Research
Research
Attempt limits are widely used in standardized tests for personnel selection and licensing. Using data from the U.S. Bar Exam, I provide descriptive evidence that introducing attempt limits increases passing rates. This is consistent with the rationale that limiting attempts imposes a test-taking cost that induces self-selection: candidates sit only when they expect a sufficiently high probability of passing. In this sense, attempt limits can improve the average quality of successful candidates. Motivated by this evidence, I develop a principal–agent model revealing a trade-off: attempt limits deter weaker candidates but also prevent some stronger ones from eventually succeeding, so governments strongly averse to human capital loss may oppose them. Thus, I provide a normative framework for the design of attempt limits. First, I show that the limit is optimal only if the test is not difficult. Second, when the limit is chosen jointly with the passing score, the optimal design depends on the test’s information structure: if the test is better at identifying strong candidates, the attempt limit is optimal; if instead the test is better at identifying weak candidates, the principal prefers a difficult exam with unlimited retakes. Finally, when the information structure itself can be chosen, the principal should adopt a test that is better at identifying strong agents if candidate's responsiveness to the attempt limit is high.
Unintended Effects of Vaccine Mandates on the News Market
This paper examines how Italy’s COVID-19 vaccine mandate affected the production of “dog-whistle” articles—news stories reporting sudden cardiac or unexplained deaths—by Italian media outlets. Following the introduction of the mandate, the volume of such articles increased substantially. I propose a model in which profit-maximizing media outlets anticipate the behavior of vaccine-skeptical readers, who oppose the mandate and actively search for anecdotal evidence that appears to undermine it. In response, outlets strategically amplify content that caters to these readers’ beliefs in order to increase online traffic. The analysis suggests that vaccination policies can generate unintended informational distortions when media incentives interact with heterogeneous audience preferences.
Media Bias in Local News Markets
This paper examines how shifts in media revenue structures affect information quality and political coverage. I show that declining advertising revenues do not necessarily degrade information quality. Instead, falling advertiser demand incentivizes outlets to transition toward subscription-based business models. As reliance on subscription revenue increases, outlets raise subscription prices and cater to the preferences of their most engaged readers, who demand more informative content. This mechanism is consistent with empirical evidence showing that political coverage contracts as advertising revenues fall: as information quality improves, politicians receive less favorable coverage.