Loading field map
Health capital is a way of thinking about health as a productive asset that individuals carry with them through life. The idea, introduced systematically by the economist Michael Grossman in the early 1970s, treats each person as the owner of a stock of health that depreciates over time and can be augmented through investment. This stock yields a flow of "healthy time"—days available for work, leisure, and the ordinary activities of living. The demand for health, in this framework, is not primarily a demand for medical care. It is a demand for the underlying stock of health itself, with medical care serving as one input among several that people use to produce it.
The subfield of health economics that grew from this idea asks a distinctive set of questions. Why do people of different ages, incomes, and education levels invest differently in their health? Why does health decline with age even when medical spending rises? Why do the rich and well-educated live longer and healthier lives than the poor and less educated? Why do people smoke, drink, or skip exercise when they know these behaviors damage their health? These questions share a common structure: they treat health-related behavior as the outcome of deliberate choices made under constraints, rather than as the passive result of biology, luck, or medical care alone.
The Grossman model, first presented in 1972, is the foundational framework of this subfield. It adapts the economic theory of human capital—originally developed to explain investments in education and job training—to the domain of health. In the standard human capital model, education raises a person's productivity in the labor market. Grossman's innovation was to argue that health similarly raises productivity, but in a broader sense: it increases the total time available for all productive activities, including work and leisure. A person who is sick loses days to illness; a person who is healthy has more time to allocate to whatever they value.
The model has several moving parts. Each person is born with an initial stock of health. This stock depreciates at a rate that increases with age; the body naturally wears down over time. The person can offset depreciation by investing in health, using inputs such as medical care, diet, exercise, and sleep. These inputs are combined through a "health production function"—a technical relationship describing how much health improvement results from a given combination of inputs. The person's total time is divided among three uses: work, leisure, and the production of health itself. Time spent visiting doctors, exercising, or recovering from illness is time not spent earning income or enjoying leisure.
The central result of the model is that people demand health not for its own sake but because it gives them more healthy time. This means the demand for health is a derived demand: it derives from the value of the time that health makes available. The model predicts that the optimal stock of health declines with age, because as depreciation accelerates, maintaining a given stock becomes more expensive. It predicts that people with higher wages demand more health, because their time is more valuable and illness is therefore more costly. It predicts that education increases the efficiency of health production—educated people get more health from the same inputs—which helps explain why education and health are strongly correlated in the data.
The model also yields a sharp distinction between the demand for health and the demand for medical care. Medical care is an input, not the object of desire. This distinction matters for policy. If the demand for medical care is derived from the demand for health, then simply expanding the supply of medical care may not improve health if people do not use it effectively, or if the underlying causes of poor health lie elsewhere—in poverty, stress, or low education. The model redirects attention from the healthcare system to the broader conditions that shape health investment.
Grossman's model belongs to a larger intellectual movement in economics that began in the 1960s: the "new home economics," associated with Gary Becker and others. This movement argued that households are not passive consumers of goods but active producers of the things they ultimately value. A household does not buy "meals"; it buys groceries and combines them with cooking time, kitchen equipment, and skill to produce meals. Similarly, a person does not buy "health"; they buy medical care, food, and exercise equipment, and combine these with time and knowledge to produce health.
This framework has important implications for how health behavior is understood. It explains why two people with identical incomes and access to the same medical care can have very different health outcomes: they may differ in the efficiency with which they convert inputs into health, or in the value they place on healthy time. It also explains why time is a crucial input in health production. A person with a demanding job and long working hours may have the money to join a gym but not the time to use it. The opportunity cost of time—what must be given up to spend an hour exercising or preparing a healthy meal—is as important as the monetary cost of health inputs.
The household production approach also illuminates the relationship between health and other investments. Education, in this view, is not just correlated with health; it may cause better health by improving the efficiency of health production. A more educated person may better understand medical instructions, process health information, or adopt new health technologies more quickly. This idea has generated a large empirical literature attempting to determine whether education actually causes better health or merely correlates with it through other factors such as family background or time preference.
A distinctive feature of the Grossman model is that health serves two purposes simultaneously. It is an investment good because it increases the time available for work and leisure, and it is a consumption good because being healthy feels good in itself. The model allows both motives to operate, but the investment motive is what gives the model its analytical power. It explains why health is demanded even by people who do not expect to work in the future: the retired, the disabled, and the wealthy all value healthy time for leisure and daily functioning.
This dual nature creates a tension in the model's predictions. If health were purely an investment, people would demand less of it as they age, because the remaining working life over which to recoup the investment shrinks. If health were purely a consumption good, people would demand more of it as they age, because the marginal utility of healthy time might rise with age. The actual prediction of the model—that health declines with age—reflects the dominance of the investment motive and the rising cost of maintaining health as depreciation accelerates.
