When there are not enough jobs for everyone wanting to work in a particular city, region, or industry, economists look for the underlying causes, which can include technological change, rising resource costs, stiffer competition from imports, or a fundamental change in consumer preferences. Any of these may explain a localized gap between jobs offered and jobs wanted. At the national or macroeconomic level, however, unemployment generally cannot be traced to any single cause; rather, it is a feature of the business cycle. Every economy is subject to cyclical expansions and contractions. The U.S. economy, for example, has experienced 11 recessions since World War II, with the national unemployment rate topping 10 percent during 2 of them. (By comparison, at the depths of the Great Depression, unemployment reached 25 percent.) Unemployment takes an economic toll in terms of foregone output—goods and services that could have been produced if all available workers had held jobs. It also entails serious social costs for those out of work and for their families. Government can ease the financial burden of joblessness through such programs as unemployment insurance, job retraining, and tax credits for companies that hire new workers.
I. Four Types of Unemployment
II. How Unemployment Is Measured
III. The Uneven Impact of Unemployment
IV. Short- and Long-Term Unemployment
V. “Natural” Unemployment, the Phillips Curve, and Okun’s Law
VI. Unemployment Insurance
VII. Unemployment and Poverty
VIII. The Political Economy of Unemployment
Four Types of Unemployment
Economists have described four different types or categories of unemployment. Frictional unemployment refers to the normal job switching that occurs in a market economy; at any given moment, there are bound to be some workers who are temporarily between jobs. Seasonal unemployment refers to those who are out of work because of normal seasonal variations in employment. This can be seen in resort communities, where, for example, jobs may be plentiful during the winter in a ski town or during the summer in a seaside town but scarce during the rest of the year. Certain industries, like agriculture and construction, also exhibit high seasonal unemployment rates. Structural unemployment occurs when jobs disappear due to changes in the structure of the economy, as, for example, when companies move their operations overseas in order to lower their wage costs. Finally, cyclical unemployment is the type that rises and falls according to the overall condition of the national economy.
When the U.S. Bureau of Labor Statistics (BLS) announces the latest unemployment rate on the first Friday of each month, it does not distinguish among these four types. One may safely assume, however, that any large movements in the official unemployment rate are mainly due to the ups and downs of cyclical unemployment. The unemployment rate is inversely related to the national output rate, but it does not track it precisely. Unemployment is a so-called lagging indicator, meaning that when the economy starts to shrink, or go into a recession, it may take several months before the unemployment rate begins to rise, and when the economy hits bottom and starts to make a recovery, it may take some time before the unemployment rate begins to fall in response.
How Unemployment Is Measured
The BLS estimates the number of people unemployed in the United States by sampling the population rather than attempting to actually count every person who is looking for work and not finding it. The sample consists of 60,000 randomly selected households, representing about 110,000 individuals. Each month, trained interviewers check up on these households—some by phone and some in person—to determine their work activity in the previous month. Every individual who is 16 years of age or older is potentially a participant in the civilian labor force. Many, however, do not participate, including those who are retired, on active duty in the armed forces, in school full-time, totally disabled, institutionalized, or, for a variety of other reasons, not looking for employment. These individuals are not counted as part of the labor force. Only those who have jobs or are available for work and actively trying to find it get counted in the labor force, as defined by the BLS. The official unemployment rate is computed by dividing the number of people who are jobless but actively looking for work by the total number who are in the labor force. This fraction is then multiplied by 100 so that the unemployment rate can be expressed as a percentage.
An important group of people excluded from the labor force are those known as “discouraged workers.” These are individuals who tell the interviewer that they want work, are available for work, have looked for work during the previous year, but are not currently looking for a job. They have become discouraged because they feel they have the wrong skills, are too young or too old, are likely to experience hiring discrimination, or just do not fit the jobs that currently exist. Beyond this group of workers, who are discouraged for job-related reasons, there is an even larger group of workers called “other marginally attached,” who say they are not job searching because of health problems, child care issues, or difficulties arranging transportation. These individuals say they want work and have looked for a job in the past year, but they have not looked in the past four weeks.
