The U.S. unemployment rate in October 2014 was 5.8%. That was a substantial improvement from the situation a few years earlier. In late 2009, after the Great Recession, unemployment peaked at 10%. But unemployment was still well above pre-recession levels; it was only 4.7% in November 2007.

Figure 8-1 shows the U.S. unemployment rate from 1948 to late 2014; as you can see, unemployment soared during the Great Recession of 2007–2009 and fell only slowly in the years that followed. What did the elevated unemployment rate mean, and why was it such a big factor in people’s lives? To understand why policy makers pay so much attention to employment and unemployment, we need to understand how they are both defined and measured.

It’s easy to define employment: you’re employed if and only if you have a job. Employment is the total number of people currently employed, either full time or part time.

Unemployment, however, is a more subtle concept. Just because a person isn’t working doesn’t mean that we consider that person unemployed. For example, as of October 2014, there were 41.8 million retired workers in the United States receiving Social Security checks. Most of them were probably happy that they were no longer working, so we wouldn’t consider someone who has settled into a comfortable, well-earned retirement to be unemployed. There were also 11 million disabled U.S. workers receiving benefits because they were unable to work. Again, although they weren’t working, we wouldn’t normally consider them to be unemployed.

The U.S. Census Bureau, the federal agency tasked with collecting data on unemployment, considers the unemployed to be those who are “jobless, looking for jobs, and available for work.” Retired people don’t count because they aren’t looking for jobs; the disabled don’t count because they aren’t available for work. More specifically, an individual is considered unemployed if he or she doesn’t currently have a job and has been actively seeking a job during the past four weeks. So unemployment is defined as the total number of people who are actively looking for work but aren’t currently employed.

A country’s labor force is the sum of employment and unemployment—that is, of people who are currently working and people who are currently looking for work, respectively. The labor force participation rate, defined as the percentage of the working-age population that is in the labor force, is calculated as follows:

The unemployment rate, defined as the percentage of the total number of people in the labor force who are unemployed, is calculated as follows:

To estimate the numbers that go into calculating the unemployment rate, the U.S. Census Bureau carries out a monthly survey called the Current Population Survey, which involves interviewing a random sample of 60,000 American families. People are asked whether they are currently employed. If they are not employed, they are asked whether they have been looking for a job during the past four weeks. The results are then scaled up, using estimates of the total population, to estimate the total number of employed and unemployed Americans.

In general, the unemployment rate is a good indicator of how easy or difficult it is to find a job given the current state of the economy. When the unemployment rate is low, nearly everyone who wants a job can find one. In 2000, when the unemployment rate averaged just 4%, jobs were so abundant that employers spoke of a “mirror test” for getting a job: if you were breathing (therefore your breath would fog a mirror), you could find work. By contrast, in 2010, with the unemployment rate above 9% all year, it was very hard to find work. In fact, there were almost five times as many Americans seeking work as there were job openings.

Although the unemployment rate is a good indicator of current labor market conditions, it’s not a literal measure of the percentage of people who want a job but can’t find one. That’s because in some ways the unemployment rate exaggerates the difficulty people have in finding jobs. But in other ways, the opposite is true—a low unemployment rate can conceal deep frustration over the lack of job opportunities.

How the Unemployment Rate Can Overstate the True Level of Unemployment If you are searching for work, it’s normal to take at least a few weeks to find a suitable job. Yet a worker who is quite confident of finding a job, but has not yet accepted a position, is counted as unemployed. As a consequence, the unemployment rate never falls to zero, even in boom times when jobs are plentiful. Even in the buoyant labor market of 2000, when it was easy to find work, the unemployment rate was still 4%. Later in this chapter, we’ll discuss in greater depth the reasons that measured unemployment persists even when jobs are abundant.

How the Unemployment Rate Can Understate the True Level of Unemployment Frequently, people who would like to work but aren’t working still don’t get counted as unemployed. In particular, an individual who has given up looking for a job for the time being because there are no jobs available—say, a laid-off steelworker in a deeply depressed steel town—isn’t counted as unemployed because he or she has not been searching for a job during the previous four weeks. Individuals who want to work but have told government researchers that they aren’t currently searching because they see little prospect of finding a job given the state of the job market are called discouraged workers. Because it does not count discouraged workers, the measured unemployment rate may understate the percentage of people who want to work but are unable to find jobs.

Discouraged workers are part of a larger group—marginally attached workers. These are people who say they would like to have a job and have looked for work in the recent past but are not currently looking for work. They, too, are not included when calculating the unemployment rate. Finally, another category of workers who are frustrated in their ability to find work but aren’t counted as unemployed are the underemployed: workers who would like to find full-time jobs but are currently working part time “for economic reasons”—that is, they can’t find a full-time job. Again, they aren’t counted in the unemployment rate.

