March 9, 2027 - 3:00 pm to 4:15 pm (Paris time)
Brishti Guha (Jawaharlal Nehru University, India): State Capacity, Access to Justice, and the Design of Domestic-Violence Enforcement, co-authored with Parimal Bag (National University of Singapore).
Abstract: Domestic-violence enforcement in developing countries must balance access to justice against limited investigative capacity. We develop a model in which victims choose whether and how to report violence, some complainants have an incentive to strategically misreport, and courts receive imperfect signals of guilt. We compare mandatory arrest laws, under which every complaint reaches the court, with warrant-based screening, under which a costly police investigation filters complaints before adjudication. Screening reduces strategic misreporting and makes severe violence easier to deter, because the probability that an investigation yields incriminating evidence rises disproportionately with the intensity of violence. Mandatory arrest, however, avoids the screening stage and so can improve access to justice for low-intensity violence, at the cost of admitting more unsubstantiated complaints and weakening the deterrence of severe violence.
We show that an unconstrained planner does better under arrest with warrants. She prefers graduated penalties under mandatory arrest, but a flat penalty under warrant-based screening. Then we characterize how binding statutory maxima constrain deterrence differently under the two regimes. Finally, we use institutional and administrative evidence from India, where the standard for arrest in domestic-violence complaints changed in 2017, to describe the empirical relevance of the mechanisms.
June 8, 2027 - 3:00 pm to 4:15 pm (Paris time)
Abraham Wickelgren (University of Texas at Austin, USA): The Alarmist's Dilemma, co-authored with Frances Xu Lee (Loyola University Chicago, USA), and Shmuel Leshem.
Abstract: We study situations where a single alarm suffices to avert harm, yet sounding a single false alarm is individually costly. Each of n players receives a private binary signal about the state of the world (safe or dangerous) and must independently decide whether to sound the alarm (act). Because individual action is payoff-relevant only if all others refrain from acting, players rationally discount a danger signal. These pivotality inferences intensify with group size and the relative cost of inaction to action. Higher-quality information can counterintuitively reduce incentives to act. Relative to a single player, larger groups never reduce, but may increase, both error rates. Our model provides a unified explanation for puzzling collective inaction in emergency response (the murder of Kitty Genovese), organizational reporting (whistle-blowing), and intelligence alarm (Pearl Harbor, October 7).
October 12, 2027 - 3:00 pm to 4:15 pm (Paris time)
Danisz Okulicz (Karlsruhe Institute of Technology, Germany).
Abstract:.
December 14, 2027 - 3:00 pm to 4:15 pm (Paris time)
Gunnar Norden USN School of Business, Norway).
Abstract:.