Surveys are often used to determine the benefits from public programs such as cleaning up the environment. Many of these surveys employ referendum questions such as, "If the cost of preserving this resource would result in an increase in your taxes of $X per year, would you vote for the program?" In recent years, there has been a great deal of debate regarding the validity of the use of referendum surveys to estimate these benefits. One issue is whether respondents' answers to referendum questions reflect their true valuation of the good or program in question. Recent theoretical work describes individuals' optimal responses to different types of survey questions. These results demonstrate that it is not always optimal for subjects to respond truthfully to referendum survey questions. These results also demonstrate that there are some types of surveys for which it is not possible to determine an optimal response because the surveys do not provide any incentives to the respondents. Unfortunately, these types of surveys are often the types used to determine the benefits from public projects. In order to interpret the results of these surveys, it is necessary to know if subjects respond truthfully or not.
This research will use the tools of experimental economics to test the theoretical results described above to determine if individuals behave in accordance with the theory. Experimental economics is the use of simulated markets to observe the responses of individuals to pure economic incentives without the noise that occurs in the real world. The experiments for this research will test the theory about two different types of referendum surveys, single referendum and double referendum surveys. The experiments will also test versions of single and double referendum surveys that provide no economic incentives. A comparison the results of experimental surveys that provide no economic incentives to the results of those that do will provide insight into how to interpret field survey results and design better surveys.
