Data, Privacy and Markets
Paper Session
Friday, Jan. 6, 2023 8:00 AM - 10:00 AM (CST)
- Chair: Vasiliki Skreta, UT Austin and University College London
Data, Competition, and Digital Platforms
Abstract
We propose a comprehensive model of digital commerce in which data and heterogeneity are defining features. A digital platform matches consumers and advertisers online. Each consumer has heterogenous preferences for each advertiser's brand, and the advertisers can tailor their product lines to the preferences of the consumer. Each consumer can access each seller's products online or offline. The digital platform can improve the quality of the match through its data collection, and monetizes its data by selling digital advertising space in (generalized) second-price auctions. We derive the equilibrium surplus sharing between consumer, advertisers and the digital platform. We evaluate how different data-governance rules affect the creation and distribution of the surplus. We contrast the unrestricted use of data with contextual and cohort-restricted uses of data. We show that privacy-enhancing data-governance rules, such as those corresponding to federated learning, can increase the competition among the advertisers and lead to welfare for the digital platform and for the consumers.Buying Data from Consumers: The Impact of Monitoring Programs in U.S. Auto Insurance
Abstract
New technologies have enabled firms to elicit granular behavioral data from consumers in exchange for lower prices and better experiences. This data can mitigate asymmetric information and moral hazard, but it may also increase firms' market power if kept proprietary. We study a voluntary monitoring program by a major U.S. auto insurer, in which drivers accept short-term tracking in exchange for potential discounts on future premiums. Using a proprietary dataset matched with competitor price menus, we document that safer drivers self-select into monitoring, and those who opt-in become yet 30% safer while monitored. Using an equilibrium model of consumer choice and firm pricing for insurance and monitoring, we find that the monitoring program generates large profit and welfare gains. However, large demand frictions hurt monitoring adoption, forcing the firm to offer large discounts to induce opt-in while preventing the unmonitored pool from unraveling given the competitive environment. A counterfactual policy requiring the firm to make monitoring data public would thus further reduce the firm's incentive to elicit monitoring data, leading to less monitoring and lower consumer welfare in equilibrium.Information Design in Consumer Credit Markets
Abstract
Over 30m US adults do not use formal consumer credit. How many of these are inefficiently excluded because they lack a credit history or have a poor credit score? We develop a framework to characterize the efficiency-maximizing system of credit histories and credit scoring, subject to the constraints imposed by the severity of adverse selection, and by the ability of credit histories to predict future risk. We find US consumer credit features a moderate amount of adverse selection and persistent consumer types. This adverse selection generates substantial welfare loss: a majority of today's non-borrowers would be first-best efficient to lend to. Credit reporting helps alleviate the costs of adverse selection, with the current US system recovering roughly two-thirds of the welfare that would be lost in a no-credit-reporting counterfactual, relative to a full-information first-best. We find that requiring histories to be shorter -- or to forget past default sooner -- would induce some market unraveling but also would help non-borrowing consumers escape the `"no history trap.'"Quality Disclosure and Regulation: Scoring Design in Medicare Advantage
Abstract
Policymakers and market intermediaries often use quality scores to alleviate asymmetric information about product quality. Scores affect the demand for quality and, in equilibrium, its supply. Equilibrium effects break the rule whereby more information is always better, and the optimal design of scores must account for them. In the context of Medicare Advantage, I find that consumers' information is limited, and quality is inefficiently low. A simple design alleviates these issues and increases consumer surplus by 2.4 monthly premiums. More than half of the gains stem from scores' effect on quality rather than information. Scores can outperform full-information outcomes by regulating inefficient oligopolistic quality provision, and a binary certification of quality attains 94% of this welfare. Scores are informative even when coarse; firms' incentives are to produce quality at the scoring threshold, which consumers know. The primary design challenge of scores is to dictate thresholds and thus regulate quality.JEL Classifications
- D82 - Asymmetric and Private Information; Mechanism Design
- L86 - Information and Internet Services; Computer Software