Introduction to Experimental Economics

Home | Introduction | Vernon Smith | Ernst Fehr | 15 Country Survey | Other Important Surveys | Conclusion | Works Cited

Why Do Economists Conduct Experiments?

Before we answer this question, it is necessary to define certain terms important to the study of experimental economics. Every experiment is defined by the environment, or the monetary rewards of the experiment. These are the things that motivate exchange. Another important term that experiments use is institutions which define language such as bids, offer, or acceptances as well as the rules about the exchange of information and contracts. Finally the last term is behavior which is how the participant reacts to the experiment.

Reasons Why

1. Test a theory, or discriminate between theories

In order to test a theory, the experimenter must compare what the theory is trying to prove to the observations that are conducted by and experiment. A good theory is when the predictions are more right than they are wrong. Theories are generally subjected to may different tests, and therefore are almost always in need of improvement.

2. Explore the causes of a theory’s failure

After the testing of the theory shows that there is something about it that needs improvement, the experimenter has to make sure that there is something wrong with the theory and not the design of the experiment. When it is confirmed that it is not in fact the design that is causing the incorrect data, the experimenter must find out what is wrong with the theory. This part is vital to anyone who needs to adjust a theory. General things that could improve a theory are if subjects are given more experience or the payoffs are increased, but it is not guaranteed that these will improve every theory.

3. Establish empirical regularities as a basis for new theory

Good theories come about by much observation and hard work. This inspires curiosity as to why there were such regularities. Because of increasing technology and the use of computers, it is just as easy to perform an experiment with complicated rules as to one that is simple. This allows experimenters to really look at the theory from every angle possible and see where the regularities lie. The double auction experiment, which is where buyers announce bid prices and sellers announce asking prices and a binding contract occurs when the buyer accepts a seller’s ask, is one of the theories where it was experimented for a long time in the laboratory before it was released for others to model it.

4. Compare environments

When the experimenter compares the environment of an experiment using the same rules, they find what makes the theory break down if anything. This is important because if one doesn’t do this, then they won’t find the nit picky things that are wrong with the theory until it is already released and being modeled and then someone else might find something wrong with it. The experimenter also finds out more about how people react in a certain environment.

5. Compare institutions

When the experimenter uses the same environments, but they change the rules of exchange is also another way to break down a theory to make it more useful. Some examples of this “include the comparison of the English, Dutch, first and second price sealed bid auctions, the comparison of uniform and discriminative price multiple unit auctions, and the comparison of posted pricing with double auction trading.” [4] People learn more about which rules cause certain reactions.

6. Evaluate policy proposals

A theory can tell someone whether a policy proposal is good or not. For example, “Friedman’s original proposal the Treasury auction securities in one-price auctions led to their comparison with the discriminative rules.”[4]

7. The laboratory as a testing ground for institutional design

The laboratory use for conducting experiments of theories is growing. If it wasn’t for experiments, we would not be able to understand many different games about economics.

What Have Economists Learned from Experiments?

One of the things that economists have learned is that institutions matter. That is, the rules to the rules of the game matter. The rules allow the players to either have access to information or not, and they also determine the incentives of individuals in trading games. “Institutions matter because incentives and information matter.” [4]

Another thing that economists have learned is that unconscious decisions matter. A key principle that was mentioned by Adam Smith was that agents “can achieve efficient outcomes which are not part of their intention.” [4] There are many agents who are uneducated or who do not understand, but they still make decisions where they can maximize their wealth. In the past, it was thought that it was impossible for these types of people to gain anything from the experiments. But it has been proved that an unconscious decision, without any information, can still be a productive one.

In most experiments, the less information, the better. In one of Smith’s experiments, he did a continuous double auction experiment using both private and public information. It turned out that the convergence to the equilibrium was faster. This has also been shown in other experiments. In general, when people are given more information, they will be able to find more self-interested outcomes than Nash Equilibrium, and they will use strategies such as punishing in order to get the outcomes that they desire. Therefore equilibrium will take longer to achieve. For example, there is a game where there is a division of 100 lottery tickets where the two players could either win a small or a large prize and the prizes are usually different between the two players. When it comes time to bargain, and the two players do not know each other’s prizes, then it reaches Nash Equilibrium. If the players know what prizes they both have, then the Nash theory is false. When both players have all the information, this is known as the “curse of knowledge.”[4]

It is said in game theory that there is a common knowledge requirement. This makes it hard for there to be experimental data. Experimenters make common knowledge known by announcing information of the game such as instructions, payoffs and other conditions. It should be known that an experimenter can announce this information as clearly or as many times as they want, but it is not guaranteed that the players will use the information properly. Common information does not imply common knowledge. The expectations of players will still be different. Common expectations come from experience with the game. In one experiment, first time players produced bell-shaped bubbles for their results, but as they played the game more times the volume of the bubble was reduced. The more one plays the game, the better their expectations will be.

Because of this, Vernon Smith believes that game theory cannot advance separate from experimental economics. Game theory has a “predictive value” meaning that there are theories of games where you can predict what the outcome will be. But experimental economics goes further and actually tests these theories on real people. This is “operationally.” [4] Subjects gain knowledge by experience, and this is why Smith believes that game theory will not advance. People need to understand the process by which things happen and this can only be done through experimental.

Another thing that we have learned from experiments is the truth about “whether a buyer or seller can tilt the conditions of the transaction toward personal gain and away from market efficiency, by not revealing true willingness to trade.” [4] A counter example comes from a study by McCabe, Rassenti, and Smith of the uniform price double auction mechanism. At equilibrium, there are many bids and offers that are tied. If a buyer tries to lower the market price by asking for a lower price, his previous bid will simply be replaced so that the market price will stay the same.

The Endowment Effect was also learned from experiments. This is when it is more important for someone to own something than for them to get money for it, even if it will make them better off. For example, there was a man who paid $5 for a case of wine. Years later he was offered $100 for each bottle, but he did not want to give up his wine because it was too important to him. The loss of his wine would be more than getting the money for this man. Kahneman, Knetsch and Thaler say that the endowment effect does not apply to good that are going to be resold, but only with goods that are to be consumed. Their results show that the endowment effect remains statistically significant.

One final thing we have learned from experiments is fairness. People say that it is unfair for companies to raise prices of things because there is a greater need for that thing at the time. For example, it is unfair to raise the price of a snow shovel after a snowstorm or raise the cost of wood after a hurricane. As a result, economic theories predict shortages, but it does not predict what people say about what is happening. Fairness is also brought up in the ultimatum game. If Player 1 has $10 and must give an offer to Player 2, Player 2 can either accept or reject the offer. If he rejects then no one gets any money. If he accepts then they both get the money. Game theory says that Player 1 will offer the smallest amount of money possible and Player 2 will always accept because he will always be better off. However, in experimental economics, Player 2 would reject this offer because it is unfair.

Home | Introduction | Vernon Smith | Ernst Fehr | 15 Country Survey | Other Important Surveys | Conclusion | Works Cited