PART 8
CHAPTER 24
Unemployment among construction workers rises substantially during recessions.©Ingram Publishing
LEARNING OBJECTIVES
After reading this chapter, you should be able to:
1. LO1List the four phases of the business cycle and explain the primary characteristics of recessions and expansions.
2. LO2Use potential output and the output gap to analyze an economy’s position in the business cycle.
3. LO3Define the natural rate of unemployment and show how it is related to cyclical unemployment.
4. LO4Apply Okun’s law to analyze the relationship between the output gap and cyclical unemployment.
5. LO5Discuss the basic differences between how the economy operates in the short run versus the long run.
“Home Sales and Prices Continue to Plummet”
“As Jobs Vanish, Motel Rooms Become Home”
“Global Stock Markets Plummet”
“Steep Slide in Economy as Unsold Goods Pile Up”
“Fed Plans to Inject Another $1 Trillion to Aid the Economy”
“World Bank Says Global Economy Will Shrink in ’09”
These headlines from The New York Times tell the story: from late 2007 to mid-2009, the U.S. economy passed through its worst economic downturn since the Great Depression of the 1930s. Average incomes fell; millions of Americans lost their jobs, many lost their health insurance, and even their homes; and governments at all levels struggled to deal with falling tax collections colliding with increased demands for public services like unemployment benefits and health care.
Other economic downturns between the Great Depression of the 1930s and the Great Recession—as the 2007–2009 downturn has come to be called—were generally milder. But they too inflicted great economic cost, most importantly lost jobs. And in some cases they had important political consequences.
In preceding chapters, we discussed the factors that determine long-run economic growth. Over the broad sweep of history, those factors determine the economic success of a society. Indeed, over a span of 30, 50, or 100 years, relatively small differences in the rate of economic growth can have an enormous effect on the average person’s standard of living. But even though the economic “climate” (long-run economic conditions) is the ultimate determinant of living standards, changes in the economic “weather” (short-run fluctuations in economic conditions) are also important. A good long-run growth record is not much consolation to a worker who has lost her job due to a recession. The bearing that short-term macroeconomic performance has on election results is one indicator of the importance the average person attaches to it (see The Economic Naturalist 24.1).
In this chapter, we begin our study of short-term fluctuations in economic activity. Commonly known as business cycles, these fluctuations consist of expansions and recessions. We will start with some background on the history and characteristics of these economic ups and downs. We next develop concepts that allow us to measure the severity of business cycles. These concepts allow us to analyze short-run economic activity from different perspectives and to link fluctuations in output to changes in unemployment. Finally, we introduce a verbal description of a basic model of expansions and recessions. Throughout this chapter, we will connect the theory to real-world examples, focusing on the most recent recession, the Great Recession of 2007–2009.
The Economic Naturalist 24.1
Do economic fluctuations affect presidential elections?
In early 1991, following the defeat of Iraq in the Gulf War by the United States and its allies, one poll showed that 89 percent of the American public approved of the job George H. W. Bush was doing as president. Prior to Bush, the last U.S. president to enjoy such a high approval rating was Harry Truman in 1945, shortly after World War II ended with the U.S. a victorious global superpower. The Gulf War victory followed a number of other popular developments in the foreign policy sphere, including the ouster of the corrupt leader General Manuel Noriega from Panama in December 1989, improved relations with China, apparent progress in Middle East peace talks, and the end of apartheid in South Africa. The collapse of the Soviet Union in December 1991—a stunning event that signaled the end of the Cold War—also occurred during Bush’s term. Yet despite these political pluses, in the months following the Gulf War, Bush’s sky-high approval rating declined sharply. According to the same poll, by the time of the Republican National Convention in the summer of 1992, only 29 percent of the public approved of Bush’s performance. Although the president’s ratings improved during the campaign, Bush and his running mate, Dan Quayle, lost the 1992 general election to Bill Clinton and Al Gore, receiving only 39 million of the 104 million votes cast. A third-party candidate, Ross Perot, received nearly 20 million votes. What caused this turnaround in (the first) President Bush’s political fortunes?
Despite his high marks from voters in foreign policy, the president’s domestic economic policies were widely viewed as ineffective. Bush received much criticism for breaking his campaign pledge not to raise taxes. More important, the economy weakened significantly in 1990–1991 and then recovered only slowly. Although inflation was low, by mid-1992, unemployment had reached 7.8 percent of the labor force—2.5 percentage points higher than in the first year of Bush’s term and the highest level since 1984. A sign in Democratic candidate Bill Clinton’s campaign headquarters summarized Clinton’s strategy for winning the White House: “It’s the economy, stupid.” Clinton realized the importance of the nation’s economic problems and pounded away at the Republican administration’s inability to pull the country out of the doldrums. Clinton’s focus on the economy was the key to his election.
Clinton's ability to parlay criticism of economic conditions into electoral success is not unusual in U.S. political history. Weakness in the economy played a decisive role in helping Franklin D. Roosevelt to beat Herbert Hoover in 1932, John F. Kennedy to best Richard Nixon in 1960, and Ronald Reagan to defeat Jimmy Carter in 1980. And in an echo of his father’s experience, President George W. Bush found the political popularity he enjoyed after the 9/11 attacks in 2001—a record 90 percent approval rating—eroded by an economic downturn and a slow subsequent recovery. His approval rating as president hit a record low of 25 percent in October 2008, at the height of the financial crisis. A few weeks later, Barack Obama was elected president, defeating the Republican candidate (and war hero) John McCain.
On the other hand, strong economic conditions have often helped incumbent presidents (or the incumbent’s party) to retain office, including Nixon in 1972, Reagan in 1984, and Clinton in 1996. Indeed, a number of empirical studies have suggested that economic performance in the year preceding the election is among the most important determinants of whether an incumbent president is likely to win reelection.
Finally, it is important to remember that the economic conditions as measured by macroeconomic indicators (such as the rate of unemployment) do not necessarily reflect the economic conditions as perceived by all voters. For example, recall from Chapter 16, Macroeconomics: The Bird’s-Eye View of the Economy, that the construction of macroeconomic indicators involves aggregation and averaging. These indicators can therefore hide economic differences and inequalities across regions, economic sectors, and demographic groups. Indeed, in spite of low and decreasing unemployment rate under Obama, frustration among populations that felt left behind economically was a frequent theme in the 2016 elections and may help explain, among other reasons, why the Democrats lost the presidency.
