East and Central Africa: A Legacy of Deadly Political Violence and the Risks of its Recurrence

Ted Robert Gurr
Distinguished University Professor Emeritus
University of Maryland

For purposes of this research note the East and Central African region is broadly defined as extending west-east from Chad to Somalia and north-south from Sudan to the Democratic Republic of Congo. For decades it has been a protracted conflict region characterized by violent political contention within and among states and between communal groups.  Insurgents, refugees, and munitions move easily across borders, armed conflict propagates with them. Governments, in response to pervasive insecurity, invest disproportionately in armies and militias and are quick to intervene in neighboring states.  Socioeconomic investment suffers, communal tensions and interethnic rivalries flourish.  These are common characteristics of protracted conflict zones elsewhere, not just in East and Central Africa.[1] The people of the region's 13 countries have been more victimized by political violence in the past half-century than those any other world conflict region except Southeast Asia. 

The human costs in this region have been very high.  Fourteen of the world's 42 episodes of genocide and politicide between 1955 and 2005 occurred in East and Central Africa.[2]  Mass killings, a concept that includes but is broader than genocide, are equally common. A recent study commissioned by the US Government's Political Instability Task Force identified all episodes of mass killing globally from 1945 to 2006, defining such events as those in which "the actions of state agents result in the intentional death of at least 1,000 noncombatants from a discrete group in a period of sustained violence."[emphasis added]  By this less restrictive definition - no "intention to destroy" a predefined collectivity is assumed - between 1955 and 2006 there were 95 mass killings of which one in four - 22 - were perpetrated by authorities, or groups claiming authority, in these 13 states. Some occurred when civilians were targeted by government forces during rebellions and civil wars, others were the result of sustained campaigns of repression and genocide. Djibouti, Kenya, and Tanzania are the only countries in the region in which no such killings took place. [3]

The roster, at the end of this note, gives dates and low and high estimates of noncombatant fatalities.  Of course such estimates are imprecise, as are distinctions between combatants and noncombatants.  Some episodes of mass killing may have been overlooked, others overcounted.  Such concerns ought not deflect attention from the fact that between 4 and 8 million ordinary people died in this region in the last half century, targeted by military, security, and militia forces - and continue to die. 

We have two bases for assessing risks of future genocides and mass killings in East and Central Africa.  First is Barbara Harff's risk list, last updated in 2009, that flags countries at high risk of genocidal violence in the near future.[4]  Seven had three or more of six genocide risk factors, as shown below:

Sudan had six risk factors that portend future episodes:  past geno/politicides, groups subject to state-led discrimination, a polarized elite, exclusionary ideology, autocratic governance, and very low trade openness.  The most likely precursor of renewed geno/politicide? A breakdown of the North-South peace process and civil war targeting Southerners.

Rwanda has four risk factors: past genocides, a Tutsi-dominated elite, autocratic governance, and low trade openness.  But the government strongly favors inter-ethnic reconciliation and prospects for a return to interethnic violence are small.

Burundi has three risk factors: past genocides, a Tutsi-dominated elite, and low trade openness.  But, as in Rwanda, the government is committed to political integration of Hutus; and again like Rwanda, there is a high degree of international political and developmental engagement that offsets limited economic interactions.  Unreconstructed Hutu rebels remain an external threat.

Somalia had three risk factors: a past geno/politicide, exclusionary ideology (among the Islamists), and very low trade openness.  The most likely precursor of genocide? A takeover by Islamist rebels who already control much of the country.

Ethiopia has three risk factors: a past genocide, an elite dominated by the Tigrean minority, and relatively low trade openness.  The most likely precursor to geno/politicide? An armed challenge by one or more marginalized communal groups that triggers a politicidal response.

The Democratic Republic of Congo and Uganda also have several of the above risk factors but there is not clear and likely path toward new geno/politicides in either country - repression yes, in response to the predations of rebels and dissidents, but not attempts to eradicate threatening groups as such

Another approach to assessing risks of future atrocities in this region is to identify risks of future political instability. All 14 genocides and politicides in East and Central Africa occurred in response to rebellions or separatist political movements that challenged state authority.  Mass killing also was a strategy used by weak post-colonial regimes to suppress political opposition. It may not be possible to forecast risks of mass killings or repression, but we are able to say what the chances are that any given state will experience their common precondition, violent instability, in the near future.   Five risk variables go into a recent global assessment by J. Joseph Hewitt: [5]

  • Major instability events in the recent past (analogous to the role that past genocides play in Barbara Harff's genocide risk analysis)
  • High infant mortality (signifying widespread poverty and lack of social services)
  • High levels of militarization (indicates a diversion of scarce resources and often a readiness to use coercion against internal opposition - a common characteristic of countries in protracted conflict regions)
  • Low levels of economic integration into the global economy, signifying both poverty and a lack of external economic influence that might mitigate political conflict
  • Lack of regional security, with one or more neighboring countries involved in armed conflict, domestic or international

These conditions have been shown by global empirical research to be precursors of instability in the recent past.  When countries in East and Central African are analyzed in this framework, 33 countries world-wide have very high short-term risks of instability - and of these, nine are in East and Central Africa.  In descending order of risk in 2010 they are


            Democratic Republic of Congo








Five of these nine states, shown in bold, also had the highest increases of any countries in the world in risks of instability between 2005 and 2008.

