Downloads: 1 available

Available in

Mapping conflict across Liberia and Sierra Leone

Download Article
Caitriona Dowd and Clionadh Raleigh describe patterns of violence in Liberia and Sierra Leone between 1997 and 2011 using analysis, trends and data from the Armed Conflict Location and Events Dataset (ACLED).

Using data derived from media reports, humanitarian agencies and research publications, Caitriona Dowd and Clionadh Raleigh use analysis, trends and data from the Armed Conflict Location and Events Dataset (ACLED) to describe patterns of violence in Liberia and Sierra Leone between 1997 and 2011. The patterns suggest that post-war violence has peaked around election periods and carries a strong cross-border dynamic.

Levels of violence have dropped in Liberia and Sierra Leone following the official end of the wars. But violence persists and, if left unaddressed, still poses a significant risk to peace.

Caitriona Dowd and Clionadh Raleigh



Mapping conflict

This article describes patterns of conflict and violence in Liberia and Sierra Leone from 1997 to 2011, a period that coincides with the onset of the second Liberian civil war (1999–2003) and the second phase of the Sierra Leonean civil war (1999–2002). 

Each state experienced an intensification of conflict events to devastating levels before the official end to hostilities led to de-escalation. But different forms of violent conflict persist post-war: Sierra Leone has continued to experience relatively high numbers of battles between armed combat-ants, while in Liberia there has been more violence against civilians.
Analysis, trends and data below draw on the Armed Conflict Location and Events Dataset (ACLED), which collected information on the geographic location of all reported conflict events between January 1997 and 2011 in 50 countries including Liberia and Sierra Leone. Data include the date and location of conflict events, the type of event, the perpetrators and victims of violence, and changes in territorial control over time. Data are derived from media reports from war zones, humanitarian agencies and research publications. There is no minimum number of fatalities per event for inclusion in the dataset.

Conflict across time: patterns of violence



Residual rebel and militia groups from the first civil war (1989–96) created a precarious security context in the years preceding the outbreak of the second civil war, which began when the rebel group Liberians United for Reconciliation and Democracy (LURD) launched attacks in Lofa County, north-west Liberia in 1999. In early 2002, then President Charles Taylor imposed a state of emergency as LURD fighters moved closer to the capital. Although government forces regained territory in 2002, the following year witnessed the opening of a second front as a new rebel group, the Movement for Democracy in Liberia (MODEL), emerged in the south-east. 

The Accra Peace Agreement in 2003 marked the end of the Taylor government and UN peacekeeping forces were able to secure the Liberian state relatively quickly: there was a dramatic drop in the recorded incidence of conflict events between 2003 and 2004 (see Graph 1). Several non-rebel groups were also active in Liberia’s conflict. Political supporters of the Unity Party and the Congress for Democratic Change participated in violent clashes throughout the war. Ethnic militia violence primarily involving Gio and Mano ethnic groups spiked in 2003, linked for reasons to do with their traditional support for Taylor’s government.
At the peak of the war in 2003, over 240 violent events were recorded across Liberia. Since the official cessation of hostilities, violence has continued at a much lower level. Post-settlement, Liberia has experienced an average of 19 violent conflict events per year (2004–10), generally following pro- and anti-Taylor stances.

Sierra Leone

Levels of violence were highest in 1998 due to continued activity by the rebel Revolutionary United Front (RUF), who were allied to the Armed Forces Revolutionary Council (AFRC). The signing of the Lomé Peace Agreement and the establishment of the UN Mission in Sierra Leone (UNAMSIL) in 1999 had a limited impact on violence. But the implementation of the peace agreement and international military support led to a sharp decrease in fighting in mid-2001. The civil war was officially declared over in January 2002.
Several political groups also engaged in violence through political militia activity throughout the recorded period, including the All People’s Congress, the Sierra Leone People’s Party and the People’s Movement for Democratic Change. Violence increased in 2002–03, and again in 2007–08, corresponding with national and local elections, respectively. Violence remained relatively high for peacetime through 2008 and 2009 (see Graph 2). Local council elections may have provided opportunities for political militia attacks on opponents.

Conflict across space: the geographic spread of violence

In Liberia, the majority of battles during the civil war took place in north-west Lofa County, concentrated in Kolahun district. Montserrado County also experienced very high levels of battles, with a predictable concentration in the capital, Monrovia. These areas also have distinct violence profiles. In Lofa, violent conflict primarily involved military forces fighting rebel groups over territory as well as fighting among armed groups. Such battles were more than twice as common in Lofa than in Montserrado, which experienced greater numbers of riots and protests (46 events overall compared to just two in Lofa). Battles in Montserrado mainly involved militant political party supporters. 

