Experimental Economics

Experimental economics is the practice of testing economic ideas by putting real people into controlled, usually paid, decision situations and observing what they do.

10 frameworks, from 1950 to today · Behavioral Economics · Economics

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

Overview

Experimental economics is the practice of testing economic ideas by putting real people into controlled, usually paid, decision situations and observing what they do1. Its practitioners build miniature markets, auctions, bargaining games and lotteries in laboratories and, increasingly, in the field, and they use the results both to check economic theory and to discover regularities theory missed1. What is at stake is the empirical status of economics itself: whether claims about markets, rationality and institutions rest on evidence or on assumption. The field's history is a story of a method becoming a discipline—of small, sceptical demonstrations growing into a standard tool of the profession, and of the arguments over what such experiments can and cannot show.

Precursors: Individual Choice Experiments before Mid-Century

Economists had long reasoned about choice without observing it. The first sustained experimental work on individual decisions came from psychologists. In the 1940s and 1950s, Ward Edwards and others began testing expected utility theory—the model that says a person chooses between gambles by weighing each outcome by its probability and its value—against how people actually behaved. Around the same time, Maurice Allais in 1953 described choice patterns that violated the theory's axioms, and Daniel Ellsberg in 1961 showed that people treat ambiguous probabilities differently from known ones23. These were thought experiments and questionnaire demonstrations rather than controlled laboratory studies, but they established the central puzzle: people's choices among risky prospects systematically depart from the standard model.

The other precursor was the theory of games4. John von Neumann and Oskar Morgenstern's Theory of Games and Economic Behavior (1944) and John Nash's equilibrium concept (1950) gave economists precise predictions about interdependent decisions. But game theory's predictions were notoriously hard to test with field data, because the theorist cannot observe the payoffs and information players actually faced. A laboratory could control exactly those things.

The 1960s: Markets in the Laboratory

The decisive move came from Vernon Smith, then at Purdue. In a series of papers beginning in 1962, Smith ran experimental markets with student subjects trading a fictitious commodity under rules borrowed from real exchanges—double auctions, in which both buyers and sellers call out bids and offers. The question was whether the competitive equilibrium price that theory predicts would emerge even when traders had no idea of the underlying supply and demand1. It did, and remarkably quickly. Smith's finding was that market institutions could aggregate dispersed private information into efficient prices, a result that supported Friedrich Hayek's intuition about markets as information processors but gave it experimental teeth.

Smith's work founded what became known as market institution experiments: the systematic comparison of trading rules—double auction, posted offer, sealed bid—and how each performs under different information conditions. The design principle that made this possible was induced value theory, articulated by Smith in 1976. The idea is that an experimenter can make a laboratory commodity mean whatever she wants by paying subjects a reward that depends on their performance: if a subject earns more by buying low and selling high, she will behave as if she valued the good accordingly. Induced values let the experimenter control preferences, which is exactly what field data never allow. The theory rests on assumptions—nonsatiation (subjects prefer more money), salience (rewards depend on the decisions in question), and dominance (the reward outweighs other motives)—and much later methodological debate concerned when these assumptions fail.

Auctions and the Theory-Experiment Dialogue

In the late 1970s and 1980s, experimental work turned to auctions, where theory made sharp predictions and real stakes were large. The motivation was practical: governments were auctioning radio spectrum, timber and oil leases, and the design of auction rules mattered enormously. Experiments compared auction formats—English (ascending open outcry), Dutch (descending clock), first-price sealed bid, second-price sealed bid—and tested the "revenue equivalence" theorem, which says certain formats should yield the same expected revenue5. They often did not, and the deviations were informative: bidders in first-price auctions tended to overbid relative to the risk-neutral prediction, a pattern consistent with risk aversion. This work fed directly into the design of real auctions, most famously the US spectrum auctions of the 1990s, where experimenters tested proposed rules before they were adopted.

