what the fighters say - ropv~what_the...what the fighters say: a survey of ex-combatants in liberia...

What the Fighters Say: A Survey of Ex-combatants in Liberiai

February-March 2006

James Pugelii

A UNDP-funded empirical study conducted in partnership with:

African Network for the Prevention and Protection against Child Abuse and Neglect (ANPPCAN)

April 2007

i The UNDP is most appreciative of the assistance rendered by Dr. Macartan Humphreys to this

project and for the subsequent permission provided by him and colleague Dr. Jeremy Weinstein to allow publication of this project’s findings with a naming convention and format similar to those of their study from Sierra Leone (August 2004) —“What the Fighters Say: A Survey of Ex-combatants in Sierra Leone” — download at http://www.earthinstitute.columbia.edu/cgsd/documents/humphreys_combatantsurvey.pdf.

ii This project would not have been possible with the foresight, personal involvement, and

sponsorship of Mr. Charles Achodo, UNDP Liberia and the stalwart support of the head of the Monitoring and Evaluation Unit – Mr. Shecku Silla. The author is most grateful for the tireless efforts of Eleonora Koeb, who not only established and organized the project and logistics but spent countless hours assisting in the refinement of the study’s design. James Pugel is a research associate with the Centre for the Study of Civil War (CSCW), International Peace Research Institute (PRIO), Oslo, Norway (www.prio.no). The research for this study was done while the author was pursuing a graduate degree from the National Defense Intelligence College, Washington, DC. The views expressed in this report are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government.

JOINT IMPLEMENTATION UNIT (JIU)

ABSTRACT

Disarmament, Demobilization, and Reintegration (DDR) programs have been studied significantly

at the macro-level but little has been done to validate these findings at the micro-level. This paper

complements a study that was conducted in Sierra Leone, the only other countrywide micro-level

investigation of its kind, by introducing an examination of a new data set from ex-combatants in

Liberia. A randomized sample, drawn from across Liberia in February and March 2006, of 590

adult former fighters submitted to an interview that serves as the tool for measuring reintegration

and reconciliation in this study. The interviews collected information on socio-economic

demographics, ex-combatant geographic locations during the 14-year civil war and their actions

while serving with their former factions, respondent knowledge of and participation in international

intervention programs, and captured social integration and political expressions. The initial

findings demonstrate a high degree of convergent validity with the Sierra Leone study and

illuminate both strengths and weakness in the current intervention program in Liberia. The initial

findings also establish a foundation of evidence that can be used to assess the interim impact of

DDR program participation as well as to aid the direction of the ongoing intervention efforts.

Significantly, empirical evidence supports the finding that those former combatants who

registered with the national DDR program and completed a course of reintegration training have

reintegrated more successfully than those ex-combatants who chose not to participate and

reintegrate on their own.

What the Fighter’s Say: A Survey of Ex-combatants in Liberia February-March 2006 3

EXECUTIVE SUMMARY

Micro-level combatant research is a growing area of interest with regard to organizational

structure, management, and recruitment as well as for exploration into the post-conflict realities of

Disarmament, Demobilization, Rehabilitation, and Reintegration (DDRR) and their program and

policy impacts. Groundbreaking field research conducted by Macartan Humphreys and Jeremy

Weinstein in 2003 resulted in a report of initial findings entitled “What the Fighters Say: A Survey

of Ex-combatants in Sierra Leone.” The data set has subsequently yielded multiple, very

important empirically based analytical papers on DDR and combatant recruitment. Through a

close association with the investigators of the Sierra Leone case study, a replication of the

methodology employed for examining micro-level impact on the macro-level outcomes was

heavily relied upon for this study in Liberia. The research covered in this report builds upon the

approach advanced by Humphreys and Weinstein, but also takes a number of important steps

that go beyond the original design.

This report endeavors to meet three broad objectives:

• Provide the initial findings of a survey of a representative sample of 590 ex-combatants from Liberia’s civil war in a manner that allows policy makers and researchers to better understand the nature of the conflict and the subsequent intricacies of the reintegration process in post-conflict Liberia

• Present findings that identify and illuminate areas of convergence and divergence with the initial findings as reported in a similar large-scale ex-combatant study in Sierra Leone3

• Provide the foundation for follow-on empirical investigations and validations that offer alternative perspectives and constructs for consideration

Against the backdrop of a nation emerging from 14 years of civil war and a peaceful round of

national democratic elections, the following major findings4 are recognized two and a half years

removed from the signing of the Comprehensive Peace Agreement (CPA) on 18 August 2003:


Intervention Program Perspectives

The study segregated the respondents into two groups – DDRR program participants and non-

participants. The program participants were further divided into their respective categories of

stage of intervention – Disarmed and Demobilized (DD)-only participants (those yet to receive

reintegration training); participants who had enrolled but had not completed reintegration training,

and those participants who had completed reintegration training (“graduates”). Those

respondents who had received program (intervention) benefits were drawn from the same

communities as the group of ex-combatants who had not received benefits.

• Registered Ex-combatants Yet to Receive Training Benefits. There is a major risk of leaving behind a very vulnerable grouping of NCDDRR registered ex-combatants – those who have disarmed and demobilized but have yet to receive training. This category of former fighters is the least educated, most agriculturally oriented, and poorest of the four classes under investigation. Most important, they have been shown to be the least reintegrated of all categories under investigation.

• Ex-combatants Currently Enrolled in a Course of Reintegration Training. The situation of this category of ex-combatants is only slightly better than the situation of those who chose not to participate in the DDRR. These combatants are challenged by their economic situation (42% live on less than US$2/day) and their physical separation from their home communities (52%). The reason for the separation was unclear in the study but may be related to limitations regarding training program availability.

• Ex-combatants Who Have Completed a Course of Reintegration Training. While this group was found to be the most capable of reintegrating back into their respective communities, the international community must not forget them. Many of the former fighters are moving on economically, but some are still having a hard time reintegrating socially within their respective communities. The low self-reported acceptance rates, lack of faith in the communities’ judges of conflict resolution, and large proportion of these former combatants spending their free time on their own indicate that much more progress is needed.

• Non-NCDDRR-Registered Ex-combatants (Program Non-Participants). While this group appears to be reintegrating to some degree, there are a couple of significant areas of concern. Members of this group are the least likely to resettle their home communities and have also indicated an overwhelming concern about being identified as ex-combatants. The survey found a DDRR non-participation rate of 12%, which was nearly identical to the finding in the Sierra Leone case (13%). The significance of the convergence between the two country cases is quite important to policy makers in itself as it may now be able to be utilized as a planning factor for future cases. Non-participation is really a non-measurable action, but convergence should now enable a bit more confidence in the planning stages.


Post-conflict Reintegration

• Acceptance (self-reported). 94% of the sample of ex-combatants reported that they did not have any problems gaining acceptance from their neighbors today. The percentage of DDRR program completers claiming major problems (7% of sub-sample) more than doubles when compared against any other category.

• Acceptance (perceived). When the sample of ex-combatants was asked to indicate how they believed the community viewed the population of ex-combatants, only 66% believed that acceptance was the norm. While program completers perceive acceptance from the community at a higher rate than the non-participants (62% vs. 44%), they also are doubly disposed to perceive that the community lives in fear of the ex-combatants (20% vs. 10%).

• Social Networks. 63% report spending the majority of their free time with family. Reliance upon family and the primacy of their position in the lives of the sample of former fighters are quite apparent and significant in achieving an overall better socio-economic status. Interesting, however, is the finding that DDRR program completers spend their free time alone at a much higher rate (22%) than any other category of ex-combatants. This finding is troubling as there is a highly significant negative relationship between being alone and reintegrating successfully that was also uncovered.

• Conflict Resolution. Over 96% of the former fighters indicated that their home community’s mechanism for conflict resolution was either “mostly fair” or “almost always fair.” Among the ex-combatants who were dissatisfied with a perceived impartiality the DDRR program completers were clearly the predominant group in both raw numbers and sample proportions.

• Judges of Conflict Resolution. 88% of all conflict was reportedly judged by traditionally based mechanisms (family, elders, and chiefs). There was a definite distinction between what was reported in the Greater Monrovia area versus the rest of Liberia, however. With regard to the traditionally based mechanisms, the chiefs were looked to 68% of the time to settle disputes outside Monrovia versus only 2% in the metropolitan area. The elders appear to accept the role in Greater Monrovia at a rate of 52%. Utilization of the courts system for both locality types stands at an almost nonexistent rate of less than 2%.

• Home Community Resettlement. 58% of the ex-combatants in the sample group have resettled in their home communities. This high rate of home community resettlement by all categories of DDRR program participants is a good indication of a positive progression towards reconciliation. The tendency of the non-participants, in stark contrast, to shy away from their pre-war home communities (56%) may be indicative of many confounding factors, but the worst among them may be the result of abusive or immoral actions committed during the conflict that keeps them away. But the most serious factors that keep them away may be the abusive or immoral actions committed during the conflict. Formal identification as an ex-combatant was the reason most often cited by non-participants (47%) for declining entry into the DDRR program. Whether the decision of the non-participants was due to a general social stigma or the result of individual conduct cannot be explained here.


• Employment Status. 25% of the sample of ex-combatants reported that they were unemployed. The newly introduced construct for determining employment status appears to have provided the variation desired for analysis. In a country like Liberia, where the nationally reported unemployment rate is around 80 percent, researchers and policy makers require a construct that provides a more precise definition. The findings show a strong entry into the formal employment sector by the DDRR program completers as compared with the non-participants. The construct was also useful in multivariate analysis as it reinforced the importance of income-earning employment in post-conflict reintegration and amplified the troubles that face the sick and disabled in Liberian society. Respondents who were categorized as sick/disabled were almost 20% worse off than those former combatants who were unemployed.

• Poverty Reduction. 82% of the respondents live on less than US$5/day. DDRR program graduates indicate a better ability to stay above the US$1/day poverty line (93% versus 85% for the non-participants). Alarmingly, 54% of the sample of respondents who have registered for DDRR benefits but have not received training live on less than US$2/day.

• Breaking Ties with Former Faction Leaders. On average, more than 75% of the sample population reported that they never go to their former faction leaders for financial aid or assistance. Troubling, however, is the finding that 22% of the DDRR program completers sustain routine ties whereas an average of only 14% of the sample’s population does so.

• Arms Presence. Only 2.2% of the respondents believed that small arms were a problem in their respective communities today with the conclusion of the Disarmament and Demobilization phase. The strength of this finding should not be understated as it has been corroborated with a finding from a parallel survey of the country’s non-combatants taken from the same communities which indicated that only 2.5% believed that small arms were a problem.

Impact of the DDRR Program

• Ex-combatants who formally disarmed, registered, and completed a program of internationally funded reintegration training show, on average, an overall 8% increase in their socio-economic situation when compared against those former fighters who never formally disarmed or registered with the National Commission on DDRR (NCDDRR).

• While the results support a markedly positive impact of the DDRR program upon the lives of those who complete the program, registering alone does not necessarily correlate with an improved overall socio-economic situation. In fact, those former fighters who have registered but have yet to receive formal training benefits are the least successfully reintegrated class, faring, on average, more than 9 percentage points worse than those who went on to receive training and complete the program. This class of former fighters is the least reintegrated of all categories.


Gender Perspectives

• While no evidence of problems associated with female fighter reintegration was uncovered in this study, many differences between genders did become apparent when male and female responses were disaggregated.

