Document Type


Publication Date



We consider the elicitation of subjective belief distributions over continuous events using scoring rules with incentives. The theoretical literature suggests that risk attitudes have a surprisingly small role in distorting reports from true beliefs. We use this theoretical prediction to test the effect of eliciting subjective belief distributions using a binary lottery procedure that should, in theory, lead to truthful reporting irrespective of the risk attitudes of the subject. In this instance this procedure leads to a prediction of “no effect” compared to using direct monetary payoffs to rewards subjects. Of course, it is always possible that there is a behavioral effect from using the binary lottery procedure, contrary to the theoretical prediction. We demonstrate that the available controlled laboratory evidence is consistent with theory in this instance. If this result is true in general, then it expands the applicability of tools for eliciting subjective belief distributions.


To learn more about the Andrew Young School of Policy Studies and ExCEN Working Papers Series, visit and