I exploit a natural experiment in Indian schools to study how being integrated with poor students affects the social behaviors and academic outcomes of rich students. Using administrative data, lab and field experiments to measure outcomes, I find that having poor classmates makes rich students (i) more prosocial, generous, and egalitarian; and (ii) less likely to discriminate against poor students, and more willing to socialize with them. These effects are driven by personal interactions between rich and poor students. In contrast, I find mixed but overall modest impacts on rich students’ academic achievement.

Can research findings change political leaders’ beliefs and policies? We use experiments with 2,150 Brazilian municipalities to measure mayors’ demand for and response to research information. In one experiment, we find that mayors are willing to pay to learn the results of evaluation studies, and update their beliefs when informed of the findings. They value larger-sample studies more, while not distinguishing between studies in rich and poor countries. In a second experiment, we find that informing mayors about research on a simple and effective policy, taxpayer reminder letters, increases the probability the policy is implemented by 10 percentage points.

We design three field experiments to estimate how workers' social preferences toward their employer motivates their work effort. We vary the pay rates offered to workers, the return to the employer, and employer generosity demonstrated via unexpected gifts. Workers exert effort even without private incentives, but their effort is insensitive to the return to the employer. This is consistent with “warm glow” but not pure altruism. The gifts have no effect on productivity, but engender extra work. This difference is explained partly by the finding that extra work is much more responsive to incentives than is productivity.

Do men and women have different social preferences? Previous findings are contradictory. We provide a potential explanation using evidence from a field experiment. In a door-to-door solicitation, men and women are equally generous, but women become less generous when it becomes easy to avoid the solicitor. Our structural estimates of the social preference parameters suggest an explanation: women are more likely to be on the margin of giving, partly because of a less dispersed distribution of altruism. We find similar results for the willingness to complete an unpaid survey; women are more likely to be on the margin of participation.