more highly skilled workers. Therefore, the jobs available for less-skilled workers are not as attractive in terms of earning power. Murphy and Topel point out that these workers would be more likely to seek employment if their wages were higher.

Another major change has been the duration of jobless spells. Nonparticipation in the labor force used to be a temporary event, with workers experiencing short periods of unemployment between jobs.

"Now, nonparticipation is a much more permanent state of affairs," says Topel. "Once you enter nonparticipation, you basically don't come out."

Joblessness among less-skilled men has increasingly become time spent out of the labor force, rather than time spent unemployed. It is also important to note that despite rising wages and the increased labor force participation for women, the high rate of joblessness among less-skilled men is not the result of improved opportunities for their working wives.

Attractive Alternatives

The authors find that more than 40 percent of the growth in nonparticipation during this time period is associated with an increase in men claiming to be ill or disabled, indicating a shift in labor supply.

After 1984, eligibility standards for disability benefits were substantially liberalized, which led to an increase in disability payments. The authors note that the interaction of disability benefits and labor market shocks are of key importance in understanding rising rates of labor force withdrawal. Changing government policies made it easier to collect disability payments at a time when wages for low-skilled men were falling.

Another facet of the equation is the duration of jobless spells. The study shows that the amount of joblessness for those who did not work at all over a given year more than tripled, from 1.8 percent in the 1960s to 6.1 percent in 1999-2000, even with two of the longest economic expansions on record. By the end of the 1990s, the average duration of nonemployment spells was over fifteen months, more than double the length of such spells in the late 1960s. Jobless spells were longer in 1999-2000 than any previous cyclical low of unemployment.

"The labor market is a less rewarding place than it was before," says Topel. "Many of the people affected by those changes in wages have found it to be such a less rewarding place that it's no longer worth participating."

A Long-Term Story

Given the focus on the current economy and frequent reports of the shifting unemployment rate, it is natural to ask how a historical view of unemployment relates to existing economic woes.

Murphy and Topel point out that the results of their study have little to do with the economic policies of the Bush administration, or any of the administrations in the past three decades. In addition, they emphasize that the unemployment rate does not accurately reflect labor market conditions over long periods of time.

"These are really structural changes in the labor market," says Topel. "The trends we see here represent a long-term phenomenon."

Murphy notes that profound industrial and technological changes have created a very different labor market for high-skilled workers relative to low-skilled workers, which is an important distinction for putting recent events in context.

"It may seem like a tough market for high-skilled workers because of the tech boom, tech bust, and current market for MBAs," says Murphy. "However, focusing on high-skilled workers is really missing the point."