Behavioral Decision Experiments and Prospect Theory

Meanwhile the psychology of choice was being transformed. In 1979 Daniel Kahneman and Amos Tversky published prospect theory, which described how people actually evaluate risky prospects: they weigh losses more heavily than equivalent gains (loss aversion), they evaluate outcomes relative to a reference point rather than absolute wealth, and they distort probabilities—overweighting small ones and underweighting moderate and large ones. Prospect theory was built partly on laboratory choice experiments, and it in turn generated a large experimental literature testing its predictions. The classic demonstration is the endowment effect: people demand more to give up a mug they own than they would pay to acquire an identical one, a gap that standard theory cannot explain because ownership should not change valuation6. Experimental economists adopted these methods and, in the 1980s and 1990s, increasingly ran their own decision experiments to test whether the patterns held under real monetary incentives—a question that mattered because psychologists often paid flat fees rather than performance-based rewards.

Social Preferences: The Other-Regarding Turn

Through the 1980s, the experimental literature had largely confirmed that markets worked and that individual choice was imperfect. A third strand asked whether people cared only about their own payoffs. The ultimatum game, introduced by Werner Güth and colleagues in 1982, was the catalyst: one player proposes a split of a sum, the other can accept or reject, and rejection leaves both with nothing. A purely self-interested responder should accept any positive offer, and a self-interested proposer should offer the smallest possible amount. In experiments, proposers typically offered around 40–50 percent, and responders frequently rejected low offers at cost to themselves. The dictator game, developed by Kahneman, Jack Knetsch and Richard Thaler in 1986, removed the responder's power and still found substantial giving.

These results launched the study of social preferences—the ways people care about others' outcomes, not just their own1. The theoretical response came in the 1990s: Ernst Fehr and Klaus Schmidt's inequity aversion model (1999) and Gary Charness and Matthew Rabin's model of reciprocity (2002) formalized the idea that people dislike unequal outcomes and reward kindness with kindness. The experimental literature also documented conditional cooperation: in public-goods games, where players contribute to a shared pool, many people contribute substantially but reduce contributions when others free-ride. The finding that punishment of free-riders is common—and that people will pay to punish even at cost to themselves—became a major theme, extended by cross-cultural field experiments such as the fifteen-society study led by Joseph Henrich and colleagues, published in Science in 2001, which found that fairness norms varied systematically with market integration and cooperation in everyday life.

Experimental Asset Markets and Bubbles

Smith, together with Gerry Suchanek and Arlington Williams, published in 1988 a study of experimental asset markets that became a landmark7. Subjects traded a risky asset with a known finite life and known dividend distribution, so the fundamental value of the asset was computable. Standard theory predicted prices near fundamental value. Instead, prices often rose far above it and then crashed—bubbles and crashes in the laboratory. The result was striking because it showed that mispricing could arise from the trading institution and the traders' psychology rather than from any real-world complexity. Subsequent work explored what reduced bubbles: experience, short-selling, futures markets, and the presence of traders who had seen bubbles before. This line of research connected experimental economics to behavioral finance and gave regulators a laboratory for studying market instability.

Methodology, Incentives and Replication

As the field grew, so did methodological self-scrutiny. The 1970s and 1980s saw debate over whether student subjects were representative, whether hypothetical choices were valid, and how to avoid experimenter demand effects—subjects behaving as they think the experimenter wants. The induced-value framework was the field's answer to the incentive problem, and the standard practice became paying subjects according to performance. In the 1990s and 2000s, the field codified its practices: pre-registration of designs, reporting of all conditions, and attention to replication. A notable episode was the replication crisis that swept psychology and, to a lesser degree, economics in the 2010s; experimental economists responded with large-scale replication projects, such as the Many Labs efforts, and with greater emphasis on statistical power and transparency. The debate over whether laboratory results generalize to the field—"external validity"—became a central methodological question1.

