Edward Chamberlin begins the article ‘It is a commonplace that, in its choice of method, economics is limited to the fact that resort cannot be had to the laboratory techniques of the natural sciences … The purpose of this article is to make very tiny breach in this position: to describe an actual experiment with a “market” under laboratory conditions and to set forth some of the conclusions indicated by it’.

The main conclusion he drew was that market outcomes can deviate from market equilibrium. While the average price was close to the equilibrium price, individual prices were volatile. The law of one price did not hold (see figure 2.20 in the textbook). Vernon Smith went on to challenge this conclusion somewhat by showing that double auction markets are efficient. The inefficiency observed in the negotiated price market used by Chamberlin can be put down to search costs and imperfect information.

Chamberlin advocated the experiments as useful in teaching. He also highlighted that the experiments had yielded “scientific” results. This is a pioneering idea in economics. The fact that we had to wait over ten years for the next article to be published on experiments shows how ahead of the game Chamberlin was.

www.jstor.org/stable/182638

In the article Vernon Smith reports on a series of market experiments that were performed between 1955 and 1961. These were some of the earliest experiments in economics and the results of some of them are summarized in the textbook (see figure 2.17).

The direct contribution of the article was to show that ‘even when numbers are “small” there are strong tendencies for a supply and demand competitive equilibrium to be attained as long as one is able to prohibit collusion and to maintain absolute publicity of all bids, offers, and transactions.’ This finding has stood the test of time and is the one most important results in experimental economics; it shows the allocative efficiency of markets.

The key to Smith’s results was to use a double auction. The only previous experiments, by Edward Chamberlin, suggested markets were not fully efficient. These experiments did not, however, maintain absolute publicity of all bids, offers, and transactions. Smith wanted to see what would happen in a double auction like that used on the stock exchange. “In January 1956 I carried out my plan. To my amazement the experimental market converged 'quickly' to near the predicted equilibrium price and exchange volume, although there were 'only' 22 buyers and sellers, none of whom had any information on supply and demand except their own private cost or value. I thought perhaps that it was an accident of symmetry in the buyer and seller surpluses. I shot that idea down with an experiment later using a design in which the seller surplus was much greater than that of the buyers. Thus, did I seem to have stumbled upon an engine for testing ideas inside and outside traditional economic theory.”

And the indirect contribution of the article was to show the benefits of experimental studies. The article highlighted that different market institutions may be less efficient and that the adjustment to equilibrium was not as many has predicted. This demonstrates the power of experiments. It is an ideal place to test what theories work, and what don’t. Vernon Smith continued to be a pioneer in experimental methods and won the Nobel Prize in Economics in 2002 “for having established laboratory experiments as a tool in empirical economic analysis, especially in the study of alternative market mechanisms”.

www.jstor.org/stable/1861810

The article ‘On Second-Best National Saving and Game-Equilibrium Growth’, was published in the Review of Economic Studies and looked at inter-generational transfer. A starting point was provided the Ramsey growth model – a workhorse of macroeconomic theory. The authors of the present study explain:

‘Ramsey made the remarkable postulate that each generation possesses what we shall call perfect altruism. By this we mean that each generation's preference for their own consumption relative to the next generation's consumption is no different from their preference for any future generation's consumption relative to the succeeding generation. This is a stationarity postulate: the present generation's preference ordering of consumption streams is invariant to changes in their timing. Thus Ramsey did not admit the possibility that the current generation would assign its own consumption a place of importance somewhat out of proportion to its proximity. … But what if people do not subscribe to this ethic? Then the rate of national saving that is optimal from the standpoint of the present generation is not the Ramsey solution.’

To take account of present bias they introduced, what we know call, beta-delta preferences. The beta measures the ‘altruism’ of the current generation. A value of beta equals one equates with the perfect altruism assumed by Ramsey. They, however, suggest a beta less than one meaning the current generation is myopic. This results in the present generation saving less for the future than it would like future generations to do so. Moreover there is recognition that unless the current generation can commit future generations to save more they will not do so. So, we get time-inconsistency, the future may not do as the present planned it should do. We get less than optimal saving for the future. We also get the idea of sophistication, the present realizes that the future may not stick to the plan and takes this into account. It took thirty years for beta-delta preferences to reappear through the work of David Laibson, Ted O’Donoghue and Mathew Rabin. When it did there was an appreciation that present bias, time-inconsistency and beta-delta preferences do not just apply to inter-generational transfers. An individual can have present bias. An individual can plan things for the future that he subsequently does not carry through because of procrastination or preproperation.

