Neuroeconomics

Neuroeconomics Neuroeconomics is the study of how the brain makes economic decisions.

8 frameworks, from 1994 to today · Behavioral Economics · Economics

still a working view ended · each leads to its entry below

Overview

Neuroeconomics

Neuroeconomics is the study of how the brain makes economic decisions1. It asks what happens, physically, between seeing an option and choosing it: how the brain assigns value to a reward, weighs a delayed gain against an immediate one, guesses what another person will do, and turns all of that into a movement of the hand. The field grew out of behavioral economics, which had already shown that people depart from the rational-actor model in systematic ways12. Neuroeconomics took the next step and asked what mechanism produces those departures. Its practitioners combine three toolkits: economic theory (especially game theory and models of choice under risk), psychological experiments, and brain measurement—functional magnetic resonance imaging (fMRI), which tracks blood flow as a proxy for neural activity; electroencephalography (EEG), which records electrical activity at the scalp; and studies of patients with specific brain damage. What is at stake is whether economics should treat the brain as a black box or as a constraint that shapes which theories are plausible. If a single neural signal encodes the subjective value of very different rewards, that is evidence for a common currency of choice; if damage to one small region changes how people bargain or gamble, that is evidence that the standard model is not just descriptively wrong but biologically impossible.

The behavioral foundation and the first brain-imaging experiments

The immediate parent of neuroeconomics is behavioral economics, and two of its results set the agenda1. Prospect theory, developed by Daniel Kahneman and Amos Tversky in 1979, showed that people evaluate outcomes as gains or losses relative to a reference point and are loss-averse—losing hurts more than an equal gain pleases. Separately, experiments on intertemporal choice found that people discount delayed rewards steeply in the near term and more gently later, a pattern known as hyperbolic discounting, which is inconsistent with the constant exponential discount rate of standard theory3. These were behavioral facts without a mechanism.

The mechanism began to appear in the 1990s, when cognitive neuroscience developed the tools to look at valuation1. A landmark study by Hans Breiter and colleagues, published in Neuron in 2001, scanned people while they played a simple gambling task and found that the brain's reward circuitry—the dopamine-rich ventral striatum and related regions—responded to both gains and losses, with losses producing a distinct pattern of activity. That was a direct neural echo of prospect theory's claim that losses are not simply negative gains. Around the same time, Wolfram Schultz's work on dopamine neurons in monkeys, begun in the 1980s and 1990s, gave the field its most influential single idea: the reward prediction error.

Reinforcement learning and the prediction error

Schultz recorded from dopamine neurons while monkeys received juice rewards. The neurons did not simply fire when reward arrived. They fired when reward was better than expected, barely fired when it was as expected, and dipped below baseline when an expected reward failed to arrive. This is exactly the signal that reinforcement learning algorithms call a prediction error: the difference between the reward received and the reward predicted. The algorithm—temporal difference learning, formalized by Richard Sutton and Andrew Barto—updates the predicted value of a situation by that error, so that predictions converge on reality. The discovery, reported in the late 1990s, was powerful because it connected a mathematical learning rule to a specific biological signal. It also gave neuroeconomics a template: take a formal model from economics or computer science, find the brain signal that implements it, and use the fit to test the model. A later and more contested extension, by Read Montague and colleagues, used fMRI to show that the human striatum carries a similar prediction-error signal during a simple investment game, which became one of the founding papers of the field.

The somatic marker hypothesis

A different route into the brain came from the study of patients with damage to the ventromedial prefrontal cortex, a region at the front of the brain involved in integrating emotion with decision-making4. Antonio Damasio and colleagues, in the 1990s, studied patients with such damage who were intelligent and articulate but made disastrous real-life decisions. In a card game designed to test this—the Iowa Gambling Task, introduced by Bechara, Damasio, Damasio and Anderson in 1994—participants chose from decks that paid off in different, initially hidden ways15. Healthy participants learned to avoid the bad decks, and crucially, they began to show anticipatory skin-conductance responses—a bodily sign of arousal—before they could consciously explain which decks were bad. The patients did not develop these signals and kept choosing badly. The somatic marker hypothesis proposed that bodily states tag options with emotional value and that these "somatic markers" guide choice, especially when the calculation is too complex to do explicitly4. The hypothesis was influential and also heavily debated; critics argued that the evidence was correlational and that the patients' deficits could be explained by other means. But it established a lasting theme: emotion is not the enemy of rational choice but part of its machinery.

Value-based decision-making and neural currency

The central organizing idea of modern neuroeconomics is that the brain computes a common value signal for options, and that this signal guides choice1. The question is where and in what form. A series of studies in the 2000s, notably by Paul Glimcher and colleagues, recorded from neurons in the parietal cortex and later the orbitofrontal cortex of monkeys choosing between juice rewards of different sizes and probabilities. They found neurons whose firing tracked the subjective value of the offered option—the expected utility—rather than its physical properties1. In humans, fMRI studies found that activity in the ventromedial prefrontal cortex and ventral striatum scaled with the value of chosen options across very different kinds of goods: money, food, and social rewards. The idea of a "neural currency" is that these regions compute a common scale on which otherwise incomparable things can be compared, which is what a choice requires1. The claim is not that there is one neuron for value but that a population code in these regions represents it. This line of work made neuroeconomics a theory-driven enterprise: it tested whether the brain's valuation signals match the predictions of economic models such as expected utility, and where they diverge.

