Loading field map
Unemployment theory is the branch of labor economics that seeks to explain why, in a market economy, there are always people who want to work at the going wage but cannot find a job. The central puzzle is not that unemployment exists—it is that it persists. If wages are prices, and prices adjust to clear markets, then the labor market should clear too: wages should fall until everyone who wants a job has one. The fact that they do not, even in prosperous times, is the anomaly that unemployment theory exists to explain.
The field is organized around a small set of enduring questions. Why do wages not fall to clear the labor market? Why does unemployment fluctuate so strongly over the business cycle, rising sharply in recessions and falling slowly in recoveries? Why does unemployment vary so much across countries, demographic groups, and time periods? And why does a large share of unemployment appear to be "involuntary"—that is, experienced by people who are actively searching and willing to accept a job at the going rate, yet unable to find one?
The starting point for all modern unemployment theory is the classical model, in which the labor market is like any other market. The wage adjusts to equate the supply of labor (workers' willingness to work at different wages) with the demand for labor (firms' willingness to hire at different wages). In this framework, unemployment is either voluntary—workers choosing not to work at the market-clearing wage—or a temporary disequilibrium that will disappear as wages adjust. The classical model does not deny that unemployment exists; it denies that it can persist.
The first major challenge to this view came from John Maynard Keynes in the 1930s. Keynes argued that wages do not adjust quickly enough to clear the labor market, and that the economy could settle at an equilibrium with high unemployment. In his account, the problem was a shortfall in aggregate demand: if households and firms are not spending enough, firms do not need to hire, and wages cannot fall fast enough to restore full employment. Keynes's theory was not primarily a theory of the labor market; it was a theory of the macroeconomy in which unemployment was a symptom of insufficient demand. But it established the idea that unemployment could be a persistent, involuntary state, and it motivated a generation of economists to think about why wages might be "sticky"—slow to adjust downward.
The Keynesian tradition, however, had a weakness that the next generation of theorists would exploit: it treated wage stickiness as an assumption rather than an explanation. If wages are sticky, why are they sticky? The classical model said they should not be. The task of modern unemployment theory, as it emerged in the 1960s and 1970s, was to provide microeconomic foundations for the wage stickiness that Keynes took as given.
The most influential modern framework for understanding unemployment is the search and matching model, developed in the 1970s and 1980s by Peter Diamond, Dale Mortensen, and Christopher Pissarides. The model begins with a simple observation: the labor market is not a single auction where all buyers and sellers meet at once. It is a decentralized, frictional market where workers and firms must find each other. Searching for a job takes time and effort; posting a vacancy takes time and money. Because of these frictions, there is always some unemployment and some vacancies simultaneously, even in equilibrium.
The model's central concept is the matching function, which describes how many matches are formed given the number of unemployed workers and the number of vacancies. The matching function is not derived from deeper principles; it is a reduced-form description of the search process, analogous to a production function for matches. The model then asks: what wage will a worker and a firm agree on when they meet? The answer is not the market-clearing wage, because the worker and the firm are in a bilateral monopoly: the worker has no other offer in hand, and the firm has no other candidate. The wage is determined by a bargaining rule, typically the Nash bargaining solution, which splits the surplus of the match between the two parties.
The key result of the model is that the equilibrium wage is above the "reservation wage"—the wage at which a worker would be indifferent between working and remaining unemployed. This is because the worker's outside option is not zero; it is the value of continuing to search, which includes the chance of finding a better job later. The firm must pay enough to make the worker prefer the job to continued search. But the wage is also below the value of the worker's output, because the firm must be compensated for the cost of posting the vacancy. The result is an equilibrium in which there are both unemployed workers and unfilled vacancies, and the wage does not clear the market. Unemployment is not a temporary disequilibrium; it is a permanent feature of the equilibrium.
