Research

This paper investigates whether enduring authoritarian regimes are rooted, in part, in the population's lack of accurate information about their social and economic costs. We explore this question using field and online experiments in the context of Turkey's May 2023 presidential and parliamentary elections. We confirm that voters, especially those supporting the incumbent authoritarian government, systematically understate both the extent to which democracy and media freedom have been eroded in Turkey and their usefulness in dealing with natural disasters and corruption—two salient issues in Turkey. In the field, we find that providing (accurate) information about the state and implications of media freedom increases the ballot-box-level opposition vote share by about 0.8 percentage points (1.5 percent). Online, similar treatment messages have significant effects on beliefs and increase the likelihood of voting for the opposition by about 3.7 percentage points (6.2 percent). Both in the field and online, the results are driven by influencing those likely to vote for the governing coalition and those with inaccurate beliefs about institutional changes since the incumbent came to power—not by mobilizing opposition supporters. The evidence indicates that information provision about the state and effectiveness of democratic institutions can persuade voters in a weak democracy to vote against an authoritarian government.

Gender inequality in conservative contexts may persist partly because of weak citizen demand for progressive policy. Estimating this demand requires eliciting truthful responses on a culturally sensitive issue and isolating voters' reactions to a single policy position. To overcome these obstacles, we use official voting records and two independently implemented field experiments in Turkish elections, in which opposition parties disseminated policy proposals on gender equality or conventional issues. The gender-equality proposals increased the opposition bloc's vote share by 1.1 percentage points (2.1 percent), an effect comparable to that of conventional proposals. This gain accrued not to the campaigning parties but to the opposition party with the strongest gender-equality record, suggesting that content and issue-based sorting, not partisan contact alone, shaped voting behavior. Post-election surveys reveal the opposite pattern: the gender-equality proposals reduced self-reported opposition support by an order of magnitude more than the administrative gain and led to more regressive stated gender attitudes. The divergence in self-reported voting is not explained by survey implementation; most plausibly, it reflects misreporting. Our findings indicate that elite advocacy can activate genuine demand for gender equality in conservative contexts even when stated preferences and attitudes suggest backlash.

We use unique data on daily labor-market outcomes for Indian casual workers to study labor reallocation between agricultural and non-agricultural activities within rural areas. Controlling for both individual time-invariant attributes and time-varying shocks, we find that workers who switch sectors across years or even within a week can obtain 23% higher wages by taking non-agricultural jobs. We then estimate a discrete choice model of daily labor allocation that decomposes preferences for jobs into two types of disamenities: i) those associated with job characteristics and ii) those associated with location. We find that the first type of disamenity is 23% of wages for men and 38% for women, and the second type is 36% of wages for men and 31% for women.

I evaluate randomly varied neighborhood exposure to information campaigns regarding either executive performance, or increases in executive power, prior to a Turkish referendum on weakening checks and balances on the executive. The campaigns increased voter polarization over the referendum, and subsequently changed party affiliation in national and local elections over the next two years, leading to partisan polarization. My results suggest that, when voters disagree on whether increasing executive power is a good policy, more information can increase voter polarization. Finally, I conclude that because potential polarization is often ignored, the impact of information campaigns on civil society is underestimated.

We evaluate the impact of an educational program that aims to build social cohesion in ethnically mixed schools by developing perspective-taking ability in children. The program is implemented in Turkish elementary schools affected by a large influx of Syrian refugee children. We measure a comprehensive set of outcomes that characterize a cohesive school environment, including peer violence incidents, the prevalence of inter-ethnic social ties, and prosocial behavior. Using randomized variation in program implementation, we find that the program significantly lowers peer violence and victimization on school grounds. The program also reduces the likelihood of social exclusion and increases inter-ethnic social ties in the classroom. We find that the program significantly improves prosocial behavior, measured by incentivized tasks: treated students exhibit significantly higher trust, reciprocity, and altruism toward each other as well as toward anonymous out-school peers. We show that this enhanced prosociality is welfare improving from the ex-post payoff perspective. We investigate multiple channels that could explain the results, including ethnic bias, impulsivity, empathetic concern, emotional intelligence, behavioral norms, and perspective-taking. Children's increased effort to take others' perspectives emerges as the most robust mechanism to explain our results.

Organized intergroup violence is almost universally modeled as a calculated act motivated by economic factors. In contrast, it is generally assumed that non-economic factors, such as an individual's emotional state, play a role in many types of interpersonal violence, such as "crimes of passion." We ask whether economic or non-economic factors better explain the well-established relationship between temperature and violence in a unique context where intergroup killings by drug-trafficking organizations (DTOs) and "normal" interpersonal homicides are separately documented. A constellation of evidence, including the limited influence of a cash transfer program as well as comparison with both other DTO crime and suicides, indicate that economic factors only partially explain the observed relationship between temperature and violence. We argue that noneconomic psychological and physiological factors that are affected by temperature, modeled here as a "taste for violence," likely play an important role in causing both interpersonal and intergroup violence.

Linkages between climate and mental health are often theorized but remain poorly quantified. In particular, it is unknown whether the rate of suicide, a leading cause of death globally, is systematically affected by climatic conditions. Using comprehensive data from multiple decades for both the United States and Mexico, we find that suicide rates rise 0.7% in US counties and 2.1% in Mexican municipalities for a 1 °C increase in monthly average temperature. This effect is similar in hotter versus cooler regions and has not diminished over time, indicating limited historical adaptation. Analysis of depressive language in more than 600 million social media updates further suggests that mental well-being deteriorates during warmer periods. We project that unmitigated climate change (RCP8.5) could result in a combined 9-40 thousand additional suicides (95% confidence interval) across the United States and Mexico by 2050, representing a change in suicide rates comparable to the estimated impact of economic recessions, suicide prevention programmes or gun restriction laws.