Working papers


Engendering Empathy, Begetting Backlash: American Attitudes toward Syrian Refugees (with Claire Adida and Adeline Lo)

Existing research has shown how easily individuals are moved to harbor exclusionary attitudes toward out-group members. Can we foster inclusion instead? This paper leverages the ongoing Syrian refugee crisis – one of the most significant humanitarian crises of our time – to test whether and under what conditions American citizens adopt more inclusionary attitudes and behaviors toward Syrian refugees. We conduct a nationally representative survey of American citizens in the weeks leading up to the 2016 presidential election and experimentally test two mechanisms hypothesized to promote inclusion: information and empathy. We examine attitudinal measures of acceptance of refugees, as well as a substantively important behavioral measure – writing a letter to the 45th president of the United States in support of refugees. Our results unveil significant effects on attitudes and behavior of both empathy and information treatments that are mediated by partisanship. The empathy treatment resulted in an increase in the likelihood of writing a letter in support of refugees. An examination of heterogeneous effects by party reveals that the empathy treatment engendered inclusionary attitudes among Independents, and the increase in letter writing was driven primarily by Democrats, whose underlying attitudes did not change, but also by Republicans. The information treatment, on the other hand, did not robustly improve attitudes or behavior of Democrats or Independents, and may have induced a backlash among Republicans. We discuss implications for understanding what kinds of interventions increase inclusion and which create backlash.
Crowdsourcing Accountability: ICT for Service Delivery (with Guy Grossman and Jonathan Rodden)

We examine the effect on service delivery outcomes of U-Bridge, a new information communication technology (ICT) platform that allows citizens to send free and anonymous messages to local government officials, thus reducing the cost and increasing the efficiency of communication about service delivery. In particular, we assess the extent to which U-Bridge improved monitoring by the district, effort by service providers, and inputs at service points in the areas of health, education and water in Arua district, Uganda. Despite relatively high levels of system uptake, enthusiasm of district officials, and anecdotal success stories, we found evidence only of marginal and uneven short- term improvements in the outcomes we measured, and no discernible long-term effects. A crucial issue is that relatively few messages from citizens provided novel, specific, and actionable information about service provision within the purview and resource constraints of the district officials. Our findings suggest that for crowd-sourced ICT programs to move from isolated success stories to long-term accountability enhancement, the quality and specific content of information provided by users is centrally important.

Social Networks and Community Reporting (with Romain Ferrali, Guy Grossman, and Jonathan Rodden)

Do social networks matter for the adoption of technological innovations for political communication? Recent work has highlighted the role of social networks in the adoption of agricultural and health products, but it is not clear whether these findings apply for goods for which substantial collective action problems may be present. We conduct a social network mapping in sixteen Ugandan villages where a new communication technology was introduced to allow citizens to report service delivery problems to local government officials, a technology with potential large benefits to the village, but with the potential for free-riding. We find that the number of ties to adopters of the technology had a large and significant effect on adoption. The relationship between number of neighbors that have adopted and an individual’s adoption of the technology was driven by ties with friends and family rather than by ties with village leaders, suggesting that people may view political communication by peers as a complement to their own efforts, but by local leaders as a substitute. We also find that while both social learning and social influence affect the likelihood of adoption, social influence matters relatively more. Together, our results demonstrate the important role of social networks in the adoption of a type of technological innovation that is becoming increasingly common – technology for political communication – and highlight the ways in which the adoption of this kind of technology may be distinct from those studied previously.

The Muslim-Christian Education Gap in Africa

While the study of politics in Africa has long focused on ethnicity as the predominant cleavage affecting political and economic outcomes, large inequalities in socio- economic status exist across religious groups, including among co-ethnics. In particular, Muslims, who comprise one-third of the population of sub-Saharan Africa and are the fastest growing Muslim population in the world, are far less educated on average than Christians. This paper documents this widespread inequality and provides an empirical investigation of the correlates of the Muslim-Christian education gap at the country level. There is a strong and non-linear relationship between percentage Muslim of a country and the education gap. The gap increases sharply among both men and women as the Muslim population grows from being a very small minority to a larger minority, after which the relationship is flat. The patterns are consistent with a long-term effect of the initial distribution of colonial investments in formal education being concentrated in non-Muslim areas, with large Muslims minorities more easily excluded from colonial investments than small Muslim minorities. Wealthier countries tend to have larger education gaps, while the percentage of Muslims living in urban areas is negatively correlated with the education gap. The findings point toward new directions for the study of African politics, the politics of education, and the persistence of inequality across identity groups.