The most comprehensive measure of unemployment goes beyond what has just been described to include individuals who work part-time because they are unable to get the additional hours of work they would like. Under the standard BLS definition, a person who works for pay as little as 1 hour per week is categorized as employed, even though underemployed would be a more accurate term if the worker would prefer to be working 10, 20, or 40 hours a week. Millions of workers in the United States can be described as holding their jobs “part-time for economic reasons,” not by choice. When we take into account all the forms of subemployment outlined above, we get a much higher number than the conventional, and most widely publicized, unemployment rate. In April 2010, for example, the standard unemployment rate was 9.9 percent, while the broadest, most encompassing measure of underemployment stood at 17.1 percent. Both numbers were swollen by what some called the Great Recession of 2008–2010, yet even when the economy is in a healthier state, the gap persists. In April 2000, for example, the official unemployment rate was 3.8 percent, while the broader gauge of labor-force underutilization reached 6.9 percent.
The Uneven Impact of Unemployment
Americans experience unemployment in different ways and at different rates, depending on their age, gender, education level, and race or ethnicity. The monthly BLS unemployment report sheds light on all of these differences. For the month of April 2010, the unemployment rate for teenaged workers, ages 16 to 19, was a whopping 25.4 percent. This implied that for every four teenagers who were counted in the labor force (meaning they were either employed or looking for a job), only three were able to find work. For adults age 20 and older, unemployment was far lower, at a little over 9 percent.
Gender disparities in unemployment are much smaller than disparities by age. For most of the 2000–2010 decade, there was no consistent difference between men’s and women’s jobless rates. That changed near the end of the decade, however, as the economy fell into recession. More jobs were lost in male-dominated occupations, like construction and manufacturing, than in the rest of the economy, causing men’s unemployment to rise faster than women’s. In April 2010, the rate for all men age 16 and older stood at 10.8 percent, compared to 8.8 percent for women. If teenaged workers were excluded, the rates were 10.1 and 8.2 percent for men and women, respectively.
Education strongly influences one’s chances of being unemployed. More highly educated Americans experience less joblessness than those with fewer educational credentials. The April 2010 figures were illustrative. Among workers aged 25 or older, those lacking a high school diploma had a 14.7 percent unemployment rate. The comparable rates for those with a high school diploma but no college, those with some college or an associate’s degree, and those with a bachelor’s degree or higher were 10.6 percent, 8.3 percent, and 4.9 percent, respectively. The advantage enjoyed by those with more education can also be seen in their higher rates of participation in paid work. Fewer than half the individuals who lacked a high school education participated in the labor force in April 2010. Those with more education had higher participation rates. College graduates, for example, participated at a 77.3 percent rate.
Race and ethnicity also affect one’s likelihood of being unemployed. Historically, African Americans have been jobless at higher rates than whites or Hispanics. In April 2010, the black unemployment rate stood at 16.5 percent, compared to 12.5 percent for Hispanics, 9.0 percent for whites, and 6.8 percent for Asians. When high-unemployment categories overlap, the resulting unemployment rates can be extraordinarily high. In December 2009, for example, the unemployment rate for black teenagers (ages 16 to 19) reached an astonishing 48 percent. (For white teens, the rate was a little under 24 percent.)
Short- and Long-Term Unemployment
The longer a person remains without work, the deeper the impact on finances, family well-being, sense of self-worth, and even health. Some workers are out of a job so briefly they do not even bother to file for unemployment compensation. At the other extreme, some workers spend long months and even years looking in vain for work. Workers who lose their jobs in middle age can be especially devastated, since they often shoulder heavy financial responsibilities and do not have time to start a new career in the working years they have left.
During the recession of 2008–2010, observers noted a considerable increase in the amount of time workers typically spent unemployed. So not only were there an unusually large number of people out of work—even for a recession—but they were staying unemployed much longer than usual. Data from the BLS showed that, in April 2010, more than three-fifths of the unemployed had been out of work for 15 weeks or more, and fully 46 percent met the definition of “long-term unemployed” by being out of work for more than six months. The median period spent unemployed was about 22 weeks, meaning that as many people were unemployed longer than 22 weeks as were unemployed for less than that length of time. In historical data going back to 1967, there is nothing comparable to this figure; only once previously had the median length of unemployment even reached 12 weeks (in March 1983).
The social consequences of unemployment are manifold. The unemployed face a greater risk of clinical depression, hospitalization, and other adverse effects, as studies have shown in the United States and abroad. The damage can be straightforwardly physical: a study conducted under the auspices of the Harvard School of Public Health showed that even short periods of unemployment lead to more diagnoses of high blood pressure, heart disease, and diabetes (Strully 2009). A Swedish study found that individuals experiencing unemployment face higher mortality risk, partly due to increased rates of suicide (Gerdtham and Johannesson 2003). It has also been shown that domestic violence, divorce, and child abuse increase when fathers are unemployed (Schiller 2008, 91). Given the well-established physical and mental health consequences of unemployment, it should hardly come as a surprise that unemployment has also been linked, in a British study, to lower levels of happiness. Somewhat less expected were the findings that unemployment produces less unhappiness among the young, among those who have been unemployed for longer periods of time, and in those regions where the overall unemployment rates are high (Clark and Oswald 1994).