The conflict risks are compounded by the fragility of most states in this region.  An index of state fragility has been developed by Monty G. Marshall and applied globally to information on 14 aspects of state's capacity to deal with political challenges, maintain legitimacy, and deliver economic and social goods to their citizens. Six of the 20 most fragile states in the world are in East and Central Africa: they are Somalia, Sudan, the Democratic Republic of Congo, Rwanda, Burundi, and Uganda.  The only country in the region that is below the global mid-point on the fragility index is Tanzania. [6]

In summary, three different comparative analyses highlight East and Central Africa as a region with a volatile combination of high potential for violent instability, weak regimes, and mass violence. None of the three analyses say when instability and genocidal killings might begin. There are stabilizing political factors and external support in some of these countries that may insulate them against exposure to risk, particularly in Tanzania and Kenya, and also - as noted above - in Rwanda and Burundi. The global comparative results nonetheless highlight once again that the peoples and governments of this region are at great risk of future political and humanitarian disasters and are in equally great need of support from regional and international organizations to help state authorities to recognize and counteract those malign conditions.


Mass Killings of Noncombatants by State Agents in East and Central Africa 1955- 2005

from Ulfelder and Valentino, note 3



Estimated Deaths



Sudan: First north-south civil war




Congo: Kasai rebellion




Ethiopia-Eritrea civil war




Rwanda: ethnic killings




Zanzibar: political repression




Congo: CNL rebellion in eastern provinces




Uganda: Idi Amin's repression/genocide




Ethiopia: Derge repression, Tigre civil war




Ethiopia: Ogaden rebellion




Uganda: civil war




Somalia: Barre repression of SNM/Issaqs




Chad: political repression and civil war

   (Habre regime)




Sudan: Second north-south civil war




Uganda: rebellions by LRA, others

1986 - present



Burundi: ethnic killings/genocide




Rwanda: ethnic killings/genocide




Chad: political repression and civil war

  (Deby regime)




DRCongo: Kabila/Tutsi rebellion




Rwanda: ethnic killings post-genocide




*DR Congo: local uprisings/civil war in east




*Sudan: Darfur




*fatalities from different sources than those used by Ulfelder and Valentino

Mass killings in bold coincide approximately with episodes of genocides and politicides on Barbara Harff's roster. Dates and estimates of fatalities differ somewhat because the two studies use different definitions.  For example the episode of mass killings in Burundi from 1988-2005, above, is analyzed by Harff as two distinct episodes of geno/politicide, one in 1988 and a second in 1993. The mass killings list also omits her 1965-73 case of geno/politicide in Burundi.


[1] The concept of protracted conflict regions and evidence about their characteristics are developed by Monty G. Marshall, Third World War: System, Process, and Conflict Dynamics (Lanham, MD: Rowman & Littlefield, 1999), esp. chaps. 4 and 5.

[2] Barbara Harff, "Assessing Risks of Genocide and Politicide," in Monty G. Marshall and T. R. Gurr, Peace and Conflict 2005 (College Park: Center of International Development and Conflict Management, University of Maryland, 2005), pp. 57-58, updated to include episodes two historical cases she has subsequently added to her roster, in Nigeria-Biafra, and Zimbabwe.

[3] The study and data are reported by Jay Ulfelder and Benjamin Valentino, "Assessing the Risks of State Sponsored Mass Killings," and was released by the Political Instability Task Force in February 2008.  The full set of events is listed in its' Appendix, pp. i-vi.  The report is/will be posted on the http://GPANet.org website under Resources.

[4] This is a summary of a more nuanced analysis posted at GPANet.org under Risks.

[5] J. Joseph Hewitt, "The Peace and Conflict Instability Ledger: Ranking States on Future Risks,' in J. Joseph Hewitt, Jonathan Wilkenfeld, and T. R Gurr, Peace and Conflict 2010 (Boulder, CO: Paradigm Publishers, 2009), pp. 7-26. The global ranking of states is on pp. 21-24 with technical notes.

[6] Monty G. Marshall and Benjamin R. Cole, Global Report 2009: Conflict, Governance, and State Fragility.  Center for Systemic Peace and George Mason University's Center for Global Policy, 2009, www.systemicpeace.org.



GurrEAfricanLegacyofPoliticalViolence.810.docx25.3 KB