• Female combatants were found to serve in logistical roles more often than their male counterparts and were also less likely to be injured while participating as an active member. Interestingly, females reported a rate of “abduction” into factions that was also less than that for their male comrades (12% versus 21%). Female combatants also showed an experience rate for participating in cross-border conflicts that was commensurate with that of males.

• Notably, males appear to be progressing through the DDRR process ahead of the females. Indicators of this include a DDRR training completion rate of 10% (female) versus 18% (male) and a percentage of those registrants that are enrolled in a reintegration training course at 42% (female) versus 51% (male).

• Strikingly, females reported not having ever attended formal education at a rate twice that of the males. Additionally, only 28% of the females in the sample had an education above the elementary school level in contrast with the males who reported achieving this same level at a rate 57%.

• Few differences surface with respect to gender in the examination of conflict and its mitigation in Liberian communities. A disaggregation by gender of the frequency of sources of conflict yielded little variation. This possibly indicates that a generally common perception of the sources and frequency of conflict is present across gender.


CONTENTS

ABSTRACT ...........................................................................................................................................2

EXECUTIVE SUMMARY......................................................................................................................3

LIST OF FIGURES..............................................................................................................................10

LIST OF TABLES ...............................................................................................................................11

ACKNOWLEDGMENT.......................................................................................................................12

ACKNOWLEDGMENT.......................................................................................................................12

1 BACKGROUND............................................................................................................................13

2 PURPOSE AND METHODOLOGY............................................................................................16

2.1 PURPOSE OF THE STUDY.............................................................................................................16 2.2 THE APPROACH – RESEARCH DESIGN........................................................................................17 2.3 ACCEPTED PROTOCOLS AND VALIDITY.......................................................................................17 2.4 SAMPLING STRATEGY..................................................................................................................18

2.4.1 DEFINING THE POPULATION.........................................................................................18 2.4.2 NATIONWIDE RANDOM SAMPLING ...............................................................................20 2.4.3 SAMPLING FOR THE LIBERIAN INTERIOR ......................................................................21 2.4.4 SAMPLING FOR GREATER MONROVIA..........................................................................21 2.4.5 OBSERVED CLUSTER VARIATIONS...............................................................................22 2.4.6 POTENTIAL SOURCES OF BIAS.....................................................................................22

2.5 ENUMERATION STRATEGY...........................................................................................................23 2.6 CAPTURING VARIATION IN THE COURSE OF THE CONFLICT ........................................................23 2.7 THE RESEARCH TEAM .................................................................................................................24


3 FINDINGS......................................................................................................................................26

3.1 EX-COMBATANT PROFILES..........................................................................................................26 3.1.1 FORMER FACTION REPRESENTATION..........................................................................26 3.1.2 BASIC DEMOGRAPHICS................................................................................................27 3.1.3 GENDER DIFFERENCES ...............................................................................................31 3.1.4 KEY PATTERNS............................................................................................................32

3.2 THE DYNAMICS OF THE WARRING FACTIONS..............................................................................32 3.2.1 SOCIAL AND POLITICAL ORIGINS..................................................................................32 3.2.2 RECRUITMENT AND INCENTIVES ..................................................................................34 3.2.3 INTERACTION WITH CIVILIANS ......................................................................................37 3.2.4 INTERSTATE INFLUENCE AND COMBATANT ACTIVITIES ...............................................39 3.2.5 GENDER DIFFERENCES ...............................................................................................40 3.2.6 KEY PATTERNS............................................................................................................42

3.3 INTERVENTION PROGRAM PERSPECTIVES...................................................................................43 3.3.1 DDRR PARTICIPATION RATES.....................................................................................44 3.3.2 DDRR TRAINING PROGRAM PERSPECTIVES...............................................................45 3.3.3 GENDER DIFFERENCES ...............................................................................................46 3.3.4 KEY PATTERNS............................................................................................................47

3.4 POST-CONFLICT REINTEGRATION DETERMINANTS.....................................................................48 3.4.1 SOCIAL REINTEGRATION..............................................................................................48 3.4.2 ECONOMIC REINTEGRATION ........................................................................................52 3.4.3 POLITICAL EXPRESSIONS AND REINTEGRATION ..........................................................56 3.4.4 PSYCHOLOGICAL REINTEGRATION ..............................................................................58 3.4.5 GENDER DIFFERENCES ...............................................................................................58 3.4.6 KEY PATTERNS............................................................................................................60

3.5 CONFLICT AND CONFLICT RESOLUTION......................................................................................61 3.5.1 SOURCES OF CONFLICT...............................................................................................61 3.5.2 MECHANISMS FOR CONFLICT RESOLUTION.................................................................62 3.5.3 GENDER DIFFERENCES ...............................................................................................63 3.5.4 KEY PATTERNS............................................................................................................64

3.6 THE IMPACT OF THE DDRR PROGRAM .......................................................................................64 3.6.1 SOCIO-ECONOMIC REINTEGRATION ............................................................................64 3.6.2 ASSESSMENT BY INTERVENTION CATEGORY...............................................................65 3.6.3 GENDER DIFFERENCES ...............................................................................................68

4 CONCLUSIONS............................................................................................................................69

5 ACRONYMS .................................................................................................................................70


List of Figures

FIGURE 1: DISTRIBUTION OF THE SAMPLE BY AGE..................................................................................................20 FIGURE 2: ETHNIC COMPOSITION, BY MAJOR FORMER FACTION ............................................................................28 FIGURE 3: EX-COMBATANT RELIGIOUS AFFILIATIONS .............................................................................................28 FIGURE 4: PRE-WAR (1989) POLITICAL ALLEGIANCES...........................................................................................29 FIGURE 5: FORMER FIGHTER PRE-WAR COMMUNITIES..........................................................................................29 FIGURE 6: MILITARY RANK HELD IN FACTION .........................................................................................................30 FIGURE 7: ROLES OF COMBATANTS WITHIN FACTIONS ...........................................................................................30 FIGURE 8: NUMBER OF TIMES WOUNDED...............................................................................................................31 FIGURE 9: GENDER – ROLES (WITHIN FORMER FACTIONS) FIGURE 10: GENDER - TIMES WOUNDED .............31 FIGURE 11: MEMBERS OF FACTIONS WHO WERE PREVIOUSLY DISPLACED ...........................................................33 FIGURE 12: PRE-WAR FAMILY STRUCTURES .........................................................................................................33 FIGURE 13: HOW COMBATANTS WERE INTRODUCED TO THEIR FORMER FACTIONS...............................................34 FIGURE 14: TRIBES NOT TRUSTED, BY FACTION ....................................................................................................34 FIGURE 15: YEAR OF RECRUITMENT TO FACTION...................................................................................................35 FIGURE 16: REASONS FOR JOINING .......................................................................................................................36 FIGURE 17: INCENTIVES FOR PARTICIPATION IN FACTIONS .....................................................................................36 FIGURE 18: DIRECTION ON TREATMENT OF CIVILIANS ON THE BATTLEFIELD...........................................................37 FIGURE 19: FORMER FACTIONAL EMPLOYMENT AND IMPLEMENTATION OF CIVILIAN FORCED LABOR......................38 FIGURE 20: NON-COMBATANT FREEDOM OF CHOICE IN SUPPORTING FORMER FACTIONS.....................................38 FIGURE 21: CIVILIAN RAMIFICATIONS OF REFUSING TO SUBMIT TO FORCED LABOR DURING CONFLICT..................39 FIGURE 22: EX-COMBATANTS WITH FOREIGN FIGHTER EXPERIENCE......................................................................39 FIGURE 23: FOREIGN FIGHTERS IN THE RANKS OF LIBERIAN FACTIONS..................................................................40 FIGURE 24: GENDER – REASONS FOR JOINING (FOR FORMER FACTIONS) .............................................................41 FIGURE 25: GENDER – FOREIGN FIGHTER EXPERIENCE ........................................................................................41 FIGURE 26: GENDER – FORMERLY DISPLACED BEFORE JOINING............................................................................41 FIGURE 27: DDRR PROGRAM PARTICIPATION.......................................................................................................45 FIGURE 28: REASONS FOR PROGRAM NON-PARTICIPATION, BY FORMER FACTION.................................................45 FIGURE 29: PARTICIPATION IN TYPES OF REINTEGRATION INTERVENTION PROGRAMS, BY MAJOR FACTION...........46 FIGURE 30: GENDER – PARTICIPATION RATES (DDRR).........................................................................................47 FIGURE 31: GENDER – REASONS FOR NON-PARTICIPATION (DDRR) ....................................................................47 FIGURE 32: COMMUNITY ACCEPTANCE OF EX-COMBATANTS, BY CATEGORY OF PROGRAM PARTICIPANT ..............49 FIGURE 33: PERCEIVED COMMUNITY ACCEPTANCE LEVEL, BY CATEGORY OF PROGRAM PARTICIPANT .................50 FIGURE 34: HOME COMMUNITY RESETTLEMENT, BY MAJOR FACTION....................................................................51 FIGURE 35: HOME COMMUNITY RESETTLEMENT, BY CATEGORY OF PROGRAM PARTICIPANT.................................51 FIGURE 36: HOME COMMUNITY RESETTLEMENT, BY GEOGRAPHY .........................................................................51 FIGURE 37: SOCIAL NETWORKS, BY CATEGORY OF PROGRAM PARTICIPANT .........................................................52 FIGURE 38: COMMUNITY PARTICIPATION, BY CATEGORY OF PROGRAM PARTICIPANT ............................................52 FIGURE 39: WAGE AND OCCUPATION OF THE SAMPLE POPULATION OF EX-COMBATANTS......................................53 FIGURE 40: EMPLOYMENT STATUS, BY CATEGORY OF PROGRAM PARTICIPANT AND LOCALITY..............................54 FIGURE 41: DAILY INCOME LEVEL, BY CATEGORY OF PROGRAM PARTICIPANT.......................................................55 FIGURE 42: HOME OWNERSHIP, BY CATEGORY OF PROGRAM PARTICIPANT ..........................................................55 FIGURE 43: EDUCATIONAL LEVELS, BY CATEGORY OF PARTICIPATION ...................................................................56 FIGURE 44: FORMER COMBATANT POLITICAL PARTY AFFILIATIONS (FEBRUARY/MARCH 2006)..............................57 FIGURE 45: BREAKING TIES WITH FACTION LEADERS.............................................................................................57 FIGURE 46: GENDER - SOCIAL (ACCEPTANCE).......................................................................................................59 FIGURE 47: GENDER - ECONOMIC (EDUCATION) ....................................................................................................59 FIGURE 48: GENDER - POLITICAL (BREAKING TIES)................................................................................................60 FIGURE 49: SOURCES OF CONFLICT IN THE COMMUNITY........................................................................................61 FIGURE 50: JUDGES FOR COMMUNITY CONFLICT RESOLUTION, BY CATEGORY OF PROGRAM PARTICIPANT...........62 FIGURE 51: IMPARTIALITY OF JUDGES OF CONFLICT RESOLUTION, BY CATEGORY OF PROGRAM PARTICIPANT ......63 FIGURE 52: GENDER – PERCEPTION OF IMPARTIALITY ...........................................................................................63 FIGURE 53: OVERALL SITUATION COMPARED WITH ONE YEAR AGO, BY CATEGORY OF PROGRAM PARTICIPANT...65 FIGURE 54: GENDER – OVERALL SOCIO-ECONOMIC SITUATION.............................................................................68 FIGURE 55: MAP OF LIBERIA ..................................................................................................................................71