The search and matching model has been enormously influential because it provides a coherent account of why unemployment exists and why it fluctuates. It explains the Beveridge curve—the empirical relationship between unemployment and vacancies—and it can be used to analyze how unemployment responds to changes in productivity, unemployment benefits, and other policies. It also provides a natural account of the business cycle: a negative shock to productivity reduces the value of a match, which reduces the incentive to post vacancies, which reduces the rate at which workers find jobs, which raises unemployment. The model is not without its critics, however. One persistent criticism is that it cannot easily explain the magnitude of unemployment fluctuations: the model predicts that a small productivity shock should produce a small change in unemployment, but in reality unemployment swings are large. This is known as the "Shimer puzzle," after the economist Robert Shimer, who documented the discrepancy. Another criticism is that the bargaining rule is an assumption, not a derivation; the model does not explain why the wage is split in the way it is.
A second major approach to unemployment theory is the efficiency wage literature, which asks a different question: why would a firm deliberately pay a wage above the market-clearing level? The answer is that the wage is not just a price; it is a tool for managing the workforce. If workers are paid more than they could earn elsewhere, they have an incentive to work harder, because they do not want to lose the job. This is the "shirking" model, associated with Carl Shapiro and Joseph Stiglitz. In their model, workers can choose to work or to shirk, and the firm cannot perfectly monitor them. If the firm pays the market-clearing wage, workers have no reason to work hard, because they can find another job at the same wage. But if the firm pays a wage above the market-clearing level, workers have a reason to work hard, because losing the job is costly. The firm chooses the wage that balances the cost of paying more against the benefit of higher productivity. The result is that the wage is set above the market-clearing level, and unemployment is the mechanism that makes the threat of firing credible. Without unemployment, the threat of firing would be empty, because the worker could find another job immediately. Unemployment is not a byproduct of the model; it is a necessary condition for the wage to be effective.
Other efficiency wage models emphasize different mechanisms. The "gift exchange" model, associated with George Akerlof, argues that workers reciprocate higher wages with higher effort, so the firm pays a wage above the market-clearing level to elicit loyalty and productivity. The "adverse selection" model, associated with Andrew Weiss, argues that the wage affects the quality of the applicant pool: a higher wage attracts more productive workers, so the firm pays a higher wage to select better workers. The "turnover" model, associated with Steven Salop, argues that a higher wage reduces worker turnover, which saves the firm the cost of hiring and training new workers. All of these models share a common structure: the wage is not a price that clears the market; it is a tool that the firm uses to solve a problem of worker management. The result is that the wage is set above the market-clearing level, and unemployment is the consequence.
The efficiency wage approach differs from the search and matching approach in a fundamental way. In the search and matching model, unemployment is a byproduct of the frictions of the search process; it exists because it takes time for workers and firms to find each other. In the efficiency wage model, unemployment is a deliberate consequence of the firm's wage-setting decision; it exists because the firm wants to create a pool of unemployed workers to discipline the employed ones. The two approaches are not mutually exclusive, and many economists see them as complementary: search frictions explain why there is always some unemployment, and efficiency wages explain why the wage does not adjust to eliminate it.
A third approach to unemployment theory focuses on the role of institutions and power in the labor market. The "insider-outsider" model, associated with Assar Lindbeck and Dennis Snower, begins with the observation that the labor market is not a single market but a market with two groups: insiders, who are employed and have some power over the wage-setting process, and outsiders, who are unemployed and have no power. The insiders can use their power to push wages up, because the firm cannot easily replace them: hiring and firing costs, training costs, and the threat of worker resistance make it costly to replace insiders with outsiders. The insiders know this, and they use their power to raise wages above the market-clearing level. The result is that the outsiders remain unemployed, not because they are unwilling to work at the market wage, but because the insiders have pushed the wage above the market-clearing level.
The insider-outsider model is closely related to the union model, which treats the labor market as a bargaining game between a union and a firm. The union represents the workers and negotiates the wage with the firm. The union's objective is not necessarily to maximize the wage; it may be to maximize the total income of its members, or to maximize employment, or to balance the two. The outcome of the bargaining depends on the relative power of the union and the firm, and on the union's objectives. If the union has power and cares more about wages than employment, it will push the wage above the market-clearing level, and the result will be unemployment. If the union cares more about employment, it will accept a lower wage to preserve jobs.
The insider-outsider and union models are often grouped together as "institutional" approaches to unemployment, because they emphasize the role of institutions—unions, firing costs, unemployment insurance, minimum wages—in shaping the labor market. These models are particularly influential in Europe, where labor market institutions are stronger than in the United States, and where unemployment has historically been higher. They are also the basis for the "labor market rigidity" hypothesis, which argues that European unemployment is high because labor market institutions prevent wages from adjusting to clear the market.
