This is a web-only article from Dollars & Sense: Real World Economics, available at

issue 292 cover

This article is a web-only article.

Subscribe Now

at a 30% discount.

Why Is Unemployment Still So High?

By Alejandro Reuss

If the recession is supposedly over, why is there still so much unemployment?

Overall demand (in the U.S. economy as a whole) is not enough to buy all the goods and services that U.S. workers could produce, if all the economy’s resources were fully employed. If the managers of firms do not think they will be able to sell the goods that could be produced when running at full capacity, they will lay off workers and idle plant and equipment. That’s what’s happened in the United States and other countries over the last few years. Millions of workers are unemployed, in economic language, due to insufficient “aggregate demand.”

I’ve heard that some economists believe a lot of current unemployment is “structural.” What does that mean?

Economists have traditionally classified unemployment into four categories: cyclical, structural, frictional, and seasonal.

Cyclical unemployment is caused by recessions. Recessions can have various kinds of causes, but here we will focus on a decline in total demand for goods and services. That is certainly the main cause of the recent recession.

Structural unemployment is caused by a mismatch between workers’ skills and the skills employers are seeking, or between workers’ location and the places where employers are hiring. The classic case would be workers in a declining industry, who have skills appropriate to that industry, but do not have skills needed by new, growing industries. Unemployed auto workers, for instance, undoubtedly have skills in assembly-line production, but employers are not looking for these skills (at least not in the areas where these workers live). On the other hand, workers who have lost their manufacturing jobs may not have training in, say, information technology, which employers are seeking.

Frictional unemployment is caused by the time gap between someone leaving one job (quitting, being laid off, getting fired) or starting to look for paid work (for example, after a period of stay-at-home parenting, after finishing college, etc.) and actually finding a new job. It takes time to answer job ads, for the employer to process applications and call applicants for interviews, for interviews to happen, and for the employer to make a hiring decision. In the meantime, the worker is “frictionally” unemployed.

Seasonal unemployment is caused by regular fluctuations in employment, in some industries, in the course of the year. For example, employment in tourism (in some areas) may be high during warm-weather months, but low the rest of the year.

In recent months, some prominent economists have, indeed, claimed that “structural” unemployment may be the main problem in the U.S. economy. Minneapolis Federal Reserve Bank President Narayana Kocherlakota has argued, “What does this change in the relationship between job openings and unemployment connote? In a word, mismatch. Firms have jobs, but can’t find appropriate workers. The workers want to work, but can’t find appropriate jobs. There are many possible sources of mismatch—geography, skills, demography—and they are probably all at work…. Most of the existing unemployment represents mismatch ….”

Whenever an economist attributes unemployment to a skills or geographical “mismatch,” this is the same as claiming that unemployment is mostly “structural.”

How can we tell if unemployment is structural or cyclical?

One way is by looking at the relationship between unemployment (that is, the number of people who are currently not employed, are looking for work, and are available to work) and job openings. On the graph below, the number of job openings is plotted on the vertical axis and the number of unemployed people is plotted on the horizontal axis. If we string together observations, over time, of different combinations of unemployment or job openings, we get a downward-sloping curve called the “Beveridge curve.” We can imagine that, over the course of the business cycle, we move along the Beveridge curve.

Figure: Beveridge Curve

Close to the “peak” in the business cycle, we expect the combination of unemployment and job openings to be near the upper left-hand corner of the graph. There’s high and growing demand for goods and services, employers are looking to expand output, and they want to hire more workers (so there are lots of job openings), but most workers already have jobs (so there are few unemployed workers looking for jobs). For the peak of the last boom, for example, we see the highest job openings rate (over 3%) and the lowest unemployment rates (less than 5%). During a recession, we expect the combination of unemployment and job openings to be near the lower right-hand corner of the graph. There are large numbers of unemployed workers, but not many job openings. During the recent recession, we see a steady decline in the job openings rate (to less than 2%), along with a dramatic increase in the unemployment rate (to over 10%).

Today, there are nearly 14 million unemployed workers in the United States (not counting those who have given up looking or those who are employed part time but are looking for full-time work). Meanwhile, there are a little more than 3 million job openings. In other words, there are more than four times as many people unemployed as job openings. Employers are not looking to hire large numbers of workers (nowhere near the number looking for jobs).

Strictly speaking, we can’t tell just from these numbers whether most of the unemployed workers have the skills employers are looking for or not. Is it possible that none of the unemployed workers have the jobs the employers are looking for? Sure, though there’s little evidence of that. If this were true, we would expect to see the wages of much sought-after types of workers rising sharply, which is not the case. What we know for sure is that, even if the millions of unemployed workers all had precisely the skills employers wanted, most of them still will not get hired. The problem cannot possible be just a lack of appropriate skills.

