Work in Progress

Natural Disasters, Aid Distribution, and Social Unrest –  Micro-Level Evidence from the 2015 Earthquake in Nepal

with Jan Pierskalla and Elisa Schwarz

How do natural disasters influence social unrest? We argue that the inconclusiveness of previous research can be partly traced back to a conflation of two countervailing mechanisms. On the one hand, direct disaster exposure increases social cohesion and reduces social conflict. Government responses to disaster damage, on the other hand, can engender new conflict over the distribution of resources. We test this argument with a local-level analysis of the effects of the 2015 earthquake in Nepal: we rely on village-level data on the exogenous spatial distribution of earthquake intensity, the number of violent events, and the patterns of post-disaster aid distribution. We show that earthquake exposure reduced unrest but governmental aid distribution mediated this effect.


Heterogeneous Effects of Development Aid on Violent Unrest in Post-War Countries – Village-Level Evidence from Nepal

Many countries experience massive aid surges when civil wars end. However, operational contexts tend to remain particularly sensitive due to a combination of persisting local-level cleavages and weakened state institutions. Consequently, aid provision risks inciting distributional conflicts and violent unrest – most notably when resources are being injected into areas of high social heterogeneity and particularly low state capacity. I investigate this argument in the case of Nepal. Focusing on the first five post-war years, I combine geocoded aid data with village-level information on various forms of violent unrest, as well as on social demographics and administrative performance. The panel analyses yield positive correlations between aid and unrest. More fine-grained estimations reveal that this net effect results from a reduction of anti-state violence and a more pronounced escalation of non-state violence. The latter is mainly driven by aid provided to ethnically fractionalized villages under the administration of weak local-level state institutions.


Climate, Food Riots and Adaptation:  A Long-Term Historical Analysis of England

with Tim Wegenast

A large body of research indicates that environmental conditions can influence the risk of social unrest. However, we know little about how these effects may change over time – for example, in the context of long-term climate change. Are the effects likely to remain constant or do they change as a consequence of human adaptation to different weather conditions? To investigate this question, we rely on a disaggregated analysis of England over a period of more than 300 years. We first assess the basic assumption that adverse weather conditions negatively impact socioeconomic living conditions by increasing agricultural commodity prices. Combining data on geo-referenced food riots with reconstructed climate data, we then assess the impact of annual temperatures on social unrest over the period 1500–1817. Finally, we use our long-term time-series dataset to assess potential conditioning effects of adaptation on year-to-year associations between temperatures and social conflict. Our models show a substantive and consistent negative correlation between temperatures and wheat prices as well as between temperatures and food riots. In addition, we find evidence of decadal processes of adaptation: past exposure to adverse weather conditions dampens the effect of current exposure. Moreover, our results tentatively indicate that regions facing high weather vulnerability exhibit the strongest adaptation effects.


Pre-Colonial Nation-Building and Contemporary Social Capital – Evidence from the Bushi Kingdom in Eastern Congo

with Carlo Koos

Previous research finds that social capital is an important determinant of the quality of political institutions, economic growth and social conditions. However, little is known about the determinants of social capital itself: why do some societies display higher levels of social capital than others? While previous research has highlighted long-term processes of social capital formation, empirical studies have mainly focused on potentially endogenous current determinants. We contribute to this research by analyzing the historical roots of social capital. We highlight the role of early processes of rudimentary forms of nation-building by arguing that strong horizontal and vertical integration processes grounded in foundational myths of pre-colonial states have strengthened people’s communal bonds to an imagined community that continues to influence pro-social norms and behavior nowadays. We investigate this argument in the Eastern Democratic Republic of Congo (DRC), combining historical information on the location and the main features of the precolonial Bushi Kingdom with original geo-referenced survey data to investigate variation in social capital within and outside of the boundaries of the pre-colonial “nation.” We exploit information on people’s awareness of proverbs associated with the original foundational myths of the kingdom to assess the plausibility of the suggested mechanism of long-term norm persistence. We find evidence in line with our argument that shared reminiscence of the myths contributes to explaining higher levels of present-day social capital.


Conditional Effects of Aid on Political Perceptions – Mixed Methods Evidence from North East Afghanistan

with Jan Koehler and Kristóf Gosztonyi

Development agencies claim that international support to public service delivery can contribute to strengthening state-society relations. The rationale is that improved access to education, health care or drinking water increases output legitimacy and people’s trust in state institutions. Previous survey-based investigations lend some support to this argument. Our paper builds on this research and takes it a step further: in addition to analyzing if development projects affect political perceptions, we investigate under what circumstances such effects can being realized. We focus on one main contextual factor deemed particularly relevant in situations of fragility and violence: the level of instability in people’s direct impact zones. Our analysis makes use of original opinion survey data collected across four provinces of north-east Afghanistan. Our quantitative analysis is systematically “grounded” by qualitative analyses based on several months of intensive field-work. Our findings lend support to the argument that aid matters for political trust. Our ethnographic analysis shows that violence affects this association in a highly context-dependent and multi-causal way. It specifically highlights the need to incorporate variations across local systems of governance into theoretical arguments on associations between aid, violence and political trust.


Efficacy, Blame Attribution, and Protest Scope: Findings from a survey experiment in South Africa

with Miquel Pellicer and Eva Wegner

Similar grievances can motivate different types of protests in terms of demanding very specific policies within a given economic or political system or calling for a change of the system itself. Little is known about the factors that shape people’s preferences for different protest issue scopes. We investigate the role of perceptions of efficacy. Our hypothesis, based on findings in the social psychology literature, posits that higher efficacy/powerfulness reduces the psychological need to justify the system which, in turn, increases preference for broad protest issue scope. To test this hypothesis, we conduct a survey experiment in two South African townships. We find that individuals asked to recall a successful protest (vs. an unsuccessful one and vs. a control group), are not only more likely to feel powerful and have higher levels of efficacy. They are also more likely to attribute common social grievances to the system and to support protests calling for system change (social inequality) rather than specific policies (service provision). These results can help understanding protest dynamics as well as the political choices of the poor.