This article, published in the Journal of Experimental Psychology, was one of a series that provided evidence of intransitive preferences over risky choices. They report experiments that compared different ways in which one could elicit preferences. One way is to ask how much a person is willing to bid for a gamble. Another is to ask them to compare gambles and choose amongst them. The authors explain:

‘The notion that the information describing a gamble is processed differently for bids than for choices suggested that it might be possible to construct a pair of gambles such that the subject would choose one of them but bid more for the other. For example, consider the pair consisting of Bet P (.99 to win $4 and .01 to lose $1) and Bet $ (.33 to win $16 and .67 to lose $2). Bet P has a much better probability of winning but Bet $ offers more to win. If choices tend to be determined by probabilities, while bids are most influenced by payoffs, one might expect that subjects would choose Bet P over Bet $, but bid more for Bet $.’

Consistent evidence was found of the kind of preference reversals expected. This is a finding that strikes at the heart of economics. As the authors write ‘These reversals clearly constitute inconsistent behavior and violate every existing theory of decision making.’ One of the most basic assumptions of economics is that preferences are transitive and the kind of preference reversals observed suggests preferences are, in fact, intransitive.

You might think that such a shocking finding would be a pivotal moment in the development of behavioral economics. For some reason, however, it was not. Work on preference reversals has continued apace till this day. We now have, for instance, more direct evidence of intransitive preferences and things like regret aversion that can explain them. Preference reversals, however, has get much less attention than other areas of behavioral economics.

Richard Easterlin begins the article: ‘This paper … brings together the results of surveys of human happiness that have been conducted in nineteen countries, developed and less-developed, during the period since World War II, to see what evidence there is of an association between income and happiness. Are the wealth members of society usually happier that the poor? What of rich versus poor countries – are the more developed nations typically happier? As a country’s income grows during the course of economic development, does human happiness advance – does economic growth improve the human lot?'

This was the first systematic study of the relationship between happiness and economic output. What did he find: In every country studied those in the highest income group were happier than those in the lowest income group. But, this did not translate into any difference in happiness between countries. Or any change in happiness over time as average income increased. The simple interpretation of these results is that relative income rather than absolute income is what matters for happiness. Economic growth of itself may not improve the human lot! This leads to some difficult questions about the goal of social policy.

While controversial the findings of this first study have stood the test of time. The study also helped spark a more general interest within economics on the study of happiness. And a recognition that the goal of economics is to improve social welfare and not national output.

The breadth of Herbert Simon’s influence was acknowledged by the Nobel Committee in citing his contribution: ‘Work in a number of fields, including the methodology of science, applied mathematical statistics, operations analysis, economics, and business administration. His work is synthesized in a new theory of organizational decision-making.’ In this theory of organizational decision making we find pioneering work in behavioral economics.

Herbert Simon rejected the idea that firms are run by rational, profit maximizing entrepreneurs. Instead he argued firms are run by a team of cooperating managers who make decisions without enough information or computational capacity to behave rationally. The most lasting contribution from this work is the idea of bounded rationality: Individuals behave subject to information, cognitive and computational limitations and we should take account of that. This led to the idea of satisficing whereby managers, or people more general, accept outcomes that are satisfactory or ‘good enough’.

Truth be told, it is hard to see direct evidence of Herbert Simon’s contribution in behavioral economics. For example, little of behavioral economics takes explicit account of computational limitations, or looks to simulate the decision making process. It is clear, however, that Hebert Simon’s contribution has had a huge indirect influence on moving economics away from the standard economic model. He set the stage for others to follow.

www.nobelprize.org/prizes/economic-sciences/1978/summary/

In this article Kahneman and Tversky criticize expected utility theory and propose the alternative of prospect theory.

The article starts with the critique of expected utility theory. Kahneman and Tversk provide a number of examples where people’s choices appear inconsistent with expected utility maximization. For example, they discuss the reflection effect; this is where a person is risk averse over gains and risk loving over losses. They also discuss the certainty effect; this is where a person favors outcomes that are near certain.

