Evaluating Marginal Internalities: A New Approach

This paper develops a new sufficient statistic approach for estimating the marginal internality from sin good consumption. It models a biased consumer who faces uncertain health harms and receives mandatory health insurance. I show that the marginal internality can be identified by observing how sin good demand reacts to changes in health insurance coverage. The method does not require to recover the true willingness to pay for the sin good or to elicit consumers’ biases using surveys. I calibrate the model to sugary drinks consumption. My results are consistent with studies that use survey-based measures of biases.