RECESSIONS AND EXPANSIONS
As background to the study of short-term economic fluctuations, let’s review the historical record of the fluctuations in the U.S. economy. Figure 24.1 shows the path of real GDP in the United States since 1929. As you can see, the growth path of real GDP is not always smooth; the bumps and wiggles correspond to short periods of faster or slower growth.
FIGURE 24.1 Fluctuations in U.S. Real GDP, 1929–2016.Real GDP does not grow smoothly but has speedups (expansions or booms) and slowdowns (recessions or depressions).Source: Federal Reserve of St. Louis Economic Data (FRED), https://fred.stlouisfed.org/series/GDPCA.
A period in which the economy is growing at a rate significantly below normal is called a recession or a contraction. An extremely severe or protracted recession is called a depression. You should be able to pick out the Great Depression in Figure 24.1, particularly the sharp initial decline between 1929 and 1933. But you can also see that the U.S. economy was volatile in the mid-1970s and the early 1980s, with serious recessions in 1973–1975 and 1981–1982. A moderate recession (but not moderate enough for the first President Bush) occurred in 1990–1991. The next recession began in March 2001, exactly 10 years after the end of the 1990–1991 recession, which was declared over as of March 1991. This 10-year period without a recession was the longest such period in U.S. history, and the 2001 recession that ended it was again short and relatively mild, lasting eight months. In contrast, the latest recession—the Great Recession—was long and severe. Its beginning in 2007 and its end in 2009 are clearly visible in Figure 24.1.
A more informal definition of a recession, often cited by reporters, is a period during which real GDP falls for at least two consecutive quarters. This definition is not a bad rule of thumb because real GDP usually does fall during recessions. However, many economists would argue that periods in which real GDP growth is well below normal, though not actually negative, should be counted as recessions. Indeed, real GDP fell in only one quarter during the 2001 recession. Another problem with relying on GDP figures for dating recessions is that GDP data can be substantially revised, sometimes years after the fact. In practice, when trying to determine whether a recession is in progress, economists look at a variety of economic data, not just GDP.
Table 24.1 lists the beginning and ending dates of U.S. recessions since 1929, as well as the duration (length, in months) of each. The table also gives the highest unemployment rate recorded during each recession and the percentage change in real GDP. (Ignore the last column of the table for now.) The beginning of a recession is called the peak, because it represents the high point of economic activity prior to a downturn. The end of a recession, which marks the low point of economic activity prior to a recovery, is called the trough. The dates of peaks and troughs reported in Table 24.1 were determined by the National Bureau of Economic Research (NBER), a nonprofit organization of economists that has been a major source of research on short-term economic fluctuations since its founding in 1920 (see The Economic Naturalist 24.2). The NBER is not a government agency, but it is usually treated by the news media and the government as the “official” arbiter of the dates of peaks and troughs.
Table 24.1 shows that since 1929, by far the longest and most severe recession in the United States was the 43-month economic collapse that began in August 1929 and lasted until March 1933, initiating what became known as the Great Depression. Between 1933 and 1937, the economy grew fairly rapidly, so technically the period was not a recession, although unemployment remained very high at close to 20 percent of the workforce. In 1937–1938, the nation was hit by another significant recession. Full economic recovery from the Depression did not come until U.S. entry into World War II at the end of 1941. The economy boomed from 1941 to 1945 (see Figure 24.1), reflecting the enormous wartime production of military equipment and supplies.
In sharp contrast to the 1930s, U.S. recessions since World War II have generally been short—between 6 and 18 months, from peak to trough. As Table 24.1 shows, the two most severe postwar recessions prior to 2007, 1973–1975 and 1981–1982, lasted just 16 months. And, though unemployment rates during those two recessions were quite high by today’s standards, they were low compared to the Great Depression. During the quarter century (25 years) from 1982 to 2007, the U.S. economy has experienced only two relatively mild recessions, in 1990–1991 and in 2001. The decline in macroeconomic volatility during those years was dubbed the Great Moderation, and some economists and other observers wondered whether we were witnessing “the end of the business cycle.” But then came the 2007–2009 recession, the longest and deepest since the end of World War II, lasting 18 months with annual real GDP falling 3.1 percent from peak year to trough year and the annual unemployment rate reaching 9.6 percent. Such events warn us to guard against overconfidence. Prosperity and economic stability can never be guaranteed.
The opposite of a recession is an expansion—a period in which the economy is growing at a rate that is significantly above normal. A particularly strong and protracted expansion is called a boom. In the United States, strong expansions occurred during 1933–1937, 1961–1969, 1982–1990, and 1991–2001, with exceptionally strong growth during 1995–2000 (see Figure 24.1). On average, expansions have been much longer than recessions. The final column of Table 24.1 shows the duration, in months, of U.S. expansions since 1929. As you can see in the table, the 1961–1969 expansion lasted 106 months; the 1982–1990 expansion, 92 months. The longest expansion of all began in March 1991, at the trough of the 1990–1991 recession. This expansion lasted 120 months, a full 10 years, until a new recession began in March 2001.
The Economic Naturalist 24.2
How was the 2007 recession called?
The Business Cycle Dating Committee of the National Bureau of Economic Research determined that a recession began in December 2007. What led the committee to choose that date?
The seven economists who form the Business Cycle Dating Committee met by conference call on Friday, November 28, 2008, and announced on December 1, 2008, that a recession had begun one year earlier.
The determination of whether and when a recession has begun involves intensive statistical analysis, mixed in with a significant amount of human judgment. Indeed, it took a full year’s worth of economic data before the committee called the recession. The Business Cycle Dating Committee typically relies heavily on a small set of statistical indicators that measure the overall strength of the economy. The committee prefers indicators that are available monthly because they are available quickly and may provide relatively precise information about the timing of peaks and troughs. Four of the most important indicators used by the committee are as follows:
· Industrial production, which measures the output of factories and mines.
· Total sales in manufacturing, wholesale trade, and retail trade.
· Nonfarm employment (the number of people at work outside of agriculture).
· Real after-tax income received by households, excluding transfers like Social Security payments.
Each of these indicators measures a different aspect of the economy. Because their movements tend to coincide with the overall movements in the economy, they are called coincident indicators.
In its long and detailed statement calling the recession (available at www.nber.org/cycles/dec2008.pdf), the committee included the following text:
Because a recession is a broad contraction of the economy, not confined to one sector, the committee emphasizes economy-wide measures of economic activity. The committee believes that domestic production and employment are the primary conceptual measures of economic activity. The committee views the payroll employment measure, which is based on a large survey of employers, as the most reliable comprehensive estimate of employment. This series reached a peak in December 2007 and has declined every month since then.
