Armed Conflicts, Violence, Terrorism, Sub-Saharan Africa


Why victimize civilians during civil war? Inspired by my survival of two civil wars in Liberia, I offer a theory of violence against noncombatants during armed conflicts in sub-Saharan Africa. Civil wars in post-colonial sub-Saharan Africa have a typical pattern that can be distinguished with three distinct stages: the "onset" period at the start of conflicts; the "intervention" period, marked by third-party involvement and negotiations; and the "settlement" stage, marked by an interim government with cabinet positions rationed among belligerents. I argue that the causes and motives for violence against civilians vary across these three stages of civil war. In the first "onset" stage, violence against civilians is mainly perpetuated by foreign recruits who lack ties with local populations. Principle-agent and moral hazard problems emerge as foreign recruits, young of age and disconnected from local populations, want to loot and rape civilian populations. In contrast, commanders responsible for these populations have difficulty monitoring and controlling these recruits. The second stage of "intervention" changes the calculus for combating groups, as the bargaining power of each depends on its control of strategic territory, such as ports and airports. Groups with minor territories and weak capacities are incentivized to attack the soft targets of noncombatants, who are sometimes forced to take sides at this stage. At the third "settlement" stage, the distribution of cabinet positions creates rivalries within groups, with disgruntled factions breaking away and resuming conflict. Civilians are caught in the middle, frequently accused of collaborating with the other side, leading to violence against noncombatants, including hostage-taking, as part of the bargaining process.

To examine the stage theory of civilian victimization in sub-Saharan Africa's civil wars, I surveyed Liberia's two civil wars that occurred from 1989 to 1996 and 1999 to 2003. I drew on three sources of data to triangulate the causes and motives of violence against civilians: (a) archives, (b) intensive interviews of combatants and civilian victims of these wars, and (c) witnesses' testimonies from the Liberia Truth and Reconciliation Commission (TRC) hearings. These data are used to qualitatively study Liberia's two civil wars, which may be generalizable to other parts of sub-Saharan Africa. It is found that foreign recruits, strategic territories, and political positions in a power-sharing government may have contributed to violence against civilians during the onset, intervention, and settlement phases of Liberia's two civil wars.

The research contributes to the general literature regarding violence against civilians during armed conflicts. According to popular opinion, violence against noncombatants is a natural outcome, but this research points out that the causes and motives of violence against civilians can vary in identifiable ways. It shows how civil wars in post-colonial sub-Saharan Africa have a unique pattern associated with foreign intervention and settlements, and these phases may account for varying motives for violence against civilians. Knowing the motivating factors for violence against civilians can help protect vulnerable populations during armed conflicts, including refugees, internally displaced persons, migrants, and victims of human trafficking. Intervening governments can make more significant efforts to protect civilian populations during the intervention phases and be more sensitive to the dangers sub-group actors pose in distributing government positions during the settlement phases of these wars.

Completion Date




Committee Chair

Mousseau, Michael


Doctor of Philosophy (Ph.D.)


College of Sciences


School of Politics, Security, and International Affairs

Degree Program

Security Studies








Release Date

December 2026

Length of Campus-only Access

3 years

Access Status

Doctoral Dissertation (Campus-only Access)

Campus Location

Orlando (Main) Campus

Restricted to the UCF community until December 2026; it will then be open access.