In Sierra Leone, battles tended to be concentrated in the northern territories, with 357 battles between armed combatants reported, compared to 204 in the east, the region with the second highest levels of violence between 1999 and 2011. The majority of violent changes in territorial control took place in the north. Some concentrations coincide with the presence of natural resources, but not all.
Violent events in Liberia spread geographically in the months before peace agreements. Parts of the country witnessed violence for the first time in the run-up to the Accra talks in June 1999. The proximity of peace talks may have provided incentives for territorial acquisition, either to strengthen conflict actors’ bargaining power at critical junctures or to seize resources before legitimating gains in the peace talks. 
In Sierra Leone, on the other hand, this pattern is not as evident in the run-up to the Lomé Peace Agreement in July 1999, although new contestations did take place in Songo in the west and Yele in the centre. The months before the ceasefire that led to the decision to implement Lomé in 2001 are also not characterised by a marked geographic expansion in violence. 

Cross-border violence

Cross-border conflict activity between Liberia and Sierra Leone peaked in 2001, coinciding with a sudden increase in violent events in Liberia, though violent incidents in Sierra Leone were declining year-on-year at this point (see Graph 3). 2008 saw another small increase in cross-border activity, although the overall level of regional conflict has dropped dramatically since the end of both civil wars.

Cross-border conflict actors between Liberia and Sierra Leone included LURD and RUF. The official military forces most active across borders were Guinean and Liberian. Many reports of cross-border activity identify mercenaries active in neighbouring states; the continuation of this phenomenon post-war – most recently with Liberian mercenaries participating in the Côte d’Ivoire conflict – poses a threat to regional stability.

Peace agreements and beyond

Despite escalations in fighting in the run-up to the talks, the 2003 peace agreements in Liberia brought an immediate reduction in levels of violence. In Sierra Leone, peace agreements had a less dramatic impact: reduced violence here is more closely linked to disarmament from mid-2001 (Map 1 illustrates how violence reduces spatially and temporally in response to peace agreements).

Violence against civilians constitutes around 30 per cent of post-war conflict events in both countries. This reflects wartime patterns; for example, in Liberia, LURD perpetrated a high proportion of violent acts against civilians – possibly the result of a deliberate strategy or of a lack of control over troops. In the post-settlement context, this form of violence, along with riots and protests, links closely with electoral cycles, suggesting that democratic contestation remains a potential trigger for violence, and that civilians will continue to bear the brunt.
Former combatants are remobilised around political contestations in patterns of organisation similar to wartime. The role of remobilised former combatants in electoral ‘task forces’ or ‘squads’ suggests that militias – affiliated explicitly or implicitly with parent political organisations – play a central role in mobilising voters and intimidating opponents through violence. 
Although violence in Sierra Leone was low relative to wartime, conflict event reports in 2002 indicate that increased violence coincided with elections and was concentrated in Freetown. Similarly, in Liberia during the election year of 2005, violent riots and protests constituted over one third of overall violent events, mainly in Monrovia.

Conflict data-mapping and peacebuilding

Tracking conflict patterns over time and space helps to see more clearly how violence mutates and develops: 

  • What types of groups are involved and what are their objectives?
  • Who are the victims?
  • What events or other stimuli cause violence to intensify, relocate or to diminish?
Levels of violence have dropped in Liberia and Sierra Leone following the official end of the wars. But violence persists and, if left unaddressed, still poses a significant risk to peace. Of primary importance is addressing the violence that surrounds electoral cycles at all levels. In Sierra Leone, for instance, high levels of violence against civilians in 2008 coincided with local council elections, which serve as an explosive arena for ‘decentralised’ violent contestation.
A second security threat relates to regional conflict activity. Although levels of cross-border violence have reduced dramatically since the end of the civil wars, several reports of cross-border activity have recorded the presence of regional mercenaries in neighbouring countries. Persistent pressures emanating from neighbouring states mean that both Sierra Leone and Liberia are vulnerable to external threats (although domestic and regional threats are not mutually exclusive), not just through violence, but also through the economic and institutional impact of neighbouring instability.
Patterns of domestic and regional violence have involved the remobilisation of former combatants, either in the military wings of established political parties or through involvement in cross-border mercenary activity. Post‑war disarmament and demobilisation programmes have contributed to a substantial reduction in overall levels of violence. But preventing future instability relies on implementing them fully – especially among specific target groups. There also needs to be much more emphasis on reintegrating former fighters back into society to reduce the likelihood that ex-combatants will be drawn back into fighting. An effective peacebuilding policy needs to pay attention to the whole process of disarmament, demobilisation – and not least reintegration for former combatants.

Box 3: Terminology


ACLED collects information on politically motivated violent and select non-violent events. Violent events include battles between political groups (eg state and rebel; political party militias), base establishment by a conflict actor, violence against civilians by a political actor, and riots. Battles involve combat between groups. Non-violent events include protests, territorial takeovers without combat, and noted arrests of political figures. Political actors are limited to state, rebel, militia, rioters and protest actors, and include civilians only when the victim of violence is from one of the aforementioned. Events are coded for each day, location, political actor and event type. An event therefore refers to a specific interaction occurring on a single date, in a noted location, between named political actors, and is further specified by the type of event. No crime information is collected. 
For more information, please see