In its deliberations, the committee also looked at quarterly domestic production measures (including GDP), which, as the committee determined, did not speak clearly about the date of the peak in activity. Other indicator series considered by the committee reached peaks in the few months before or after December 2007, providing evidence consistent with an economywide December peak, in support of the committee’s dating decision.
CONCEPT CHECK 24.1
Using the National Bureau of Economic Research website (www.nber.org/cycles.html), is the U.S. economy currently in recession or expansion? How much time has elapsed since the last peak or trough? Explore the NBER website to find additional useful information about current conditions in the U.S. economy.
SOME FACTS ABOUT SHORT-TERM ECONOMIC FLUCTUATIONS
Although Figure 24.1 and Table 24.1 show data starting only in 1929, periods of expansion and recession have been a feature of industrial economies since at least the late eighteenth century. Karl Marx and Friedrich Engels referred to these fluctuations, which they called “commercial crises,” in their Communist Manifesto of 1848. In the United States, economists have been studying short-term fluctuations for at least a century. The traditional term for these fluctuations is business cycles, and they are still often referred to as cyclical fluctuations. Neither term is accurate though; as Figure 24.1 shows, economic fluctuations are not “cyclical” at all in the sense that they recur at predictable intervals, but instead are irregular in their length and severity. This irregularity makes the dates of peaks and troughs extremely hard to predict, despite the fact that professional forecasters have devoted a great deal of effort and brainpower to the task.
Expansions and recessions usually are not limited to a few industries or regions but, as noted in The Economic Naturalist 24.2, are felt throughout the economy. Indeed, the largest fluctuations may have a global impact. For instance, the Great Depression of the 1930s affected nearly all the world’s economies, and the 1973–1975 and 1981–1982 recessions were also widely felt outside the United States. When East Asia suffered a major slowdown in the late 1990s, the effects of that slowdown spilled over into many other regions (although not so much the United States).
Recessions are very difficult to forecast.
As you already know, the 2007–2009 recession quickly became worldwide in scope, and some of its effects are still being felt around the world today. But even a relatively moderate recession, like the one that occurred in 2001, can have global effects. Figure 24.2, which shows annual growth rates of real GDP over the period 1999–2014 for China, Germany, Japan, the United Kingdom, and the United States, illustrates this point. (The figure’s shaded areas show U.S. recession dates, taken from Table 24.1.) You can see that all five economies—the world’s largest by GDP—slowed significantly in 2008 and, except for China, they all contracted rather significantly in 2009. All five economies also started recovering together, but after a promising 2010, they all slowed again in 2011 and have, in general, been growing more slowly in recent years than in the years just before the crisis. Figure 24.2 also shows that all five economies slowed at least somewhat from 2000 to 2001.
FIGURE 24.2 Real GDP Growth in Five Major Countries, 1999–2014.Annual growth rates (measured as the change in real GDP over the past four quarters) for the world’s five largest economies show that all the countries slowed somewhat in 2001—the year of the previous recession—and slowed significantly in 2008 and 2009—during the latest, much more severe, recession.Source: Federal Reserve of St. Louis Economic Data (FRED), https://research.stlouisfed.org/fred2
Unemployment is a key indicator of short-term economic fluctuations. The unemployment rate typically rises sharply during recessions and recovers (although more slowly) during expansions. Figure 16.3 showed the U.S. unemployment rate since 1929, and Figure 17.4 showed the rate since 1965. You should be able to identify, most recently, the recessions that began in 1969, 1973, 1981, 1990, 2001, and 2007 by noting the sharp peaks in the unemployment rate in those or the following years. Recall that the part of unemployment that is associated with recessions is called cyclical unemployment. Beyond this increase in unemployment, labor market conditions generally worsen during recessions. For example, during recessions real wages grow more slowly, workers are less likely to receive promotions or bonuses, and new entrants to the labor force (such as college graduates) have a much tougher time finding attractive jobs.
Generally, industries that produce durable goods, such as cars, houses, and capital equipment, are more affected than others by recessions and booms. In contrast, industries that provide services and nondurable goods like food are much less sensitive to short-term fluctuations. Thus an automobile worker or a construction worker is far more likely to lose his or her job in a recession than is a barber or a baker.
Like unemployment, inflation follows a typical pattern in recessions and expansions, though it is not so sharply defined. Figure 16.4 showed the U.S. inflation rate since 1929, and Figure 24.3 shows the rate since 1960 (periods of recession are again indicated by shaded vertical bars). As you can see, recessions tend to be followed soon after by a decline in the rate of inflation. For example, the recession of 1981–1982 was followed by a sharp reduction in inflation, and the recession of 2007–2009 ended with slightly negative inflation. Furthermore, many—though not all—postwar recessions have been preceded by increases in inflation, as Figure 24.3 shows. The behavior of inflation during expansions and recessions will be discussed more fully in the next two chapters.
FIGURE 24.3 U.S. Inflation, 1960–2016.U.S. inflation since 1960 is measured by the change in the CPI, and periods of recession are indicated by the shaded vertical bars. Note that inflation declined during or following each of those recessions and rose prior to many of those recessions.Source: U.S. Bureau of Labor Statistics, www.bls.gov
RECAP
RECESSIONS, EXPANSIONS, AND SHORT-TERM ECONOMIC FLUCTUATIONS
· A recession is a period in which output is growing more slowly than normal. An expansion, or boom, is a period in which output is growing more quickly than normal
· The beginning of a recession is called the peak, and its end (which corresponds to the beginning of the subsequent expansion) is called the trough.
· The sharpest recession in the history of the United States was the initial phase of the Great Depression in 1929–1933. Severe recessions also occurred in 1973–1975, 1981–1982, and 2007–2009. Two relatively mild recessions occurred in 1990–1991 and 2001.
· Short-term economic fluctuations (recessions and expansions) are irregular in length and severity, and thus are difficult to predict.
· Expansions and recessions have widespread (and sometimes global) impacts, affecting most regions and industries.
· Unemployment rises sharply during a recession and falls, usually more slowly, during an expansion.
· Durable goods industries are more affected by expansions and recessions than other industries. Services and nondurable goods industries are less sensitive to ups and downs in the economy.
· Recessions tend to be followed by a decline in inflation and are often preceded by an increase in inflation.