The investment-consumption distinction also matters for interpreting empirical evidence. When researchers observe that health spending rises with age, they cannot immediately conclude that older people "demand more health." The Grossman model predicts that older people demand less health but spend more on medical care, because they are trying to offset a faster rate of depreciation. The rising spending is a symptom of declining health, not a sign of increased health demand. This counterintuitive prediction is one of the model's most important contributions: it separates the observable quantity (medical spending) from the underlying variable of interest (health stock).
The basic Grossman model assumes that people know their health depreciation rate and the effectiveness of health investments. In reality, both are uncertain. A person does not know whether they will develop heart disease, how quickly their joints will deteriorate, or whether a new medication will work for them. This uncertainty has led to extensions of the model that incorporate risk and information.
One important extension treats health as a form of precautionary savings. Just as people save money to protect against income shocks, they may invest in health to protect against future illness. This creates a demand for health that is not captured by the simple investment-consumption framework. It also creates a role for health insurance, which can be understood as a way of pooling the financial risk of illness. The demand for health insurance is thus closely related to the demand for health, but it is not the same thing: insurance protects the financial resources needed to purchase health care, not the health stock itself.
Information plays a complex role in health demand. People learn about their health through symptoms, screening tests, and medical checkups. This learning can change their behavior. A person who discovers they have high blood pressure may invest more in exercise and diet; a person who receives a clean bill of health may feel licensed to continue unhealthy behaviors. The economics of information in health has developed into a substantial literature on how people update their beliefs about their health and how these beliefs affect their investment decisions.
The model also raises questions about the limits of rational choice. The Grossman framework assumes that people make informed, deliberate decisions about their health. But much health behavior appears inconsistent with this assumption. People smoke despite knowing the risks, skip medications because of forgetfulness or cost, and procrastinate on preventive care. Behavioral economists have documented systematic biases in health decisions: present bias (valuing immediate pleasure over future health), optimism bias (underestimating personal risk), and status quo bias (sticking with current behaviors even when change would help). These findings do not necessarily overturn the Grossman model, but they complicate it. Some researchers have incorporated behavioral biases into the model, treating them as additional factors that shift the effective cost or benefit of health investment.
Testing the Grossman model requires measuring health capital, which is not directly observable. Researchers have used a variety of proxies: self-reported health status, mortality rates, disability measures, and biomarkers such as blood pressure or cholesterol levels. Each measure captures a different aspect of health, and the choice of measure can affect the conclusions drawn.
The most influential empirical finding in this subfield is the strong, persistent relationship between socioeconomic status and health. People with higher income, more education, and higher-status occupations live longer and healthier lives. This "gradient" in health is observed across countries, across time periods, and across the entire socioeconomic spectrum—it is not just a difference between the very poor and everyone else, but a continuous relationship in which each step up the socioeconomic ladder is associated with better health. The Grossman model offers one explanation for this gradient: people with higher wages have a higher opportunity cost of illness and therefore invest more in health; people with more education produce health more efficiently. But the gradient is so robust that it has generated competing explanations, including the possibility that health determines socioeconomic status (the sick become poor) rather than the reverse, or that both are driven by a third factor such as childhood conditions or genetic endowment.
Empirical work in this subfield has also examined the relationship between medical care and health outcomes. The Grossman model predicts that medical care is only one input among many, and that its marginal effect on health should be smaller when other inputs are inadequate. This prediction has been tested in studies of health care access, health insurance expansion, and public health interventions. The results generally support the view that medical care matters, but that its effect is limited by behavioral and environmental factors. The famous RAND Health Insurance Experiment of the 1970s and 1980s, which randomly assigned families to different levels of health insurance generosity, found that more generous insurance increased the use of medical care but had little effect on most health outcomes. This finding is consistent with the Grossman model's emphasis on the distinction between medical care inputs and health outcomes.
The Grossman model is fundamentally a life-cycle model: it describes how health evolves from birth to death and how investment decisions change with age. This life-cycle perspective has generated a rich literature on health trajectories and their determinants.
Childhood is a critical period for health capital formation. The health stock with which a person enters adulthood depends on nutrition, disease exposure, and family resources during childhood. This has led to the concept of "health capital at birth" and the idea that early-life conditions cast a long shadow over adult health. The fetal origins hypothesis, associated with the economist David Barker, argues that conditions in the womb program later health outcomes, particularly cardiovascular disease and diabetes. This hypothesis has been integrated into health capital theory as an explanation for why the initial stock of health varies across individuals and why early interventions may have large long-term returns.