List of Tables

TABLE 1: SAMPLING FRAMEWORK AND DISTRIBUTION OF EX-COMBATANTS OVER 18 YEARS OLD..........................19 TABLE 2: SAMPLING FRAMEWORK AND DISTRIBUTION OF FEMALE EX-COMBATANTS OVER 18 YEARS OLD ............19 TABLE 3: NATIONWIDE CLUSTER REPRESENTATION................................................................................................21 TABLE 4: GREATER MONROVIA CLUSTERS REPRESENTATION................................................................................22 TABLE 5: RANDOM SAMPLE OF FORMER FACTION REPRESENTATION (1989-2003) ...............................................27 TABLE 6: REINTEGRATION OUTCOMES BY PARTICIPATION CATEGORY (MEAN SCORE COMPARISON MATRIX)........66 TABLE 7: IMPACT OF DDR PROGRAM PARTICIPATION (COMPARED AGAINST PROGRAM NON-PARTICIPANTS)..........67


ACKNOWLEDGMENT

The UNDP sponsored this countrywide evaluation for the period December 2005 to February

2007. Extensive planning and coordination was done with many mission partners, and collection

took place during February-March 2006. The author is thankful for the support of Mr. Charles

Achodo, UNDP DDRR Program and Policy Advisor, and the JIU’s Monitoring and Evaluation Unit

headed by Mr. Shecku Silla. Thanks are also in order for the tremendous work done by the two

UNDP M&E enumerator team supervisors, John Newmah and Bernard Mijoni. The data entry

interface for the project was programmed superbly by Mr. Raj Karikol, and the DDRR data entry

clerks of the JIU tirelessly typed in over 120,000 entries from the extensive interview tool used in

the research. The author would also like to recognize Mr. Clement O. Stephens, Executive

Director of the Liberian chapter of African Network for the Prevention and Protection against Child

Abuse and Neglect (ANPPCAN), Alvin Winford, project director, and the rest of his local NGO for

providing an excellent team of enumerator and supervisors for the study. Additionally, I would like

to acknowledge the advice and wisdom provided by Mr. Johnson Kerr, Director for Cartography

at the Liberian National Statistics Bureau, in formulating questions and in the designing the field

collection efforts. I am also very grateful to Jamie Fuller for providing an exceptional editorial

review of this report. The study was founded upon the methodologies and techniques advanced

by Macartan Humphreys and Jeremy Weinstein in their study on Sierra Leone, and without their

instruments, advice, and wisdom, this project would not have properly materialized. Finally, once

again, the author is deeply grateful for the tenacity of project manager and fellow investigator

Eleonora Koeb.


1 BACKGROUND

The Republic of Liberia is emerging from fourteen years of horrific civil war. The conflict has been

traced back, by some, to the initial landing of the former American slaves in 1822.5 It is ironic that

a country conceived as a humanitarian effort to repatriate former slaves became such a

repressive oligarchic state. The elitist class of Americo-Liberians represented only a fraction of

the population in this tiny parcel of African land but ruled tyrannically over the indigenous Africans

from its independence in 1847 until the regime was finally toppled in a military coup in 1980. “The

coup was the manifestation of 158 years of pent-up settler-native and civil society-government

(post-1950s) hostilities, spawned by a sociopolitical order that never evolved fully to

accommodate Liberia’s diverse and dynamic population.”6 Samuel Doe, the master sergeant

turned Liberian president, continued the oppression of his people and introduced a new level of

violence – ethnocide.7 In the wake of an abortive coup, Doe, an ethnic Krahn, ordered a purging

of any possible conspirators in Nimba County who were of Gio or Mano tribal decent. This parlay

by Doe into ethnic dividing would be one of the catalysts that Charles Taylor would rely upon in

the sparking of his insurgency in 1989, which would also begin in Nimba County.

The “war” as most contemporary Liberians understand it began in 1989 when Charles Taylor

crossed into Liberia from neighboring Ivory Coast. The war can be characterized by three distinct

phases. First was the Taylor insurgency, which included the surrounding of Monrovia,

international intervention, the killing of Samuel Doe, and an eventual electoral process that

installed Charles Taylor as president in 1997. Second was a brief interlude where Taylor

continued the legacy of authoritarian rule in Liberia. President Taylor’s “poor governance,

administrative malfeasance, corruption, intimidation, threats, torture, terrorist acts against his

population and summary executions of civilians”8 set the stage for the final phase of the “war.”

In 1999 Liberia spawned a new rebel organization, the Liberians United for Reconciliation and

Democracy (LURD), backed by neighboring Guinean president Lansana Conte. Fighting in

Liberia’s northern Lofa County intensified by 2001 and reached Monrovia in June 2003. By this

time, a second rebel faction intent upon the removal of Taylor from power emerged from the

eastern part of the country – the Movement for Democracy in Liberia (MODEL) – and was said to

be backed by the government of the Côte d’Ivoire.9 With the capital city, Monrovia, surrounded,

Charles Taylor accepted an invitation for asylum from the Nigerian government and departed the

county on 11 August 2003. The UN mission in Liberia (UNMIL) then was established by

September 2003, and the process of negotiating the peace and placing Liberia on a track of

renewal began.


With the signing of the Comprehensive Peace Agreement (CPA) in August 2003, all former

warring factions including the government troops began the post-conflict process of

Disarmament, Demobilization, Reintegration, and Rehabilitation (DDRR). On 19 September 2003

the United Nations Mission in Liberia (UNMIL) was established by Security Council resolution

1509 with a mandate to support the implementation of the cease-fire agreement. The National

Transitional Government of Liberia (NGTL) was organized with Monrovian businessman Charles

Gyude Bryant assuming leadership duties as chairman. Multi-party elections were peacefully

charted for and carried out in October 2005, and a constitutionally mandated two-candidate runoff

occurred in November 2005. The inauguration of the first female African head of state took place

in Monrovia on 16 January 2006, and the republic has since begun the process of reconciling and

rebuilding.

The transition to democratically elected officials has gone quite peacefully to date. Much of the

peacefulness can be attributed to the focus on the disarmament process, which has been noted

as the point of failure in the last round of international interventions (1996-1997).10 By December

2004 the UNMIL was able to disarm over 100,000 former combatants and secure approximately

28,000 individual weapons. Gun violence has been almost non-existent11 since the completion of

the Disarmament and Demobilization phase and the election process occurred in a quite peaceful

manner. The Congress for Democratic Change (CDC) and their standard bearer football great

George Weah contested fraud until well into the new year with highly charged rhetoric but only

minor violence. Ultimately, the CDC submitted and political conflict has subsided. The NTGL

succeeded in its mandate of moving Liberia to free and fair elections and with President Sirleaf’s

inauguration, the responsibility of the transitional government was terminated.

It is against this backdrop of a nation emerging from civil war and a peaceful round of national

democratic elections that this large-scale nationwide study was conducted. Field collection

occurred in February and March of 2006 just over two and a half years from the signing of the

Comprehensive Peace Agreement (CPA) on 18 August 2003. The survey was directed in order

to provide an initial assessment of the progress towards the objectives of national reconciliation,

conflict prevention, and social and economic reintegration, as well as the sustainable

development made by the DDRR interventions.

Micro-level ex-combatant research is a growing area of interest with regard to organizational

structure, management, and recruitment as well as exploration into the post-conflict realities of

disarmament, demobilization, rehabilitation, and reintegration (DDRR) and their program and

policy impacts. The research design, implementation, and methods for this study were derived


from the groundbreaking field research conducted by Macartan Humphreys and Jeremy

Weinstein in 2003 that resulted in a report of initial findings entitled “What the Fighters Say: A

Survey of Ex-combatants in Sierra Leone.” The data set has subsequently yielded multiple, very

important empirically based analytical papers in the topical focus area.

Every study endeavors to ensure the validity of the research though a sound methodological

design that enables others to replicate any of the procedures implemented. While attention to this

component of validity is often strictly adhered to, actual replication of large-scale studies is usually

the exception. Experimental validation is often sacrificed in the quest for the next big theory, and

nuanced areas of research go unattended. In the area of empirical research, a divide has been

identified between the United Nations (UN) and the academic community.12 The International

Peace Academy (IPA) suggests that a lack of empirically based research results in ill-informed

decisions by policy makers which subsequently impact the international program intervention

programs. In order to overcome the safety and logistical issues associated with data collection for

academics in post-conflict countries and the resource constraints faced by the UN, a closer

partnership has been identified as essential to moving ahead.

This report endeavors to build upon the important micro-level research advanced by Humphreys

and Weinstein. Also, the author hopes to close the divide between the UN and academic

community as well as to offer unique individual level (micro) insight into the issues that the ex-

combatants in Liberia faced during the war as well as within their respective communities today. It

is also the purpose of this paper to present findings that can be compared with the case in Sierra

Leone.


2 PURPOSE AND METHODOLOGY

2.1 Purpose of the Study

The study was conceived in order to determine the status of ongoing international intervention

reintegration and reconciliation efforts at the individual level and to obtain a mid-point assessment

of the impact of the DDRR program. To accomplish this, a number of goals were established for

the survey to ensure a comprehensive analysis.

• Collect demographic data on age, gender, tribal and factional affiliations, education level, DDRR program participation status, and other attributes that will enable respondent groupings and allow generalizations of the findings

• Gather information that informs on the factional activities of ex-combatants during the hostilities as well as data that profiles their dispositions prior to the onset of the war

• Collect information on how the former fighters experienced the post-conflict DDRR intervention process, how successfully they reconciled and reintegrated into Liberian communities, and the perspectives that the former fighters have with respect to their political expressions.

• Explore the linkage between reintegration success and the psychological construct of self-esteem

The interview intake questionnaire, which was based upon the instrument designed by

Humphreys and Weinstein for ex-combatants in Sierra Leone, was modified for use in Liberia.

While this study has focused on replicating the approach implemented by Humphreys and

Weinstein in Sierra Leone, a number of important steps that go beyond the original design have

also been implemented. A few of the original survey instrument’s periphery questions were

removed for logistical reasons. A number of new constructs and variables were also formulated

and introduced to amplify discovered nuances and modify perceived methodological

weaknesses. The instrument consisted of 185 questions and took approximately one hour to

administer. Eight sections collected the following information:

1. Background information that collected socio-economic demographics 2. Factional and sub-factional affiliations and geographic locations 3. Combatant recruitment techniques and motivations and incentives for participation 4. Organizational structures and associated patterns of military operations 5. Factional interactions with civilians during conflict 6. Former combatant experience with international intervention programs 7. Individual feelings – self-esteem 8. Perspectives on social integration and political expression

The survey results address many of the unanswered questions that face analysts and policy

makers with regard to the study of African civil wars and post-conflict integration of former

combatants. While the findings of this survey can be generalized to the population of ex-


combatants in Liberia, the survey also complements the Humphreys and Weinstein study in

Sierra Leone and can be used as the foundation for cross-country case analysis at the micro or

individual level.

2.2 The Approach – Research Design

Since a baseline study had not been conducted and the DDRR program intervention was already

in progress, a non-equivalent control group, posttest-only, quasi-experiment was considered the

best available research design. Assignment of experimental groups was selected randomly from

the same clusters of geographic locations. Ex post facto, the respondents were segregated into

two groups – DDRR program participants and non-participants. The program participants were

further divided into their respective categories of stage of intervention – Disarmed and

Demobilized (DD) only participants, participants who had enrolled but had not completed

reintegration training, and those participants who had completed reintegration training

(“graduates”). Those respondents who had received program (intervention) benefits were drawn

from the same communities as the group of ex-combatants who had not received benefits.