OUTPUT GAPS AND CYCLICAL UNEMPLOYMENT
How can we tell whether a particular recession or expansion is “big” or “small”? The answer to this question is important to both economists who study business cycles and policymakers who must formulate responses to economic fluctuations. Intuitively, a “big” recession or expansion is one in which output and the unemployment rate deviate significantly from their normal or trend levels. In this section, we will attempt to be more precise about this idea by introducing the concept of the output gap, which measures how far output is from its normal level at a particular time. We will also revisit the idea of cyclical unemployment, or the deviation of unemployment from its normal level. Finally, we will examine how these two concepts are related.
POTENTIAL OUTPUT
The concept of potential output is a useful starting point for thinking about the measurement of expansions and recessions. Potential output, also called potential GDP or full-employment output, is the amount of output (real GDP) that an economy can produce when using its resources, such as capital and labor, at normal rates. The term “potential output” is slightly misleading, in that potential output is not the same as maximum output. Because capital and labor can be utilized at greater-than-normal rates, at least for a time, a country’s actual output can exceed its potential output. These greater-than-normal utilization rates, however, cannot be sustained indefinitely, partly because workers cannot work overtime every week and machinery must occasionally be shut down for maintenance and repairs.
Potential output is not a fixed number but grows over time, reflecting increases in both the amounts of available capital and labor and their productivity. We discussed the sources of growth in potential output (the economy’s productive capacity) in Chapter 19, Economic Growth, Productivity, and Living Standards. We will use the symbol Y* to signify the economy’s potential output at a given point in time. Figure 24.4 presents estimated potential output for the United States from 1949 to 2016. Compare this graph with the data on actual real GDP shown in Figure 24.1. Notice that the estimate for potential output, Y*, is much smoother than actual output, Y. This reflects the fact that increases in the economy’s productive capacity are due to factors (such as human capital) that grow relatively smoothly over time. Potential output is therefore predicted to grow relatively smoothly as well.
FIGURE 24.4 U.S. Potential Output, 1949–2016.Estimated potential output Y* grows more smoothly than actual output Y. Compare these data with Figure 24.1.Source: U.S. Congressional Budget Office, Real Potential Gross Domestic Product [GDPPOT], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/GDPPOT, November 1, 2017.
Why does a nation’s output sometimes grow quickly and sometimes slowly, as shown for the United States in Figure 24.1? Logically, there are two possibilities: First, changes in the rate of output growth may reflect changes in the rate at which the country’s potential output is increasing. For example, unfavorable weather conditions, such as a severe drought, would reduce the rate of potential output growth in an agricultural economy, and a decline in the rate of technological innovation might reduce the rate of potential output growth in an industrial economy. Under the assumption that the country is using its resources at normal rates, so that actual output equals potential output, a significant slowdown in potential output growth would tend to result in recession. Similarly, new technologies, increased capital investment, or a surge in immigration that swells the labor force could produce unusually brisk growth in potential output, and hence an economic boom.
Undoubtedly, changes in the rate of growth of potential output are part of the explanation for expansions and recessions. In the United States, for example, the economic boom of the second half of the 1990s was propelled in part by new information technologies, such as the Internet. And the slow recovery since 2009 from the financial crisis seems to reflect, at least in part, a slowdown in potential output caused by demographic changes and slow productivity growth. When changes in the rate of GDP growth reflect changes in the growth rate of potential output, the appropriate policy responses are those discussed in Chapter 19, Economic Growth, Productivity, and Living Standards. In particular, when a recession results from slowing growth in potential output, the government’s best response is to try to promote saving, investment, technological innovation, human capital formation, and other activities that support growth.
THE OUTPUT GAP
A second possible explanation for short-term economic fluctuations is that actual output does not always equal potential output. For example, potential output may be growing normally, but for some reason the economy’s capital and labor resources may not be fully utilized, so that actual output is significantly below the level of potential output. This low level of output, resulting from underutilization of economic resources, would generally be interpreted as a recession. Alternatively, capital and labor may be working much harder than normal—firms may put workers on overtime, for example—so that actual output expands beyond potential output, creating a boom.
At any point in time, the difference between actual output and potential output is called the output gap. Recalling that Y* is the symbol for potential output and that Y stands for actual output (real GDP), we can express the output gap as Y − Y*. A negative output gap—when actual output is below potential, and resources are not being fully utilized—is called a recessionary gap. A positive output gap—when actual output is above potential, and resources are being utilized at above-normal rates—is referred to as an expansionary gap.
How big is the output gap at a point in time? The difference Y − Y* does not provide a complete answer because we often want to compare it to the potential size of the economy, which grows over time. For instance, a difference of $100 billion between actual and potential output is large compared to potential output of $2 trillion (roughly the size of potential GDP in the early 1950s, in 2009 dollars), but much smaller compared to $15 trillion of potential output (about the level of potential GDP in 2009). To compare the difference between actual and potential output with the economy’s potential output, we often calculate the output gap as percentage of potential output:
Figure 24.5 shows the output gap in percent for the U.S. from 1949 to 2016. You’ll notice both expansionary gaps and recessionary gaps. In particular, notice the large recessionary gaps in the early 1980s and in the late 2000s, reflecting the severe recessions the U.S. economy experienced from 1980 to 1982 and from 2007 to 2009.
FIGURE 24.5 The Output Gap in the U.S., 1949–2016.Source: Authors’ calculations using data from Figures 24.1 and 24.4.
Policymakers generally view both recessionary gaps and expansionary gaps as problems. It is not difficult to see why a recessionary gap is bad news for the economy: When there is a recessionary gap, capital and labor resources are not being fully utilized, and output and employment are below normal levels (that is, they are below maximum sustainable levels). This is the sort of situation that poses problems for politicians’ reelection prospects, as discussed in The Economic Naturalist 24.1. An expansionary gap is considered a problem by policymakers for a more subtle reason: What’s wrong, after all, with having higher output and employment than normal? A prolonged expansionary gap is problematic because, when faced with a demand for their products that significantly exceeds their normal capacity, firms tend to raise prices. Thus an expansionary gap typically results in increased inflation, which reduces the efficiency of the economy in the longer run.
Thus, whenever an output gap exists, whether it is recessionary or expansionary, policymakers have an incentive to try to eliminate the gap by returning actual output to potential. In this and the next chapters we will discuss both how output gaps arise and the tools that policymakers have for stabilizing the economy—that is, bringing actual output into line with potential output.
