Search and matching models of unemployment explain why jobs and workers do not pair instantly, even when firms want to hire and people want to work. In labor economics, these models treat hiring as a process with frictions: workers must search for vacancies, employers must screen applicants, and both sides need time, information, and resources to decide whether a match is worthwhile. That basic insight changed how economists understand unemployment because it moved analysis beyond the simple idea that wages alone clear the labor market. I have used these models in policy briefs and labor market analysis, and they remain the most practical framework for explaining why vacancies and unemployment can rise together.
The core terms are straightforward. A vacancy is an unfilled job opening actively being recruited for. A job seeker is a worker searching while unemployed or employed. A match occurs when a firm and worker form an employment relationship. Separation is the end of that relationship, whether through layoffs, quits, contract completion, or business closure. Search frictions are the costs and delays that prevent instant coordination. The matching function is the formal tool economists use to summarize how unemployed workers and vacancies combine into new jobs over time. Together, these concepts support modern analysis of unemployment duration, hiring rates, labor market tightness, and the effects of recessions, regulations, and technology.
This topic matters because unemployment is not only about the number of jobs available. It is also about fit, timing, geography, skills, bargaining power, information, and expectations. Two economies can have the same unemployment rate but very different labor market health if one has rapid reemployment and the other has long spells of joblessness. Search and matching models help explain youth unemployment, regional mismatch, recruiting bottlenecks, the Beveridge curve, and why stimulus or labor reforms can have delayed effects. They also provide the backbone for many central bank, treasury, and international institution models when policymakers assess labor slack and inflation pressure.
As a hub within economics, this article covers the miscellaneous but essential ideas connected to search and matching unemployment: the classic framework, the Beveridge curve, wage bargaining, job destruction and creation, policy design, business cycle dynamics, and the limits of the model. Each section answers a common question directly and gives practical examples. If you are studying labor economics, evaluating employment policy, or building a broader understanding of macroeconomic fluctuations, this framework is one of the most useful places to start because it links individual decisions to economy-wide unemployment outcomes with unusual clarity.
What are search and matching models?
Search and matching models are labor market models in which unemployed workers and vacant jobs meet through a decentralized process rather than a frictionless auction. The landmark formulation is associated with Diamond, Mortensen, and Pissarides, whose work showed how unemployment can persist in equilibrium even without wage rigidities. In these models, workers decide how intensively to search, firms decide whether to post vacancies, and successful hiring depends on a matching technology. The result is an equilibrium unemployment rate determined by both inflows into unemployment and outflows from it.
The practical value of this approach is that it fits observed labor markets far better than older textbook stories. In real hiring data, firms often keep positions open for weeks or months, applicants send many resumes before receiving offers, and employers reject candidates who are technically available because the productivity match is poor. During the recovery after the 2008 financial crisis, the United States had elevated unemployment and substantial vacancies at the same time. A frictionless model struggles with that pattern. A search model explains it through weak matching efficiency, uncertainty, skill mismatch, and cautious vacancy creation.
How does the matching function work?
The matching function is the engine of the framework. It summarizes how many new hires are created from the stock of unemployed workers and open vacancies. Economists often write it as M(U,V), where matches increase with unemployment U and vacancies V, but usually at a diminishing rate. A common form is Cobb-Douglas, which lets researchers estimate how efficient the labor market is at turning search effort into hires. If vacancies double while unemployed workers stay fixed, hires rise, but not proportionally, because firms begin competing for the same workers.
One of the most useful derived measures is labor market tightness, usually written as vacancies divided by unemployment. High tightness means many vacancies relative to job seekers, which tends to raise job-finding rates and strengthen worker bargaining power. Low tightness means many searchers chasing few jobs, which lengthens unemployment spells. Analysts often combine vacancy data from the Job Openings and Labor Turnover Survey with unemployment data from labor force surveys to estimate tightness and infer whether hiring conditions are easing or overheating.
| Concept | What it measures | Why it matters |
|---|---|---|
| Unemployment | Workers actively seeking jobs | Captures available labor supply |
| Vacancies | Open jobs firms are trying to fill | Shows labor demand not yet satisfied |
| Matching efficiency | How effectively workers and firms connect | Explains changes in hiring beyond counts alone |
| Labor market tightness | Vacancies divided by unemployment | Predicts job-finding rates and wage pressure |
| Separation rate | Frequency of jobs ending | Determines inflow into unemployment |
Why do unemployment and vacancies coexist?