The Bureau of Labor Statistics is the federal agency that calculates the official unemployment rate. It also calculates broader “measures of labor underutilization” that include the three categories of frustrated workers. Figure 8-2 shows what happens to the measured unemployment rate once discouraged workers, other marginally attached workers, and the underemployed are counted. The broadest measure of unemployment and underemployment, known as U-6, is the sum of these three measures plus the unemployed. It is substantially higher than the rate usually quoted by the news media. But U-6 and the unemployment rate move very much in parallel, so changes in the unemployment rate remain a good guide to what’s happening in the overall labor market, including frustrated workers.

Finally, it’s important to realize that the unemployment rate varies greatly among demographic groups. Other things equal, jobs are generally easier to find for more experienced workers and for workers during their “prime” working years, from ages 25 to 54. For younger workers, as well as workers nearing retirement age, jobs are typically harder to find, other things equal.

Figure 8-3 shows unemployment rates for different groups in 2007, when the overall unemployment rate was low by historical standards, in 2010, when the rate was high in the aftermath of the Great Recession, and in 2014, when it had come down much of, but not all, the way to pre-crisis levels. As you can see, the unemployment rate for African-American workers is consistently much higher than the national average; the unemployment rate for White teenagers (ages 16–19) is normally even higher; and the unemployment rate for African-American teenagers is higher still. (Bear in mind that a teenager isn’t considered unemployed, even if he or she isn’t working, unless that teenager is looking for work but can’t find it.) So even at times when the overall unemployment rate is relatively low, jobs are hard to find for some groups.

So you should interpret the unemployment rate as an indicator of overall labor market conditions, not as an exact, literal measure of the percentage of people unable to find jobs. The unemployment rate is, however, a very good indicator: its ups and downs closely reflect economic changes that have a significant impact on people’s lives. Let’s turn now to the causes of these fluctuations.

Compared to Figure 8-1, Figure 8-4 shows the U.S. unemployment rate over a somewhat shorter period, the 35 years from 1979 through late 2014. The shaded bars represent periods of recession. As you can see, during every recession, without exception, the unemployment rate rose. The severe recession of 2007–2009, like the earlier one of 1981–1982, led to a huge rise in unemployment.

Correspondingly, during periods of economic expansion the unemployment rate usually falls. The long economic expansion of the 1990s eventually brought the unemployment rate to 4.0%, and the expansion of the mid-2000s brought the rate down to 4.7%. However, it’s important to recognize that economic expansions aren’t always periods of falling unemployment. Look at the periods immediately following the recessions of 1990–1991 and 2001 in Figure 8-4. In each case the unemployment rate continued to rise for more than a year after the recession was officially over. The explanation in both cases is that although the economy was growing, it was not growing fast enough to reduce the unemployment rate.

Figure 8-5 is a scatter diagram showing U.S. data for the period from 1949 to 2013. The horizontal axis measures the annual rate of growth in real GDP—the percent by which each year’s real GDP changed compared to the previous year’s real GDP. (Notice that there were ten years in which growth was negative—that is, real GDP shrank.) The vertical axis measures the change in the unemployment rate over the previous year in percentage points—last year’s unemployment rate minus this year’s unemployment rate. Each dot represents the observed growth rate of real GDP and change in the unemployment rate for a given year. For example, in 2000 the average unemployment rate fell to 4.0% from 4.2% in 1999; this is shown as a value of −0.2 along the vertical axis for the year 2000. Over the same period, real GDP grew by 3.7%; this is the value shown along the horizontal axis for the year 2000.

The downward trend of the scatter diagram in Figure 8-5 shows that there is a generally strong negative relationship between growth in the economy and the rate of unemployment. Years of high growth in real GDP were also years in which the unemployment rate fell, and years of low or negative growth in real GDP were years in which the unemployment rate rose.

The green vertical line in Figure 8-5 at the value of 3.25% indicates the average growth rate of real GDP over the period from 1949 to 2013. Points lying to the right of the vertical line are years of above-average growth. In these years, the value on the vertical axis is usually negative, meaning that the unemployment rate fell. That is, years of above-average growth were usually years in which the unemployment rate was falling. Conversely, points lying to the left of the green vertical line were years of below-average growth. In these years, the value on the vertical axis is usually positive, meaning that the unemployment rate rose. That is, years of below-average growth were usually years in which the unemployment rate was rising.

A period in which real GDP is growing at a below-average rate and unemployment is rising is called a jobless recovery or a “growth recession.” Since 1990, there have been three recessions, each of which was followed by a period of jobless recovery. But true recessions, periods when real GDP falls, are especially painful for workers. As illustrated by the points to the left of the purple vertical line in Figure 8-5 (representing years in which the real GDP growth rate is negative), falling real GDP is always associated with a rising rate of unemployment, causing a great deal of hardship to families.