Muslim Majority Disadvantage in Africa

Religion is increasingly recognized as an important determinant of political and economic outcomes, from political participation to innovation. This paper identifies a new and related empirical relationship between religious identity and human capital in the context of sub-Saharan Africa. Using census and survey data from eleven African countries, I find that Muslims have lower educational attainment than Christians, and educational attainment is lowest where Muslims live as a local majority, a phenomenon I refer to as Muslim majority disadvantage. This pattern is found for men and women, adults and children. What explains Muslim majority disadvantage? Using administrative and survey data from the case of Malawi, I show that physical access to schools is lower in Muslim majority areas, and that Muslims living as a majority tend to be poorer than others. However, a large residual Muslim majority effect remains after taking into account both physical access and poverty. Using data from a new household survey with embedded experiments conducted in two districts in Malawi, one Muslim minority and one majority, I investigate a set of possible mechanisms that may explain the residual effect of Muslim majority status, controlling for access. I find similar levels of demand for education across Muslim minority areas, and no evidence that religious education is displacing formal education. Muslims living as a majority have more religiously homogeneous social networks, and lower access to formal sector jobs through these networks. The evidence is consistent with a “ethnic capital” explanation for the persistence of educational inequality in Muslim majority areas.

Ebola, Elections, and Immigration (with Claire Adida and Kim Yi Dionne)
Abstract: Do infectious disease threats affect attitudes toward immigration? Scholarship investigating the determinants of individual attitudes toward immigration has found that, across a wide range of contexts, cultural threats increase exclusionary attitudes toward immigrants. Yet much of this literature ignores the link between immigrants and disease that has characterized immigrant exclusion for decades. Further, existing work that does investigate the relationship between immigrants and disease does not consider how political entrepreneurs might exploit such threats to move public opinion on immigration. Our paper leverages the Ebola crisis in the United States, which coincided with the 2014 midterm elections, and which politicians sought to manipulate for political gain, to assess the impact of the politicization of a public health crisis. We ran a survey experiment between November and December 2014 to measure the American public’s response toward Ebola and immigration. Our results confirm that public opinion on immigration is easily swayed by political entrepreneurs, and that this effect is not merely driven by partisanship.
Religion, Patriarchy and the Perpetuation of Harmful Social Conventions: The Case of Female Genital Cutting in Egypt. (with Lisa Blaydes)
Abstract: How are harmful social practices brought to an end? Female genital cutting (FGC) — also known as female genital mutilation or female circumcision — is among the most widespread forms of physical violence committed against children, worldwide. This paper describes the social and religious determinants of FGC in Egypt as well as differential trends in abandonment of the practice across different religious and societal groups. While FGC remains a nearly ubiquitous practice among ever-married women in Egypt, significant declines in the practice have been witnessed among a cohort of younger women and girls. We find that although FGC was nearly universal among a previous generation of Coptic Christian women, the Coptic community has seen much steeper declines in the practice over time compared to Muslims despite a narrowing of the educational attainment gap between Christians and Muslims in Egypt. We also find that women who have more sons relative to daughters — an indicator with an exogenous component — are more likely to circumcise their daughters and to state their support for the continuation of the practice in the future. Women with more sons relative to daughters are also less likely to “change course” across daughters with regard to the practice; in other words, while some families have older daughters who were circumcised and younger daughters who were not, this is pattern is observed less frequently in families with a larger percentage of sons. We hypothesize that having sons invests mothers, to some degree, in patriarchal values that impact outcomes for female siblings in the family.
Africa’s Health Tragedy? Ethnic Diversity and Health Outcomes.
Abstract: Does ethnic diversity affect health outcomes? This paper examines the effect of ethnic diversity on a broad range of health outcomes in a global sample of countries, as well as by region and income level, with particular emphasis on the sample of sub-Saharan African countries. The paper also includes an analysis of the relationship between ethnic diversity and health care provision at the sub-national level within Uganda. In the global model, with a wide range of controls, greater ethnic diversity is associated with poorer health outcomes, including higher infant and child mortality, and lower public health expenditure. However, the determinants of health outcomes, and the relationship between ethnic diversity and health outcomes, varies by region and income level. Within Africa, variation in health outcomes is explained primarily by access to health facilities, as well as the quality of institutions. Within Uganda, more ethnically diverse districts tend to have lower provision of health services but the same amount of physical health infrastructure as their homogeneous counterparts.


Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s