Adolescence and young adulthood are periods of active health investment, but also of health-damaging behavior. The model predicts that young people should invest heavily in health because they have a long future over which to reap the returns. Yet many young people engage in behaviors—smoking, drinking, drug use, risky sex—that damage their health capital. This apparent contradiction has led to extensions of the model that incorporate time preference: people who discount the future heavily will invest less in health, even when the long-term returns are high. The economics of time preference has become a major theme in health capital research, with studies showing that measures of patience and self-control predict health behaviors and outcomes.
Middle age is when the depreciation of health capital becomes noticeable. Chronic conditions such as hypertension, diabetes, and arthritis begin to emerge, and medical spending rises. The model predicts that people in middle age should respond to this depreciation by increasing their health investment, but the response is often inadequate. This is the period when the gap between actual and optimal health investment is largest, and it is the period when many of the behavioral biases documented by behavioral economists are most consequential.
Old age is characterized by rapid depreciation and high medical spending. The model predicts that the optimal health stock declines with age, but this does not mean that health investment should cease. Even at advanced ages, investments in health—whether through medical care, exercise, or social engagement—can improve quality of life, even if they cannot prevent eventual decline. The economics of aging and health has become increasingly important as populations in developed countries age and the demand for long-term care grows.
The Grossman model has been enormously influential, but it has also been criticized on several grounds. The most fundamental critique is that it assumes a degree of rationality and deliberateness in health behavior that may not exist. People do not generally think of themselves as managing a stock of health capital; they eat, drink, smoke, and exercise for reasons that have little to do with investment calculus. The model may describe the aggregate patterns of health behavior without accurately describing the decision processes of individuals.
A related critique concerns the model's treatment of preferences. The model assumes that people have stable preferences over health and other goods, and that these preferences do not change over time. But preferences over health may themselves be shaped by experience. A person who has been seriously ill may value health more than someone who has never been sick; a person who has watched a parent die of lung cancer may be more averse to smoking. The model has difficulty accommodating such endogenous preference change.
The model has also been criticized for its treatment of the health production function. The idea that people "produce" health by combining inputs assumes a degree of technical knowledge and control that may be unrealistic. Many health inputs are not chosen but imposed by circumstances: a person living in a polluted city cannot easily choose clean air; a person working a physically demanding job cannot easily choose safer working conditions. The model tends to emphasize individual choice over structural constraint, which some critics argue leads to an overemphasis on personal responsibility in health policy.
Finally, the model has been criticized for its treatment of time. The model treats time as a resource to be allocated, but time is also a site of meaning and experience. The time spent caring for a sick child, cooking a meal, or walking in nature is not just an input into health production; it is part of what makes life worth living. The model's instrumental view of time may miss important aspects of health and well-being.
The subfield of health capital and demand for health has evolved considerably since Grossman's original formulation. The core insight—that health is a produced good, not a given—remains central, but the framework has been extended and modified in several directions.
One major development is the integration of behavioral economics. Modern health capital research increasingly incorporates insights from psychology about how people actually make decisions. This has led to models that combine the Grossman framework with behavioral biases, and to policy recommendations that use "nudges" and other choice architecture to improve health decisions. The behavioral turn has not replaced the Grossman model; rather, it has enriched it by making the assumptions about decision-making more realistic.
Another development is the emphasis on the social determinants of health. While the Grossman model focuses on individual choice, contemporary research recognizes that health is shaped by social, economic, and environmental factors that individuals do not fully control. This has led to a broader conception of health capital that includes not just individual investments but also community resources, social networks, and public goods. The relationship between individual health investment and social context is an active area of research.
A third development is the use of new data and methods. The availability of large longitudinal datasets, genetic data, and natural experiments has allowed researchers to test the predictions of the Grossman model more rigorously than was possible in the 1970s. The use of instrumental variables, regression discontinuity designs, and other causal inference methods has sharpened the empirical literature on the determinants of health.
The subfield also continues to grapple with its central normative question: how much of the responsibility for health lies with individuals, and how much with society? The Grossman model, with its emphasis on individual investment, tends to support policies that improve information, reduce barriers to healthy behavior, and align incentives. But the model's own logic implies that people with fewer resources—less money, less education, less time—will invest less in health, and that this is a rational response to their circumstances. This creates a tension between the model's positive analysis (what people do) and its normative implications (what policy should do). The resolution of this tension remains an open question in the field.
The demand for health, understood as the demand for a stock of productive time, remains a powerful lens for understanding why some people are healthy and others are not. It explains the gradient, the age pattern of health, and the limited effect of medical care on health outcomes. It also provides a framework for thinking about policy: if health is produced, then policy can affect health by changing the inputs, the efficiency of production, or the constraints under which people make their choices. The subfield has moved beyond the simple model, but the model's central question—what determines how much health people demand—continues to organize research and debate.