Abstract
In a theoretical model, we compare the signaling power of a tax policy and an information campaign targeting the consumption of unhealthy goods when the side effects are not observable to consumers. The tax policy is modeled as a costly signal, whereas the information campaign as a costless one (cheap talk). While the information campaign only transmits partial information, we show that the tax policy can transmit accurate information to consumers via a fully-revealing equilibrium. Our empirical findings confirm theoretical predictions. The analysis is based on home scanner data from a panel of more than 7,000 households purchasing high levels of sugar-sweetened beverages. By exploiting the announcement (on March 16, 2016) of the two-tiered structure of the UK soft drink tax and a difference-in-differences specification, we show that the tax policy is an efficient signaling device. Despite the tax not being implemented yet, the consumption of soft drinks announced to enter the high tax rate decreased relatively more than that of soft drinks announced to enter the low tax rate. In contrast, consuming sugar-added soft drinks exempted from the tax did not respond to the announcement. Instead, a previous information campaign launched by star chef Jamie Oliver six months before the announcement of the tax policy had a similar impact on all sugar-sweetened beverages, independently from their sugar content.