Loading field map
Labor demand theory is the branch of labor economics that studies how employers decide how many workers to hire and at what wage. It asks why firms employ the number of people they do, why employment rises and falls with wages and with economic conditions, and how policy interventions—minimum wages, payroll taxes, employment protection laws—change hiring decisions. The field is built on the insight that labor is not a fixed resource but a purchased input, and that the demand for workers is therefore derived from the demand for the goods and services those workers produce.
The central concept in labor demand theory is that the demand for labor is a derived demand. Firms do not hire workers for their own sake; they hire because workers contribute to producing output that can be sold. This simple observation has far-reaching consequences. It means that the demand for labor depends on three things: the productivity of workers, the price of the output they produce, and the cost of other inputs that might substitute for or complement their work.
The most basic model of labor demand treats the firm as choosing how many workers to hire, taking the wage as given. The firm compares the cost of an additional worker—the wage—with the additional revenue that worker generates, called the value of the marginal product. The value of the marginal product is the extra output produced by one more worker, multiplied by the price at which that output sells. A profit-maximizing firm hires workers up to the point where the value of the marginal product equals the wage. If a worker generates more revenue than they cost, the firm should hire them; if less, it should not.
This simple rule generates the most important prediction of the field: the labor demand curve slopes downward. When wages rise, firms want to hire fewer workers; when wages fall, they want to hire more. The strength of this response is measured by the elasticity of labor demand, which answers the question: if wages rise by one percent, by what percentage does employment fall? This elasticity is not a fixed number; it depends on several factors that the theory identifies.
The first factor is the ease of substitution. If firms can readily replace workers with machines, or with workers of different skill levels, then labor demand is more elastic—a wage increase leads to a large drop in employment. If workers are hard to replace, demand is less elastic. The second factor is the elasticity of demand for the final product. If consumers are very sensitive to price changes, then a wage increase that raises production costs will lead to a large fall in output and hence in employment. The third factor is the share of labor costs in total production costs. If labor is a small part of total costs, a wage increase has little effect on the price of the final product and thus little effect on employment. The fourth factor is the supply of other inputs: if the supply of capital or other factors is fixed, firms cannot easily substitute away from labor, making demand less elastic.
These four factors are known as the Marshall–Hicks rules of derived demand, after Alfred Marshall and John Hicks, who formalized them. They remain a central organizing framework in the field, not because they are surprising, but because they identify precisely which empirical facts matter for predicting the employment effects of wage changes.
The basic model just described assumes a perfectly competitive labor market: many firms hiring identical workers, no firm large enough to influence the wage, and no frictions preventing adjustment. This model is the workhorse of the field, and it is surprisingly powerful. It explains why employment in a particular industry tends to fall when wages in that industry rise, why employment is more responsive to wages in the long run than in the short run (because firms have more time to adjust their capital and technology), and why the employment effects of a minimum wage depend on how easily firms can substitute capital for labor.
But the competitive model has well-known limits, and much of the field's development has been a response to them. The model assumes that firms can hire any number of workers at the going wage, but in reality firms often face adjustment costs: recruiting, training, and firing workers all cost money. These costs mean that firms do not immediately adjust employment to every change in the wage; they wait until the gap between the wage and the value of the marginal product is large enough to justify the adjustment cost. This insight gave rise to the idea of labor demand as a dynamic process, in which employment adjusts gradually toward its desired level, and in which temporary shocks can have persistent effects.
The competitive model also assumes that all workers are identical, but in reality workers differ in skills, education, and experience. The field handles this by distinguishing among different types of labor—skilled and unskilled, for example—and treating them as separate inputs. This leads to the question of substitutability and complementarity between labor types. If skilled and unskilled workers are substitutes, then an increase in the supply of skilled workers (perhaps due to increased college attendance) will reduce the wages of skilled workers and increase the employment of unskilled workers. If they are complements, the opposite happens. The empirical finding that skilled and unskilled workers are imperfect substitutes, and that capital is often complementary with skilled labor but substitutable for unskilled labor, has been central to explaining the rise in wage inequality in many countries.
A further limit of the competitive model is that it treats the wage as determined outside the firm's decision. But in many labor markets, firms have some power to set wages. This is the domain of monopsony models, which have become increasingly important in the field. In a monopsonistic labor market, a firm faces an upward-sloping labor supply curve: to hire more workers, it must raise the wage it offers to all its workers, not just the new ones. This gives the firm market power over wages. The key implication is that the wage can be set below the value of the marginal product, and that a minimum wage can actually increase employment by pushing the wage closer to the competitive level. The empirical relevance of monopsony is hotly debated, but the theoretical possibility has reshaped how economists think about minimum wage policy.
The intellectual roots of labor demand theory lie in the marginalist revolution of the late nineteenth century, when economists such as Marshall and John Bates Clark developed the idea that factor prices are determined by marginal productivity. For much of the early twentieth century, labor demand was treated as a straightforward application of the theory of the firm, with little separate attention paid to the peculiarities of labor as an input.
The modern field took shape in the mid-twentieth century, driven by two developments. The first was the formalization of production theory, which allowed economists to describe precisely how firms combine labor, capital, and other inputs, and to derive the demand for each input from the firm's cost-minimization problem. This produced a rigorous framework for thinking about substitution between inputs and about how changes in the price of one input affect the demand for others. The second development was the growth of empirical labor economics, which began to test the predictions of the theory using data on wages, employment, and output.