To account for such behavior Kahneman and Tversky introduced Prospect Theory. This says that people go though two phases when facing a choice with risk. In the editing phase they code, combine and cancel details of the choice to make it simpler. In doing this probabilities are converted into decision weights; in particular, small probabilities are overweighted. Outcomes are also converted into gains and losses relative to a reference point. This is done using a value function.

Prospect Theory has stood the test of time and is one of the most important ideas in behavioral economics. Indeed this article by Kahneman and Tversky is one of the most widely cited papers in economics. As one might expect the theory has been refined over the years. Indeed, it was not until a parameterized version of the theory was published in 1992 that it could readily be applied. The key ideas, however, were all there in this pioneering article.

www.jstor.org/stable/1914185

The study ‘An experimental analysis of ultimatum bargaining’ was published in the Journal of Economic Behavior and Organization. At the time there had been experimental studies of bargaining games but little on, so-called, ultimatum bargaining games. An ultimatum bargaining game is essentially characterized by some last stage in which a person is given an offer that she has to either accept or reject. For example, in the simplest two stage ultimatum game the proposer makes an offer of how to split $10 and the receiver either accepts of rejects the offer.

This final stage makes ultimatum games relatively simple to study. In particular, they lend themselves to backward induction. For example, in the simple ultimatum game we ask what the receiver should accept in the final stage. We then ask what the proposer should offer. The sub-game perfect equilibrium is for the receiver to accept any positive offer and the proposer to offer the smallest possible amount. This simple prediction is the ideal thing to test in the experimental lab. Do people behave as predicted? Because if not we are probably a long way from being able to model more complex bargaining games.

The authors found that behavior in ultimatum games was not consistent with sub-game perfection. More often than not proposers offered a fair split. And receivers were willing to reject less than fair offers. This was a ground breaking finding that sparked considerable interest in ultimatum games, dictator games, and fairness more generally.

In this article the fundamental ideas of mental accounting were developed. The article starts by looking at the interpretation of outcomes. For instance, how gains and losses are coded and organized through hedonic editing. It then turns to the question of how this interpretation of outcomes feeds through into choice through transaction utility. Finally, it looks at the implications of mental accounting for marketing and economic behavior.

The discussion is spiced up with some thought provoking examples. For instance in the section on interpreting outcomes we have the following example:

Mr. A was given tickets to lotteries involving the World Series. He won $50 in one lottery and $25 in the other. Mr. B was given a ticket to a single, larger World Series lottery. He won $75. Who was happier? Of those surveyed 64 percent thought that A would be happier and 18 percent thought that B would be happier. The lesson seems to be to segregate gains. In the section on transaction utility we have the following example Imagine that you are going to a sold-out Cornell hockey playoff game, and you have an extra ticket to sell or give away. The price marked on the ticket is $5 which is what you paid for each ticket. You get to the game early to make sure you get rid of the ticket. An informal survey of people selling tickets indicates that the going price is $5. You find someone who wants the ticket and takes out his wallet to pay you. He asks how much you want for the ticket. Assume that there is no law against charging a price higher than that marked on the ticket. What price do you ask for if 1. he is a friend 2. he is a stranger. What would you have said if instead you found the going market price was $10? 3. friend 4. stranger.

For some people the following wording was changed - but you were given your tickets for free by a friend. For others the wording was - but you paid $10 each for your tickets when you bought them from another student. The question of interest is whether price is anchored on seller cost, market price or the price on the ticket. The lesson was charge friends seller cost and charge strangers market price. The paper is one of the most widely cited in behavioral economics and set the agenda for future work on mental accounting. Its main contribution was to show that things like loss aversion and reference dependence matter even for choices that do not involve risk. As such, it essentially took a step back from prospect theory. This is crucial because it shows that the consequences of mental accounting are likely to be felt in just about any economic decision.

The efficiency properties of double auction markets were by now well known in situations where buyers and sellers know the value the product they are trading. But, as of yet little was known about the efficiency of markets in which the value of the product is unknown to traders. Such uncertainty is crucial in asset markets like the stock exchange. In the paper ‘Bubbles, crashes and endogenous expectations in experimental spot asset markets’ published in Econometrica, Vernon Smith, Gerry Suchanek and Arlington Williams describe the first experiments that look at this issue.