“Natural” Unemployment, the Phillips Curve, and Okun’s Law
At various times from the 1960s onward, economists in the United States have offered opinions about the rate of unemployment that should be considered normal or “natural.” If the actual unemployment rate fell below this natural level, it would signal an overheated economy likely to be experiencing inflation. If it rose above the natural level, it would signal slack in the system, suggesting that the economy was performing below its potential. The cumbersome, technical term for the natural rate of unemployment is nonaccelerating inflation rate of unemployment, or NAIRU. Economists have tried to determine what the natural rate of unemployment would be in a given economy. For the United States, some thought 4 percent, some 5 percent, some even thought 6 percent unemployment might be the natural level. To date, no consensus has emerged, and in fact the search for the true NAIRU appears, for now, to have been suspended.
Implicit in discussions of NAIRU was the assumption that for any given economy there was a stable relationship between unemployment rates and inflation rates. Empirical research in the 1950s and 1960s appeared to support this notion. Soon economics students were finding something new in their principles textbooks—the Phillips Curve, which graphed an inverse relationship between the inflation rate and the unemployment rate. The Phillips Curve seemed to offer policymakers the possibility of choosing from a variety of inflation–unemployment combinations. If they wanted to lower the unemployment rate, the cost would be higher inflation; if they wanted to lower inflation, the cost would be higher unemployment.
In the same era that produced NAIRU and the Phillips Curve, another empirical relationship involving unemployment was put forth: Okun’s Law. Named for the economist Arthur Okun, this law posited a stable inverse relationship between changes in the unemployment rate and the economy’s real growth rate. Growth in this case referred to increases in the real, inflation-adjusted gross domestic product, or GDP. Okun’s Law builds on the obvious fact that, if national output is to be increased, more labor will have to be employed, and that will reduce unemployment. A widely accepted version of Okun’s Law states that, for every 2 percent growth in GDP above its long-term trend rate, unemployment falls by 1 percent. For example, if the long-term GDP growth trend is 3 percent and GDP actually grows by 5 percent, the unemployment rate would be expected to fall by 1 percent. (In this example, therefore, if unemployment had previously been 6 percent, it would fall to 5 percent.) As with the Phillips Curve, Okun’s Law has fallen into disfavor among most economists; neither appears to have strong enough empirical underpinnings to be relied upon as a macro policy tool.
When workers lose their jobs not through any fault of their own but because their industry—or the entire economy—is going through a cyclical downturn, most people believe they need and deserve some kind of temporary financial assistance from the government. Such was not always the case. Historically, state assistance to the unemployed could not be legislated as long as the public viewed joblessness as a matter of individual responsibility. During the periodic depressions that characterized the 19th century, even private charity for the unemployed was sometimes decried as “pauperizing”—that is, liable to turn temporary hardship into permanent dependency. But gradually opinion shifted toward a more positive view of public responsibility to aid the jobless, whether in the form of emergency public relief programs during times of mass unemployment or through employment bureaus operated by cities and states on a model borrowed from Great Britain. (Such bureaus could be found in more than half the states by 1915, according to Roy Lubove 1968.)
Compulsory unemployment compensation plans began to be introduced into state legislatures during the 1920s, but only Wisconsin enacted a program before Congress passed the Social Security Act of 1935. Under that landmark law, all states must operate unemployment insurance (UI) programs. The features that are common to all state programs are: UI applicants must have some work experience; they must have lost their jobs through no fault of their own; and they must be ready and willing to work. Benefit levels vary widely and generally make up only a fraction of the worker’s normal wage income. The benefits paid during the standard 26-week period of eligibility are financed out of state payroll taxes. In times of unusual economic distress, Congress can authorize a 13-week extension of UI benefits, in which case the funding is shared by the states and the federal government.
The common view that almost any unemployed person can collect unemployment benefits is far from true. Of those counted as unemployed by the BLS at any given time, only about one-third are able to receive unemployment compensation. The jobless who are just entering the labor force—college graduates, for example—or who are reentering the labor force after a period spent out of work are ineligible to receive UI benefits. Also ineligible are those who have been dismissed for cause from their jobs.