Neuroeconomic theory and the dual-system debate

As the empirical work accumulated, a theoretical wing emerged. Neuroeconomic theory tries to build formal models in which the brain's constraints—limited attention, costly computation, separate learning systems—are part of the model rather than assumed away. One influential strand is the dual-system account of intertemporal choice, associated with Samuel McClure, David Laibson, George Loewenstein and Jonathan Cohen. Their 2004 fMRI study found that immediate rewards activated limbic regions associated with emotion and reward, while all rewards, delayed or not, engaged lateral prefrontal regions associated with deliberation. They proposed that choices between sooner and later rewards reflect a competition between a fast, impulsive system and a slow, patient one, and that this competition explains why discounting is hyperbolic. The dual-system model is intuitive and widely taught, but it is also contested: later work suggested that the "impulsive" and "patient" signals are not as cleanly separable as the original study implied, and that the same valuation regions respond to both immediate and delayed rewards. The debate is a good example of how neuroeconomics works: a striking neural result motivates a theory, and further experiments test and often complicate it.

Social and strategic neuroscience

Economic decisions are often made against other people, and game theory provided ready-made models for testing1. The "trust game" and the "ultimatum game" became laboratory workhorses16. In the trust game, one player sends money to another, and the amount returned determines both payoffs; in the ultimatum game, one player proposes a split of a sum and the other can accept or reject it16. Brain imaging during these games found that activity in regions such as the anterior insula and dorsolateral prefrontal cortex tracked the tension between self-interest and fairness, and that the insula was especially active when people rejected unfair offers—suggesting that rejection is driven by an emotional response to unfairness. Studies of patients with damage to the ventromedial prefrontal cortex, notably by Michael Koenigs and colleagues in 2007, found that they rejected unfair offers more often than healthy controls, which complicated the simple story that emotion drives rejection and reason restrains it. The social strand also connected to the oxytocin system, with experiments suggesting that intranasal oxytocin can increase trust and generosity in some settings, though the effects proved fragile and context-dependent. This area is where neuroeconomics most directly overlaps with behavioral game theory, and it remains one of the field's most active and most debated frontiers.

Neuroforecasting

A more recent development asks whether brain activity can predict not just an individual's choice but the choices of a population1. The idea, sometimes called neuroforecasting, is that the average response of a small group of brains to a stimulus—a song, a crowdfunding pitch, a movie trailer—may predict how a much larger market will respond. In a well-known study, Gregory Berns and Sara Moore reported in 2012 that activity in the ventral striatum of a small group of adolescents listening to new songs predicted the songs' later commercial performance better than the adolescents' own stated preferences did. The interpretation is that the striatal signal captures a more direct, less socially filtered valuation than a verbal rating. The result is striking but the evidence base is small, and the field is cautious about how far it generalizes. Neuroforecasting is best understood as an application of the value-signal idea to aggregate outcomes, and it remains an open research question rather than an established tool.

The present

Neuroeconomics today is a mature but still contested field. Its main approaches descend directly from the history above: value-based decision-making, built on the neural-currency idea and tested with economic models; reinforcement learning and prediction error, which remains the most successful bridge between a formal algorithm and a biological signal; social and strategic neuroscience, which uses game theory to design experiments; and the study of intertemporal choice, where the dual-system model is still the reference point even as it is revised. The field's central open questions are whether a single value signal really underlies all choice or whether there are multiple, competing valuation systems; how much of the brain's decision-making is best described by economic models versus simpler associative learning; and whether neural data can improve predictions or policy in ways that behavioral data alone cannot. The field has not replaced behavioral economics; it has given it a mechanistic foundation and, in doing so, made the departures from the rational-actor model harder to dismiss as mere noise.

References

  1. Neuroeconomics (introduction, History, Decision-making under risk and ambiguity, Major research areas, Social decision-making, Applications), Wikipedia
  2. Behavioral economics, Wikipedia
  3. Hyperbolic discounting, Wikipedia
  4. Somatic marker hypothesis (introduction, Background), Wikipedia
  5. Iowa gambling task, Wikipedia
  6. Dictator game (Variants), Wikipedia

The frameworks of Neuroeconomics

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1994-2010

    Somatic Marker Hypothesis

    Do bodily emotional signals guide risky choices?

    Fly to its tower in Nousfarer

  2. 1997-Present

    Reinforcement Learning and Reward Prediction Error

    How does the brain learn from reward prediction errors?

    Fly to its tower in Nousfarer

  3. 2000-Present

    Behavioral Neuroeconomics

    What brain processes produce economic decisions?

    Fly to its tower in Nousfarer

  4. 2001-Present
  5. 2003-Present

    Social and Strategic Neuroscience

    How does the brain handle fairness and strategy?

    Fly to its tower in Nousfarer

  6. 2004-Present

    Dual-System Theory of Intertemporal Choice

    Why do people prefer immediate rewards to delayed ones?

    Fly to its tower in Nousfarer

  7. 2005-Present

    Neuroeconomic Theory

    Which formal models fit the brain's constraints?

    Fly to its tower in Nousfarer

  8. 2012-Present

    Neuroforecasting

    Can brain activity predict market-level choices?

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Somatic Marker HypothesisReinforcement Learning and Reward Pre…Behavioral NeuroeconomicsValue-Based Decision-Making and Neura…Social and Strategic NeuroscienceDual-System Theory of Intertemporal C…Neuroeconomic TheoryNeuroforecasting

Beyond the field

Where Neuroeconomics's ideas came from in other fields, and what they led to there: the strongest 12 of 21 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Neuroeconomics is a planet in Nousfarer

Its city has a tower for each of these 8 frameworks. Land, walk in, and climb one concept at a time.

Land on Neuroeconomics

Related fields

Fields its ideas are linked to

More in Behavioral Economics

See where Neuroeconomics sits on the Map of Human Knowledge