Three Papers Accepted for Presentation at CODE@MIT

Our paper on the welfare properties of recommender systems will be presented at CODE@MIT. This paper shows that firms may hurt consumers when they manipulate recommender systems in ways that maximize profits. We use results from a randomized experiment in which we changed price and search cost simultaneously to measure demand. Then we resort to simulations to claim our results. Our paper on the effects of Time-Shift TV (TSTV) on how users consume media will also be presented at CODE@MIT. In this paper we show that giving TV channels for free to consumers with TSTV reduces Internet consumption, VoD revenues and theatrical revenues. We claim these substitution effects by looking at outcomes of a randomized experiment in which we awarded TV channels to consumers. Finally, our paper on the effects of pro-active churn management will also be presented at CODE@MIT. In this paper, we resort to outcomes of a randomized experiment in a network setting to show that firms can benefit largely from pro-actively calling likely churners and more so when social network data is taken into account.

Read More