Unemployment and Poverty
The correlation between unemployment and poverty in the United States is not hard to understand. When the economy goes into a recession and jobs are lost or hours are cut, families that had been managing to stay just above the poverty line can easily drop below it. When the economy recovers and laid-off workers go back to work, many families have a chance to earn their way out of poverty. A graph showing the trend lines for unemployment and poverty in the United States from 1960 to 2004 strongly confirms the correlation between these two rates (Schiller 2008, 99). The data since 2004 offer further confirmation. Poverty rates stayed in the 12 percent range from 2004 to 2007, while the unemployment rate trended downward from 5.6 to 4.4 percent. But in 2008, as the recession got under way, the poverty rate shot up to 13.2 percent—its highest level in more than a decade—while the unemployment rate also rose significantly. Poverty statistics are never released in as timely a fashion as the monthly unemployment statistics, but it is a safe bet that, when the poverty rates for 2009 and 2010 are released, they will mirror the dramatic increases seen in the U.S. unemployment rate during the recession.
The Political Economy of Unemployment
People have very different takes on what, if anything, should be done about unemployment and the hardships it produces. Those on the conservative end of the spectrum have long argued that unemployment is an inevitable feature of the dynamic capitalist economy. The 20th century’s most outspoken advocate of free enterprise, Milton Friedman, asserted in Capitalism and Freedom (1962) that severe unemployment was almost always the result of government ineptitude, not ordinary business cycles. He also blamed some unemployment on minimum wage laws, which, therefore, he wanted to see abolished. At the other (left) end of the political spectrum, radical thinkers from Marx onward have depicted unemployment as both a systemic and a necessary feature of capitalism. Periodic bouts of mass unemployment, in their view, serve to weaken labor unions, restrain wages, and bolster corporate profits. In Marx’s view, capitalist crises could be expected to not only continue but grow in intensity, cutting into production and expanding the “reserve army of the unemployed,” until at last the working classes threw off their shackles and seized control of the system.
The centrist view of unemployment since the Great Depression has been to see the macro economy as subject to periodic (though not worsening) downturns, which the government can moderate but not eliminate. Unemployment is therefore accepted as a fact of the modern mixed economy. By a proper use of fiscal and monetary tools, economic fluctuations can be held within acceptable bounds, and unemployment insurance— part of a broader safety net—can keep the consequences for individual workers tolerable. It would be hard to find a prominent conservative or libertarian thinker today who would advocate abolishing the UI system that Americans have had since 1935. Debate about how best to help the unemployed centers instead on where the benefit levels should be set and how long the unemployed should be able to collect their benefits. If benefit levels are set too low, the safety net will not meet the basic test of adequacy. There may be another problem as well: low benefit levels will tend to cause the unemployed to shorten their job searches and accept the first job that comes along, even if it does not make appropriate use of their human capital. The end result is inefficiency in the economy. If benefit levels are set too high, it allows workers to extend their job searches beyond what would otherwise be the case. This, too, can be a source of economic inefficiency and waste.
- Clark, Andrew E., and Andrew J. Oswald, “Unhappiness and Unemployment.” Economic Journal 104, no. 424 (May 1994), 648–659.
- Friedman, Milton, Capitalism and Freedom. Chicago: University of Chicago Press, 1962.
- Gerdtham, Ulf-G., and Magnus Johannesson, “A Note on the Effect of Unemployment on Mortality.” Journal of Health Economics 22, no. 3 (May 2003), 505–518.
- Lubove, Roy, The Struggle for Social Security, 1900-1935. Cambridge, MA: Harvard University Press, 1968.
- Schiller, Bradley, The Economics of Poverty and Discrimination, 10th ed. Upper Saddle River, NJ: Pearson Prentice Hall, 2008.
- Sharp, Ansel, Charles Register, and Paul Grimes, Economics of Social Issues, 19th ed. New York: McGraw-Hill/Irwin, 2009.
- Strully, Kate W., “Job Loss and Health in the U.S. Labor Market.” Demography 46, no. 2 (May 2009), 221–246.
- “Symposium: The Natural Rate of Unemployment.” Journal of Economic Perspectives 11, no. 1 (1997): 3–108.
- U.S. Department of Labor, Bureau of Labor Statistics, “How the Government Measures Unemployment.” http://www.bls.gov/cps/cps_htgm.htm
- U.S. Department of Labor, Bureau of Labor Statistics, “Economic News Release: The Employment Situation—April 2010.” http://www.bls.gov/news.release/empsit.nr0.htm