Implementation of the design was conducted in a blended approach that incorporated a

quantitative interview employment alongside a qualitative participatory method – focus group

discussion.13 The findings of this paper are for the most part conclusions of the quantitative

portion of the study only.

2.3 Accepted Protocols and Validity

In order to measure all the variables in the study, an existing instrument used in a previous micro-

level ex-combatant reintegration study was selected. However, as this study is the first of its kind

to address the issues associated with post-conflict Liberia, the investigator developed and/or

modified the existing instrument according to the theoretical definitions of those variables. Two

pilot tests of the modified Liberia survey instrument were carried out before a final version was

established.

Humphreys and Weinstein investigate the Disarmament, Demobilization, and Reintegration

(DDR) programs at a micro-level in order to discriminate components of a successful program.

Little research has been done in this area, and the authors utilize extensive research interviews

that follow the former combatants of Sierra Leone’s civil war. While the study established a

baseline for investigating the micro-level determinants of successful ex-combatant reintegration, it

failed to uncover any evidence of impact variance between program participants and non-

participants. Humphreys and Weinstein provide the framework and instrument for an excellent

quantitative study of micro-level relationships, and their work was heavily relied upon in the


foundational construction of the methodology for this research. This research endeavored to

improve upon some of the weaker protocols and introduce new constructs that might enable

more variation.

The validity of the survey instrument was established by testing face validity through feedback

from five Liberian ex-combatants and multiple UNDP Liberia staff members. The reliability of the

survey instrument was also established through two successive pilot studies that were conducted

in January-February 2006. As for external validity, congruency of protocols with the Sierra Leone

study should enable findings that can be generalized “across conflicts”14 to a broader population

of ex-combatant reintegration demographics. Replication of the Sierra Leone survey’s

methodological design will allow broader regional analysis and context as well as the testing of a

seamless data set.

2.4 Sampling Strategy

2.4.1 Defining the Population

The survey targeted a sample of 550 of an estimated ex-combatant population of 105,669.15 Field

collection and validation resulted in 590 usable interviews. While the study hopes to inform on

many issues associated with conflict, conflict analysis, reintegration, and reconciliation, the overall

sample size was primarily established as a function of determining the impact of the DDRR

program’s reintegration training. Statistical power analysis16 for sample size estimation utilized the

following parameters: significance (α) = 0.05, power = 0.9, and a population proportion=0.81

(Reintegration Participants/DD Participants). The estimated sample size was selected in order to

control type I and II errors while rejecting the null hypothesis (reintegration training has no impact)

and accepting an alternative that captures a greater than 10 percent improvement in the degree

of reintegration. An additional 20% was added to the study to account for DDRR program non-

participants and problematic/incomplete interviews.

The sampling framework at Table 1 was based upon NCDDRR database numbers for ex-

combatants 18 years of age or older and includes the distribution of the respondents. The study

divided the population into four primary categories – DDRR program non-participants, DDRR

program participants (disarmed and demobilized only), DDRR program participants (currently

enrolled in training), and DDRR participants who had completed reintegration training The study

also sought to obtain a representative sample by category and gender. Table 2 presents the

framework as a function of female ex-combatants. Both tables illustrate that the samples obtained


are largely representative of the targeted population with the exception of the grouping of females

who completed reintegration training, which appears to be slightly under sampled.

Table 1: Sampling Framework and Distribution of Ex-combatants over 18 Years Old

NCDDRR Totals Survey Sample Frequency Percent Frequency Percent

Disarmed/Demobilized Participants

46,152 43.7% 265

44.9%

Reintegration Participants 26,536 25.1% 170

28.8%

Completed DDRR Reintegration Training

19,244 18.2% 86 14.6%

DDRR Non-participants

13,7371 13.0% 69 11.7%

Total 105,6692 100% 590 100%

Table 2: Sampling Framework and Distribution of Female Ex-combatants over 18 Years Old

NCDDRR Totals Survey Sample Frequency Percent Frequency Percent

Disarmed/Demobilized Participants

13,081 12.4% 63

10.7%

Reintegration Participants 5,465 5.2% 34

5.8%

Completed DDRR Reintegration Training

3,896 3.7% 11 1.9%

DDRR Non-participants

996 0.9% 11 1.9%

Total 23,438 22.2% 119 20.2%

Of secondary importance to population generalizations is the distribution of respondents by their

former factions. While these generalizations are important, a suitable population frame proves to

be quite problematic. A breakdown and discussion of factional representation is included in

section 3.1.1, Former Faction Representation.

For logistical and focus reasons, the decision was made to intentionally exclude all former

combatants currently under the age of 18. Of note, however, is that approximately 108 of the 590

respondents included in this survey would’ve been under the age of 18 at the time of the CPA in

2003. Figure 1 provides a histogram of the age distribution for this study. The mean age for this

study was 26.6 years old. The NCDDRR database reports a mean population age 25.317 years

old. The population included registrants as young as seven.

1 Not a number reported by the NCDDRR; represents an estimated rate of 13%.

2 Exact NCDDRR database total was 81,368.


Figure 1: Distribution of the Sample by Age

020

40

60

Nu

mbe

r o

f R

esp

on

de

nts

18 20 25 30 35 40 45 50Age of Respondents

NB: Not illustrated on the graph are two respondents ages 56 and 78

Age Distribution of Sample

2.4.2 Nationwide Random Sampling

The survey targeted a sample population of 550 of the estimated 105,699 ex-combatants over

the age of 18 in Liberia. In order to obtain a representative selection of respondents for this

countrywide pool, a multi-stage approach of random sampling methodologies was implemented.

To ensure that every potential respondent residing in Liberia had the opportunity to be included in

the study, a cluster random sampling method18 was used.

For efficiency of survey administration, first-stage sampling narrowed the population to thirty

clusters of twenty subjects each. Table 3 illustrates how seventeen of the selected clusters fell

within nine Liberian administrative districts and thirteen fell within Greater Monrovia. The

percentage of clusters in Monrovia (33%) also closely matches the percentage of the population’s

preferred resettlement choice (37%). This method utilized information that was entered into the

DDRR Joint Implementation Unit’s (JIU) ex-combatant database.19 The cluster random sampling

was run using all 66 Liberian administrative districts (includes Greater Monrovia) weighted for the

number of ex-combatant preferred resettlement location choices.

Two refining approaches were employed for the second stage and took into account the

administrative logistical limitations unique to both the urban Greater Monrovia and the rural

interior. The following details the methodology utilized for selecting respondents in both.


Table 3: Nationwide cluster representation

County District Name

Estimated Ex-com

Resettlement Population

Number of 20

person clusters

Sample Size

Bomi Klay 2891 2 40

Bong Jorquelleh 4849 2 40

Bong Suakoko 879 1 20

Grand Bassa District#3 1881 1 20

Grand Cape Mount Tewor 654 1 20

Grand Gedeh Tchien 3339 3 60

Lofa Voinjama 5378 2 40

Montserrado Greater Monrovia 30468 13 260

Montserrado St. Paul River 1014 2 40

Nimba Sanniquelleh-Mahn 5183 2 40

Sinoe Juarzon 152 1 20

Total 600

2.4.3 Sampling for the Liberian Interior

For the rural interior of Liberia—i.e., the 17 clusters selected in nine districts—the second-stage

sampling method employed a second round of cluster random sampling from communities with a

population above 40 households weighted by the number of inhabitants. The National Village

mapping Project data20 from the Humanitarian Information Center, Liberia, was utilized for this

final macro sampling stage.

With the community clusters now selected, the UNDP field team supervisors approached the

local/traditional leaders to organize the ex-combatants residing in the area for the final stage of

sampling. The third stage of the sampling strategy produced the targeted respondents. Potential

respondent pools were built up to two to three times the required numbers and a local method of

random sampling was employed to identify the respondents. In some rural areas however, the

ex-combatant populations were quite small and the field teams could not obtain robust pools on

every occasion. A proportional quota sampling method was utilized to ensure a 20 percent

representation of female ex-combatants.

2.4.4 Sampling for Greater Monrovia

The second-stage method employed for the interior of Liberia could not be easily implemented

within Greater Monrovia because the database listing did not allow suitable geographic

refinement for community definition—i.e., no unique address or location with Greater Monrovia.

So utilizing a listing of the sixteen metropolitan zones of Greater Monrovia and using equal


weighting for all zones, the 13 clusters designated in the first stage were refined once again

through cluster random sampling. The sampling method yielded a distribution of the 13 clusters

within ten metropolitan zones (Table 4). Within these ten zones, the third-stage sampling strategy

ultimately identified the required sample population.

Table 4: Greater Monrovia clusters representation

With teams of enumerators set up within the targeted

Monrovian communities (zones), the UNDP field team

supervisors approached the local/traditional leaders to

organize the ex-combatants residing in the area for the final

stage of sampling. The third stage of the sampling strategy

produced the targeted respondents. Potential respondent

pools were built up to two to three times the required

numbers, and a local method of random sampling was

employed to identify the respondents. A proportional quota

sampling method was utilized to ensure a 20 percent

representation of female ex-combatants.

2.4.5 Observed Cluster Variations

The random selection of clusters resulted in targeted areas that paralleled the variation found in

the Sierra Leone study and provided a uniquely random sample. As reported in the Sierra Leone

study and presented here redundantly21 are a few of those variations:

• Presence or absence of DDRR programs, government, and social services;

• Levels of ex-combatant presence, and variation in faction membership;

• Diversity of ethnic groups, language groups, and economic activity.

Unlike the survey conducted in Sierra Leone, data collection in Liberia took place before the rainy

season, which negated any possible impacts of the weather. This helped to ensure access to

numerous remote locations.

2.4.6 Potential Sources of Bias

The survey introduces a number of sources of bias that must be considered when evaluating the

findings. While total elimination of bias was the goal, numerous challenges are inherently

encountered when surveying in a post-conflict situation. This section will illuminate the issues that

readers should be aware of when analyzing the findings of the paper. From the administrative

side of collecting the data, to a minor degree, item non-response was encountered in this study

Greater Monrovian Zones

Number of 20 person clusters

01 New Kru Town 2

04 West Point 1

06 Central Monrovia B 1

07 Sinkor 2

08 Lakpazee 1

10 Congo Town 1

11 Paynesville 2

13 New Georgia 1

14 Barnersville 1

16 Caldwell 1

13


through some enumerator error and possible respondent fatigue, sensitivity, lack of knowledge, or

other factors. Item non-response occurs when respondents are not asked or decline to give any

response during interviews. With regard to the context of the environment, a few additional factors

may possibly have affected the respondents’ answering of questions.

While many levels of randomization were incorporated into the planning of the geographic

clusters that were to be drawn upon for the study, the final selection of respondents was most

definitely influenced by the community leaders who provided the study with the pool of

candidates. This effect was mitigated by selecting random candidates from the pool that was

presented on location.

Truth telling/self-reporting is always a concern when interviewing former combatants. While

Liberia was a quite conciliatory environment during the time of field collection and the

overwhelming majority of the respondents were more than willing to participate, the Liberian Truth

and Reconciliation Commission was inaugurated (21 February 2006) at almost the same time

that the research began field collections. It’s hard to be certain what effect this had on the

respondents’ answers, but it is assumed to have been quite minimal.

An important component of the population that was not accessible to the enumerators was

refugees of the conflict. As of 1 January 2006, UNHCR reported that the Liberian refugee count

was an estimated 231,100.22 The refugees of the Liberian conflict were said to be located in

Sierra Leone, Guinea, Côte d'Ivoire, Ghana, and the United States. The population of refugees

most definitely contains an element of former fighters. The perspectives of this group in the

presentation of findings on warring faction dynamics and DDR perspectives are not included.

