Research

Working Papers

Job Market Paper

• Behavioral responses to fines: direct and indirect deterrence (pdf)

  Abstract: Governments use fines to deter socially unwanted behavior on a wide range of contexts. It is commonly assumed that, fixing the probability of sanction, the higher the price, the more deterrent a fine is. But with limited government capacity for effective enforcement, the fine is often only an invitation for behavioral change and perceived fairness of the fine price might affect deterrence. I explore this behavioral mechanism with data from automated speed cameras in Bogotá, Colombia. I use the fact that the speed threshold for issuing a ticket differs from the posted speed limit (and is unknown to drivers) as a source of exogenous variation, and the high visibility of speed cameras as a way to differentiate direct (speeding at cameras sites) and indirect deterrence (crashes/speeding elsewhere). With a regression discontinuity design I compare drivers whose speed just exceeded the unknown real limit, and thus get a ticket, with those driving just below the real limit, who do not get a ticket. I find that a ticket causally changes the direct behavior of drivers: ticketed drivers reduce all forms of speeding (direct deterrence). However, I find no change in the probability of crashes (indirect deterrence). Using the commercial price of the vehicle as a proxy for wealth, I also estimate heterogeneous effects based on the relative price of the fine. Drivers at higher wealth levels both reduce speeding and crash involvement, while drivers on lower wealth levels show no change in speeding, but an increase in crash participation. My findings are consistent with fines being an invitation by the government to broadly change behavior that is only accepted if it is perceived to be fair. This has specific public policy implications: speeding tickets might not be a well targeted policy for preventing road fatalities. I hypothesize that in many contexts lower fines might be a more behaviorally effective policy tool.

• Reading Others as Individuals: A Field Experiment on Fiction and Prosociality (with Jose-Alberto Guerra) (pdf) Presented at 2026 Advances with Field Experiments Conference at University of Chicago, 2026 JBEE–LACEA BRAIN–Tec Workshop at Tecnológico de Monterrey, Mexico City. (With funding from III TREES Research Grant Fund and Universidad de los Andes)

  Abstract: What drives prosocial behavior, and how can it be shaped through educational interventions? We examine whether guided reading of literary fiction changes pro-social behavior. In a preregistered field experiment, university students were randomly assigned to one of two four-week extracurricular courses: guided fiction reading or behavioral science, with comparable duration and reading requirements. Following the intervention, participants completed incentivized economic games and psychological scales. We find that the fiction course increases altruism and trustworthiness. We find no robust effects on trust or costly norm-enforcement, and some prosocial norms become more permissive. While empathy does not change, we find suggestive evidence of greater narrative transportation, richer attributions about human behavior, and greater attention to individual differences, alongside lower self–other overlap. These findings are consistent with a social-cognitive mechanism centered on valuing individuation: representing others as distinct individuals rather than interchangeable members of a category. A model of coarse reasoning illustrates how finer representations can increase average prosocial behavior without requiring stronger other-regarding preferences. Our paper contributes to the growing literature on how narrative experiences shape social behavior and suggests paths for educational interventions to increase prosociality by changing social perception, without strengthening prosocial norms.

• Mal Parqueo Campaign Held in Bogotá, 2025: Individual Direct and Indirect Effects (with Rafael Unda and María Ortíz Polanía) Selected fot the presentation at the 2027 Transportation Research Board Annual Meeting.

  Abstract: This article evaluates Bogotá’s 2025 Mal Parqueo Campaign, comparing the effectiveness of formal monetary sanctions and pedagogical civic-culture actions. Using administrative data and a staggered difference-in-differences framework (Callaway and Sant’Anna, 2021), it shows that while both measures correlate with improved behavior, only formal fines deliver robust causal impacts —increasing ZPP usage intensity by 21% and significantly reducing the probability of other traffic infractions. Furthermore, formal sanctions on illegal parking trigger more persistent reductions in speeding than pedagogical warnings, particularly among motorcyclists. These findings suggest that in low-compliance environments, formal enforcement acts as a critical “information shock” that resets risk perceptions and reinforces institutional legitimacy. We conclude that the success of urban mobility norms is fundamentally tied to the credibility and legitimacy of the authority. Since these behavioral shifts occur without an aggregate increase in other sanctions, we argue that fines modify behavior primarily by triggering a cognitive re-evaluation of enforcement risk.

• Police legitimacy and deterrence: an exploration from traffic offenses and crashes (pdf)

  Abstract: Increasing the deterrent effect of sanctions is a pressing public policy issue in many areas where the state has limited capacity to punish all wrongdoers. If institutions in charge of enforcing the law are perceived as illegitimate, sanctions might lose their capacity to deter behavior. This paper uses local media, traffic tickets and road crashes in Bogotá, Colombia to study whether sanctions imposed by more illegitimate authorities are less deterrent. Using the fact that traffic control is done by the national police, I consider police scandals that are highly reported in the media as an exogeneous negative shock to the perceived legitimacy of the institution in charge of road traffic enforcement. My estimation compares the behavior of drivers that get their first tickets on days just after a police scandal with drivers that get their first tickets just before a scandal. Drivers ticketed just after a police scandal are more likely to not pay their fine and get a second ticket. Furthermore, they are 21% (11%) more likely to be in a crash in the following six months than drivers that get their first ticket just before a scandal (any other day). Finding the opposite effect when news shows the police as the victim also supports my hypothesis of loss of legitimacy causing state sanctions to be less deterrent. My results suggest that in this policy issue increasing the legitimacy of existing sanctions might be more effective than increasing the number of sanctions issued.

• The role of legitimacy on deterrence: a speed-control intervention (pdf)

  Abstract: The deterrence effect of legal sanctions has been widely studied and measured. Less attention has been given to the mechanisms by which some sanctions are more deterrent than others. This paper suggests one possible mechanism: the legitimacy of legal sanctions. I present a model in which individuals respond to a legal sanction for previously non-regulated behavior. An implication of the model is that more legitimate sanctions, those with higher social acceptance of the mandated behavior, cause a larger change in the targeted behavior. I test this hypothesis on the Cámaras Salvavidas Program, a speed-control intervention in Bogotá, Colombia. I estimate a Non-linear Difference in Difference equation, using the staggered implementation of the program as source of exogeneity. Legitimacy of legal sanctions is the source of heterogeneity. The program has a bigger effect in locations where the speed limit is more legitimate. It is shown that the result is robust to different ways of measuring legitimacy. Policy implications are discussed, suggesting that less strict but more legitimate sanctions might have a larger impact in reducing socially unwanted behavior.