In the experiments subjects could trade in an asset over a number of periods. In each period the asset yielded an uncertain dividend. The two main results can be summarized as follows: (i) In the majority of experiments a price bubble was observed. This means the trading price went above the highest level that could possibly be justified given the expected dividend stream. (ii) Over time the price almost always converged to the rational expectations equilibrium. How to interpret these results?

The authors chose to focus on convergence to equilibrium. It is not uncommon for subjects to take time to learn and so one could arguably say that bubbles were a consequence of inexperience. Many, however, took from these experiments the ease with which bubbles could occur. It is not clear how readily such systematic deviations from reasonable price could be seen as consistent with any kind of equilibrium behavior. It seems more consistent with irrational exuberance. The seed was, therefore, set for a long line of work that looks at bubbles in the lab.

The article ‘Allocative efficiency of markets with zero-intelligence traders: Markets as a partial substitute for individual rationality’ showed that efficient outcomes are obtained in a double auction market even if traders have zero-intelligence. They conclude ‘Adam Smith’s invisible hand may be more powerful than some may have thought: when embodied in market mechanisms such as a double auction, it may generate aggregate rationality not only from individual rationality but also from individual irrationality’.

But why is meant by zero-intelligence traders? They consider a market in which traders randomly come up with bid and ask prices to buy and sell. In one condition traders can bid whatever they like even if it means they incur a loss. In a constrained condition traders can bid whatever they like as long as trading at that price would not mean they incur a loss. Efficiency when traders can make a loss is not good. Efficiency when traders are constrained to not make a loss is high and as good as with human traders. So, truly zero-intelligence does not give efficiency. Dumb, but not really dumb, is enough to obtain efficiency.

This nicely illustrates the role of heuristics and ecological rationality. As long as traders use the heuristic ‘do not make a loss’ the outcome will be relatively efficient. So, simple heuristics can work. But they only work when matched with the right environment. The ‘do not make a loss’ heuristic works well in double auction markets but may not work so well in other situations. Ecological rationality is about a good match between heuristic and environment and is a key concept in behavioral economics.

Models of social preferences go back a long way but models until this date were pretty rudimentary. They essentially assumed people were altruistic or envious. In this article Matthew Rabin proposed a model of fairness that was a lot more subtle.

He built his framework around the following stylized facts: ‘(A) People are willing to sacrifice their own material well-being to help those who are being kind. (B) People are willing to sacrifice their own material well-being to punish those who are being unkind. (C) Both motivations (A) and (B) have a greater effect on behavior as the material cost of sacrificing becomes smaller.’ This, therefore, is a model of reciprocity – giving to those who are king and punishing those who are unkind. He applied the model to look at two issues. First he showed that consumers may object to paying too high a price. Second he showed that workers may exert more effort if paid a high wage.

A crucial thing the model takes account of is that judgments of kindness and unkindness are made relative to some beliefs. To capture this Matthew Rabin used the idea of psychological game theory recently proposed by John Geanakopolos, David Pearce and Ennio Stacchetti. This approach allows one to make behavior conditional on beliefs of what others will do, and on what you believe they believe. In equilibrium beliefs are consistent with reality.

The model proposed by Matthew Rabin set a high standard for others to follow. He showed how psychological game theory could be applied to give a subtle and realistic model of fairness. He also showed how the model can be put to good use – not only in explaining behavior in experiments but also explaining behavior in the marketplace. Many models of social preferences have come since that have built upon the foundation set by this article.

This Nobel Prize recognized the importance of non-cooperative game theory in economic modeling and was awarded to the three researchers most influential in its development. John Nash came up with the fundamental equilibrium concept of Nash equilibrium. John Harsanyi showed how equilibrium analysis can be applied in games of incomplete information. And Reinhard Selten showed how equilibria can be refined through concepts like sub-game perfection.

Game theory primarily studies rational choice. Nash equilibrium, for example, is a situation where every person maximizes their payoff given the strategies of others. Game theory also, however, naturally raises questions of behavior. This is because there are typically multiple equilibria in a game and so to say that people play a Nash equilibrium is not enough – which equilibrium are they going to play. We can see behavioral insights in the works of John Nash and John Harsanyi. Reinhard Selten, though, is the most known of the three for the contribution made to behavioral economics. He was a pioneer in experimental economics and one of the first to run experiments looking at strategic interaction.

www.nobelprize.org/prizes/economic-sciences/1994/summary/

In 1995 experiments were still something of a fringe activity in economics. Some great economists were doing experiments but experimental studies were still very much the exception rather than the rule. The Handbook of Experimental Economics can be seen as marking a turning point where experimental methods become more mainstream, more routine.