2.5 Enumeration Strategy

Twenty candidates were interviewed and tested for positions as enumerators of the survey. A

three-day training and testing event resulted in sixteen certified enumerators, who were

subsequently broken down into four teams of four each. Each team was organized with requisite

language skills for their designated cluster areas. UN field officers served as the team leads and

controlled and facilitated movements throughout their assigned zones.

2.6 Capturing Variation in the Course of the Conflict

The civil war in Liberia lasted for over 14 years and involved many warring factions. Many of the

former combatants had participated in multiple factions. A thorough understanding of the conflict

also requires an examination of the war through the lens of Charles Taylor the insurgent warlord


and then Charles Taylor, Liberian head of state. In order to ensure randomness in data collection,

multiple randomization forms were introduced for focusing the detailed collection of information

associated with factional alliances. The extended temporal nature of the conflict potentially serves

as a bias that can be introduced through a respondent’s “averaging” of the totality of his or her

wartime experiences.

In order to mitigate this bias, the same protocol used in the Sierra Leone study was employed in

Liberia. Respondents were asked to report their geographic locations, factional affiliations, and

level of activity during eleven discrete periods of time between 1989 and 2003. The periods of

time were selected for their ability to trigger a precise memory of the events and activities that

occurred during the period in question. Event recognition for the randomization protocol was

piloted in Liberia prior to inclusion in the final instrument. During the survey, once the enumerator

completed this section that mapped the respondents’ locations, affiliations, and activities, a

separate sheet (randomization form) was used to select one period of time for further

examination. The randomization form drove which faction and period would be addressed in the

sections pertaining to “Recruitment and Participation,” “Inside the Organization & Patterns of

Military Operations,” and “Interaction with Civilians.” For the purpose of this report, sections 3.1,

Ex-combatant Profiles, and 3.2, The Dynamics of the Warring Factions, are driven by this

randomization protocol.

2.7 The Research Team

James Pugel served as the principal investigator for this ex-combatant research effort. Together

with Eleonora Koeb, lead for the qualitative, focus group discussion portion of the project and co-

investigator for the non-combatant parallel project, and the executive committee of the Liberian

chapter of African Network for the Prevention and Protection against Child Abuse and Neglect

(ANPPCAN), Pugel conceptualized the project, designed the survey, and led the research in the

field. The final report was written and coordinated with the assistance of the JIU’s Monitoring and

Evaluation Unit headed by Mr. Shecku Silla.

James Pugel is a Research Associate with the Centre for the Study of Civil War (CSCW),

International Peace Research Institute, Oslo. The data collection effort was completed under the

direction of Mr. Charles Achodo, Senior Program and Policy Advisor, UNDP Liberia, as part of a

graduate thesis with the National Defense Intelligence College, Washington, DC.

The enumerators for this survey included: Franklin Varmah, Armah Winford, Tawah Fayiah, Joe

Bautz, Albert Tokpa, Joseph Sims, Comfort Gayon, Jonathan Roberts, Michael Edwards,


Laurene Deermi, Cynthia Momo, Benjamin Alete, Masadeh Collins, Rose Dowee, Tiah Diagbe,

and Ruby Folly.


3 FINDINGS

There were 590 validated surveys used in the construction of the findings. The primary findings

are presented according to the former combatant’s program participation category. Secondary in

importance is identifying and validating a sample frame along factional lines. Understanding how

the former factions acted and interacted on the battlefield during the 14 years of conflict will also

help us to understand the nature of the reintegration issues that face the international intervention

programs today. Unfortunately, the NCDDRR database has proven to be quite problematic when

it comes to requiring a benchmark for former faction identification. The purpose of the sampling

frame is to serve as a check for a representative sample. Initial estimates of faction strengths

totaled 38,00023 combatants. In February 2004, the National Transitional Government of Liberia

increased the numbers to 53,000.24 By August of 2004 the DDRR program had demobilized

68,85125 and by February of 2005 over 100,00026 combatants were disarmed.

3.1 Ex-combatant Profiles

Important to understanding the nature of conflict and interpreting indicators associated with

reintegration and reconciliation is ensuring a proper profile for the sample of ex-combatants. As

this paper addresses both the wartime and the post-conflict time periods, two unique profiles will

be presented for discussion. This section introduces the general demographic profile of the

sample of former fighters and then proceeds to present the respondents as a function of their

wartime participatory role. Subsequent sections will then examine the intervention treatments

categorically as pertaining to their association and status in/with the NCDDRR program.

3.1.1 Former Faction Representation

During the 14-year course of the civil war in Liberia, there were many different armed factions.

The difficulty in sorting out the players to the Liberia civil war lies with a transitional period in which

Charles Taylor the “warlord” became Charles Taylor the “president.” Many of the respondents

fought during both segments of the conflict and maintained numerous factional affiliations –

sometimes even jumping from supporting the state to joining the insurgent movements. The

survey was designed to capture the experiences of respondents who had participated in any of

the factions from the war’s onset. Many of the factions that were initially formed have since been

demobilized or renamed. Table 5 represents the breakdown of respondents by former factional

representation. Important to note here is that this breakdown is a random selection of only one of

the factions that the respondents participated in. The distribution given at table 5 constitutes the

sample population that will be drawn upon for subsequent analysis. For the purposes of this


report, only the four major factions – which also have a significant number of respondents –

NPFL, GoL (Taylor), LURD, and MODEL, will be used for detailed examinations.

Table 5: Random Sample of Former Faction Representation (1989-2003)

Warring Faction

Frequency Sample

Percentage

Armed Forces of Liberia (AFL) 8 1.36

Government of Liberia Militia (GoL/militia) – Taylor 186 31.53

Liberians United for Reconciliation and Democracy (LURD) 126 21.36

Movement for Democracy in Liberia (MODEL) 76 12.88

National Patriotic Front of Liberia (NPFL) 150 25.42

United Liberation Movement for Democracy in Liberia (ULIMO) 7 1.19

ULIMO-K (Alhaji G.V. Kromah) 14 2.37

ULIMO-J (Roosevelt Johnson) 5 0.85

Liberian Peace Council (LPC) 14 2.37

Independent National Patriotic Front of Liberia (INPFL) 4 0.68

TOTALS 590 100.00

3.1.2 Basic Demographics

This section provides a profile of the combatants with regard to some of their pre-war

demographics as well as information that characterizes their status and level of participation

within their reported factions. Figures 2-5 provide a snapshot of the pre-war profiles of the former

combatants. Throughout the illustrations, it is important to note the correlations between the

NPFL and the GOL/Militia categories, for they provide a historical progression and insight into the

predilections of the fighters loyal to Taylor from 1989 to 2003. The tables illustrated here do much

to confirm the polarization of tribal and religious affiliations that are often spoken to in discussions

of the conflict. While the MODEL was found to have an overwhelmingly homogenous

membership of Krahn, the LURD’s combatants were drawn from the tribal groups most

associated with the northeastern hinterland regions where their insurgency was initiated –

Gbandi, Loma, and Mandigo. Taylor’s forces were shown to be drawn from the historically

antagonistic tribes of the LURD and MODEL – Gio and Mano as well as a strong contingent of

Bassa and Kpelle. Figure 3 highlights the predominance of Christians within the ranks of all

factions and the diminishing role that “non-believers” or those practicing traditional religions have

in Liberia. By percentage, the LURD’s population has three times more Muslims in its ranks than

any other faction.


Figure 2: Ethnic Composition, by Major Former Faction

02

04

060

80

Nu

mbe

r o

f R

esp

on

de

nts

Bassa

Gbandi

Gio

Gola

Kpelle

Krahn

Loma

Mandingo

Mano

VaiO

ther

None G

iven

by Major Faction and Tribal Affiliation

Distribution of Respondents

NPFL GOL/Militia LURD MODEL

Figure 3: Ex-combatant Religious Affiliations

93.55%

3.763%

71.43%

26.98%

90.79%

7.895%

91.33%

5.333%

GOL/Militia LURD

MODEL NPFL

None Christian Muslim

NB: 538 respondents (186-Taylor's GoL, 126-LURD, 76- MODEL, 150-NPFL)

by faction

Ex-combatant Religious Affiliations

Figure 4 illustrates the level of political activism within the ranks of the four major factions under

investigation. Almost 88% of all of the respondents did not claim a pre-war political party

affiliation. One important indicator is the showing of support—one out of every five respondents—

for the former President Doe within the ranks of the MODEL membership. Significantly in Figure

5, which highlights pre-war communities, every faction was made up of a membership primarily

from outside the greater Monrovian area.


Figure 4: Pre-War (1989) Political Allegiances

7.692%3.846%

88.46%

.8%

99.2%

20.27%

2.703%

77.03%

4%14.67%

81.33%

GOL/Militia LURD

MODEL NPFL

President Doe NPFL

No Party/Political Affiliation

NB: 582 respondents

by faction

Ex-combatant Pre-war Political Affiliations

Figure 5: Former Fighter Pre-War Communities

54.95%45.05%

57.72%

42.28%

88.89%

11.11%

72.92%

27.08%

GOL/Militia LURD

MODEL NPFL

Outside of Monrovia Greater Monrovia

NB: 521

Outside of Monrovia vs. Greater Monrovia

Pre-war Home Community Locations

A few basic demographics of the factions that characterize the wartime status and level of

participation of the respondents are detailed in figures 6-8. Fully three quarters of the respondents

indicated that they left their factions as an enlisted or “common” soldier. The sample of NPFL

respondents indicated a large proportion of officers, while few MODEL respondents considered

themselves in the category at all. Of note, numerous informal interviews conducted alongside the

collection of primary data set indicated that the MODEL’s organizational structure did not include

many “officer” billets. Figure 7 details an examination of the roles that the former combatants held

within their respective former factions. Fully 84% of the sample of ex-combatants indicated that

their primary role within the factions was as a combat soldier. Extremes on this can be seen when

the roles are disaggregated by faction – MODEL claims a 95% combat role rate while the LURD

claimed a rate of only 72%.


Figure 6: Military Rank Held in Faction

20.43%

79.57%

13.49%

86.51%

2.632%

97.37%

54%46%

GOL/Militia LURD

MODEL NPFL

Officer Enlisted

NB: 538

by faction

Ex-combatant Rank in Former Factions

Figure 7: Roles of Combatants within Factions

81.08%

7.027%

72%

16%

94.67%

1.333%

90%

3.333%

GOL/Militia LURD

MODEL NPFL

Combat Soldier Administrative Medical Corps

Communications Spying Wife

Sex Slave Camp Worker Forced Labor

NB: 535

by faction

Ex-combatant Roles in Former Factions

One final general characteristic of the former combatants speaks to the intensity of their

respective service while serving within their former factions. Figure 8 illustrates the number of

times that a respondent was injured while serving with his or her faction during a given period of

time. Overwhelmingly, both Taylor-related factions incurred the greatest frequency of injury when

compared against their opposing insurgent groups. On average, 77% of Taylor troops reported

battle injuries while the collective insurgent groups reported injuries at a rate of 46%.


Figure 8: Number of Times Wounded

37.41%

37.41%

17.99%

7.194%

46.88%

40.63%

11.46%1.042%

56.9%27.59%

8.621%6.897%

26.55%

44.25%

15.04%

14.16%

GOL/Militia LURD

MODEL NPFL

Never Once Twice Three or more times

NB: 538

by faction

Number of Times Wounded During the Conflict

3.1.3 Gender Differences

Approximately 20% of the sample of ex-combatants were female, and across the four factions,

this percentage generally held true except for MODEL. The MODEL sample indicated a

percentage of almost 40% female. Figures 9 and 10 disaggregate the sample of ex-combatants

by gender in an effort to distinguish any trends. Figure 9 shows that, as expected, the females

appeared to take on more of the administrative or supporting roles within the former factions.

