Abstract:

While the process of data collection can lead to bias, little empirical evidence investigates the role of the survey method. In this paper, we compare two survey methods: the standard face-to-face interview and an alternative method we call Human-Assisted Self-Administered survey (HASA). In the latter, respondents are guided by an enumerator reading questions, but they answer privately on an electronic device. Taking advantage of an RCT in Benin, we randomize the survey method across respondents while holding the questionnaire constant. We show that the survey method leads to different results depending on the degree of enumerator influence. Identifying this influence by quantifying how much of the variation in the outcome variable is attributable to enumerators, we document that variables that are unlikely to be influenced by enumerators do not differ significantly across survey methods. However, variables that are likely to be affected differ systematically. These variables are mainly related to norms, opinions, and beliefs. In particular, we find that respondents who answer directly on an electronic device report less gender-equal behavior and values. Investigating the mechanisms, we show that social desirability bias is more likely to affect responses in classical face-to-face interviews, where individuals' responses are less confidential.

Link to published paper; Link to Working Paper

Abstract: 

In sub-Saharan Africa, women constitute the majority of new Christian membership (including Evangelical and Pentecostal denominations), and this gender gap exceeds that of any other religion. Existing explanations for conversion to these churches emphasize demand for mutual help or informal insurance. We instead show that emancipation is central: these churches provide services that support women’s economic advancement and help them challenge patriarchal norms. Using experimental data from Benin, we find that women randomly offered an economic opportunity become more likely to convert, partly because these churches help counter witchcraft threats, a risk that disproportionately targets economically successful women. To assess external validity, we combine large African datasets with local variation in exposure to positive economic shocks and prevalence of witchcraft beliefs. Women are more likely to join new Christian churches following such shocks, especially where witchcraft beliefs are widespread. There, women work more, have fewer children, and exercise greater decision‑making power, while both women and men reject traditional beliefs, rituals, and authorities.

Link to Working Paper

Abstract:

Conventional demographic models systematically overestimate fertility decline in sub-Saharan Africa. This paper proposes a complementary explanation grounded in a prevalent but understudied belief system: ancestors influence the living and seek the continuation of their lineage, into which they may be reincarnated. In this worldview, having children becomes a moral and collective duty, rooted in the spiritual responsibility to ensure the survival of the lineage. Drawing on first-hand data, novel ethnographic information, and historical and contemporary surveys, I document a strong and quantitatively large positive relationship between ancestral beliefs and fertility across contexts and time periods. A simple model in which children are a public good for the lineage rationalizes the patterns observed in the data: the fertility effect of ancestral beliefs is concentrated in patrilineal societies, and a specific form of free-riding emerges among siblings whose children continue the same family line. These findings suggest that high fertility in sub-Saharan Africa rests on moral foundations that standard, externally designed interventions should take into account.

Link to Working Paper

Abstract:

Spatial inequality in access to schooling is large and persistent across much of the developing world, and is often traced to the colonial placement of Christian missions. We show that in the Democratic Republic of the Congo this inequality does not fade but is actively reproduced by religious competition. Combining a geolocated census of all secondary schools with newly assembled data on colonial missions, we document that schools increasingly cluster around former mission sites, leaving a quarter of the population without any local school. Exploiting the timing of openings within locations, we find that religious providers avoid opening where their own group has just opened, but respond to a rival group by opening more, a pattern that appears only after the 1990s, with the rise of new Christian churches. A simple model of competition for believers explains these patterns.

Link to Working Paper

Abstract:

Why does fertility vary widely across societies, net of income, education, and health? We argue that where kinship is organized around lineages, fertility is a partly collective decision: a lineage-size externality and pooled costs raise fertility directly and open a wedge between private and lineage benefit, from which pronatalist norms can arise, further raising fertility and dampening its responsiveness to price-based policy. Combining 147 DHS waves across sub-Saharan Africa with ethnographic data and a spatial regression discontinuity design at ethnic-homeland borders, we find lineage organization raises completed fertility by 0.19–0.28, about a third of the female-education effect. The effect attenuates along erosion margins: historical Christian missions, urban proximity, and pre-colonial political centralization. Consistent with cost-pooling, the gap widens with fostering prevalence; consistent with norm enforcement, direct pressure and stated ideals track gender-specific descent rules. The institution that raises fertility also insulates it from standard policy: a randomized contraceptive subsidy in Burkina Faso and a free-primary-education rollout both lower fertility only where lineage organization is weakest.

This paper examines the long-term impact of Catholic missionary nuns on women’s human capital in the Democratic Republic of Congo. By using newly digitized historical data on Christian missions, recent demographic surveys, and administrative data on schools and healthcare facilities, we analyze the lasting effects of the missionaries' presence, focusing on gender-specific outcomes. While both Catholic and Protestant missions influenced educational attainment, the presence of Catholic nuns significantly enhanced these effects, especially for girls. Proximity to Catholic missions is also associated with better health outcomes. Beyond education and health, exposure to missions with nuns delays marriage, reduces polygamy, and increases women’s decision-making power within households. However, the negative effects on female labor force participation likely reflect the enduring influence of the “Christian household” model promoted during the colonial period. Overall, Catholic missionary nuns played a decisive role in shaping women’s outcomes, with effects that remain visible more than a century later. 

Link to published paper