A fourth approach to unemployment theory is the "natural rate" hypothesis, which is not a theory of why unemployment exists, but a theory of how it behaves over time. The natural rate of unemployment is the rate that the economy tends to return to after a shock, and it is determined by the structural features of the labor market: the level of frictions, the wage-setting behavior of firms, the power of unions, the generosity of unemployment insurance. The natural rate is not a fixed number; it can change over time as the structural features of the labor market change. But the hypothesis is that the economy tends to return to the natural rate after a shock, so that the unemployment is a temporary deviation from the natural rate.
The natural rate hypothesis was developed by Milton Friedman and Edmund Phelps in the 1960s, in response to the Phillips curve, which suggested a stable trade-off between inflation and unemployment. Friedman and Phelps argued that the trade-off was only temporary: if the government tried to reduce unemployment by increasing inflation, the unemployment would fall below the natural rate in the short run, but in the long run, workers would adjust their expectations of inflation, and the unemployment would return to the natural rate. The natural rate hypothesis was a major influence on macroeconomic policy, and it remains the standard framework for thinking about the relationship between unemployment and inflation.
The natural rate hypothesis has been challenged by the "hysteresis" hypothesis, which argues that the natural rate is not independent of the actual rate of unemployment. The idea is that a period of high unemployment can permanently raise the natural rate, because the unemployed workers lose skills, lose contact with the labor market, and become less attractive to employers. The result is that the natural rate is not a fixed point; it is a path-dependent variable that depends on the history of the economy. The hysteresis hypothesis was developed to explain the experience of Europe in the 1980s, where unemployment rose sharply in the 1970s and then remained high even after the economy recovered. The hypothesis is controversial, but it has been influential in shaping the way economists think about the persistence of unemployment.
The modern unemployment theory is not a single unified framework; it is a set of overlapping and sometimes competing approaches. The search and matching model is the dominant framework for thinking about the labor market, and it is the standard tool for analyzing the effects of unemployment policy. The efficiency wage models are less central, but they remain important for understanding why wages are sticky and why unemployment can persist. The insider-outsider and union models are influential in Europe, where they are used to explain the high unemployment of the 1980s and 1990s. The natural rate hypothesis is the standard framework for thinking about the long-run behavior of unemployment, but it has been modified by the hysteresis hypothesis.
The field is also characterized by a number of ongoing debates. One debate is about the causes of the large fluctuations in unemployment over the business cycle. The search and matching model has difficulty explaining the magnitude of these fluctuations, and economists have proposed a variety of extensions to the model—including changes in the productivity of matches, changes in the cost of posting vacancies, and changes in the bargaining power of workers—to account for the observed fluctuations. Another debate is about the role of institutions in determining the natural rate of unemployment. Some economists argue that the labor market institutions are the main cause of high unemployment in Europe, while others argue that the institutions are not the main cause, and that the high unemployment is due to other factors, such as the macroeconomic policy or the structure of the economy.
A third debate is about the nature of unemployment itself. The search and matching model treats unemployment as a frictional phenomenon: it is the time it takes for workers to find jobs. The efficiency wage model treats unemployment as a disciplinary device: it is the threat that keeps workers working hard. The insider-outsider model treats unemployment as a result of the power of the insiders. These are not mutually exclusive, but they have different implications for policy. If unemployment is frictional, the policy should be to reduce the frictions: improve the matching process, provide better information, reduce the cost of search. If unemployment is a disciplinary device, the policy should be to reduce the need for discipline: improve the monitoring, reduce the cost of firing, or provide a better safety net. If unemployment is a result of the power of the insiders, the policy should be to reduce the power of the insiders: weaken the unions, reduce the hiring and firing costs, and make the labor market more flexible.
The field of unemployment theory is a mature field, but it is not a settled one. The central puzzle—why the labor market does not clear—remains the central puzzle, and the different approaches are different attempts to solve it. The search and matching model is the most successful approach, but it is not the only one, and it is not the final word. The field continues to evolve, and the debates are likely to continue as long as the unemployment persists.