Surprisingly, however, over 70% of the females reported that their primary role was that of a

combat soldier. This administrative or supporting role looks to have afforded the population of

females a level of protection within the units as well. Figure 10 indicates that well over 50%

percent of the population did not receive any battle injuries, compared with only 34% of their male

comrades.

Figure 9: Gender – Roles (within Former Factions) Figure 10: Gender - Times Wounded

34.17%

40.34%

17.09%

8.403%

55.17%32.18%

6.897%5.747%

Male Female

Never Once Twice Three or more times

NB: 444

by gender

Number of Times Wounded During the Conflict

87.15%

3.854%

4.283%

71.43%

4.202%

17.65%

Male Female

Combat Soldier Administrative Medical Corps

Communications Spying Wife

Sex Slave Camp Worker Forced Labor

NB: 586

by gender

Ex-combatant Roles in Former Factions


3.1.4 Key Patterns

The basic demographics presented in this first section serve as a cornerstone for understanding

the conflict that ravaged Liberia for 14 years. A characterization of the former warring factions

begins to illustrate key patterns that might potentially inform us on some of the variances

encountered during reintegration.

The former warring factions tend to sort themselves out around tribal affiliations. While not

completely homogeneous organizations, the all-Liberian factions appear to have a dominant

group or groups in their membership. Christianity, the predominant spiritual driver of the Liberian

people, was also central to all factions. In the LURD, however, Muslims comprised more than a

quarter of the LURD, which was almost four times as many as any other group. Across factions,

the overwhelming majority of fighters began the war as residents in towns and villages outside

Monrovia.

3.2 The Dynamics of the Warring Factions

In order to fully understand the scope and the challenges of reintegrating the ex-combatants in

Liberia, we must understand the dynamics of the organizations that they participated in during the

conflict. This section endeavors to highlight the dynamics of the former warring factions through

examinations under four particular topical areas – social and political origins, recruitment and

incentives, interactions with civilians, and interstate influences and activities.

3.2.1 Social and Political Origins

The civil war in Liberia is complex on many different levels. The current 14-year war under study

in this report is quite ambiguous when it comes to examining its social and political origins. It can

be said, on the one hand, that the 14-year war was really two wars – one that resulted in the

overthrow of the Doe regime and one that resulted in ousting President Taylor from power.

Others will contend, however, that the war that Charles Taylor started never really ended until his

asylum was instituted in 2003 – one long continuous struggle for power in the transition of

government with roots in the 1980 coup that Master Sergeant Samuel Doe participated. Years of

indigenous Liberian oppression by the ruling Whig Party and the minority elitist group, the

repatriated slaves or Americo-Liberians, can be looked to as one essential piece of the enigma.

Another domino of the social and political origins fell with the promotion by Samuel Doe of an

unhealthy and most hateful rivalry between select Liberian tribes. Both of these catalysts are

surely contributors to the origins that are investigated within this report.


Figures 11-14 attempt to portray some of the characteristics and demographics that help to inform

us on the social and political origins of the former warring factions. The UNHCR reported that

during the height of the conflict, over 800,000 people were displaced. The impact of this very

large number is shown in the findings presented in figure 11. Significantly, at least 50% of the

membership from each faction had suffered a forced displacement prior to joining. The pre-war

family structures are examined in figure 12. The disaggregation by faction shows little variance in

pre-war dispositions of the former combatants, and almost 86% of the respondents maintained

that at least one parent was alive at the war’s onset.

Figure 11: Members of Factions Who Were Previously Displaced

35%

65%

50%50%

42.47%

57.53%

43.75%

56.25%

GOL/Militia LURD

MODEL NPFL

Not Displaced Displaced

NB: 568 responses

Were you ever displaced?

Pre-Recruitment Disposition

Figure 12: Pre-War Family Structures

13.44%

15.05%

11.83%59.68%

15.87%

12.7%

9.524%61.9%

6.579%

19.74%

11.84%61.84%

13.33%

12%

17.33%57.33%

GoL/m LURD

MODEL NPFL

Neither Mother only

Father only Mother and Father

538 respondents

Which parents were alive?

Family Structure Before the War

The introductions to the factions that the sample of ex-combatants received prior to joining varied

tremendously. Figure 13 illustrates the differences in techniques utilized by each faction. While

the factions affiliated with Taylor both pre- and post-presidency relied heavily upon friend and


family camaraderie, the LURD focused on more violent introductions – village attacks and

ambushes—for filling their ranks. The MODEL, on the other hand, appears to exhibit a uniquely

high percentage of members who actively sought out an introduction.

Figure 13: How Combatants Were Introduced to Their Former Factions

53.3%

13.19%

24.18%

5.495%

39.68%

27.78%

24.6%

5.556%

25.68%

22.97%16.22%

33.78%44.97%

28.19%

10.07%

12.75%

GOL/Militia LURD

MODEL NPFL

Friend/Relative joined group Group attacked my village

Group ambushed me on road Group attacked my faction

A member/family approached me Went looking for faction

NB: 533

by faction

How Introduced to Former Faction

As mentioned within the discussion of demographics of the preceding section, tribal affiliation,

although completely homogeneous, proved to divide along “allied” lines. Figure 14 seeks to shed

some light on the ethnic/tribal conflict that the ex-combatants encountered in their experiences

with the factions. While the majority of respondents did not report distrusting any other tribe, those

tribes that respondents did report being unable to trust include the Mandigo, Gio, and the Krahn.

Figure 14: Tribes Not Trusted, by Faction

020

40

60

80

Nu

mbe

r o

f R

esp

on

de

nts

Americo-Liberian/Congo

Gbandi

Gio

Krahn

Mandingo

Mano

Other

None G

iven

by major faction

Tribes not Trusted by Competing Faction

NPFL GOL/Militia LURD MODEL

3.2.2 Recruitment and Incentives

Insight into the recruitment techniques of the former warring factions provide crucial indicators that

may possibly assist an understanding of the reintegration challenges that the former combatants


may face. Figures 15-17 illustrate the recruitment schemes, and the findings show some

important variations in approach.

Figure 15 provides a simple overview of the recruitment profiles of the former warring factions

over time. For ease of illustration, the forces affiliated with Taylor have been grouped into one

category and present a profile that begins in 1989. The first stage of the war had a formal

demobilization that ended in 1997. At this point, most of Liberia’s former warring factions were

disbanded – including the NPFL. In many instances, however, demobilization during this interwar

period in Liberia appeared to have no real effect upon the force structures of many of the factional

organizations, specifically the NPFL. Sub-factional (or the units subordinate to the major factions)

tracking analysis surfaces numerous pre- and post-1997 duplicate organizational names. Notably,

the recruitment profile provides a very good picture of how the sample population was drawn into

the final or post-1997 stage of the conflict. The rumblings of the LURD incursion in 2000 reflect

the first post-1997 recruitment spike. The years 2001 and 2002 show a relatively flat profile that

increases significantly during the last year of hostilities. In 2003 we see significant emphasize on

recruitment by all three factions. At this point, both insurgent groups had enveloped Monrovia and

were engaged in intense urban warfare in and around the capital during June and July.

Figure 15: Year of Recruitment to Faction

020

40

60

19891990

19911992

19931994

19951996

19971999

20002001

20022003

2004

by 2003 CPA signatories

Recruitment Years

Taylor's Forces (NPFL/GoL) LURD MODEL

Figure 16 addresses the reasons for joining the faction. Many rationales were cited for

participation in the conflict, and a few distinctions surface when the factions are queried.

Overwhelmingly, family protection was cited most often by the respondents of all groups. Only a

paltry 10% of the sample gave ideological reasons which is fairly consistent across the factions.

MODEL, which reported the highest percentage of those citing “family protection,” also had the

lowest instances of abduction. An important distinction requiring further investigation lies within


coercion as a means of recruitment. The survey recorded a similar rate of responses from former

fighters with regard to joining because they were “abducted” and those who joined because they

were “scared.” These two methods of coercive recruitment accounted for almost 40% of the

responses. Notably, only 4% of the overall responses cited “money” as reason for joining.

Figure 16: Reasons for Joining

18.8%

22.22%

33.76% 21.47%

19.63%

33.74%

12%

23%

42% 18.46%

20%

34.87%

GoL/m LURD

MODEL NPFL

Supported Political Goals People in Group Lived Better

Abducted Money

Scared Protect Family

Other Soldiers Killed my Family

Respondents could give multiple reasons - 692 responses

by Faction

Reasons for Joining

The incentives used by the factions to encourage participation are presented in figure 17. The

incentives offered by commanders to their membership prove to significantly inform us on the

dynamics of their particular organization. While pecuniary incentives (food, money, job) were the

primary rewards for both of Taylor’s factions and the LURD, the MODEL was distinctly oriented

on a non-pecuniary incentive scheme. Family protection and improving the situation in Liberia

accounted for over 57% of the responses given by former MODEL participants. With an additional

14% of the respondents citing “revenge” as a motivating factor (the only group to really

acknowledge this incentive), the orientation of MODEL is shown to be distinctly different.

Figure 17: Incentives for Participation in Factions

32.81%

16.93%15.1%

17.19%

15.63% 28.51%

16.67%10.09%

21.05%

20.18%

16.13%4.839%5.645%

26.61%30.65%

26.07%

14.52%13.2%

19.8%

23.43%

GoL/m LURD

MODEL NPFL

Money Food

A Job Improve Situation in Liberia

Family would be Protected A way to get Revenge

Other

Respondents were allowed to indicate multiple incentives (1,039 responses)

responses by faction

Incentives Promised for Joining Group


3.2.3 Interaction with Civilians

Historically on the African continent, civilians have always been forced into participating in conflict

through the mechanism of forced labor.27 This participation and the psychological and emotional

effects attached are often overlooked by the international community in the aftermath of the

conflict. While former combatants can be offered rehabilitation treatment for talking through their

traumatic wartime experiences, the non-combatants are usually left to work through their issues

on their own. Figures 18-21 illustrate the interactions that each of the major factions had with the

civilian population with regard to formal instruction given on the treatment of non-combatants and

also the employment of forced labor by their respective organizations. Figure 18 shows that the

majority of ex-combatants were given directions by a senior commander on how non-combatants

were to be treated. Taylor’s post-1997 supporting factions were found to be the least informed on

civilian treatment, with 1 in 4 soldiers operating without guidance.

Figure 18: Direction on Treatment of Civilians on the Battlefield

25.16%

74.84%

14.58%

85.42%

5.882%

94.12%

12.3%

87.7%

GOL/Militia LURD

MODEL NPFL

No Yes

NB: 424

by faction

Ever Given Instructions on How to Treat Civilian Population?

In order to precisely characterize the employment of non-combatants in forced labor roles during

the Liberian civil war, figures 19-21 will place the phenomenon within context through a

progressive examination of issues as reported by the combatants themselves. Figure 19 displays

the frequency that civilians were reportedly used for forced labor operations – hauling

ammunition, equipment, and the spoils of war. While civilian impressments appears to have been

the norm for both of the Taylor-affiliated factions and the LURD (fewer than 10% of these factions

reported that civilians were never used), again, the MODEL appears to have been less

dominating of the battlefield’s non-combatants, citing a non-usage rate of over 34%.