A major turning point was the introduction of the translog production function and related flexible functional forms in the 1970s. These allowed economists to estimate the degree of substitutability between different types of labor and capital without imposing restrictive assumptions about the shape of the production function. This led to a large empirical literature on the elasticity of substitution between skilled and unskilled labor, and between labor and capital, which remains an active area of research.
Another important development was the recognition that labor demand cannot be understood without considering the institutional and legal environment. Employment protection laws, which make it costly to fire workers, affect hiring decisions because firms anticipate the future cost of dismissal. Payroll taxes, which are levied on wages, affect the cost of labor and hence the quantity demanded. Minimum wage laws set a floor on wages and can push firms up their labor demand curve. The field has developed a rich set of models for analyzing how these policies affect employment, and the empirical literature on their effects is one of the most active in economics.
Contemporary labor demand theory is characterized by several overlapping developments that have moved the field well beyond the simple competitive model.
The first is a focus on heterogeneity. Workers differ in skills, and firms differ in productivity, technology, and market power. The demand for labor is not a single curve but a distribution of curves across firms and worker types. Modern models incorporate this heterogeneity explicitly, allowing researchers to ask questions that the aggregate model cannot address: How do the employment effects of a minimum wage differ across firms with different productivity levels? How does technological change affect the demand for different skill groups? How does the demand for labor vary across regions with different industrial structures?
The second development is the integration of search and matching theory. In the standard competitive model, the labor market clears instantly: anyone who wants a job at the going wage can find one. In reality, finding a job takes time and effort, and firms cannot instantly fill vacancies. Search and matching models treat the labor market as a process of matching workers to jobs, with frictions on both sides. In these models, the demand for labor is not simply a matter of hiring up to the point where the value of the marginal product equals the wage; it also involves the decision to post vacancies, which depends on the expected cost of searching and the probability of finding a suitable worker. This framework has become the standard tool for analyzing unemployment, job creation, and the effects of labor market policies.
The third development is the growing importance of dynamic models. Firms do not make a single hiring decision; they make a sequence of decisions over time, anticipating future conditions. Dynamic models of labor demand incorporate adjustment costs, uncertainty about future demand and wages, and the option value of waiting. These models can explain why employment responds sluggishly to economic shocks, why firms hoard labor during downturns, and why the employment effects of policies may take years to fully materialize.
A fourth development is the increased attention to market power and concentration. The competitive model assumes that firms are price-takers in the labor market, but a growing body of evidence suggests that many labor markets are concentrated, with a small number of firms employing a large share of workers. This has revived interest in monopsony models and has led to new questions about how market power affects wages and employment, and about the role of antitrust policy in labor markets.
These different approaches are not rival schools in the sense of mutually exclusive paradigms; they are complementary tools for addressing different questions. The competitive model remains the baseline because it is simple and generates clear predictions. Search and matching models are used when the focus is on unemployment and the process of job creation, which the competitive model cannot address. Dynamic models are used when the timing of adjustment matters. Monopsony models are used when firms have wage-setting power.
The field is unified by a common method: specify a model of firm behavior, derive the demand for labor, and test the predictions against data. The disagreements are about which assumptions are appropriate for which questions, not about the fundamental framework. There is, however, a genuine and ongoing debate about the empirical importance of different mechanisms. For example, the question of whether minimum wages reduce employment is ultimately an empirical question, and the answer depends on the elasticity of labor demand, the degree of monopsony power, and the speed of adjustment—all of which are contested.
Several questions continue to define the field's research frontier. The first is the effect of technological change on labor demand. The rise of automation, artificial intelligence, and digital technologies has raised the question of whether machines are substitutes or complements for human labor, and for which types of labor. The field has developed the concept of task-based models, in which production is described as a set of tasks that can be performed by workers or machines, and the demand for labor depends on the comparative advantage of humans versus machines in each task. These models are used to analyze which jobs are at risk of automation and how technological change affects the wage distribution.
The second is the effect of globalization on labor demand. International trade changes the demand for labor by shifting production toward industries in which a country has a comparative advantage and away from industries that face import competition. The field has developed models that link trade to labor demand through the prices of final goods and through the offshoring of production tasks. These models are used to analyze the employment and wage effects of trade liberalization and of the growth of global supply chains.
The third is the design of labor market institutions. Minimum wages, payroll taxes, employment protection, and unemployment insurance all affect the demand for labor, and the field is centrally concerned with quantifying these effects. The empirical literature has become increasingly sophisticated, using natural experiments and quasi-experimental methods to identify causal effects. The results are often nuanced: the employment effects of minimum wages are small or negligible in many settings, but larger in others; payroll taxes are partly shifted to workers in the form of lower wages; employment protection reduces both hiring and firing, with ambiguous effects on net employment.
The fourth is the measurement of labor demand itself. The field has developed a variety of empirical tools for estimating the elasticity of labor demand, the degree of substitution between inputs, and the speed of adjustment. These estimates vary widely across countries, industries, and time periods, and a major ongoing effort is to understand the sources of this variation.
Labor demand theory is thus a mature field with a well-established core and an active research frontier. Its central insight—that the demand for labor is derived from the demand for output and shaped by the possibilities for substitution—remains the foundation on which all more complex models are built. The field's enduring contribution is to provide a rigorous framework for thinking about how wages, technology, and policy affect employment, and to generate empirical predictions that can be tested against the evidence.