Why would some economists think that structural unemployment is the main problem now?

Figure:Beveridge Curve: Changes in cyclical unemployment are associated with movements “along” the curve

The relationship between job openings and unemployment, represented by the Beveridge curve, is not fixed in one position for all time. It can move around. Increases in structural unemployment are associated with a shift “out” in the Beveridge Curve (more job vacancies along with more unemployment). Some kind of mismatch explains why workers who want jobs are not getting hired, even though there are plenty of openings.

Figure: Beveridge Curve: Changes in structural unemployment are associated with “shifts” in the curve

Supporters of the structural-unemployment story point to the uptick late last year in job openings with no reduction in unemployment. It looks like employers are starting to look for workers again (since job openings are going up), but employers are not hiring (since unemployment is not coming down). So some economists have concluded that there must be some kind of mismatch between workers’ skills or location and employers’ needs.

They argue that this indicates an outward shift in the Beveridge curve, suggesting an increase in structural unemployment.

If the main issue is cyclical unemployment, not structural unemployment, what’s behind the “uptick”?

There may be several different things going on. One important factor, according to Murat Tasci and John Linder of the Cleveland Federal Reserve Bank, may be that some people are starting to look for work again (after giving up for a while). To be officially counted as unemployed, a person cannot have a job, must be looking for work, and must be available to work. Therefore, workers who have stopped looking for work—perhaps because they looked for a while, couldn’t find a job, and got discouraged—are not counted as unemployed. As hiring picks up in the beginnings of an economic recovery, workers who have given up the job search, and are therefore not counted as unemployed, start looking again and are added to the official count of the unemployed. This keeps the official unemployment rate from dropping, even as employers have started hiring again. During past recessions and recoveries, Tasci and Linder note, the combination of job vacancies and unemployment has typically followed a sort of “counterclockwise” pattern, with an uptick as the recovery in hiring begins. (They also note that a spike in Census hiring earlier this year was responsible for much of the increase in job openings.)

The uptick may also reflect “frictional” unemployment. Frictional unemployment is caused by the delay between new job vacancies (when employers first advertise a job) and new employment, since it takes time for people to apply for jobs, for employers to interview candidates and make a hiring decision, and for the new worker to start on the job. Finally, we cannot rule out that the uptick may include some structural unemployment—because long-unemployed workers have (or are perceived by employers as having) lost skills, have not be able to keep up-to-date with changes in their industry or occupation, or are not living in the same areas where new hiring is happening. It would not be the first time a major recession delivered the coup de grace to declining industries, and signaled a long-term jobs shortage in some regions.

You said current unemployment was mainly caused by low demand. So why is demand so low?

Demand is currently low for a number of reasons. The immediate causes center on the housing “bubble” and its collapse. When the housing market was booming, home owners felt richer with each passing day, and many were spending money freely (in many cases, even borrowing against the value of their houses). When the bubble burst, people who had been spending suddenly cut back—“consumption demand” fell dramatically. As this component of demand decreased, employers cut back on output. (No sense in producing goods that people are not going to buy.) They canceled orders for materials, machinery, new buildings, and so on—“investment demand” plummeted as well. Employers also laid off workers, who cut back on their spending. Millions of other people grew fearful of losing their jobs, and cut their spending as well.

This collapse had its roots in longer-term structural problems in the U.S. economy. In the quarter century after the Second World War, real wages of U.S. workers grew at about the same rate as labor productivity (output produced per hour of labor). Rising real wages fueled demand for consumer goods, and overall economic growth. Employers, meanwhile, benefited from demand for their products, and from fast productivity growth. (Wage growth eventually outpaced productivity growth, in the late 1960s and early 1970s, “squeezing” profits.) Since the late 1970s, in contrast, real hourly wages have stagnated, even as productivity has continued to grow. For people in the lower two-thirds of the income scale, growing consumption has depended increasingly on two factors: 1) household members doing more and more hours of paid work and 2) rising household indebtedness. (Borrowing against home equity to finance consumption, as many people did during the housing boom, was just one part of this growing reliance on debt.) Meanwhile, the distribution of income has become much more unequal. The growing gap between what workers produce per hour and what they get paid per hour has meant rising profits (which mostly go to people at the top of the income ladder). Both sides of this coin made the economy vulnerable to crisis. (See Rick Wolff, “Capitalism Hits the Fan,” Dollars & Sense, November/December 2008). Since the beginning of the recession, wages have fallen and credit has tightened. Low- and middle-income people, depending on credit to continue spending, have cut back their consumption spending dramatically. Meanwhile, high-income people can choose to spend or to save. As profitable investment opportunities have vanished, they have cut back on their investment spending.

