The consolidation of the experimental branch of economics was given a considerable boost with the Nobel Prize awarded in 2002 to Vernon Smith, the founding father of economic experiments on markets; and in 2012 to Alvin E. Roth for his theoretical, empirical and experimental work on stable allocations and market design.¹

In parallel to this, there has been an intensification of the publication of experimental articles in the most prestigious scientific journals in economics. This suggests that experiments are now established as a valid explanatory tool in the profession toolkit. All things considered, we want to argue that experimental economics is now entering a “golden age.”

This introductory chapter of the book is structured in two blocks. Block one briefly presents the beginnings and some fundamental milestones of experimental and behavioral economics. The second block explains to the reader, by means of a few basic notions, how to carry out an experiment. It also gives a series of indications as to how mistakes can be avoided when designing an experiment.

To test this hypothesis, Bernoulli asked the people around him. From his survey he concluded that most people are not willing to pay even small amounts of money to participate in this lottery. This very simple experiment was the first to draw attention to the divergence between how humans behave when making economic decisions and how they are expected to behave in accordance with a given theory. This situation reflects that the mere calculation of the mathematical expected value of a problem does not necessarily predict (human) behavior. This result indicated, therefore, that it was necessary to modify and improve the theory to explain this phenomenon. This sort of feedback between theory and experiment is what has led much science to substantially progress, and it is what we economists equally should aspire to.

In the same vein, of individual decision-making, Thurstone’s (1931) experiments were the first to explore preferences over different goods, hats, shoes and coats, attempting to construct authentic indifference curves. In 1942, Wallis and Friedman would criticize this experiment due to the hypothetical character of the choices made. This was the beginning of a long debate on the validity of choices under hypothetical or real motivation, which allowed social psychology and experimental economics to be marked out as providing distinct insights, up to the present.

In 1944, Von Neumann and Morgenstern developed a fundamental theoretical contribution to the analysis of individual decision-making. Their theory of expected utility could be applied to problems similar to Bernoulli’s problem and attempted to explain decision-making under uncertainty and risk. Their ideas had considerable impact on game theory and on experimental economics. In 1953, Allais began a systematic exploration of the violations of expected utility theory. In his famous paradox he would reformulate a problem and observe the incoherence between the original decisions and those made for an equivalent formulation of the same problem.

More recently, other theories, such as prospect theory by Kahneman and Tversky (1979) and generalized expected utility by Quiggin (1993), have attempted to go beyond the model of expected utility.

Hence today, thanks to experimental evidence, it is acknowledged that humans are not neutral to risk, not even when facing small financial inducements, and also that there is an asymmetry between a situation where an individual is facing losses and one in which he can achieve gains; in addition, each individual tends to interpret the same probabilities differently.

The analysis of game theoretic models in the laboratory is another important line of study in the discipline. There are now hundreds of experimental papers studying problems of strategic interaction between multiple players. A paradigmatic example is the “prisoner’s dilemma,” which captures a situation where acting selfishly results in a worse outcome for the actors involved than if they had cooperated. The origin of this game is found in a study developed for the RAND corporation in 1950 by Desden and Flood,² in which subjects participating in the experiment – where the game was repeated a 100 times – would cooperate more frequently than predicted by the theory. The experimental literature is filled with variations of the prisoner’s dilemma, with different systems of payoffs, conditions of information, communication, repetition, etc.

Other pioneering experiments – such as those by Lakisch, Milnor, Nash, and Nering in 1954, which explored situations with multiple players – influenced the design of posterior experiments in an important way. Laboratory work by Schelling (1957) on coordination games also drew attention to certain parameters of experimental design which could generate effects that were not controlled (a priori) by the experimenter. From all this we have learned that following a rigorous methodology in the laboratory is fundamental to any experimental science. The process of analysis and revision of the methods employed is continuous and sets out what is, and what is not, desirable in experimental practice.

The analysis of markets was another important area of research in which game theory and experimental economics were applied with excellent results. In 1948, Harvard professor Edward H. Chamberlin had the idea of studying markets under experimental control. Using students who were able to sell fictitious products in the market, while others bought, he wanted to observe if the prediction that markets will reach equilibrium would occur. His laboratory “market” – where students could go around the desks bargaining until the allotted time was up – led (for economists) to a very disappointing result: the amount of goods sold was notably larger than predicted and the prices did not converge to equilibrium (see chapter 11, volume II).

However, the full impact of this work in the academic world did not come immediately from its publication, Chamberlin (1948), but from one of the actual participants in the experiment, who was not convinced as to the interpretation (nor the experimental design) that his professor had derived from the study. Fifteen years later, this same student, one Vernon Smith, published two works – Smith (1962, 1964) – in which he showed that, when the information about the asks and bids (of sellers and buyers) was public and the agents were able to interact repeatedly in the market, both prices and quantities converged rapidly to equilibrium. For this pioneering work and its subsequent extension, Smith was awarded the Nobel Prize in Economics in 2002.

Another source for the development of experimental economics came from psychology in the 1950s and 1960s. A fruitful body of literature developed, around the already mentioned prisoner’s dilemma, studying the capacity people have to play in strategic environments and of calculating the Nash equilibrium of a game.³ Simultaneously, from inside the discipline of economics, researchers started applying game theoretic concepts to non-competitive markets. A notable work in this line is the first experiment on oligopolies (Sauermann and Selten, 1959) developed in Europe by the German researcher Reinhard Selten, who would receive the Nobel Prize in Economics in 1994 together with Harsanyi and Nash himself.

In 1960, Siegel and Fouraker investigated cooperation in oligopolies and attempted to control different methodological aspects such as anonymity or monetary motivation, which have become, together with the rule of not deceiving subjects, hallmarks of experimental economics, and have survived till today.⁴

In parallel to experimental economics, has come the development of behavioral economics.⁵ The fortunes of both disciplines have been generally intertwined, although strictly speaking they are not identical. The former, experimental economics, is a tool (a methodology) while the latter is a discipline whose objective is to develop theoretical models of human behavior. These models incorporate the results obtained in empirical research of “neighboring” sciences (as defined by Camerer and Weber, 2006): social psychology, sociology and anthropology. The aim is to better inform economics of what other disciplines have already learned, although, as said by Binmore and Shaked (2010), this line of thinking is not new, and had already been proposed by Selten in 1978.

Experiments in these allied disciplines had already made evident the limitations of humans when making calculations (bounded rationality), their limited desire to face priority tasks (dilation) and the fact that subjects do not always aim to get what is best for them: that is, they do not maximize utility (see Camerer and Lowenstein, 2003; Weber and Dawes, 2005). As Brandts (2009) aptly summarizes it, the aim should be to understand how “normal” people function.

There are at least three broad research topics that have developed within behavioral economics. The first topic is social preferences: that is, whether the wellbeing of other individuals influences our own wellbeing. Results from multiple experimental settings – for example, the dictator game and the ultimatum game (see Chapters 3, 6 and 8) – show how a significant proportion of people are concerned with the earnings of others: they are willing to give up a significant part of their earnings to either benefit or hurt the other players.

Other examples, such as the trust game (see Chapter 3 and 6), also show that people trust others and behave reciprocally: that is, they respond to kind actions in a kind manner. The work by Charness and Rabin (2002) is a good example of a model that incorporates social aspects into economic decision-making.

The second topic of interest in behavioral economics is that of biases in preferences (see Chapter 2). A key example can be found in the issue of dynamic preferences: gauging present against future outcomes. The influential work by Laibson (1997) set off a series of reflections as to how...