VoxEU Column Development

The spread of civil war

Civil wars are devastating to a country’s development perspectives. What’s more, they often spread across borders. But this column argues that only ethnic civil wars pose a significant threat to neighbouring countries’ stability. Countries with ethnic links to a neighbouring ethnic conflict see their chances of experiencing civil conflict increase by six percentage points.

With the publication of last year’s World Bank Development Report (2011) that focuses entirely on the role of conflict in the development process, the World Bank has put conflict at centre stage in the international economic development debate.

Civil wars are devastating to a country’s development prospects, not only in the short run by destroying lives, resources, and infrastructure, but also in the long run by affecting people’s trust in society or government, generating outflows of foreign direct investment and expertise, and creating dormant grievances that increase the likelihood of a future recurrence of the conflict.

Besides wreaking havoc in the countries directly affected by conflict, civil wars also often carry substantial negative effects on the neighbours of the war-torn country. These effects range from lost trading opportunities, an influx of refugees, or even a spread of the violence itself onto its own territory.

Abundant anecdotal evidence supports the idea that neighbouring conflict poses severe threats to a country’s own stability. Liberia’s, Sierra Leone’s and Guinea’s civil wars were heavily intertwined. Burundi’s civil war spread to Rwanda, the Congo, and Uganda. Following the breakup of Yugoslavia civil war swept across the Balkans, and Afghanistan’s civil war has now spread to Pakistan’s (North-) Western provinces.

Indeed, the figure below shows that outbreaks of civil war are mainly concentrated in parts of Africa, Asia, and Latin America, suggesting a possible role for contagion effects of neighbouring civil wars.

Figure 1. Number of civil war onsets, 1945–2000

Notes: A civil war is defined as an armed conflict between an internationally recognised state and (mainly) domestic challengers able to mount an organised military opposition to the state. A war must have caused more than 1,000 battle deaths in total and within at least a three-year period.

Although clearly compelling, the above figure plus the anecdotal evidence is not conclusive in establishing just how important these spillovers are. Are these anecdotes the exception rather than the rule? Rwanda’s civil war did not spill over into Tanzania, nor did Mozambique’s civil war spread into Malawi or South Africa. Maybe only particular types of conflict tend to spread? Or maybe the geographical clustering of civil wars is simply explained by a similar clustering of the underlying country-specific causes of conflict such as poverty, corruption, or natural-resource dependence?

In a recent study (Bosker and de Ree 2010), we consider all outbreaks of civil war since World War II and find that there is no evidence that all civil wars have the same tendency to spread across international boundaries. Taking explicit account of the country-specific factors associated with a higher risk of civil war1, we find that only ethnic civil wars pose a significant threat to neighbouring countries’ stability.

Also, not all countries are at the same risk of contagion. Ethnic civil wars only spread along ethnic lines. That is, only countries with ethnic links to a neighbouring ethnic conflict are at significant risk of contagion. It increases their chances on also experiencing civil conflict by six percentage points. Moreover, we find some (weaker) evidence that richer countries are better able to prevent a neighbouring conflict from spreading to their own territory.

Based on these findings we identify those regions in the world where ethnic conflict has the highest risk of spreading, ie regions where many ethnic groups are present in multiple countries. The figure below shows that sub-Saharan Africa is the region most prone to conflict spillovers (a legacy of the artificial borders drawn by European colonisers; see the recent Vox column by Michalopoulos and Papaioannou 2011). But Central America, the Middle East, Europe, and Central Asia are also prone to ‘ignite’ (and have ignited in the past) when ethnic violence erupts in only one of its countries.

Figure 2. Average number of ethnic links to a neighbouring country

Given the devastating, and often long-lasting, impact of civil wars on a country’s economic development prospects, understanding the causes and consequences of civil war is of vital importance in identifying precautionary measures that prevent conflict from breaking out, or from escalating further once the violence has started.

Our findings show that some of the important causes of civil war lie beyond the reach of individual countries themselves. Besides policies targeting the causes and consequences of civil wars within each country, taking preventive action to stop conflict, and ethnic conflict in particular, from spreading across international boundaries is important. If effective, it decreases the chances of one ethnic conflict dragging a whole region into a vicious cycle of conflict and underdevelopment.


Bosker, Maarten and Joppe de Ree (2010) “Ethnicity and the spread of civil war”, CEPR Discussion Paper 8055.

Blattman, Chris and Edward Miguel (2010) “Civil War”, Journal of Economic Literature 48: 3-57.

Michalopoulos, Stelios and Elias Papaioannou (2011), “The long-run effects of the Scramble for Africa”, VoxEU.org, 6 January.

World Bank (2011). World Development Report 2011: Conflict, Security and Development. Washington, DC: World Bank.

1 See Blattman and Miguel (2010) for a nice overview of the empirical literature on civil war that largely focuses on identifying country-specific factors influencing the chances of civil war breaking out.

4,829 Reads