Unemployment and vacancies coexist because labor markets are characterized by time-consuming search and heterogeneity. Workers differ in skills, location, experience, and wage expectations. Jobs differ in tasks, schedules, compensation, and training requirements. Even when there are enough jobs in aggregate, specific workers may not fit specific openings immediately. This coexistence is not a market failure by itself; it is a normal implication of decentralized hiring. The important question is whether the degree of mismatch and delay is temporary, cyclical, or structural.
The best-known way to visualize this relationship is the Beveridge curve, which plots unemployment against vacancies. In typical downturns, unemployment rises and vacancies fall, moving the economy down the curve. In expansions, vacancies rise and unemployment falls, moving up the curve. When the curve shifts outward, the same vacancy rate is associated with higher unemployment, suggesting lower matching efficiency. That happened in several countries after major sectoral disruptions, when workers displaced from construction or manufacturing did not move quickly into growing service or technology roles.
How are wages determined in these models?
Wages in search and matching models are usually determined through bargaining rather than by a single market-clearing wage. The standard approach is Nash bargaining, where the worker and firm split the surplus generated by a successful match. The size of that surplus depends on worker productivity, unemployment benefits, expected job duration, recruiting costs, and outside options for both sides. This matters because wages affect vacancy posting. If firms expect most productivity gains to be absorbed by wages, they create fewer vacancies, reducing job-finding rates.
In practice, wage setting is more complicated than the baseline model suggests. Some firms use posted wages, internal pay bands, or collective bargaining agreements. Others respond slowly to labor market tightness because of fairness norms, retention concerns, or budget cycles. That is one reason the simple model can underpredict unemployment volatility: if wages do not fall much in recessions, the surplus from hiring drops sharply, and vacancy creation contracts more than a flexible-wage benchmark would imply. Modern versions therefore incorporate wage rigidity, bargaining delay, or on-the-job search.
What drives unemployment persistence?
Unemployment persistence comes from both separations and weak job-finding. A recession usually begins with a rise in job destruction as firms cut staff, but prolonged unemployment often reflects a slow recovery in matching and vacancy creation. Workers lose current income, firm-specific skills, networks, and confidence as joblessness lengthens. Employers may statistically discriminate against long unemployment spells, interpreting them as signals of lower productivity even when that inference is unjustified. Search and matching models capture persistence by showing how today’s weak hiring reduces tomorrow’s employability and bargaining position.
Hysteresis is the broader idea that temporary shocks can leave permanent labor market scars. Search frameworks make that mechanism concrete. If a region loses a major industry, displaced workers may not relocate, retrain, or reconnect quickly. New firms may be reluctant to post vacancies if they doubt local matching quality. I have seen this pattern in regional labor market reviews where headline unemployment eventually improved, yet participation remained depressed and vacancy durations stayed elevated. The model helps separate a short cyclical shortfall from a deeper structural adjustment problem.
How do policy and institutions affect matching?
Policy matters because institutions shape search incentives, recruiting costs, and job stability. Unemployment insurance can lengthen search by raising reservation wages, but it can also improve match quality by allowing workers to avoid poor fits that end quickly. Well-designed public employment services, vacancy databases, and counseling can reduce information frictions and improve matching efficiency. Training subsidies can help when unemployment reflects genuine skill gaps, though many programs fail when they are disconnected from employer demand. The policy lesson is not simply more or less intervention, but better alignment between search support and actual vacancies.
Employment protection rules, minimum wages, licensing, and collective bargaining also enter the picture. Strong dismissal protections may reduce separations but can make firms more cautious about posting vacancies. Minimum wages can compress pay at the bottom, which may reduce low-productivity matches but can also improve retention and search intensity if workers view jobs as worth accepting. Licensing can preserve standards in health or legal services while restricting entry and geographic mobility. The net effect depends on how each rule changes both the expected surplus from a match and the cost of a failed one.
How do these models explain business cycles?
Search and matching models became central in macroeconomics because they connect labor market flows to recessions and recoveries. In a downturn, lower demand or higher uncertainty reduces the expected payoff from opening vacancies. Because posting a vacancy is costly, firms cut recruiting quickly. Job-finding rates then fall, even before separation rates peak. This mechanism explains why unemployment can rise fast once hiring freezes begin. It also clarifies why recoveries can feel weak: output may improve before vacancy creation and matching fully recover, leaving unemployment elevated for longer than headline growth suggests.
