We use data from Pakistan to establish a reciprocal exchange relationship between the judiciary and government. We document large transfers in the form of expensive real estate from the government to the judiciary, and large favors in the form of pro-government rulings from the judiciary to the government. Our estimates indicate that the allocation of houses to judges increases pro-government rulings by 50% and reduces decisions on case merits by 30%. The allocation also incurs a cumulative cost of 0.03% of GDP to the government. However, it allows the government to expropriate additional land worth 0.2% of GDP in one year. The results suggest that such transfers within the state deteriorate the rule of law.
How and when does a reform trigger a cascading effect? This paper provides evidence that a judge selection reform in Pakistan, which shifted the appointment power of judges from the government to a committee of judges, had a multiplier effect on anti-government rulings in the decade following its implementation. As the first generation of committee-appointed judges (first-degree of separation from the government) is replaced by the second generation of committee appointees (second-degree of separation from the government), the reform's effect compounds. Nevertheless, as the reform amplifies anti-government rulings, it also increases the concentration of judges hired from a few select law firms. Despite the rise in judges hiring their former colleagues from law firms where they previously practiced as attorneys, there is no observable decline in the quality of judicial decisions. Rulings based on case merits and adherence to due process of law rise with each degree of separation from government appointments. Selection effects of second-generation judges drawing talent from top law schools emerge as a key contributor behind the reform multiplier effect. Overall, our results underscore that measures increasing the independence of the judiciary can have enduring positive effects on judicial autonomy and decision quality, even when they concurrently alter the composition of the judicial elite by making it concentrated.
The strategic use of the judiciary against political rivals, often termed as lawfare, has historically characterized authoritarian regimes and is now increasingly a source of concern in democracies. The central empirical challenge lies in distinguishing between neutral enforcement of the law versus selective prosecution based on political alignment. Using a regression discontinuity design and newly assembled data from Pakistan’s corruption courts, we provide causal evidence on the existence and extent of lawfare. We document that political opponents of the government are significantly more likely to be prosecuted and convicted for corruption. These effects are large enough to crowd out other anti-corruption efforts. In contrast, government affiliates facing comparable allegations are less likely to be convicted, underscoring the asymmetric nature of law enforcement. These disparities carry profound implications for political competition. Convictions of opposition politicians eliminate the typically observed incumbency advantage, undermining the ability of opposition politicians to contest on equal footing in subsequent elections. In doing so, institutions tasked with upholding the rule of law are redeployed as instruments of political victimization, distorting both justice and electoral competition. That these patterns persist under both civilian and military regimes underscores a deeper institutional fragility: democratization alone may not be sufficient to safeguard the neutrality of courts.
Teaching in public schools is one of the most stressful jobs in the world. We work with one of the largest networks of charter schools to investigate how mental health assistance to teachers impacts their and students' mental health. We provide experimental evidence on how teacher stress can be reduced, how it transmits to student stress, and how it may hold back academic achievement. We randomly assign teachers to mindful meditation, individual cognitive behavioral therapy sessions, and pharmacological aid. Meditation and cognitive behavioral therapy reduce stress, while no impact of pharmacological assistance is observed. In a factorial design, we also cross-randomize teachers to reduce mental health stigma, increasing the acceptability of seeking pharmacological aid. We observe increased take up and impact of pharmacological aid in the cross-randomized group. Overall, we find stress from teacher to student is transmissible and hurts student learning, but these costs can be ameliorated through policy interventions. Social stigma against pharmacological assistance is a key stumbling block beyond financial constraints that hurts teachers' and students' mental health. A social action signaling social acceptance of seeking mental health reduces stigma associated with accepting pharmacological help, increase take-up of pharmacological aid, and reduces the stress in short and medium term.
This paper studies a large-scale digital reform in Pakistan that changes the entry point to justice: Women’s Virtual Police Stations (WVPS), which allow women to report gender based violence (GBV) remotely through a centralized phone-based intake system and digital case management, reducing the need for in-person visits and potentially limiting local discretion at first contact. We combine high-frequency administrative data on police calls and complaints with downstream records on case registration and escalation to examine how digital access reshapes reporting and early-stage justice delivery. Using a difference-in-differences and event-study design around the reform’s rollout, we document three core findings. First, the WVPS increases reported GBV incidents, with effects concentrated in categories where stigma and household control are plausibly most binding. Second, the reform changes the composition and pathway of reporting: remote reporting rises. In contrast, walk-in reporting falls, implying substantial substitution in how women engage the police rather than purely mechanical increases in contact. Third, the digital intake produces measurable downstream movement along the justice pipeline, increasing formal registration and follow-up actions with a delay, consistent with capacity constraints and case processing frictions. The effects are larger in places where women’s mobility and social constraints are more severe, indicating that digital access can relax binding constraints on initiating legal claims. The results show that digitizing the first interface between women and the state can expand access to justice in resource-constrained settings, while also shifting where in the pipeline constraints bind.
We study the altruistic motives behind vaccination intentions. Using data from a field experiment in Pakistan, we find that providing information about the probability of transmitting Covid-19 to others substantially increases vaccination intention. Subjects are more responsive to a treatment that focuses their attention to the loss experienced when they do not get vaccinated and infect others, relative to a treatment that focuses their attention on the gain experienced when they get vaccinated and do not infect others. We explain our findings in a theoretical model that accounts for reference-dependent preferences and loss aversion. Using the model, we estimate subjects’ loss aversion in the context of altruistic preferences to be between 2.2 and 2.9.Overall, our results suggest that loss aversion affects altruistic preferences and can be leveraged to increase vaccination take up.
This paper examines the impact of air pollution on judicial decision-making in district courts across India. Using a novel dataset that combines judicial records with satellite-based pollution and wind data, we exploit exogenous variation in pollution levels caused by upwind pollution sources to identify the causal effect of air quality on court rulings. Our findings indicate that higher levels of air pollution are associated with harsher sentences, longer delays, erratic decisions, and more reversal of decisions at higher courts, consistent with the hypothesis that pollution impairs cognitive function. These results have important implications for understanding the broader societal impacts of environmental degradation.
The public display of violence has become increasingly accessible in recent decades around the world. While such displays can embolden marginal actors to engage in violent acts – amplifying the social contagion of violence – they may also induce others to self-censor in the public sphere, thereby reinforcing silencing norms and reducing overt norm violations in areas with a higher presence of potential perpetrators. This paper examines the impact of a large-scale public support of a violent act on subsequent patterns of violence, fear, and self-censorship. Using a difference-in-differences design, we provide causal evidence on both the contagion of violent behavior and the tightening of social norms in the aftermath of publicized support for violence.