VoxEU Column Development

Do conflicts cause poverty, or vice-versa?

This column suggests that in Africa an income drop of 5%—a large but altogether common deterioration in economic conditions—increases the risk of civil conflict in the following year to nearly 30%. This suggests that aid agencies could help prevent war by targeting short-term emergency aid towards countries hard-hit by adverse commodity price movements or weather shocks.

Understanding the tangled web of cause and effect that potentially links poverty and violence is a task that has long stymied social scientists. Does war cause poverty, or vice-versa? Or perhaps other factors – such as societal hatreds or divisions – cause both economic stagnation and war. Maybe all three of these are operating at once. This is no matter for ivory-towered thumb twiddling – foreign aid priorities hinge critically on the answer.

In an earlier Vox commentary, World Bank economist Simeon Djankov and Marta Reynal-Querol of Pompeu Fabra claim that the evidence weighs against poverty-induced violence. Yet their arguments fail to account for more recent research that has most convincingly cut through the Gordian knot of causation to identify the underlying connections between economic deprivation and violence. The punchline: poverty does cause violence.

As Djankov and Reynal-Querol observe, the first to apply econometric methods to studying poverty and violence were a pair of co-authors, Collier and Hoeffler (2004) and Fearon and Laitin (2003). Both sets of researchers report a positive correlation between economic development and violence in a cross-section of countries. Yet these researchers would be the first to agree that their analyses are not the last word on the topic (Once again, did poor countries become violent, or violent countries become poor, or both?)

Djankov and Reynal-Querol rightly caution that one must be careful in the conclusions drawn from such analyses. Yet the methods on which they base their counter-arguments are themselves built on shaky statistical foundations. First, instead of looking only at the poverty-violence connection across countries, they focus on the relationship within countries (using a statistical method called fixed effects): do countries that become poorer simultaneously become more violent over time? They find that this is not the case. But their within-country analysis still suffers from the causal “dog chasing his tail” – poverty-causing-violence-causing-poverty – problem we mention above, and even worse, social upheaval may lead to both civil conflict and less economic activity in ways that obscure the link between poverty and violence. Djankov and Reynal-Querol add “historical” variables – in particular, European settler mortality during the early colonial period – to the regressions of earlier researchers and find that these are associated with civil conflict. Yet these cross-sectional associations once again lead to more questions than they answer: Did countries with high settler mortality end up with political institutions that are responsible for poverty today and also make modern violent conflicts more likely? Or did high settler mortality lead to colonial policies that impoverished nations, laying the seeds for future armed conflicts?

Economists have developed so-called instrumental variables to deal with exactly this sort of chicken-and-egg causal conundrum. Rather than looking at correlations – either across countries, or within the same country across time – we need to look at what happens when some countries become richer owing to nothing other than random chance, and some other countries become poorer simply due to bad luck. Then we could see if these unexpected changes in economic circumstances that “fall from the sky” lead to violence in the poor countries and peace in the richer ones.

Since countries aren’t Petri dishes and we can’t play with levels of global poverty from our office computers, we need to wait for nature to create changes that we can study as if they were a real laboratory experiment. In this case, it turns out that our falling-from-the-sky metaphor isn’t really a metaphor at all. For sub-Saharan Africa’s agrarian economies, poverty does fall from the sky, literally, in the form of (lack of) rain. When the rains don’t arrive, crops wither and die, leaving many without anything to eat or the means to purchase other basic necessities.

If we believe that a direct link connects poverty and violence, then when failing rains create economic hardship, war should follow. In this case, we can actually figure out whether poverty caused violence by isolating rainfall’s effects. Drought and the resulting economic hardship turn out to matter a lot for understanding conflict in Africa. In work with co-authors Shanker Satyanath and Ernest Sergenti of NYU, we find that a 1% decline in national GDP increases the likelihood of civil conflict by about 2 percentage points. So an income drop of 5%—a large but altogether common deterioration in economic conditions, especially when the rains fail—increases the risk of civil conflict in the following year to nearly 30%, up from an already-high average probability of conflict in Africa of around 20% in normal rainfall years. So we find that short-term shocks to income – exactly the type that Djankov and Reynal-Querol purport to study – do trigger violent conflict on the world’s war-prone continent.

A growing number of other studies using rigorous instrumental variable methods have reached similar conclusions. Anjali Bohlken and Sergenti also use rainfall shocks to show that communal riots erupt in India in the aftermath of economic downturns. Using coffee price movements as an economic shock to coffee-growing regions in Colombia, economists Oeindrila Dube and Juan Vargas report analogous findings: when farmers’ incomes collapse along with global coffee prices, local flare-ups in that country’s civil war follow. And in a related study in a sample of African countries, Antonio Ciccone and Markus Bruckner of Pompeu Fabra find that sudden export commodity price drops also make armed conflict more likely, especially in Africa’s less democratic countries.

This emerging stylised fact in economic development, that short-term income shocks result in violent conflict, has critical implications for how foreign aid should be designed. In our recent book, Economic Gangsters: Corruption, Violence and the Poverty of Nations, we propose a form of aid that would treat the economic symptoms of conflict – negative shocks to the incomes of the poor – before the disease of civil war takes hold, and would do so in a way that would not undermine the functioning of the market system. Under our proposal – which we call rapid conflict prevention support– organisations like the World Bank and IMF would target short-term emergency aid towards countries hard-hit by adverse commodity price movements or weather shocks like drought, and would do so before famine or war break out. These emergency funds would be scaled back when prices stabilised or the rains returned.

Economic research can be a valuable tool in guiding foreign aid decisions. But before proceeding to policy pronouncements, it is important that we get the analysis right.


Bohlken, Anjali and Ernest Sergenti, "Ethnic Violence and Economic Growth: An Empirical Investigation of Hindu-Muslim Riots in India", Revise and Resubmit, Journal of PeaceResearch

Brückner, Markus & Ciccone, Antonio, 2007. "Growth, Democracy, and Civil War," CEPR Discussion Papers 6568, C.E.P.R. Discussion Papers.

Collier, P., and A. Hoeffler "Greed and Grievances in Civil Wars." Oxford Economic Papers 56, (2004)

Djankov, S. and Reynal-Querol, M. “Poverty and Civil War: Revisiting the evidence”, CEPR No. 6980 (2008)

Fearon, J. and Laitin, D., "Ethnicity, Insurgency, and Civil War," American Political Science Review, 97 (February 2003)

Fisman, Raymond and Edward Miguel, Economic Gangsters: Violence, Corruption, and the Poverty of Nations, Princeton University Press, 2008

Oeindrila Dube & Juan F. Vargas, 2006. "Are All Resources Cursed? Coffee, Oil and Armed Confict in Colombia," Documentos de CERAC 002024, CERAC -Centro de Recursos para el Análisis de Conflictos

13,333 Reads