As consumption and investment demand have collapsed, the government has implemented policies designed to stimulate demand. These efforts, however, have fallen flat. “Expansionary” monetary policies (lowering short-term interest rates in order to encourage private borrowing and spending) have long since run up against the lower bound of 0%. The government cannot push these interest rates any lower, so there’s nothing more the government do with conventional monetary policy (though it is now pursuing policies focused on longer-term rates). Increased federal spending on real goods and services and tax cuts (or fiscal “stimulus”) have also not brought about hoped-for results, partly because the planned spending increase was nowhere near large enough to make up for the collapse in private (consumption and investment) demand, and partly because it has been offset by drastic cuts in state and local government spending.

All this resulted in the vicious circle in which we are still trapped.

So what can be done to reduce unemployment? If the problem is not enough demand, isn’t the solution more demand?

The demand for labor is mostly “derived demand.” This means that it is mainly driven by demand for something else—for the goods and services employers hire workers to produce. Textbook macroeconomics recognizes that increased demand can be a cure for cyclical unemployment. As demand increases, employers may decide to increase output, and this may force them to start hiring again. There can be slippage both between increased demand and increased new output and between increased output and new employment. If employers have large unsold inventories on hand, for example, they may not increase output even if demand has started increasing. Even if employers decide to increase output, they may find that they can do so without hiring more workers (perhaps because they can make current workers, afraid of losing their jobs in a period of high unemployment, work harder and faster).

There are several kinds of fairly conventional policies that could boost demand. Increasing demand through a combination of increased government spending and decreased taxes—“injecting” demand into the economy in the form of more spending without “leakage” of demand in the form of increased taxes—is one possible response. (This is called an “expansionary fiscal policy” or “fiscal stimulus.”) Many “progressive” economists have called, rightly, for a second stimulus as an obvious way to boost demand. Another possible response is increased transfers, especially to people hit hard by the recession, such as the long-term unemployed. That directly eases the suffering caused by unemployment while also helping to boost spending and buoy up demand. Still another approach would be direct government hiring, like the Works Progress Administration (WPA) and other New Deal programs during the Great Depression. This would directly increase demand for labor, but it would also boost demand indirectly, as those hired would likely spend most or all of the wages they received.

To the extent that there is structural unemployment, the story is more complicated. If the problem is a mismatch between workers’ skills and the skills employers are seeking, an obvious response it to help workers gain new skills (like through publicly supported job retraining and educational programs or even public hiring and on-the-job training). But increased demand can help here too. This runs counter to the textbook view that policies to increase demand can’t do anything for the structurally unemployed, since they lack the skills employers are looking for, anyway. During “boom” periods in the business cycle, labor markets can become very “tight” (with few workers available to hire) and workers previously seen as “unemployable” can finally get hired, get on-the-job training and new experience, and see their overall career prospects revived.

Obviously, however, we’re very far from that situation now.

Wouldn’t boosting demand again just create another bubble, and then another crash? Is there a longer-term solution to the problem of unemployment?

The crisis shows that the existing structure of the U.S. economy could not sustain steady economic growth or full employment. It does not prove that the U.S. economy cannot produce more output than it is currently producing. There are millions of unemployed workers and vast amounts of wasted productive capacity (shuttered facilities, unused machinery, etc.) that could be put to work producing useful goods and services. Meanwhile, there are certainly unmet needs that could be satisfied by putting unemployed workers and unused resources to good use. (To cite just one example, there is a widely recognized crisis in primary and secondary education today, and yet qualified and experienced teachers are being laid off!) Such needs remain unsatisfied not because the means to satisfy them do not exist, but because the way the economy is organized causes these resources to go unused.

A capitalist economy—in which most “means of production” (like factory buildings, machinery, mines, farmland, etc.) are privately owned, most people work (for pay) for others, and owners of means of production hire workers only to produce goods they can sell at a profit—is like a very complex circuit. We can think of the circuit as beginning with production (the supply of goods and services, which depends on decisions made by millions of different individuals and firms) and completed by spending (the demand for goods and services, likewise depending on decisions made by millions of different actors). Capitalist economies can be—and have been, in different times and place—wired in different ways. Leading up to the current crisis, the circuitry of the U.S. economy became increasingly reliant on asset “bubbles” and rising debt to generate sufficient demand (enough to buy the output the economy was capable of producing). These fraying wires eventually ended up producing a short circuit.

One possible response is to re-wire the economy, while keeping the basic underlying capitalist structure. Liberal or progressive economists, for example, have argued in favor of a more regulated capitalist system, with a more equal distribution of income and wealth, a larger role for labor unions, a more extensive social “safety net,” a stronger government role in maintaining overall economic stability, etc. Such a version of capitalism might be more like what existed in the United States in the quarter century after the Second World War, or the “social democratic” systems of Western Europe of the late 20th century (which are somewhat diminished today). These economists argue that such systems could combine greater fairness with steadier economic growth and full employment, in a way that does not depend on unsustainable bubbles.