THE NATURAL RATE OF UNEMPLOYMENT AND CYCLICAL UNEMPLOYMENT
Efficiency
Whether recessions arise because of slower growth in potential output or because actual output falls below potential, they bring bad times. In either case, output falls (or at least grows more slowly), implying reduced living standards. Recessionary output gaps are particularly frustrating for policymakers, however, because they imply that the economy has the capacity to produce more, but for some reason available resources are not being fully utilized. Recessionary gaps violate the Efficiency Principle in that they unnecessarily reduce the total economic pie, making the typical person worse off.
An important indicator of the low utilization of resources during recessions is the unemployment rate. In general, a high unemployment rate means that labor resources are not being fully utilized so that output has fallen below potential (a recessionary gap). By the same logic, an unusually low unemployment rate suggests that labor is being utilized at a rate greater than normal so that actual output exceeds potential output (an expansionary gap).
To better understand the relationship between the output gap and unemployment, recall from Chapter 20, The Labor Market: Workers, Wages, and Unemployment, the three broad types of unemployment: frictional unemployment, structural unemployment, and cyclical unemployment. Frictional unemployment is the short-term unemployment that is associated with the matching of workers and jobs. Some amount of frictional unemployment is necessary for the labor market to function efficiently in a dynamic, changing economy. Structural unemployment is the long-term and chronic unemployment that occurs even when the economy is producing at its normal rate. Structural unemployment often results when workers’ skills are outmoded and do not meet the needs of employers—so, for example, steelworkers may become structurally unemployed as the steel industry goes into a long-term decline, unless those workers can retrain to find jobs in growing industries. Finally, cyclical unemployment is the extra unemployment that occurs during periods of recession.
Unlike cyclical unemployment, which is present only during recessions, frictional unemployment and structural unemployment are always present in the labor market, even when the economy is operating normally. Economists call the part of the total unemployment rate that is attributable to frictional and structural unemployment the natural rate of unemployment. Put another way, the natural rate of unemployment is the unemployment rate that prevails when cyclical unemployment is zero, so that the economy has neither a recessionary nor an expansionary output gap. We will denote the natural rate of unemployment as u*.
Cyclical unemployment, which is the difference between the total unemployment rate and the natural rate, can thus be expressed as u − u*, where u is the actual unemployment rate and u* denotes the natural rate of unemployment. In a recession, the actual unemployment rate u exceeds the natural unemployment rate u*, so cyclical unemployment, u − u*, is positive. When the economy experiences an expansionary gap, in contrast, the actual unemployment rate is lower than the natural rate, so that cyclical unemployment is negative. Negative cyclical unemployment corresponds to a situation in which labor is being used more intensively than normal, so that actual unemployment has dipped below its usual frictional and structural levels.
The Economic Naturalist 24.3
Why has the natural rate of unemployment in the United States declined?
According to the Congressional Budget Office, which regularly estimates the natural rate of unemployment in the United States, the long-term natural rate fell steadily from 1979 to 1999, from 6.3 percent of the labor force to about 5 percent.1 The natural rate then stayed at roughly 5 percent for 10 years. It was estimated to have risen some after the financial crisis (from 2009 to 2013) but has since declined again, to below 4.8 percent in 2016, and the CBO predicts that it will keep declining in the next 10 years. Why is the U.S. natural rate of unemployment apparently so much lower nowadays than it was in the late 1970s?
The natural rate of unemployment may have fallen because of reduced frictional unemployment, reduced structural unemployment, or both. A variety of ideas have been advanced to explain declines in both types of unemployment. One promising suggestion is based on the changing age structure of the U.S. labor force.2 The average age of U.S. workers is rising, reflecting the aging of the baby boom generation. Indeed, over the past four decades, the share of the labor force aged 16–24 has fallen from about 25 percent to below 14 percent and is projected by the BLS to keep falling to about 12 percent by 2026. Since young workers are more prone to unemployment than older workers, the aging of the labor force may help to explain the overall decline in unemployment.
Why are young workers more likely to be unemployed? Compared to teenagers and workers in their twenties, older workers are much more likely to hold long-term, stable jobs. In contrast, younger workers tend to hold short-term jobs, perhaps because they are not ready to commit to a particular career or because their time in the labor market is interrupted by schooling or military service. Because they change jobs more often, younger workers are more prone than others to frictional unemployment. They also have fewer skills, on average, than older workers, so they may experience more structural unemployment. As workers age and gain experience, however, their risk of unemployment declines.
Another possible explanation for the declining natural rate of unemployment is that labor markets have become more efficient at matching workers with jobs, thereby reducing both frictional and structural unemployment. For example, agencies that arrange temporary help have become much more commonplace in the United States in recent years. Although the placements these agencies make are intended to be temporary, they often become permanent when an employer and worker discover that a particularly good match has been made. Online job services, which allow workers to search for jobs nationally and even internationally, have also become increasingly important. By reducing the time people must spend in unemployment and by creating more lasting matches between workers and jobs, temporary help agencies, online job services, job-search apps, and similar innovations may have reduced the natural rate of unemployment.3
RECAP
OUTPUT GAPS AND CYCLICAL UNEMPLOYMENT
· Potential output is the amount of output (real GDP) that an economy can produce when using its resources, such as capital and labor, at normal rates. The output gap, Y − Y*, is the difference between actual output Y and potential output Y*. The output gap in percent is:
· When actual output is below potential, the resulting output gap is called a recessionary gap. When actual output is above potential, the difference is called an expansionary gap.
· A recessionary gap reflects a waste of resources, while an expansionary gap threatens to ignite inflation; hence policymakers have an incentive to try to eliminate both types of output gaps.
· The natural rate of unemployment u* is the sum of the frictional and structural unemployment rates. It is the rate of unemployment that is observed when the economy is operating at a normal level, with no output gap.
· Cyclical unemployment, u − u*, is the difference between the actual unemployment rate u and the natural rate of unemployment u*. Cyclical unemployment is positive when there is a recessionary gap, negative when there is an expansionary gap, and zero when there is no output gap.
OKUN’S LAW
What is the relationship between an output gap and the amount of cyclical unemployment in the economy? We have already observed that by definition, cyclical unemployment is positive when the economy has a recessionary gap, negative when there is an expansionary gap, and zero when there is no output gap. A more quantitative relationship between cyclical unemployment and the output gap is given by a rule of thumb called Okun’s law, after Arthur Okun, one of President Kennedy’s chief economic advisors. According to Okun’s law , each 1 percent increase in cyclical unemployment is associated with about a 2 percent widening of a negative output gap, measured in relation to potential output.4 So, for example, if cyclical unemployment increases from 1 percent to 2 percent of the labor force, the recessionary gap will increase from −2 percent to −4 percent of potential GDP.