In the preface the editors write: ‘The impetus for a Handbook of Experimental Economics came from the great growth of interest in the results and methods of laboratory experiments in economics. This created a growing feeling, both inside and outside the experimental community, that it would be useful to have an overview of the field, to help lower the barriers to entry facing potential experimenters, as well as those facing economists and others who wish to have a critical understanding of what experiments accomplish in economics.’

In eight chapters the major contributions of experimental economics to date were explained. Interestingly the focus was very much on strategic behavior with chapters on auctions, coordination problems, bargaining and industrial organization. Individual decision making was only give one chapter. This illustrates how experimental economics is, and always has been, more than about cognitive bias and deviations from the standard economic model.

The editors finish the preface by writing: ‘One of the pleasures of participating in this project has been that it has afforded us the best seats from which to observe one of the most exciting games in town. New centers of experimental economics have sprung up continually while this work was underway, and the interaction between theorists and experimenters has increased apace. Indeed, one of the special pleasures of finally finishing this project is that it is clear that in only a few more years, a single volume will no longer be able to do even the rough justice that we manage here, to such a rapidly growing area of economic research.’

www.kuznets.fas.harvard.edu/~aroth/alroth.html

Daniel Kahneman won the prize “for having integrated insights from psychological research into economic science, especially concerning human judgment and decision-making under uncertainty”. Vernon L. Smith won the prize “for having established laboratory experiments as a tool in empirical economic analysis, especially in the study of alternative market mechanisms”.

The work of both Kahneman and Smith is discussed at length in the book and elsewhere in this Timeline. So, I think it is more interesting here to focus on the differences between the approaches of the two.

Daniel Kahneman has focused more on individual behavior and the inadequacy of the standard economic model. He showed how individual decisions can systematically depart from those predicted by the standard model. The emphasis, therefore, is on cognitive biases in decision making.

Vernon Smith has focused more on aggregate behavior and the validity of the standard economic model. He showed that markets can be efficient. The emphasis, therefore, is more on ecological rationality. People have ways of making good decisions when in the right environment.

These different approaches illustrate the wide role that behavioral economics can play. It gets most ‘publicity’ as a critique of the standard economic model. But, it also has a very important role in evaluating what is good about the standard economic model. The decision of the Nobel Prize Committee reflects that wide remit.

www.nobelprize.org/prizes/economic-sciences/2002/summary/

Virtually all economic experiments up until now had been performed in western Europe and the US. Clearly we can learn a lot from these experiments but there are many questions that we cannot answer with such a narrow subject pool. For example, ample evidence now existed that people have social preferences – they offer a fair share in the ultimatum game and reject an unfair offer. But where do such social preferences come from? To what extent are social preferences an evolutionary adaption or a consequence of culture? What counts as a fair share?

To answer these kinds of questions a large cross-culture study was set up to see how people play in games like the ultimatum game. The study passed through fifteen small scale societies including foraging societies, nomadic groups, small-scale agricultural societies and some that practice slash-and-burn- horticulture. In all fifteen societies they found evidence of social preferences. But, there was also considerable variation from one society to the next. This variation could be explained by differences in economic organization and market interaction.

What this study showed is how preferences and behavior are shaped by culture. Fairness matters. But, what is counted as fair or unfair can be very different in different cultures. For example, in some societies (e.g. Au and Gnau in Papua New Guinea) giving a large share is counted as unfair because there will be an expectation this should be reciprocated. In other societies (e.g. Machiguenga of Peru) a low offer is fair. Behavior is, therefore, shaped by the society in which we live. This finding that has been reinforced by subsequent cross-cultural comparisons.

Robert Aumann and Thomas Schelling made fundamental contributions to the development of game theory. Robert Aumann, for example, made seminal contributions in the study of repeated games and cooperative games. Thomas Schelling showed how game theory can help understand conflict and the importance of bargaining power. Much of this work assumes rational strategic behavior and so is outside the domain of behavioral economics.

