Data Privacy in Pricing: Estimation Bias and Implications

Abstract

Problem definition: We study two privacy protection mechanisms motivated by emerging privacy regulations, limited retention (the removal of private data from a firm's database) and self-protection (a fraction of privacy-aware customers hiding their private information). We study in the context of data-driven (personalized) pricing how privacy protection affects the estimation of the demand model and, thus, the offered price and consumer surplus. Methodology/Results: Assuming a linear demand curve, we first find that the product/service type determines whether consumer surplus increases and hence surplus redistribution among consumers. In particular, for inferior goods (e.g., those whose demand is negatively correlated with income as a focal private feature), offered prices would decrease for all customers due to limited retention, and counter-intuitively, under self-protection, protecting customers (instead of privacy-unconscious ones) would see a price increase. On the other hand, the magnitude of the resulting price change is greater for industries with stronger historical price personalization, but nonmonotonic in the number of protected features. We validate our theoretical findings with a real dataset of online auto loans and observe that protective customers face consistent price increases, regardless of what features such customers protect. Managerial implications: The two protection scenarios share several properties, yet the effects of comprehensive privacy laws on consumers remain subtle from a regulator's perspective. The unintended consequences should be cautioned: self-protection may inadvertently harm low-income consumers, and stricter privacy laws do not always result in improved consumer surplus despite protecting more information.

Keywords: Privacy, Personalized Pricing, Model Estimation, Revenue Management