We can also express Okun’s law as an equation. Using our expression for the output gap, we have
The following example further illustrates Okun’s law.
EXAMPLE 24.1Okun’s Law and the Output Gap in the U.S. Economy
How is Okun’s law applied to real-world data?
The table below presents data on the actual unemployment rate, the natural unemployment rate, and potential GDP (in billions of 2009 dollars) for the U.S. economy in five selected years.
In 1995, cyclical unemployment, u − u*, was 0.3 percent of the labor force (5.6% − 5.3%). Applying Okun’s law, the output gap for 1995 was −2 times that percentage, or −0.6 percent of potential output. Potential output was estimated to be $10,319.0 billion, so the value of the negative output gap for that year was $61.9 billion.
2000 was near the end of an expansion and the actual unemployment rate was below the natural rate. Specifically, cyclical unemployment was −1.0 percent; using Okun’s law, this means that the output gap was 2.0 percent and the U.S. economy’s output was $248.2 billion more than it typically would have been in 2000.
The data for 2010 give a sense of the depth of the most recent recession. Although the 2007–2009 recession had already ended, cyclical unemployment was still very high, at 4.5 percent, implying an output gap of −9.0 percent. Thus, according to Okun’s law, the U.S. economy was producing about $1,391.1 (0.09 × 15,457.0) billion less than it would produce had all resources been fully employed. There were about 309 million people in the United States in 2010, so Okun’s law implies that average incomes (i.e., per capita GDP) could have been around $4,500 higher in 2010—about $18,000 for a family of four—had the economy not been operating below potential. Thus output gaps and cyclical unemployment have significant costs, a conclusion that justifies the concern that the public and policymakers have about recessions.
Of course, Okun’s law is only a rule of thumb. The Congressional Budget Office’s estimates of the output gap for the years shown in the table above are somewhat different from the numbers we have just calculated. However, those estimates, which use better—and significantly more complicated—techniques to more accurately estimate the output gap, are in the same order of magnitude as the numbers we calculated above.
CONCEPT CHECK 24.2
According to the table in Example 24.1, in 2015 the U.S. unemployment rate was 5.3 percent. The Congressional Budget Office estimated that in 2015 the natural rate of unemployment was 4.8 percent. Applying Okun’s law, by what percentage did actual GDP differ from potential GDP in 2015?
The Economic Naturalist 24.4
Why did the Federal Reserve act to slow down the economy in 1999 and 2000?
As we have noted in earlier chapters, monetary policy decisions of the Federal Reserve—actions that change the level of the nation’s money supply—affect the performance of the U.S. economy. Why did the Federal Reserve take measures to slow down the economy in 1999 and 2000?
Throughout the 1990s, cyclical unemployment in the United States fell dramatically, becoming negative sometime in 1997, according to Congressional Budget Office estimates. Okun’s law indicates that growing negative cyclical unemployment rates signal an increasing expansionary gap and, with it, an increased risk of future inflation.
In 1997 and 1998, the Federal Reserve argued that the inflationary pressures typically caused by rapidly expanding output and falling unemployment rates were being offset by productivity gains and international competition, leaving inflation rates lower than expected. Because inflation remained low during this period—despite a small but growing expansionary gap—the Federal Reserve did little to eliminate the gap.
However, as the actual unemployment rate continued to fall throughout 1999 and early 2000, the expansionary gap continued to widen, causing the Federal Reserve to grow increasingly concerned about the growing imbalance between actual and potential GDP and the threat of increasing inflation. In response, the Federal Reserve took actions in 1999 and 2000 to slow the growth of output and bring actual and potential output closer into alignment (we will give more details in Chapter 26, Stabilizing the Economy: The Role of the Fed, and Chapter 27, Aggregate Demand, Aggregate Supply, and Inflation, about how the Fed can do this). The Fed’s actions helped to “promote overall balance in the economy”5 and restrain inflation throughout 2000. By early 2001, however, the U.S. economy stalled and fell into recession, leading the Federal Reserve to reverse course and take policy measures aimed at eliminating the growing recessionary gap.
RECAP
OKUN’S LAW
Okun’s law relates cyclical unemployment and the output gap. According to this rule of thumb, each percentage point increase in cyclical unemployment is associated with about a 2 percent widening of a negative output gap, measured in relation to potential output.
WHY DO SHORT-TERM FLUCTUATIONS OCCUR? A PREVIEW AND A TALE
What causes periods of recession and expansion? In the preceding sections, we discussed two possible reasons for slowdowns and speedups in real GDP growth. First, growth in potential output itself may slow down or speed up, reflecting changes in the growth rates of available capital and labor and in the pace of technological progress. Second, even if potential output is growing normally, actual output may be higher or lower than potential output—that is, expansionary or recessionary output gaps may develop. Earlier in this book, we discussed some of the reasons that growth in potential output can vary, and the options that policymakers have for stimulating growth in potential output. But we have not yet addressed the question of how output gaps can arise or what policymakers should do in response. The causes and cures of output gaps will be a major topic of the next three chapters. Here is a brief preview of the main conclusions of these chapters:
· 1.In a world in which prices adjusted immediately to balance the quantities supplied and demanded for all goods and services, output gaps would not exist. However, for many goods and services, the assumption that prices will adjust immediately is not realistic. Instead, many firms adjust the prices of their output only periodically. In particular, rather than changing prices with every variation in demand, firms tend to adjust to changes in demand in the short run by varying the quantity of output they produce and sell. This type of behavior is known as “meeting the demand” at a preset price.
· 2.Because, in the short run, firms tend to meet the demand for their output at preset prices, changes in the amount that customers decide to spend will affect output. When total spending is low for some reason, output may fall below potential output; conversely, when spending is high, output may rise above potential output. In other words, changes in economywide spending are the primary cause of output gaps. Thus government policies can help to eliminate output gaps by influencing total spending. For example, the government can affect total spending directly simply by changing its own level of purchases.
· 3.Although firms tend to meet demand in the short run, they will not be willing to do so indefinitely. If customer demand continues to differ from potential output, firms will eventually adjust their prices to eliminate output gaps. If demand exceeds potential output (an expansionary gap), firms will raise their prices aggressively, spurring inflation. If demand falls below potential output (a recessionary gap), firms will raise their prices less aggressively or even cut prices, reducing inflation.