The work of Thomas Schelling has, however, a clear behavioral economics flavor to it. In his widely read books The Strategy of Conflict and Micromotives and Macrobehavior he encourages the reader to think through problems and to imagine the consequences of strategy. For example, in The Strategy of Conflict he demonstrated the power of focal points and showed their importance in bargaining problems. His ideas were backed up by simple experiments. Basically, the focus was on how people do behave rather than how they should behave. This is an important step towards behavioral game theory.

https://www.nobelprize.org/prizes/economic-sciences/2005/summary/

In the book Thaler and Sunstein elaborate on a simple but very powerful point: If framing and context effects matter then we need to think very carefully about what frames to use in what context. Appropriate framing can nudge people to make better choices. An inappropriate frame can nudge people to make bad choices.

Policy makers were quick to realize the potential of nudge. Indeed, nudge was particularly appealing in times of austerity and budget cuts. It held out the promise that a simple change in the framing of policy could lead to big outcomes. This was welcome news compared to the traditional message of economists that behavior can only be changed through costly changes in taxes, benefits and the like. Nudge was thus important in bringing behavioral economics to the attention of policy makers. That, in turn, sets the challenge for behavioral economists to deliver good, evidence based advice.

Political scientist Elinor Ostrom challenged the conventional economic view that common pool resource goods will be over-exploited. She demonstrated that with the right institutions people can successful manage commons.

Elinor Ostrom first started studying common pool resource goods as a graduate student in the 1960’s when she participated in a research team studying the water industry in Southern California. In the early 1970s she set up a research centre with her husband Vincent Ostrom called the Workshop in Political Theory and Policy Analysis. Hence began a cross-disciplinary quest to uncover how common resources are best managed. Elinor Ostrom was a relative pioneer in conducting economic lab experiments. But, real insight came from in depth field study in combination with laboratory experiments. For example, in her book Governing the Commons she discusses examples of irrigation, forest management, mountain pasture, fisheries and ground water basins.

While challenging the conventional view that commons can be managed appropriately it is worth highlighting that Elinor Ostrom also found limits to what can be achieved. In some cases commons were not being managed appropriately. This lead to a checklist of things needed if a group is to successful manage a common resource. That checklist is highly invaluable for policy makers as they question whether or not to intervene and how to intervene in resource management.

www.nobelprize.org/prizes/economic-sciences/2009/ostrom/facts/

In many markets find a good match between buyer and seller is complicated. For example, matching doctors and hospitals or matching organ donors with those in need of a transplant is a challenging problem. Game theorists and economists had proposed various algorithms for solving this kind of problem, such as the Gale-Shapley algorithm. Alvin Roth put this theoretical insight to practical use. He helped redesign institutions for matching doctors with hospitals, students with schools, and organ donors with patients.

The Gale-Shapley algorithm has nothing to do with behavioral economics. Applying the algorithm does, however, require thinking through how people will react to different kinds of institutions. Alvin Roth performed many experimental studies to help inform the practice of market design. His contribution to behavioral economics extends well beyond, however, studies of matching markets. For example, he was a pioneer of bargaining experiments and has often written on the methodology of experimental economics. He was also coauthor with John Kagel of the important Handbook of Experimental Economics.

www.nobelprize.org/prizes/economic-sciences/2012/roth/facts/

The 2013 Nobel Prize award was interesting in the contrast it provides between the standard economic model and behavioral economics. The research of one of the winners, Eugene Fama, fits with the standard economic model while that of another, Robert Shiller, fits with behavioral economics.

To explain Eugene Fama’s contribution was to show that asset prices are unpredictable in the short term - days and weeks. A speculator cannot, therefore, make excess profit trading in assets. This was seen as vindicating the efficient market hypothesis that asset prices take account of all the available information.

By contrast, Robert Shiller showed that asset prices are predictable over the long term – years. He showed that asset price movements are too volatile to be consistent with the efficient market hypothesis. Instead, animal spirits or irrational exuberance leads to price volatility along with sustained asset price bubbles. The fact that these bubbles eventually burst is what allows for long term prediction. If you see stocks that are undervalued or overvalued then, while you cannot gain from trading in the short term, you can in the long term.

www.nobelprize.org/prizes/economic-sciences/2013/shiller/facts/