“Radical” economists, on the other hand, are not as confident that capitalism can be restructured in a way that eliminates the scourge of unemployment. Capitalism creates unemployment, they argue, for two principal reasons:

First, in capitalist economies, goods are produced for the sake of profit. Those who own and control the means of production will not hire workers and direct them to produce goods if they do not expect to sell those goods for a profit. As the productive capacity of the economy grows, however, demand does not always keep pace. Insufficient demand means goods go unsold. Capitalist employers cut back production and lay off workers. They stop spending on new buildings, machinery, materials, and so on. Workers who have lost their jobs, or are afraid of losing their jobs, cut back their spending as well. The economy can get caught in a vicious cycle: low spending causes low production, which causes low incomes, which causes low spending. This vicious cycle explains why official unemployment has hovered near 10 percent since mid 2009 (much higher if we count workers who have become discouraged and given up looking, or those who want to work full time but can only find part-time jobs). Unemployment on this scale is not the usual state in a capitalist economy, but the system does create these periodic eruptions of cyclical unemployment, sometimes relatively moderate and sometimes (like now) very severe.

Second, some unemployment is necessary for capitalists to make profits. When unemployment is relatively high, workers are fearful of losing their jobs and have relatively little bargaining power. Employers find it relatively easy to keep wages down (or even push them down further), and to keep the pace of work high (or even increase it). When unemployment is very low, workers are less fearful of losing their jobs and enjoy relatively high bargaining power. They may be able to win higher wages, and force down the pace of work. Higher wages and a lower work pace, however, drive up labor costs per unit of output, and can “squeeze” profits. Employers may respond by substituting machinery for workers, or even cutting back production (as long as it remains unprofitable). These responses, if taken by enough employers, can drive the unemployment rate back up, and bring down workers’ bargaining power again. For this reason, substantial unemployment is the normal state of affairs in a capitalist economy like that of the United States. Even during the “boom” years of 2006-2007 the number of unemployed workers remained consistently around 7 million (almost half the level it reached during the depths of the current crisis). Radicals do not attribute this unemployment mainly to a lack of appropriate skills (as mainstream economists explain “structural” unemployment), pointing out that even so-called “unemployable” workers do get hired when a boom lasts long enough. Rather, they argue, this “reserve army” of unemployed workers is necessary to keep workers scared of losing their jobs—and therefore willing to work hard enough, and for low enough wages, to ensure capitalist profits.

Seen this way, then, unemployment is not just a sign that the wiring of capitalism has gone wrong. Rather, it is an integral part of the wiring. This does not mean that radical economists think unemployment is good, or even that it is a regrettable, but unavoidable, fact of life. Instead, they argue that getting rid of unemployment requires a fundamental change in economic structure—not just rearranging the wiring of capitalism, but creating a new, non-capitalist economic system. Not all radical economists agree about the details of such a system, but most favor some system of democratic decision-making at all levels of the economy. This involves replacing the ownership and control of the means of production by a few capitalist owners and top managers (today, less than 5% of the population in the United States and other rich capitalist countries) with some form of collective or public ownership. At the level of the individual enterprise, radicals argue for some form of workers’ control. Instead of being run like a small dictatorship, with the capitalist owner or managers handing down commands from on high, each enterprise would be run like a small democracy, with workers collectively making decisions that affect their lives. At the level of the economy as a whole, they argue for some kind of democratic planning. The total output of the economy, the mix of goods produced, and the ways of producing them, radicals argue, should not be decided by a handful of capitalist owners and managers, with only profit in mind. These decisions should, instead, be made democratically by the members of society as a whole.

Alejandro Reuss is an historian and economist and a member of the Dollars & Sense collective.

Sources: Narayana Kocherlakota, President, Federal Reserve Bank of Minneapolis, “Inside the FOMC” (Speech), Marquette, Michigan, August 17, 2010; Job Openings and Labor Turnover Survey (JOLTS) Highlights, November 2010, Bureau of Labor Statistics; Murat Tasci and John Linder, “Economic Trends: Has the Beveridge Curve Shifted?” Federal Reserve Bank ofCleveland, August 10, 2010.
Figure 1 Sources: For job openings rate (total nonfarm, seasonally adjusted), Job Openings and Labor Turnover Survey (JOLTS), Bureau of Labor Statistics; for unemployment rate (16 years and over, seasonally adjusted), Labor Force Statistics from the Current Population Service.

Did you find this article useful? Please consider supporting our work by donating or subscribing.

« Back to Econo FAQs

end of article