· 4.Over the longer run, price changes by firms eliminate any output gap and bring production back into line with the economy’s potential output. Thus the economy is “self-correcting” in the sense that it operates to eliminate output gaps over time. Because of this self-correcting tendency, in the long run actual output equals potential output, so that output is determined by the economy’s productive capacity rather than by the rate of spending. In the long run, total spending influences only the rate of inflation.
These ideas will become clearer as we proceed through the next chapters. Before plunging into the details of the analysis, though, let’s consider an example that illustrates the links between spending and output in the short and long run.
AL’S ICE CREAM STORE: A TALE ABOUT SHORT-RUN FLUCTUATIONS
Al’s ice cream store produces gourmet ice cream on the premises and sells it directly to the public. What determines the amount of ice cream that Al produces on a daily basis? The productive capacity, or potential output, of the shop is one important factor. Specifically, Al’s potential output of ice cream depends on the amount of capital (number of ice cream makers) and labor (number of workers) that he employs and on the productivity of that capital and labor. Although Al’s potential output usually changes rather slowly, on occasion it can fluctuate significantly—for example, if an ice cream maker breaks down or Al contracts the flu.
The main source of day-to-day variations in Al’s ice cream production, however, is not changes in potential output but fluctuations in the demand for ice cream by the public. Some of these fluctuations in spending occur predictably over the course of the day (more demand in the afternoon than in the morning, for example), the week (more demand on weekends), or the year (more demand in the summer). Other changes in demand are less regular—more demand on a hot day than a cool one, or when a parade is passing by the store. Some changes in demand are hard for Al to interpret: For example, a surge in demand for rocky road ice cream on one particular Tuesday could reflect a permanent change in consumer tastes, or it might just be a random, one-time event.
How should Al react to these ebbs and flows in the demand for ice cream? The basic supply and demand model that we introduced in Chapter 3, Supply and Demand, if applied to the market for ice cream, would predict that the price of ice cream should change with every change in the demand for ice cream. For example, prices should rise just after the movie theater next door to Al’s shop lets out on Friday night, and they should fall on unusually cold, blustery days, when most people would prefer a hot cider to an ice cream cone. Indeed, taken literally, the supply and demand model predicts that ice cream prices should change almost moment to moment. Imagine Al standing in front of his shop like an auctioneer, calling out prices in an effort to determine how many people are willing to buy at each price!
Cost-Benefit
Of course, we do not expect to see this behavior by an ice cream store owner. Price setting by auction does in fact occur in some markets, such as the market for grain or the stock market, but it is not the normal procedure in most retail markets, such as the market for ice cream. Why this difference? The basic reason is that sometimes the economic benefits of hiring an auctioneer and setting up an auction exceed the costs of doing so, and sometimes they do not. In the market for grain, for example, many buyers and sellers gather together in the same place at the same time to trade large volumes of standardized goods (bushels of grain). In that kind of situation, an auction is an efficient way to determine prices and balance the quantities supplied and demanded. In an ice cream store, by contrast, customers come in by twos and threes at random times throughout the day. Some want shakes, some cones, and some sodas. With small numbers of customers and a low sales volume at any given time, the costs involved in selling ice cream by auction are much greater than the benefits of allowing prices to vary with demand.
So how does Al the ice cream store manager deal with changes in the demand for ice cream? Observation suggests that he begins by setting prices based on the best information he has about the demand for his product and the costs of production. Perhaps he prints up a menu or makes a sign announcing the prices. Then, over a period of time, he will keep his prices fixed and serve as many customers as want to buy (up to the point where he runs out of ice cream or room in the store at these prices). This behavior is what we call “meeting the demand” at preset prices, and it implies that in the short run, the amount of ice cream Al produces and sells is determined by the demand for his products.
However, in the long run, the situation is quite different. Suppose, for example, that Al’s ice cream earns a citywide reputation for its freshness and flavor. Day after day Al observes long lines in his store. His ice cream maker is overtaxed, as are his employees and his table space. There can no longer be any doubt that at current prices, the quantity of ice cream the public wants to consume exceeds what Al is able and willing to supply on a normal basis (his potential output). Expanding the store is an attractive possibility, but not one (we assume) that is immediately feasible. What will Al do?
Certainly one thing Al can do is raise his prices. At higher prices, Al will earn higher profits. Moreover, raising ice cream prices will bring the quantity of ice cream demanded closer to Al’s normal production capacity—his potential output. Indeed, when the price of Al’s ice cream finally rises to its equilibrium level, the shop’s actual output will equal its potential output. Thus, over the long run, ice cream prices adjust to their equilibrium level, and the amount that is sold is determined by potential output.
This example illustrates, in a simple way, the links between spending and output—except, of course, that we must think of this story as applying to the whole economy, not to a single business. The key point is that there is an important difference between the short run and the long run. In the short run, producers often choose not to change their prices, but rather to meet the demand at preset prices. Because output is determined by demand, in the short run total spending plays a central role in determining the level of economic activity. Thus Al’s ice cream store enjoys a boom on an unusually hot day, when the demand for ice cream is strong, while an unseasonably cold day brings an ice cream recession. But in the long run, prices adjust to their market-clearing levels, and output equals potential output. Thus the quantities of inputs and the productivity with which they are used are the primary determinants of economic activity in the long run, as we saw in Chapter 19, Economic Growth, Productivity, and Living Standards. Although total spending affects output in the short run, in the long run its main effects are on prices.