Field Experiments

The most consequential methodological development of the 2000s was the rise of field experiments: randomized controlled trials conducted in real settings with real stakes. The approach drew on development economics, where researchers such as Esther Duflo and Abhijit Banerjee tested interventions—deworming, microfinance, teacher incentives—with randomization. In behavioral economics, field experiments tested whether laboratory findings held outside the lab: do people save more when reminded, do they respond to default options, does loss aversion shape labor supply? The 2019 Nobel Prize to Banerjee, Duflo and Michael Kremer recognized the field-experiment method's rise. Field experiments blurred the boundary between experimental economics and applied microeconomics, and they raised new questions about ethics, consent and the limits of what randomization can identify.

Online and Digital Laboratories

The final turn was technological. From the 2000s onward, researchers began running experiments over the internet, first through custom platforms and then through large online labor markets such as Amazon Mechanical Turk, launched in 2005, and later Prolific. Online laboratories made it cheap to recruit large, diverse samples and to run studies quickly, and they enabled cross-cultural comparisons at scale. The trade-off was control: researchers could not monitor subjects, and the representativeness of online pools became a subject of study in itself. The COVID-19 pandemic of 2020 accelerated the shift, as in-person laboratories closed and online experiments became the default. Today, many experimental economists run hybrid programmes, combining laboratory, field and online designs.

The Present

Experimental economics today is a mature subfield with its own journals, its own methodological standards and its own Nobel laureates—Smith and Kahneman shared the 2002 prize, and Thaler won in 2017. Its main approaches descend directly from the history above: market and auction experiments test institutional design; game experiments test equilibrium and social preferences; decision experiments test prospect theory and its successors; field experiments test external validity; and online platforms extend all of these1. The open questions are methodological and substantive: when do laboratory results generalize, how should experiments handle subject pools that are not representative, and what do the persistent deviations from standard theory imply for policy1. The field's founding insight—that controlled observation can discipline economic theory—remains its organizing principle.

References

  1. Experimental economics (introduction, Market games, Social preferences, Critiques, Experimental topics), Wikipedia
  2. Allais paradox (introduction, History), Wikipedia
  3. Ellsberg paradox, Wikipedia
  4. Game theory, Wikipedia
  5. All-pay auction, Wikipedia
  6. Endowment effect, Wikipedia
  7. Experimental finance (History), Wikipedia

The frameworks of Experimental Economics

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

  1. 1950-Present

    Game-Theoretic Equilibrium Testing

    Do people play equilibrium strategies in games?

    Fly to its tower in Nousfarer

  2. 1962-Present
  3. 1976-Present

    Induced Value Theory

    How can experimenters control preferences?

    Fly to its tower in Nousfarer

  4. 1979-Present

    Auction Experiments

    Which auction rules raise the most revenue?

    Fly to its tower in Nousfarer

  5. 1979-Present

    Behavioral Decision Experiments

    How do people actually choose under risk?

    Fly to its tower in Nousfarer

  6. 1982-Present

    Social preferences

    Do people care about fairness for others?

    Fly to its tower in Nousfarer

  7. 1988-Present

    Experimental Asset Markets

    Why do asset prices deviate from fundamentals?

    Fly to its tower in Nousfarer

  8. 2000-Present

    Field experiment

    Do laboratory findings hold in the real world?

    Fly to its tower in Nousfarer

  9. 2005-Present

    Online Laboratory Experiments

    Can experiments be run online reliably?

    Fly to its tower in Nousfarer

  10. 2010-Present

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.

Game-Theoretic Equilibrium TestingMarket Institution ExperimentsInduced Value TheoryAuction ExperimentsBehavioral Decision ExperimentsSocial preferencesExperimental Asset MarketsField experimentOnline Laboratory ExperimentsReplication and Methodological Self-S…

Beyond the field

Where Experimental Economics's ideas came from in other fields, and what they led to there: the strongest 12 of 53 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.

Experimental Economics is a planet in Nousfarer

Its city has a tower for each of these 10 frameworks. Land, walk in, and climb one concept at a time.

Land on Experimental Economics

Related fields

Fields its ideas are linked to

More in Behavioral Economics

See where Experimental Economics sits on the Map of Human Knowledge