Figure 19: Former Factional Employment and Implementation of Civilian Forced Labor

41.94%

47.85%

10.22%

38.1%

52.38%

9.524%

22.37%

43.42%

34.21%

46.67%

46%

7.333%

GOL/Militia LURD

MODEL NPFL

Almost Always Sometimes Almost Never

NB: 538

by faction

Forced Labor: Use of Non-combatants

Figures 20 and 21 truly make an argument for challenging the use of “porter” as the descriptor for

the civilians who were drawn into the conflict through this facet of war. The euphemistic

characterization of non-combatant involvement is disputed through examination of their freedom

to choose and the consequences of their decisions should they choose not to participate.

Overwhelmingly across factions, non-combatants were rarely given the opportunity to decide

whether they wanted to participate or not, as illustrated in figure 20. The ramifications of non-

participation were brutal. Important variations, again, surface with respect to the MODEL. While

the other three factions under examination reported unmerciful means of coercion to sway those

civilians who would not comply (on average over 77% reported violent means), MODEL

respondents reported at a rate of 62% that “nothing” actually happened to non-combatants who

did not want to support their efforts.

Figure 20: Non-Combatant Freedom of Choice in Supporting Former Factions

11.83%

38.17%

50%

11.9%

30.16%57.94%

17.11%

38.16%

44.74%

9.396%

40.94%49.66%

GOL/Militia LURD

MODEL NPFL

Almost Always Sometimes Almost Never

NB: 537

to work for a faction

Forced Labor: Non-combatant Freedom of Choice


Figure 21: Civilian Ramifications of Refusing to Submit to Forced Labor during Conflict

80.22%

3.846%

14.84%

70.4%

10.4%

17.6%

35.53%

61.84%

74.32%

5.405%

16.89%

GOL/Militia LURD

MODEL NPFL

Killed Beaten

Reported to Elders Nothing

NB: 531

for not participating

Forced Labor: Non-combatant Consequences

3.2.4 Interstate Influence and Combatant Activities

The Mano River Region signatories (Liberia, Sierra Leone, and Guinea) have been inextricably

linked to the conflicts of their neighbors due to the porous borders and inability of the states to

control their territorial sovereignty. The recent collapse of state authority in Liberia’s western

neighbor, Côte d’Ivoire, has erased all notions that the region can be viewed with any degree of

stability. The phenomenon of foreign, regional warriors has been promulgated through the

consistent (and profitable) market need for combatants within a very small geographical area with

numerous lines of communication for moving between states. Figures 22 and 23 examine the

interstate influences and activities of the sample of ex-combatants. On average, only a little more

than 13% of the respondents indicated that they had fought in countries other than Liberia. The

majority of these fighters served in Sierra Leone and Côte d’Ivoire. Other countries mentioned

included Guinea and Burkina Faso. Figure 23 presents the foreign fighter experience by faction.

Figure 22: Ex-combatants with Foreign Fighter Experience

92.05%

7.947%

84.38%

15.63%

83.56%

16.44%

86.18%

13.82%

GOL/Militia LURD

MODEL NPFL

Fought only in Liberia Foreign Fighting Experience

NB: 443

by Faction

Ex-combatant Foreign Fighting Experience


Figure 23 attempts to gauge the influence of foreign fighters within the ranks of Liberian factions.

The findings reinforce the commonly accepted notion that interstate influence is rampant within

African civil wars. Overwhelmingly, the majority of the sample indicated the presence of foreign

fighters within their ranks; the largest percentage was acknowledged by the LURD by nearly 80%

of their membership. At least 100 respondents from the sample acknowledged participation by

citizens from Guinea and Côte d’Ivoire, but the overwhelming interstate participation came from

Sierra Leone, as indicated by 230 of the former combatants.

Figure 23: Foreign Fighters in the Ranks of Liberian Factions

38.85%

61.15%

21.5%

78.5%

37.88%

62.12%

47.66%52.34%

GOL/Militia LURD

MODEL NPFL

No Foreigners in Unit

Foreign Fighters Present in Unit

NB: 458 respondents

percentage of respondents that indicated foreigners in their unit

Foreign Fighter Presence within Liberian Factions


A few perspectives on gender are worthy of inclusion in this section on factional dynamics. The

findings here do little to show any great variance in perspective or disposition. Interestingly, when

the reasons for joining the former foreign factions are disaggregated by gender (figure 24), the

most notable finding is that only 12% of the females (compared with 21% of the males) reported

that they were abducted. The difference in this category appears evenly distributed through the

rest of the categorical rationales. Figure 25 demonstrates a near equal percentage of female ex-

combatants that had foreign fighting experience commensurate with that of their male

counterparts, and figure 26 illustrates a similarly comparable rate of forcible displacement prior to

joining the ranks of their former factions.


Figure 24: Gender – Reasons for Joining (for Former Factions)

10.25%

20.5%

20.66%

34.38%

11.76%

12.42%

21.57%

36.6%

Male Female

Supported Political Goals People in Group Lived Better

Abducted Money

Scared Protect Family

Other Soldiers Killed my Family

Respondents could give multiple reasons - 758 responses

by Gender

Reasons for Joining

Figure 25: Gender – Foreign Fighter Experience

86.39%

13.61%

88.35%

11.65%

Male Female

Fought only in Liberia Foreign Fighting Experience

NB: 485

by Gender

Ex-combatant Foreign Fighting Experience

Figure 26: Gender – Formerly Displaced before Joining

41.94%

58.06%

42.61%

57.39%

Male Female

Not Displaced Displaced

NB: 568 responses

Were you ever displaced?

Gender: Pre-Recruitment Disposition


3.2.6 Key Patterns

The dynamics of the former warring factions provide insight into the challenges that the

reconciliation and peace-building process will face as Liberia and the international community

attempt to mend the socio-economic fabric of the country. The discussions on social and political

origins, organizational recruitment techniques and incentives, interaction with non-combatants,

and interstate activities have truly characterized each warring faction as unique. A summary

understanding of each faction’s patterns and predilections is therefore presented now.

Taylor’s Government Forces and Militias

The core of Charles Taylor’s forces while he served as president was the group of fighters that

served the leader when he began his insurgency in 1989 – the NPFL. Taylor’s military force was

resident not only within the capital but also throughout the hinterland. Approximately 40% of the

ranks had lost either one or both parents before the conflict’s onset, and recruitment into Taylor’s

forces appears to have been driven by acquaintances as over 50% of its members were

introduced by their friends and family. Family protection and loyal support to Taylor’s political

ideals constituted the bulk of the reasons cited for joining, and a method of coercion (abduction or

scaring) was indicated in 41% of the responses. Pecuniary incentives were used as the primary

means of recruitment by the sub-factions under Taylor’s control. Over 80% of Taylor’s forces

indicated that they were combat troops, with a fair distribution of combat support roles identified in

the remaining 20%. In the transition from NPFL to government forces and militias, instruction on

how to treat non-combatants diminished and the sub-units were left to establish their own

operating procedures. Taylor’s troops, both NPFL and GoL, had the highest reported rate of non-

combatant forced labor and were also the most brutal when the civilians did not submit to

supporting their war efforts.

Liberians United for Reconciliation and Democracy (LURD)

The former fighters of the LURD primarily resided outside the Greater Monrovian area before the

war. On average, 60% of the fighters had both parents alive before the conflict. Violent

introduction (54%) was the predominant method employed by the LURD in securing new recruits,

and coercion (41%), through either abduction or scaring, was cited as the primary reason for

joining the group. Pecuniary incentives – job, food, money – were cited the majority of the time as

rewards promised for joining. Significant emphasis during the recruiting process was also placed

upon ensuring family protection and touting the group’s ideological goals. During operations the

LURD maintained the lowest percentage of combat soldiers, when compared with the other


competing factions, within their faction (71%). Dedicated camp workers made up over 15% of the

fighting force. Instruction as to the treatment of non-combatants on the battlefield appears to have

been routinely given. Employment of civilians in a forced labor capacity was also routine, with

very few non-combatants having a choice to participate. Over 72% of the LURD’s membership

reported that the consequences for a civilian who chose not to participate included beatings and

even death. The LURD respondents indicated the highest percentage of foreign fighters within

their sub-factions – almost 80% of the respondents indicated that a foreign fighter was present in

their respective units.

Movement for Democracy in Liberia (MODEL)

The MODEL pre-war disposition identified only 11% of its members as residing in Monrovia. Just

like the other two factions, approximately 60% of MODEL fighters had both parents living at the

commencement of hostilities. The rebel faction had an attraction that induced over 32% of its

members to actively seek them out and join. Violent introductions by the faction were made only

38% of the time for recruitment. Significantly, the fighters of this faction primarily joined to protect

their families (42%), with coercion (35%) playing a much smaller role than was observed in the

other two armed groups. Non-pecuniary incentives topped the rewards scheme for joining the

MODEL. Family protection was at 31% while ideological goals and combined pecuniary

incentives tied at 27%. The primary role during operations most cited by MODEL ex-combatants

was as a combat soldier (93%). Very few of the respondents in the sample acknowledged a

combat support role. Approximately 95% of all soldiers within the MODEL were informed on how

to treat civilians on the battlefield. This faction had the lowest reported rate of civilian

impressments during the conflict and was the most likely to release those non-combatants who

preferred not to support the faction’s efforts with their labor.

3.3 Intervention Program Perspectives

In this study, the sample is broken into four categories of ex-combatants. Three categories of

respondents were at various stages on the DDR process – disarmed and demobilized only,

currently enrolled in reintegration training, and completed reintegration training. The fourth

category includes all respondents who indicated that they did not register with the NCDDRR at all

for any benefits – the “non-participant.” This section will detail the DDR participation rates and

include some perspectives on the mechanics of the program.


3.3.1 DDRR Participation Rates

The approach advanced by this study relies upon obtaining a population of ex-combatants that fit

the distribution profile of the population of Liberian former fighters as reported by the NCDDRR at

the time of the survey. Essential to this is the understanding that to properly assess the impact of

the DDRR program, all categories of intervention must be segregated for analysis. Considering all

DDRR participants as a single homogeneous group artificially skews the targeted outcomes

desired in an assessment and obviates a proper monitoring of the participants engaged in the

process. The temporal nature of the program’s intervention and the uniqueness of not only the

benefits but the socio-economic impacts that influence the lives of the beneficiaries preclude an

analysis that does not disaggregate by intervention type. Given that a representative sample of

the population has been secured through this study, the following findings throughout the rest of

this report should be considered as generalizable to the overall population of Liberian former

fighters.

All former combatants of the Liberian civil war were accepted into this study. Participation in the

DDRR program was not a requisite criterion within the sample and nearly 12% of the respondents

had not registered with the NCDDRR. Figure 27 captures the disposition of the sample with

respect to their category of intervention and locality. As part of the DDRR program, resettlement

preferences were collected during the registration period. While many of the combatants selected

resettlement counties outside the greater Monrovian area, much of the population has been

drawn to the national capital region. The findings indicate an uneven distribution of program

opportunities outside Monrovia. Almost 70% of the sample who had registered with the NCDDRR

and are residing outside the greater Monrovian area had yet to register for a reintegration training

program - even after almost 14 months since the formal portion of the Disarmament and

Demobilization period ended in December of 2004. The disparity in reintegration training

opportunities can be observed against the Monrovian population that is yet to receive training –

only 26%.