SUMMARY
· Real GDP does not grow smoothly. Periods in which the economy is growing at a rate significantly below normal are called recessions; periods in which the economy is growing at a rate significantly above normal are called expansions. A severe or protracted recession, like the long decline that occurred between 1929 and 1933, is called a depression, while a particularly strong expansion is called a boom. (LO1)
· The beginning of a recession is called the peak because it represents the high point of economic activity prior to a downturn. The end of a recession, which marks the low point of economic activity prior to a recovery, is called the trough. Since World War II, U.S. recessions have been much shorter on average than booms, lasting between 6 and 18 months. The longest boom period in U.S. history began with the end of the 1990–1991 recession in March 1991, ending exactly 10 years later in March 2001 when a new recession began. (LO1)
· Short-term economic fluctuations are irregular in length and severity, and are thus hard to forecast. Expansions and recessions are typically felt throughout the economy and may even be global in scope. Unemployment rises sharply during recessions, while inflation tends to fall during or shortly after a recession. Durable goods industries tend to be particularly sensitive to recessions and booms, whereas services and nondurable goods industries are less sensitive. (LO1)
· Potential output, also called potential GDP or full-employment output, is the maximum sustainable amount of output (real GDP) that an economy can produce. The difference between the economy’s actual output and its potential output is called the output gap. When output is below potential, the gap is called a recessionary gap; when output is above potential, the difference is called an expansionary gap. Recessions can occur either because potential output is growing unusually slowly or because actual output is below potential. Because recessionary gaps represent wasted resources and expansionary gaps threaten to create inflation, policymakers have an incentive to try to eliminate both types of gap. (LO2)
· The natural rate of unemployment is the part of the total unemployment rate that is attributable to frictional and structural unemployment. Equivalently, the natural rate of unemployment is the rate of unemployment that exists when the output gap is zero. Cyclical unemployment, the part of unemployment that is associated with recessions and expansions, equals the total unemployment rate less the natural unemployment rate. (LO3)
· Cyclical unemployment is related to the output gap by Okun’s law, which states that each extra percentage point of cyclical unemployment is associated with about a 2 percent widening of a negative output gap, measured in relation to potential output. (LO4)
· Our further study of recessions and expansions will focus on the role of economywide spending. If firms adjust prices only periodically, and in the meantime produce enough output to meet demand, then fluctuations in spending will lead to fluctuations in output over the short run. During that short-run period, government policies that influence aggregate spending may help to eliminate output gaps. In the long run, however, firms’ price changes will eliminate output gaps—that is, the economy will “self-correct”—and total spending will influence only the rate of inflation. (LO5)
KEY TERMS
boom
business cycles
depression
durable goods
expansion
expansionary gap
natural rate of unemployment, u*
nondurable goods
Okun’s law
output gap
peak
potential output, Y* (or potential GDP or full-employment output)
recession (or contraction)
recessionary gap
trough
REVIEW QUESTIONS
1. 1.Define recession and expansion. What are the beginning and ending points of a recession called? In the postwar United States, which have been longer on average: recessions or expansions? (LO1)
2. 2.Which firm is likely to see its profits reduced the most in a recession: an automobile producer, a manufacturer of boots and shoes, or a janitorial service? Which is likely to see its profits reduced the least? Explain. (LO1)
3. 3.Define potential output. Is it possible for an economy to produce an amount greater than potential output? Explain. (LO2)
4. 4.How is each of the following likely to be affected by a recession: the natural unemployment rate, the cyclical unemployment rate, the inflation rate, the poll ratings of the president? (LO1, LO3)
5. 5.True or false: When output equals potential output, the unemployment rate is zero. Explain. (LO4)
6. 6.If the natural rate of unemployment is 5 percent, what is the total rate of unemployment according to Okun’s law if output is 2 percent below potential output? What if output is 2 percent above potential output? (LO4)
PROBLEMS
1. 1.Using Table 24.1, find the average duration, the minimum duration, and the maximum duration of expansions in the United States since 1929. Are expansions getting longer or shorter on average over time? Is there any tendency for long expansions to be followed by long recessions? (LO1)
2. 2.From the homepage of the Bureau of Economic Analysis (www.bea.gov), obtain quarterly data for U.S. real GDP from three recessions: 1981–1982, 1990–1991, 2001, and 2007–2009. (LO1)
a. How many quarters of negative real GDP growth occurred in each recession?
b. Which, if any, of the recessions satisfied the informal criterion that a recession must have two consecutive quarters of negative GDP growth?
3. 3.Given below are data on real GDP and potential GDP for the United States for the years 2005–2016, in billions of 2009 dollars. For each year, calculate the output gap as a percentage of potential GDP and state whether the gap is a recessionary gap or an expansionary gap. Also calculate the year-to-year growth rates of real GDP. Identify the recessions that occurred during this period. (LO2)
Source: Potential GDP, Federal Reserve Bank of St. Louis; real GDP, www.bea.gov.
4. 4.From the homepage of the Bureau of Labor Statistics (www.bls.gov), obtain the most recent available data on the unemployment rate for workers aged 16–19 and workers aged 20 or over. How do they differ? What are some of the reasons for the difference? How does this difference relate to the decline in the overall natural rate of unemployment since 1980? (LO3)
5. 5.Using Okun’s law, fill in the four pieces of missing data in the table below. The data are hypothetical. (LO4)
6. 6.Of the following, identify the incorrect statement. (LO5)
a. Output gaps are caused by inflationary pressures generated by the unintended side effects of government policy.
b. Low aggregate spending can make output fall below potential output.
c. When spending is high, output may rise above potential output.
d. Government policies can help to eliminate output gaps.
ANSWERS TO CONCEPT CHECKS
1. 24.1Answers will vary, depending on when the data are obtained. As of late 2017, the last recession was the Great Recession of 2007–2009, which ended in June 2009. While recessions are officially declared with a lag, it seems that the economy has been expanding for more than six years now (since the last trough). (LO1)
2. 24.2The actual unemployment rate in 2015 exceeded the natural rate by 0.5 percent. Applying Okun’s law, actual output fell below potential output by 1.0 percent. (LO4)
1U.S. Congressional Budget Office, Natural Rate of Unemployment (Long-Term) [NROU], retrieved from FRED, Federal Reserve Bank of St. Louis, https://fred.stlouisfed.org/series/NROU, November 1, 2017.
2See Robert Shimer, “Why Is the U.S. Unemployment Rate So Much Lower?” in B. Bernanke and J. Rotemberg, eds., NBER Macroeconomics Annual, 1998.
3For a detailed analysis of factors affecting the natural rate, see Lawrence Katz and Alan Krueger, “The High-Pressure U.S. Labor Market of the 1990s,” Brookings Papers on Economic Activity 1 (1999), pp. 1–88.
4This relationship between unemployment and output has weakened over time. When Arthur Okun first formulated his law in the 1960s, he suggested that each extra percentage point of unemployment was associated with about a 3 percentage point widening in the (negative) output gap. At the same time, the weakened relationship has held up surprisingly well over time. For a recent discussion, see Mary C. Daly, John Fernald, Òscar Jordà, and Fernanda Nechio, “Interpreting Deviations from Okun’s Law,” FRBSF Economic Letter 2014–12, April 21, 2014, www.frbsf.org/economic-research/publications/economic-letter/2014/april/okun-law-deviation-unemployment-recession/.
5Testimony of Chairman Alan Greenspan, The Federal Reserve’s Semiannual Report on the Economy and Monetary Policy, Committee on Banking and Financial Services, U.S. House of Representatives, February 17, 2000. Available online at www.federalreserve.gov/boarddocs/hh/2000/February/Testimony.htm.