Between 1981 and 2013, the share of the global population living in extreme poverty fell by 34 percentage points. This paper argues that such rapid reductions will become increasingly hard to achieve for two reasons. First, the majority of the poor now live in middle-income countries where the benefits of growth have often been distributed selectively and unequally. Second, a reservoir of extreme poverty remains in low-income countries where growth is erratic and aid often fails to reach the poor. If the international community is to most effectively leverage available resources to end extreme poverty, it must ensure that its investments in institutions and physical infrastructure actually provide the poor the capabilities they need to craft an effective pathway out of poverty. We term the human and social systems that are required to form this pathway "invisible infrastructure" and argue that an effective domestic state is central to building this. By corollary, ending extreme poverty will require both expanding state capacity and giving the poor power to demand reforms they need by solving agency problems between citizens, politicians, and bureaucrats.

We link over- and under-confidence in math at ages 8-11 to education and employment outcomes 22 years later among the children of PSID households. About twenty percent of children have markedly biased beliefs about their math ability, and beliefs are strongly gendered. Conditional on measured ability, childhood over- and under-confidence predict adolescent test scores, high school and college graduation, majoring or working in STEM, earnings, and unemployment. Across all metrics, higher confidence predicts better outcomes. These biased beliefs persist into adulthood and could continue to affect outcomes as respondents age, since intermediate outcomes do not fully explain these long-run correlations.

Draft available here. Online supplement available here.

Abstract. Mobilizing citizen climate lobbying among Republicans may be a key tool in building political will for bipartisan climate policy in Congress. However, record-high issue polarization—partisan gaps in beliefs—and affective polarization—animosity towards counter-partisans—may hold back efforts to expand the left-leaning citizen climate movement across party lines. We run a series of online experiments with 20,000 participants testing how polarization shapes the spread of citizen movements. When we randomly pair Democrats with Americans across the political spectrum, they are 27% more likely to invite other Democrats than Republicans to email Congress about climate change, even when all of them believe climate change is human-caused. We find three explanations for this outreach gap. First, Democrats correctly believe that their invitation will have half as much impact on Republicans’ action. Second, these strategic beliefs are driven by anticipated affective polarization, or the idea that Republicans will react badly to outreach from Democrats. Finally, Democrats’ own affective polarization matters: they prefer to reach out to co-partisans even when cross-party outreach is as effective.

Draft available here on SSRN. 

Abstract. Food systems account for approximately one-third of global greenhouse gas emissions, and carbon-footprint labeling is an increasingly common tool to shift consumers towards lower-carbon diets. In a randomized field experiment with over 200,000 customers at a major US food-services company, we find that carbon labels increase customer retention by 1.1% and company profits by 0.9%, despite reducing customers’ carbon footprints by only 0.6%. These profit effects suggest that carbon labeling may remain a common sustainability tool, despite its small environmental benefits. Moreover, label targeting is crucial: labels may increase footprints among those who do not agree with their purpose.

Draft available here on SSRN.

Abstract. Policy progress often advances incrementally, and citizen movements must sustain engagement across wins and losses. Wins could mobilize citizens by building political-efficacy beliefs—perceived government responsiveness to citizen demands—but could also deflate demand for additional policy. In an online experiment (N≈6,000), learning about a major US climate policy slightly increases political-efficacy beliefs, but reduces demand for additional policy and has no effect on climate advocacy. In contrast, combining this information with a fictional narrative linking legislation to citizen action increases political-efficacy beliefs, policy demand, information-gathering about climate marches, and lobbying donations. These findings highlight the power of storytelling to shape political movements.