Unemployment

Now let's look at the bad news, at our measure of disequilibrium in the labor market. Returning to the CPS survey, to be classified as unemployed a respondent must indicate they did not work during the survey week, but they have looked for work at least once in the past 4 weeks. By combining the employed and unemployed we have the labor force. A person is considered, 'not in the labor force' if they are less than 16 or if you are not actively seeking work. Included here would be individuals who have retired, who are in school, who are unable to work because of illness, those engaged in housework, and those discouraged from looking for work.  If we add together those in the labor force with all persons 16 years of age or older who are not institutionalized we have the civilian non institutional population.   The unemployment rate, meanwhile, is easily computed by dividing the number of unemployed by the labor force. In fact there are two 'official' measures of unemployment, one which includes the armed forces and one based only on the civilian labor force.

The unemployment track record since 1890 reveals some interesting features of the labor market. By far the most obvious feature of the graph is the peak in the 1930s. It is no surprise this period is called the Great Depression.  There has been nothing else quite like it, with the unemployment reaching 25 percent during the low point of the decade.  It is also no surprise that the Depression altered the way we viewed the economy and the appropriateness of government involvement in management of it.  Passage of the Full Employment Act of 1946 can be directly traced to unemployment during the Depression.  Following WW II there was an unmistakable upward trend in the unemployment rate that extended into the 1980s, although the variation around the trend was substantial as the economy moved from recession to expansion. 

The aggregate unemployment rate masks some substantial differences between segments of the labor market.  In the diagrams below the age and racial dimensions to unemployment are quite apparent.  The unemployment rates in each group possess similar cyclical patterns, but the rates for young males average three times the rate for older males, while the rate for Blacks and Others averages twice the rate for Whites.

One measure of the potential importance of demographics in the labor market mentioned earlier can be seen in the graph below of the ratio of unemployment rates of the young (16-19) to the not young (20+).   When the leading edge of the baby boomers came into their late teens the relative unemployment rate of the teens rose sharply so that by the late 1960s it was nearly 5.5 times higher than the rate for those 20 or older.

There are a number of other ways to slice the unemployment pie and below are a few graphs in which the many dimensions to unemployment are apparent.  There should be few surprises.  Occupation matters - the unemployment rate among highly skilled managerial and professional workers is less than a third of the rate for less skilled operators, fabricators, and laborers.  What industry you work in also matters, with unemployment rates in government less than a third of the rate in agriculture and construction. 

The chances of being unemployed also vary substantially with education.  An individual with less than a high school education is four times as likely to be unemployed as a college graduate.   In 1996, a generally good year for the economy, one in eight people without a high school degree was unemployed, while nearly one of every five teenagers could not find work. 

Finally, where you live also matters.   In the summer of 1998 the unemployment rate in the nation's 227 metropolitan areas ranged from a low of 1.2 percent in Rochester, MN and Sioux Falls, SD to 36.4 percent in Yuma, AZ and 17.5 percent in McAllen -Edinburg -Mission, TX. The chances in 1996 of being unemployed in Rhode Island was more than twice what it was in Nebraska, but less than the odds in DC and West Virginia.  The relative unemployment position of the states does, however, vary substantially over time.  During the New England Miracle in the mid 1980s the unemployment rate in New England was among the lowest in the nation, but once the real estate bubble burst in 1988, things deteriorated quickly.  We saw a similar pattern in California where the decline in the defense industry at the end of the 1980s drove them into a severe recession.  In the Midwest, the rustbelt, recessions tend to come early and hit hard, but by 1996 the region had recovered and unemployment rates were below the national average in a number of states including Michigan.  For a visual representation of the regional impact of the business cycle you should visit a slide show site.

And how does the unemployment situation in the US compare to what we see in the other industrialized countries? There are similarities in the patterns, although Japan has by far the lowest and most stable unemployment rate. In each of the countries there is an unmistakable upward trend, and in the US and UK the unemployment rate tends to be more volatile over the business cycle.  The rise of unemployment in Europe is also evident with all three European countries experiencing higher unemployment rates than the US by the mid 1990s.

There are many, however, who question the validity of the unemployment rate as a measure of underutilization of the labor force. As a measure of the malfunctioning of the labor market, the unemployment rate is viewed as incomplete because there has been no account taken of underemployment, discouraged workers, and those working in the underground economy. The underemployed represent those individuals employed in jobs where they are not working to their potential. Included in this group would be all those college grads working as taxi drivers or bar tenders as they wait to be discovered, as well as those who can only find part-time work. Those who have given up any efforts at finding work, meanwhile, are classified as discouraged workers because they are out of work but not looking because it is not worth the effort. You would be likely to find this type of behavior when unemployment rates approached one-in-three, about what they are for Black teenage males.

In an effort to adjust for some of these problems, the BLS now produces eight measures of unemployment, each with its own definition of the unemployed and the labor force. The narrowest definition would be the percentage of the civilian labor force (CLF) unemployed more than 15 weeks, the hard core unemployed which in 1992 averaged 2.6 percent. This was substantially lower than the official number of 7.1 percent. What both of these numbers miss is any measure of underemployment and discouraged workers. In an effort to account for these effects, the BLS has calculated a much broader measure of unemployment that accounts for part-time workers who are working part-time due to economic factors, which was 10 percent, and if we add in BLS's estimate for discouraged workers, the rate rises to 10.8 percent. Missing in all of these estimates is any adjustment of the official figures for those working in the underground economy.

Alternative Unemployment Measures: 1992

The BLS also publishes monthly unemployment rates by states that are derived from two sources, the Current Population Survey and the Local Area Unemployment Statistics program. The labor force, employment, and unemployment data for 11 large states (CA, FL, IL, MA, MI, NY, NJ, NC, OH and PA) are derived directly from the CPS, while the remaining 39 are derived using a standardized BLS technique.

What about a benchmark against which we can evaluate the unemployment rate? There is one, and it is most often referred to in texts as the full-employment rate of unemployment. Full employment, that illusive macroeconomic goal, would not be 0 percent, although over the years many students have offered this as a definition of full employment before any discussion begins. At any point in time, a vibrant, dynamic national economy would be experiencing unemployment as workers shifted from declining sectors to expanding sectors. In fact this could be looked at this component of unemployment as a positive sign - a sign that the economy is dynamic. We call this frictional unemployment.

What about the decline of the railroad and the rise of the automobile industry? The decline of steel industry and the rise of plastics and aluminum, the decline of agriculture and the rise of manufacturing, the decline of manufacturing and the rise of the service sector? In each instance hundreds of thousands of workers were at least temporarily displaced by these structural changes, what economists have defined as structural unemployment.

Some indication of the structural unemployment situation can be seen in the number of individuals who remain on unemployment for over 27 weeks.  A strong cyclical pattern is evident, with substantial declines in the number of long-term unemployed during the long expansions in the 1960s and 1980s and the rise during the recessions in the early 70s, 80s, and 90s.  There also appears to be an upward trend in the number of long-term unemployed with each peak and trough a bit higher than the one before it. 

While economists have been able to agree that full employment exists at some non zero rate of unemployment, they have not been able to agree on exactly what it is. President Kennedy, back in the early 1960s, seems to have been the first individual to have identified full employment as 4 percent, a number that seemed to stick until the stagflation of the 1970s. At that time there was a movement away from full employment to the natural rate of unemployment which was defined as the rate of unemployment that is sustainable without accelerating inflation. As we will see later, this became a real political issue - just what we saw with the CPI measure of inflation.