Foundations of Neuroeconomics: Bridging Neuroscience, Psychology, and Economic Theory

Foundations of Neuroeconomics: Bridging Neuroscience, Psychology, and Economic Theory

Classical economic theory treats the decision-maker as a logical agent who maximizes expected utility given stable preferences and Bayesian beliefs. For half a century this abstraction served as both descriptive model and normative benchmark, supplying a precise language in which contracts, markets, and policy interventions could be analyzed. Yet the abstraction came at a cost: the very phenomena that make human choice interesting—impulse, social context, affect, learning under uncertainty—were either ignored or relegated to "deviations" from the rational ideal. Neuroeconomics emerged in the late 1990s and early 2000s as an explicit attempt to overcome this limitation by treating choice as a biological phenomenon implemented in neural tissue, while preserving the formal precision that economics offers. This chapter builds the conceptual scaffolding for that project.

We begin with the problem the field was designed to address: why the divergence between rational-choice predictions and observed behavior demands a more substantive theory of decision-making, and why that theory cannot be constructed from economics, psychology, or neuroscience alone. We then trace the historical convergence that produced the field, examine the formal apparatus that neuroeconomics inherits and modifies, articulate its methodological commitments, and confront the mapping problem that any neural account of choice must solve. A worked example—the ultimatum game under framing manipulations—illustrates how a single experimental paradigm can simultaneously test economic predictions, identify the neural systems engaged, and constrain computational theory. The chapter closes with a preview of how subsequent chapters extend these foundations to progressively richer decision phenomena.

1. The Problem of Choice and the Limits of the Rational Model

The standard economic model of choice under risk, dating to Bernoulli (1738) and formalized by von Neumann and Morgenstern (1944), specifies that a decision-maker selects the option with the highest expected utility, where utility is a concave function of monetary outcomes and probabilities are treated as objective frequencies. Savage's (1954) subjective expected utility theory extends this to choice under uncertainty by requiring agents to hold probabilistic beliefs over states of the world and to update those beliefs in accordance with Bayes' rule. Together, these frameworks supply the axioms—completeness, transitivity, independence, continuity—that define rational choice.

The predictive success of these axioms in market settings is substantial. They account for the broad features of consumption, saving, and portfolio allocation; they underwrite contract theory, mechanism design, and most of modern macroeconomics. The hypothesis that aggregated behavior approximates the rational benchmark is, for many purposes, sufficient. Yet the descriptive adequacy of the axioms at the level of individual decisions is, and has long been, contested. The Allais paradox (1953) showed that preferences systematically violate the independence axiom: many respondents choose $a$ over $b$ and $b$ over $c$ but not $a$ over $c$, in patterns that no expected utility representation can accommodate. The Ellsberg paradox (1961) demonstrated aversion to ambiguity that cannot be captured by expected utility without modification, even though the payoffs are objectively identical. By the late 1970s, Tversky and Kahneman had assembled a catalog of systematic departures from rational choice—framing effects, anchoring, availability heuristics, base-rate neglect, endowment effects, status quo bias, hyperbolic discounting—that called into question not just the axioms but the very idea that preferences are stable inputs to choice (Kahneman & Tversky, 1979; Thaler, 1992).

Three responses are possible. The first is to defend the rational model by treating observed departures as errors of measurement, confusion, or strategic behavior; this remains the dominant posture in mainstream economics and is supported by appeals to market discipline, learning, or selection effects. The second is to relax the axioms and construct a generalized model that retains enough structure to be useful; this is the path taken by prospect theory, cumulative prospect theory, and the broader family of nonexpected utility models. The third is to ask what kind of system produces the choices we observe—what cognitive processes compute values, how those computations are implemented in neural tissue, and why evolution has produced a decision-maker that systematically deviates from the rational ideal in particular ways. This third response is the defining move of neuroeconomics.

The move is consequential because it refuses a particular kind of dualism. In the standard economic account, preferences are primitive: they are the inputs that the agent supplies to a choice function, and there is no further question about where they come from. Preferences may be revealed by choice, may be incomplete or context-dependent, but they are not, in the standard account, the objects of biological explanation. In the neuroeconomic account, preferences are themselves outputs of a biological process, and the task of the field is to characterize that process at multiple levels of analysis—from the computational algorithms that assign values to options, through the neural circuits that implement those algorithms, to the cellular and molecular mechanisms that support them. This does not mean that preferences are illusory or that the rational model is useless; it means that both must be situated within a more comprehensive theory of how biological agents make choices.

A natural objection is that neuroscience adds little: if preferences are well-defined inputs and the choice function is well-specified, the biology is irrelevant. The objection fails for two reasons. First, the inputs are not well-defined. Stated and revealed preferences often diverge because of cognitive dissonance, self-control problems, and the influence of arousal and emotion; preferences depend on framing, context, and affective state; and many choices involve ambiguity or conflict that the rational model does not address. Second, even when preferences are stable, the process by which they are translated into action is not transparent. The same stated preference can produce different choices depending on attention, working memory load, time pressure, and social context. Without a theory of the process, the rational model cannot predict which path will be taken; with a theory of the process, the predictions become substantially richer and substantially more testable.

A second objection is that biology is too messy to contribute to economic theory. The objection underestimates the structure that biological systems impose. The brain is not a tabula rasa upon which any preference can be inscribed; it is an evolved system with conserved computational motifs, characteristic time scales, and identifiable biases. The deviations from rationality are not random; they are clustered around specific dimensions (loss aversion, ambiguity aversion, hyperbolic discounting) that suggest common underlying mechanisms. The neuroeconomic program treats these regularities as evidence about the architecture of choice and seeks to characterize that architecture in ways that complement, rather than replace, formal economic modeling.

2. Historical Convergence: Three Disciplines Become One Field

Neuroeconomics did not emerge from a single discovery. It is the product of three separate research traditions that converged in the late 1990s and early 2000s, each of which had independently accumulated findings that the others could not accommodate.

The first tradition is behavioral economics. Beginning with the work of Allais, Ellsberg, and Simon (1955, 1956), and culminating in the system