Editor’s note: This post is part of a series showcasing Barcelona GSE master projects by students in the Class of 2016. The project is a required component of every master program.
Measles is currently one of the leading causes of death for young children worldwide. We analyze the impact of measles prevention on later-life human capital outcomes by taking advantage of a measles eradication campaign implemented in 1967 in the United States. We provide evidence with a difference-in-differences design from the 2000 US census micro-sample for the following statistically significant results: the campaign increased completed years of schooling by two weeks, the probability of completing high school by 0.32 per cent and decreased the probability of being unemployed by 4.26 per cent. Due to the exogenous timing of the eradication campaign, we argue that these results can be interpreted causally. To the best of our knowledge our paper is the first one to document adult human capital impacts of early-life measles exposure using a natural experiment.
Figure 1: Yearly Cases of Measles in the United States
Empirical strategy:
The 1967 measles eradication campaign led to an unprecedented drop in reported measles exposure in the US, as depicted in figure 1.
Our empirical strategy uses the fact that there is variation in measles exposure between states prior to the eradication campaign: the decrease in incidence is highest in those states with the highest incidence rates, as depicted in the first stage relationship in figure 2. This allows for a difference-in-differences design, exploring whether the states that had higher prior exposure to the disease gained more in human capital outcomes than the states with less exposure, controlling for pre-existing state-level linear time trends and state fixed-effects among other controls.
Figure 2: Decline 1966-1970
We also perform placebo interventions to test the robustness of our results. As depicted in figure 3, the only positive and statistically significant impacts are found for 1967, the actual intervention year. This lends more support to the causal interpretation of our results.
Figure 3: Placebo Interventions Around Cutoff
Conclusion:
In this paper we show suggestive evidence that exposure to a previously common childhood disease can have negative impacts on educational attainment in adulthood, although the effect sizes are not large. This finding strengthens the literature on the early-life origins of human capital.
Our results are for the most part relevant for developing countries, many of which have not yet achieved the vaccination levels required for herd immunity.
The importance of history in economic development is well-established (Nunn 2009; Spolaore and Wacziarg 2013), but less is known about the specific channels of transmission which drive this persistence in outcomes. Dell (2010) stresses the negative effect of the mita in Latin America, and Nunn and Wantchekon (2011) document the adverse impact of African slavery through decreased trust. But did other colonial arrangements lead to positive outcomes in the long run?
I address this question in my Job Market Paper by analyzing the long-term economic consequences of European missionary activity in South America. I focus on missions founded by the Jesuit Order in the Guarani lands during the seventeenth and eighteenth centuries, in modern-day Argentina, Brazil and Paraguay. This case is unique in that Jesuits were expelled from the Americas in 1767 –following European “Great Power” politics— precluding any continuation effect. While religious conversion was the official aim of the missions, they also increased human capital formation by schooling children and training adults in various crafts. My research question is whether such a one-off historical human capital intervention can have long-lasting effects.
The author at the site of one of the Jesuit missions on Guarani lands in South America.
Setup
To disentangle the national institutional effects from the human capital shock the missions supplied, I use within country variation in missionary activity in three different countries:
Note: The map shows the exact location of the Guarani Jesuit Missions (black crosses) with district level boundaries for Argentina, Brazil and Paraguay.
The area under consideration was populated by a single semi-nomadic indigenous tribe, so I can abstract from the direct effects of different pre-colonial tribes (Maloney and Valencia 2012; Michalopoulos and Papaioannou, 2013). The Guarani area also has similar geographic and weather characteristics, though I control for these variables in the estimation.
Key Findings
Using municipal level data for five states (Corrientes and Misiones in Argentina, Rio Grande do Sul in Brazil, and Itapúa and Misiones in Paraguay), I find substantial positive effects of Jesuit missions on human capital and income, 250 years after the missionaries were expelled. In municipalities where Jesuits carried out their apostolic efforts, median years of schooling and literacy levels remain higher by 10-15%. These differences in educational attainment have also translated into higher modern per capita incomes of nearly 10%. I then analyze potential cultural mechanisms that can drive the results. To do so I conduct a household survey and lab-in-the-field experiments in Southern Paraguay. I find that respondents in missionary areas have higher non-cognitive abilities and exhibit more pro-social behavior.
Endogeneity
Even though I use country and state-fixed effects as well as weather and geographic controls, Jesuit missionaries might have chosen favorable locations beyond such observable factors. Hence the positive effects might be due to this initial choice and not to the missionary treatment per se.
To address the potential endogeneity of missionary placement, I conduct two empirical tests. The first one is a placebo that looks at missions that were initially founded by the Jesuits but were abandoned early on (before 1659). I can thereby compare places that were initially picked by missionaries with those that actually received the missionary treatment. I find no effect for such “placebo” missions, which suggests that what mattered in the long run is what the missionaries did and not where they first settled.
Second, I conduct a comparison with the neighboring Guarani Franciscan Missions. The comparison is relevant as both orders wanted to convert souls to Christianity, but Jesuits emphasized education and technical training in their conversion. Contrary to the Jesuit case, I find no positive long-term impact on either education or income for Franciscan Guarani Missions. This suggests that the income differences I estimate are likely to be driven by the human capital gains the Jesuits provided.
In addition, I employ an IV strategy, where I use as instruments the distance from early exploration routes and distance to Asuncion. Distance from the exploration routes of Mendoza (1535-1537) and Cabeza de Vaca (1541-1542) serves as a proxy for the isolation of the Jesuit missions (in the spirit of Duranton et al. 2014). Asuncion, in turn, served as a base for missionary exploration during the foundational period, but became less relevant for Rio Grande do Sul after the Treaty of Madrid (1750) transferred this territory to Portuguese hands. For this reason and to avoid the direct capital –and Spanish Empire—effects, I use this variable only for the Brazilian subsample of my data (as in Becker and Woessmann 2009; Dittmar 2011). The first-stage results are strongly significant throughout (with F-statistics well above 10), and the second-stage coefficients for literacy and income retain their sign and significance –appearing slightly larger—in the IV specifications.
Extensions and Mechanisms
To complete the empirical analysis, I examine cultural outcomes and specific mechanisms that can sustain the transmission of human capital from the missionary period to the present. I find that respondents in missionary areas possess superior non-cognitive abilities, as proxied by higher “Locus of Control” scores (Heckman et al., 2006). Using standard experiments from the behavioral literature, I find that respondents in missionary areas exhibit greater altruism, more positive reciprocity, less risk seeking and more honest behavior. I use priming techniques to further investigate whether these effects are the result of greater religiosity –which appears not to be the case.
In terms of mechanisms, my results indicate that municipalities closer to historic missions have changed the sectoral composition of employment, moving away from agriculture and towards manufacturing and services (consistent with Botticini and Eckstein, 2012). In particular, I document that these places still produce more handicrafts such as embroidery, a skill introduced by the Jesuits. People closer to former Jesuit missions also seem to participate more in the labor force and work more hours, consistent with Weber (1978). I also find that indigenous knowledge —of traditional medicine and myths—was transmitted more from generation to generation in the Jesuit areas. Unsurprisingly, given their acquired skills, I find that indigenous inhabitants from missionary areas were differentially assimilated into colonial and modern societies. Additional robustness tests suggest that the results are not driven by migration, urbanization or tourism.
We use our own and third party cookies to carry out analysis and measurement of our website activities, in order to improve our services. Your continuing on this website implies your acceptance of these terms.
The following job market paper summary was contributed by 
