Behavioral Finance

CHAPTER 21

Experimental Economics

It is common in physical sciences to do a laboratory experiment to test a theory or to unearth a new and interesting hypothesis. The research activities of economics did not involve experimental methods until the second half of the twentieth century. The early work in experimental economics was met with some skepticism and even as the field matured and garnered its first Nobel Prize winner, experimental economics remained something of a stepchild among the various research methods utilized in economics.

The rise of interest in behavioral finance brought new life into experimental economics, which had been mostly devoted in earlier days to market price formation and voting behavior. Behavioral finance suggested a number of other issues to explore using experimental methods—bubbles, inefficient pricing, overconfidence, endowment and status quo effects, and others. Charles Holt distinguishes classroom experiments from economic experiments, where the latter involve actual monetary payments.1 Since real-world agents are presumed to be motivated by financial considerations only, economic experiments where actual money is involved are viewed as legitimate research experiments. Classroom experiments, where no money is involved, are viewed by Holt as learning experiments, where the students are doing the learning.

So how does an experiment work? Normally, the structure of the experiment is game theoretic. Game theoretic means that there are explicit “players” ...