Behavioral Economics Criticism

Irrationality of Economic Experiments

Experiments can be extremely valuable in the scientific world when analyzing things such as gravity and magnetism. In the lab, you always know if you drop a ball it will fall to the ground, and this will be true for the world outside the laboratory as well. Economists are trying to use experiments done with humans to explain actions of the outside world. Unfortunately, economic experiments are not the same as scientific experiments. What is shown in an economic experiment in a laboratory will not always coincide with what is really occurring in the real world. There are four major reason as to why it is not feasible to relate laboratory results to the outside world.

• Subjects know they are being watched
  • A large aspect of analyzing economic experiments is looking if subjects make moral or wealth-maximizing choices. People are easily influenced if they know they are making decisions in an "artificial environment". It is human nature to act morally if you know you are being watched and judged by an experimenter. In the real world, people are not being watched and therefore will be less inclined to act unselfishly.
• Ability for experiments to be controlled
  • Experimenters have the ability to manipulate and make slight changes to experiments to get the subjects to act in a certain way. A common example given is with regards to the prisoners dilemma game. By using differing words such as "opponent" instead of "partner", the subject will have a tendency to react different ways. Another example given was when a public goods experiment was conducted in many different communities. The results that came were different in many of the communities. Communities that had a culture of sharing returned results of people acting morally and unselfish, while other communities that were not as tight-knit produced results of self-maximizing.
• Small stakes
  • While taking part in an economic experiment, the stakes and cost of making certain decisions are very small for the subject. There is not a very large effect on the subject whether they make one decision or not in an laboratory experiment. Obviously in the real world that are serious consequences and costs when people make decisions. Therefore it is difficult to extrapolate results from laboratory experiments into the real world.
• Ability for subjects to be self-selecting
  • Not all subjects are analyzed in experiments are selected and therefore many subjects are volunteers that are interested in the subject. These people are typically "scientific do-gooders" who have a tendency to want to appease the experimenter and "seek social approval". This obviously does not coincide with the outside world. In the real world, the people making decisions in the market place environment are extremely knowledgeable and have great experience acting in this environment. Subjects performing the laboratory experiments do not have this expertise and will not be a good representation of what is actually going on in the outside world.