Random Utility Models, Wine and ExpertsJEL codes

We empirically investigate the welfare effects of providing product quality information to consumers in the form of expert opinion scores in a setting of asymmetric information. Identification of the effects of the provision of information comes from a field experiment in the retail wine market. We use a monthly-product-store panel scanner data set, collected before and during the field experiment, which involves treating a random subset of wine products by displaying expert scores in one store and comparing sales in similar non-treated stores. Using a structural random utility model of demand, we show that, on average, consumers significantly value one score point increase due to the treatment by about 0.5 to 0.7 cents. As a consequence, for a bottle featuring an average score of 83, consumers would be willing to pay additionally between twenty and sixty cents more due to the treatment. Using counterfactual scenarios, we find that adding expert opinion shelf labels increases consumer surplus. In addition, allowing for strategic price reactions by retailers would lead to an overall significant welfare improvement given that (a) consumers significantly value the score information albeit facing higher prices, and (b) the profits increase with the market power of firms.