What Experimental Protocol Influences Disparities Between Actual and Hypothetical Stated Values? Evidence from a Meta-Analysis John A. List and Craig A. Gallet Abstract. Preferences elicited in hypothetical settings have recently come under scrutiny, causing estimates from the contingent valuation method to be challenged due to perceived "hypothetical bias." Given that the received literature derives value estimates using heterogeneous experimental techniques, understanding the effects of important design parameters on the magnitude of hypothetical bias is invaluable. In this paper, we address this issue statistically by using a meta-analysis to examine data from 29 experimental studies. Our empirical findings suggest that on average subjects overstate their preferences by a factor of about 3 in hypothetical settings, and that the degree of overrevelation is influenced by the distinction between willingness-to-pay and willingness-to-accept, public versus private goods, and several elicitation methods. Key words: CVM, hypothetical bias, meta-analysis JEL classification: Q26, Q28, H41 This paper examines the effects of experimental protocols on the discrepancy between hypothetical and actual stated values. Using a meta-analysis of 29 studies, the authors find that subjects typically overstate their preferences in hypothetical settings by a factor of about 3. The degree of overstatement is influenced by factors such as the distinction between willingness-to-pay and willingness-to-accept, public versus private goods, and various elicitation methods. The study also addresses several key questions, including whether hypothetical bias exists in contingent valuation exercises, how the distinction between willingness-to-pay and willingness-to-accept affects the hypothetical/actual ratio, and whether different elicitation methods or experimental designs influence the calibration factor. The authors conclude that a more thorough examination of the technical aspects influencing calibration functions is needed. The paper is structured with a review of previous work, an empirical model, estimation results, and a summary.What Experimental Protocol Influences Disparities Between Actual and Hypothetical Stated Values? Evidence from a Meta-Analysis John A. List and Craig A. Gallet Abstract. Preferences elicited in hypothetical settings have recently come under scrutiny, causing estimates from the contingent valuation method to be challenged due to perceived "hypothetical bias." Given that the received literature derives value estimates using heterogeneous experimental techniques, understanding the effects of important design parameters on the magnitude of hypothetical bias is invaluable. In this paper, we address this issue statistically by using a meta-analysis to examine data from 29 experimental studies. Our empirical findings suggest that on average subjects overstate their preferences by a factor of about 3 in hypothetical settings, and that the degree of overrevelation is influenced by the distinction between willingness-to-pay and willingness-to-accept, public versus private goods, and several elicitation methods. Key words: CVM, hypothetical bias, meta-analysis JEL classification: Q26, Q28, H41 This paper examines the effects of experimental protocols on the discrepancy between hypothetical and actual stated values. Using a meta-analysis of 29 studies, the authors find that subjects typically overstate their preferences in hypothetical settings by a factor of about 3. The degree of overstatement is influenced by factors such as the distinction between willingness-to-pay and willingness-to-accept, public versus private goods, and various elicitation methods. The study also addresses several key questions, including whether hypothetical bias exists in contingent valuation exercises, how the distinction between willingness-to-pay and willingness-to-accept affects the hypothetical/actual ratio, and whether different elicitation methods or experimental designs influence the calibration factor. The authors conclude that a more thorough examination of the technical aspects influencing calibration functions is needed. The paper is structured with a review of previous work, an empirical model, estimation results, and a summary.