Loading field map
Health economics emerged from a puzzle: healthcare markets do not behave like textbook competitive markets. Patients lack the information to judge what they need, insurance separates payment from consumption, and many treatments are collective goods. Yet economists have never agreed on a single toolkit for analyzing these failures. The field's history is a series of competing frameworks, each emphasizing a different dimension of the problem—information, incentives, individual behavior, institutional constraints, or causal identification—and each responding to the limitations of its predecessors.
The modern field begins with Kenneth Arrow's 1963 article "Uncertainty and the Welfare Economics of Medical Care." Arrow argued that healthcare is fundamentally different from ordinary goods because uncertainty is pervasive: patients do not know what treatments they need, physicians act as imperfect agents, and insurance markets are plagued by adverse selection and moral hazard. This Market Failure Analysis in Healthcare framework provided a normative justification for government intervention, regulation, and public insurance. It did not, however, offer a theory of how individuals themselves invest in health.
That gap was filled a decade later by Michael Grossman's 1972 model of health capital. The Grossman Health Capital Model treated health as a durable stock that individuals produce and maintain through time, medical care, and lifestyle choices. Unlike Arrow's supply-side focus on market structure, Grossman placed the consumer at the center: people demand health both as a consumption good (feeling well) and as an investment good (enabling productive work). The model gave health economics a rigorous microeconomic foundation for analyzing how education, income, and age shape health-seeking behavior. Where Arrow saw systemic failure, Grossman saw rational individual choice—a tension that would resurface decades later when behavioral economists questioned the rationality assumption itself.
Arrow's analysis of uncertainty directly motivated a second line of inquiry: Health Insurance Theory. Beginning in the late 1960s and accelerating through the 1970s, economists formalized the concepts of adverse selection (sicker people buy more insurance, driving up premiums) and moral hazard (insured people use more care because they face lower out-of-pocket costs). These models showed that insurance markets could unravel or produce inefficiently high utilization, but they also raised a critical empirical question: how large is moral hazard in practice?
The RAND Health Insurance Experiment (1974–1982) was the landmark attempt to answer that question. By randomly assigning families to insurance plans with different cost-sharing levels, the experiment demonstrated that higher copayments reduced medical spending without, for most people, harming health outcomes. The RAND study became a touchstone for both Health Insurance Theory and the emerging Microeconometrics in Health Economics tradition, which prized randomized or quasi-random variation as the gold standard for causal inference.
As health spending grew, policymakers needed tools to decide which treatments and programs were worth funding. Two frameworks emerged in the 1970s, each with a different philosophical anchor. Cost-Benefit Analysis (CBA) , rooted in welfare economics, valued all outcomes—including lives saved—in monetary terms and accepted any project whose benefits exceeded costs. Cost-Effectiveness Analysis (CEA) , by contrast, measured outcomes in natural units such as life-years gained or quality-adjusted life-years (QALYs) and compared the cost per unit of health improvement across interventions.
The competition between CBA and CEA was not merely technical; it reflected a deeper disagreement about whether health should be valued like any other commodity. CBA's welfarist approach treated health gains as commensurable with consumption, while CEA's extra-welfarist stance held that health is a distinct good that should not be traded off against money in the same way. Over time, CEA—especially in the form of cost-utility analysis using QALYs—became the dominant framework for health technology assessment in countries such as the United Kingdom, Canada, and Australia. CBA survives in regulatory contexts and in environmental health, but CEA's narrower focus on health outcomes proved more politically and ethically palatable for public coverage decisions.
By the 1980s, health economists had two broad empirical strategies for studying causal questions. Microeconometrics in Health Economics emphasized reduced-form methods—instrumental variables, difference-in-differences, regression discontinuity—that could estimate treatment effects under minimal assumptions. Structural Econometric Modeling in Health , by contrast, specified explicit models of individual behavior (for example, a Grossman-style health production function or a dynamic insurance choice model) and used data to estimate deep parameters such as preferences and technology. The two approaches coexisted uneasily. Microeconometricians accused structural modelers of making untestable functional-form assumptions; structural modelers countered that reduced-form estimates could not predict the effects of counterfactual policies.
This tension came to a head with the Credibility Revolution in Health Economics , which gained momentum around 2010. Inspired by broader developments in empirical economics, the credibility revolution demanded that researchers prioritize research designs with transparent identification strategies—randomized experiments, natural experiments, and sharp discontinuities—over complex modeling. The Oregon Health Insurance Experiment (2008), which used a lottery to allocate Medicaid coverage to low-income adults, became a flagship example: it provided clean causal estimates of insurance on health care use, financial strain, and health outcomes without relying on structural assumptions.
The credibility revolution did not eliminate structural modeling, but it narrowed its domain. Today, structural models are most influential when reduced-form designs are infeasible—for example, when evaluating long-run dynamic responses or counterfactual policies that have never been observed. Microeconometric methods, meanwhile, dominate the empirical literature on health insurance, provider behavior, and public health interventions. The two traditions now operate in a productive but unresolved tension, with each side borrowing insights from the other.
Two more recent frameworks have pushed health economics beyond its traditional assumptions. Political Economy of Health Policy, emerging in the 1990s, shifted attention from normative market-failure analysis to positive questions: why do governments adopt inefficient or inequitable health policies? This framework models the behavior of politicians, interest groups, and bureaucrats, showing that policy outcomes often reflect political incentives rather than welfare-maximizing design. It challenges the benevolent-planner assumption implicit in Arrow's framework and in CEA-based recommendations, explaining, for example, why the United States maintains a fragmented, employer-based insurance system despite widespread agreement that it is inefficient.
Behavioral Health Economics, which crystallized as a distinct framework around 2013, revises the rational-agent assumptions of the Grossman model and standard insurance theory. Drawing on psychology and behavioral economics, it documents systematic deviations from rationality: present bias (people discount future health too steeply), limited attention (patients ignore deductibles or drug side effects), and framing effects (how choices are presented changes decisions). Behavioral health economists have developed interventions—such as default enrollment in insurance plans, simplified drug formularies, and commitment devices for smoking cessation—that nudge behavior without restricting choice. The framework does not reject Grossman's insight that health is an investment; rather, it argues that people invest imperfectly because of cognitive and motivational limitations.
No single framework dominates health economics today. Market Failure Analysis remains the normative foundation for justifying public intervention, but it is increasingly supplemented by Political Economy analysis that explains why interventions often fail. Health Insurance Theory continues to generate predictions about adverse selection and moral hazard, now tested with administrative data and randomized designs. The Grossman model still anchors empirical work on health disparities and the returns to medical care, though Behavioral Health Economics has complicated its rational-choice core. Cost-Effectiveness Analysis is the standard tool for coverage decisions in many health systems, but Cost-Benefit Analysis retains advocates who argue that monetary valuation is necessary for cross-sectoral resource allocation.
The deepest disagreement today concerns the role of structural assumptions. The Credibility Revolution has pushed the field toward simpler, more transparent designs, while Structural Econometric Modeling continues to argue that only explicit models can answer policy-relevant counterfactual questions. Meanwhile, Microeconometrics has absorbed many of the credibility revolution's lessons and now routinely employs quasi-experimental methods. The field's vitality lies in this pluralism: each framework captures a piece of the puzzle, and the most influential research often combines insights from multiple traditions. A student entering health economics today inherits not a settled doctrine but a set of competing tools, each with its own strengths, blind spots, and historical roots.