Shimer’s critique pushed the literature further by arguing that the basic model did not generate enough volatility in unemployment and vacancies relative to productivity shocks. Researchers responded with richer versions featuring wage stickiness, financial frictions, endogenous separations, heterogeneous firms, and varying recruiting intensity. Those additions made the framework more realistic. During the COVID-19 shock, for example, temporary layoffs, reopening frictions, health risk, remote work transitions, and sectoral reallocation all mattered. A simple one-shock model was insufficient, but the search framework still organized the facts better than a static labor supply and demand diagram.
What are the limitations and why do they still matter?
No model captures every labor market reality. Standard search and matching models often assume a representative worker, a representative firm, and a stable matching function. Real labor markets contain informal hiring channels, discrimination, part-time constraints, migration barriers, unpaid care obligations, algorithmic screening, and occupational licensing differences that vary sharply across sectors. The model also abstracts from household balance sheets, social norms, and the institutional details of wage setting that shape acceptance decisions. If used mechanically, it can make labor markets look more uniform and efficient than they really are.
Even with those limits, the framework remains essential because it provides disciplined language for discussing unemployment as a flow problem rather than a static stock. It helps analysts ask the right questions: Are separations rising? Are vacancies collapsing? Is matching efficiency deteriorating? Are wages preventing or supporting vacancy creation? Those questions guide better policy than broad claims that labor markets simply need more flexibility or more stimulus. Use this hub as a foundation, then explore related topics such as the Beveridge curve, reservation wages, labor force participation, efficiency wages, and structural unemployment to deepen your economics understanding today.
Frequently Asked Questions
What are search and matching models of unemployment?
Search and matching models of unemployment are a major framework in labor economics used to explain why unemployment can exist even when there are people who want jobs and firms that want to hire. The central idea is that labor markets do not work like a simple auction where wages instantly adjust and every worker is immediately paired with an employer. Instead, hiring takes time. Workers must look for openings, decide which jobs fit their skills and preferences, submit applications, and often wait through interviews or other screening steps. Employers must advertise vacancies, sort applicants, evaluate qualifications, compare candidates, and determine whether a hire is likely to be productive and worth the cost.
These models focus on what economists call frictions, meaning the real-world obstacles that slow down matching between workers and firms. Frictions include incomplete information, geographic distance, skill mismatch, uncertainty about job quality, recruiting costs, and the time required for bargaining over wages and conditions. Because of these frictions, unemployment is not always a sign that wages are “wrong” in a simple sense. It can also reflect the fact that forming productive employment relationships is costly and time-consuming.
One of the most important contributions of search and matching theory is that it treats both unemployment and job vacancies as normal features of the labor market that can coexist. That helps explain why an economy may have many unfilled jobs at the same time many workers remain unemployed. The issue is not necessarily a lack of labor demand or labor supply alone, but the difficulty of bringing the right worker and the right job together at the right time. This perspective has been especially influential in explaining labor market dynamics over the business cycle and in shaping how economists think about unemployment persistence, hiring incentives, and labor market policy.
Why don’t jobs and workers match instantly if both sides are willing?
Jobs and workers do not match instantly because willingness alone is not enough to create a productive employment relationship. A successful match requires information, compatibility, timing, and mutual agreement. Workers need to know which vacancies exist, what wages are offered, what tasks are involved, whether the location works for them, and whether the job fits their skills and career goals. Employers need to know whether applicants are qualified, reliable, trainable, and likely to stay long enough to justify hiring and training costs. Neither side has perfect information at the outset, so search and screening are unavoidable.
In addition, labor is not a standardized product. Workers differ in education, experience, preferences, and productivity. Jobs differ in required skills, work schedules, compensation, advancement opportunities, and workplace conditions. Even when there are many openings and many job seekers, not every worker is suitable for every role. Economists refer to this as heterogeneity, and it is one of the main reasons matching takes time. A firm may prefer to leave a vacancy open rather than hire an unsuitable worker, while a worker may reject available jobs that offer poor pay, weak stability, or a bad skill fit.
Search and matching models also emphasize costs. Looking for work takes time and effort. Recruiting and interviewing are expensive for firms. Negotiating wages and benefits can delay hiring further. Sometimes workers wait for better offers, and sometimes firms delay hiring because they expect stronger candidates to appear. These strategic decisions are rational from each side’s perspective but can still produce unemployment and vacancies in the aggregate. In short, labor markets are slow not because participants are inactive, but because matching is a complex process under uncertainty.
How do search frictions help explain unemployment and job vacancies at the same time?
One of the most important insights of search and matching models is that unemployment and vacancies can coexist naturally. In older, simpler views of the labor market, it was harder to explain why firms would report open positions while many workers were still jobless. Search theory resolves this by showing that vacancies and unemployment are jointly determined by the process of matching. Firms post vacancies because they want to hire, but filling those vacancies is not immediate. Workers search because they want employment, but finding a suitable opening is also not immediate. As a result, an economy can simultaneously have unemployed workers and unfilled jobs.
The speed at which workers and vacancies come together is often summarized with a matching function, a tool economists use to describe how the number of successful matches depends on the number of job seekers, the number of vacancies, and the efficiency of the matching process. If matching is inefficient because of poor information, skill mismatch, weak mobility, or costly recruitment, then many vacancies may remain open while many workers continue searching. This does not require irrational behavior. It simply reflects the frictions built into real labor markets.
This framework also helps explain why labor markets can feel “tight” and “slack” at the same time, depending on perspective. Employers may say it is hard to hire, while workers may say it is hard to find good jobs. Both can be true if available workers are not aligned with available openings. For example, a region may have unemployed workers from declining industries while employers in expanding sectors struggle to find people with the right training. Search and matching models therefore provide a more realistic way to understand labor market imbalance than approaches that assume all workers and all jobs are interchangeable.
What role do wages play in search and matching models of unemployment?
Wages remain important in search and matching models, but they are not the only force shaping unemployment. In a simple textbook market, wages adjust until labor supply equals labor demand, leaving little room for persistent unemployment. Search and matching theory keeps wages in the picture but adds the idea that employment relationships must first be created through a costly search process. That means wages influence incentives to search, recruit, accept offers, and keep vacancies open, but they do not eliminate frictions on their own.
In many versions of these models, wages are determined through bargaining between workers and firms after they meet. This matters because both parties typically generate a surplus from the match, meaning the worker earns more than in unemployment and the firm gains productive labor. How that surplus is divided affects behavior across the labor market. Higher wages may make jobs more attractive to workers and can improve acceptance rates, but they also raise the cost of hiring for firms. Lower wages may encourage firms to post more vacancies, yet they can cause workers to search longer or reject offers that do not meet their reservation wage, the minimum compensation they are willing to accept.
These models also highlight that wages may not adjust quickly or fully enough to erase unemployment. Firms and workers care about match quality, expectations, and long-term productivity, not just the immediate wage. Even if wages become more flexible, workers still need time to find openings and employers still need time to screen candidates. That is why search and matching models shifted the conversation: unemployment is not explained only by wage stickiness or excessive wages, but also by the structure and efficiency of the matching process itself.
Why are search and matching models important for economic policy and real-world labor markets?
Search and matching models are highly important because they give policymakers a more practical understanding of how labor markets actually function. If unemployment were caused only by wages being too high or too low, policy would focus narrowly on wage adjustment. But if unemployment also reflects search frictions, mismatch, and costly hiring, then effective policy can target those specific problems. That opens a wider range of tools, including better job placement services, improved labor market information, retraining programs, mobility support, hiring subsidies, unemployment insurance design, and measures that reduce the cost and uncertainty of recruitment.
These models are especially useful for understanding business cycles. During recessions, firms may cut vacancies sharply, making it harder for workers to find jobs even if many are willing to accept work. In recoveries, hiring may remain slow because firms are cautious, workers may need time to reallocate across sectors, and old matches may have been permanently destroyed. Search and matching theory helps explain why labor market recovery can lag behind broader economic growth. It also provides a framework for thinking about how shocks affect both job creation and job destruction, rather than treating unemployment as a single static number.
In real-world labor markets, the value of these models is easy to see. Technological change can create demand for new skills while leaving some workers mismatched. Geographic barriers can keep workers away from job-rich areas. Online platforms may reduce some search costs while increasing competition and information overload in other ways. Employers may face long vacancies in specialized roles, while workers may struggle to access stable and well-matched employment. Search and matching models bring all of these issues into a coherent analytical framework. That is why they remain central to modern labor economics and to policy debates about unemployment, hiring, and labor market efficiency.