Figure 27: DDRR Program Participation

20.39%

42.75%

22.75%

14.12%

Completed Reintegration Training Reintegration Participant

Disarmed/Demobilized Non-participant

Greater Monrovia

DDRR Program Enrollment Status

An examination of the population of former combatants who elected not to register with the

NCDDRR is presented at figure 28. Overwhelmingly, across the three contemporary former

warring factions, the reason most often cited for non-participation was fear of being identified as

an ex-combatant. There is little evidence in this study to indicate that an external barrier to

program entry—e.g., “commander took my gun”—affected program enrollment drastically.

Figure 28: Reasons for Program Non-participation, by Former Faction

0 5 10 15 20Number of Respondents

Commander took my gun

My name was not validated by the JIU

My school/project was not contracted by the JIU

Afraid to apply

Didn't know enough about program

Waiting for the right program to open

Didn't want to be identified as an Ex-combatant

Able to reintegrate without help

NB: 45 respondents

by contemporary faction

Reasons for Program Non-participation

GoL LURD MODEL

3.3.2 DDRR Training Program Perspectives

The national Liberian DDRR program provided former combatants with a wide diversity of

reintegration training alternatives. Figure 29 presents the general types of programs that were

selected by a sub-sample of the population under study. Included are the findings as separated

by the civil war’s most recent antagonists. As observed in the findings, vocational training

opportunities were the most sought after by the former combatants but were followed closely by

formal education enrollments. The MODEL, who appears to have resettled the rural areas of

10.15%

18.21%

61.79%

9.851%



Outside of Monrovia

DDRR Program Enrollment Status


Liberia at a higher rate than its counterparts, has a preference for agricultural as well as for

vocational training.

Figure 29: Participation in Types of Reintegration Intervention Programs, by Major Faction

0 20 40 60 80 100percent

MODEL

LURD

GOL/Militia

NB: 252 respondents

by 2003 faction

Type of Reintegration Training Enrolled/Completed

Vocational Formal Education

Agriculture Civil Works

Over 90% of the respondents who participated in reintegration training believed that the training

they received through the DDRR intervention programs would ensure their sustainability for many

years to come. Of the respondents who completed reintegration training, approximately half

indicated that the work they do today is directly related to the training they received, but a

significant percentage of reintegration training completers (65%) indicated that there weren’t any

jobs to be found or that the work that was available was only sporadic and temporary by nature.

This rate of frustration over employment appears to be centered around formal employment

sector opportunities as the program completers maintain formal employment (employee or

employer) at a rate of 40%, with the remainder being self-employed or actually unemployed.


The focus of this section will be on uncovering gender differences that lie within an assessment of

opportunities provided during the DDRR and will also look into reasons for non-participation.

Figure 30 separates the sample population by gender and category of intervention. Notably,

males appear to be progressing through the DDRR process ahead of the females. Indicators of

this include a training completion rate of 10% (female) versus 18% (male) and a percentage of

those registrants that are enrolled in a reintegration training course at 42% (female) versus 51%

(male). Figure 31, unfortunately, includes only 10 female respondents, but the findings appear to

suggest that ex-combatant identification isn’t as problematic with females as with the males.


Figure 30: Gender – Participation Rates (DDRR)

15.92%

28.87%

42.89%

12.31% 9.244%

28.57%

52.94%

9.244%

Male Female



NB: 590

by gender

DDRR Participation Rates

Figure 31: Gender – Reasons for Non-Participation (DDRR)

0 10 20 30Number of Respondents

Commander took my gun

My name was not validated by the JIU

My school/project was not contracted by the JIU

Afraid to apply

Didn't know enough about program

Waiting for the right program to open

Didn't want to be identified as an Ex-combatant

Able to reintegrate without help

NB: 57 respondents (10 female)

by Gender

Reasons for Program Non-participation

Male Female

3.3.4 Key Patterns

Notably, this section surfaced a few imbalances in the employment of the DDRR program. The

national DDRR program appears to have many beneficiaries who have disarmed and

demobilized outside the Greater Monrovian area and are awaiting reintegration training

opportunities. While the members of this category await enrollment opportunities, females appear

to be progressing through the national program at a much slower rate than their male

counterparts.

With regard to participant registration, the stigma of the label “ex-combatant” weighed heavily in

the minds of those who chose not to register in the program. Interestingly, the non-participation

rate in the Liberian program was found to be generally equivalent to the rate found by the study in

Sierra Leone.


3.4 Post-Conflict Reintegration Determinants

Post-conflict reintegration is a very difficult area to assess. Numerous dimensions of reintegration

are referred to in the literature when looking at post-conflict intervention scenarios. A January

2000 UNDP internal evaluation recommends that the organization “should concentrate its support

to political, social and economic reintegration of war-affected populations on restoring social and

human capital while contributing to political and economic stability.”28 Kees Kingma agrees with

the three components of reintegration highlighted by the UNDP evaluation team but goes further

by noting that there are distinct psychological aspects as well.29 Mats Berdal acknowledges the

importance of both the economic and social dimensions of reintegration but also argues that

within a post-conflict environment these two areas are inexorably linked to political and security

considerations.30 Without an agreement in the literature as to what post-conflict reintegration

should be and in the absence of a definable standard, four selected dimensions of reintegration

are framed and discussed in order to achieve a tangible understanding of the situation in Liberia.

Covered in the following sub-sections will be discussions on the determinants of reintegration as

categorized with social, economic, political, and psychological dimensions.

3.4.1 Social Reintegration

The most widely discussed dimension of reintegration is usually presented in a coupling approach

that presents a socio-economic picture of a post-conflict intervention program’s efficiency. While

the economic side of the coin can be observed in many tangible outcomes, the social aspect of

reintegration is full of more ambiguity and can be fraught with many pitfalls for those conducting

the analysis. In many respects, the social impacts that are measured in reintegration are by many

accounts influenced by economic spillover effects. Figures 32-38 present some key findings that

will hopefully illuminate the social dimension of reintegration as it pertains to the Liberian case.

In any conflict situation, “acceptance” is always a variable pursued to gauge the climate of a

community that is undergoing a period of reconciliation. In this study, two perspectives on this

variable were operationalized. Figure 32 portrays “acceptance” as a self-reported level of the

respondent’s personal feeling while figure 33 introduces a nuance into the questioning by asking

the respondents to give what they believe to be their respective community’s perception of the

population of ex-combatants. In figure 32, the self-reported acceptance rate is separated by

category of intervention. Overall, the percentage of respondents that indicated no problems with

community acceptance is 94%. International intervention appears to have little effect on the lives

of those ex-combatants who indicated experiencing “big problems” within their respective

community. All six of the respondents in the completed DDRR training category who indicated


“big problems” today had also indicted “big problems” immediately after leaving their respective

factions.

Figure 32: Community Acceptance of Ex-combatants, by Category of Program Participant

6.977%

88.37%

1.765%

94.71%

1.515%

95.83%

2.941%

92.65%



Yes, big problems Some problems No problems

NB: 588

Degree that ex-combatants report community acceptance

Ex-combatant Acceptance - Self-reported

Perceived acceptance appears to provide a bit more variance in response by the sample

population. The variable was operationalized and employed with the advice of Mr. Charles

Achodo, the senior UNDP advisor for the Liberian DDRR program. The variable looks to assess

the state of acceptance within communities through a perceived sociological path of reconciliation

that has been observed by Mr. Achodo during his many years of experience in post-conflict

environments. It is interesting to note that while the ex-combatants report that they feel accepted

at a rate of 94%, they also report that they believe their respective communities look upon the

population of ex-combatants “with acceptance” at a rate of only 66%. Two other interesting

anomalies emerge in these findings as well. First is that those ex-combatants who have

completed training programs are doubly disposed to perceiving that the community looks upon

the community of former fighters “with fear.” Second, the NCDDRR non-participants are by far

the most cautious of the community, citing only a 44% acceptance rate and characterizing 46% of

the communities as “watchful or distrustful” of the population of ex-combatants.


Figure 33: Perceived Community Acceptance Level, by Category of Program Participant

19.77%

18.6%61.63%

11.76%

22.94%

65.29%

9.091%

17.05%

73.86%

10.29%

45.59%

44.12%



With fear Watchful/Distrustful With Acceptance

NB: 588

How ex-combatants perceive community acceptance

Ex-combatant Acceptance - Perceived

Figures 34-36 examine the social facet of the physical act of reintegration, the actual rate at which

former combatants have returned to their home communities. As of the time of the survey,

approximately two and half years after the signing of the CPA and the formal cessation of

hostilities, 58% of the population of ex-combatants had returned to their home communities of

record before the war. Figure 34 illustrates significant variation in resettlement across former

factions. The MODEL appears to be the quickest to return at a rate of over 75%, while the former

fighters from the NPFL appear to be drastically averse to returning to their pre-war communities.

Figure 35 examines the issue as a function of DDRR program participation and status.

Interestingly, those respondents who completed a course of reintegration training and those who

have only disarmed and demobilized have returned at a much higher rate than those enrolled in

reintegration training and the non-participants. The non-participants show the slowest rate of

reintegration of any category – 44%. Finally, figure 36 provides insight into the general makeup of

Liberian towns, villages, and cities. In general, the findings with regard to locality and home

community resettlement indicate that about one-third of the ex-combatants outside the greater

Monrovian area could be considered “strangers” within their communities. Alternatively, the

metropolitan area holds a population of near 60% “strangers” to Monrovia.


Figure 34: Home Community Resettlement, by Major Faction

42.7%

57.3%

37.3%

62.7%

25%

75%

54.67%45.33%

GOL/Militia LURD

MODEL NPFL

Not in Home Community Resettled in Home Community

NB: 588

by faction

Ex-combatant Home Resettlement Rate

Figure 35: Home Community Resettlement, by Category of Program Participant

40.7%

59.3%51.76%48.24%

33.33%

66.67%55.88%

44.12%




NB: 588

by DDRR participation/status


Figure 36: Home Community Resettlement, by Geography

29.04%

70.96%

59.84%

40.16%

Not Monrovia Greater Monrovia


NB: 588



Figures 37 and 38 investigate the final two proposed indicators of reintegration, social networks

and community participation. An understanding of the types of people that former combatants

spend their free time with as well as their propensity for engaging in the community are posited


here for consideration. Abundantly apparent in figure 37 across categories is the primacy of family

in the lives of ex-combatants. This fact is significantly demonstrated by the population of ex-

combatants who have registered with the NCDDRR but have not yet enrolled in a reintegration

training program. Interestingly, only 4 of 588 respondents indicated that they spent their free time

with friends from their former factions or the war. Figure 38, which details community participation

levels, indicates little variance between categories of respondents, but it does illustrate that the

former fighters appear to regard community participation as a priority.

Figure 37: Social Networks, by Category of Program Participant

22.09%

51.16%

17.44%

8.824%

51.76%24.71%

4.545%

75.76%

10.61%11.76%

58.82%

14.71%



On my own Family

Friends before the war New friends after the war

Former faction friends Friends from war

NB: 588

by DDRR Participation/Status

Ex-combatant Social Networks

Figure 38: Community Participation, by Category of Program Participant

34.12%

47.06%

3.529%

15.29%

36.9%

43.45%

2.381%

17.26%

31.66%

48.26%

7.336%

12.74%28.36%

46.27%

10.45%

14.93%



Often Sometimes

Seldom Never

NB: 579


Community Participation

3.4.2 Economic Reintegration

Hand in hand with social reintegration is the second dimension considered in this assessment –

economic. Figures 39-40 endeavor to investigate the particulars of an economic perspective. The

first aspect of this reintegration dimension is addressed as a description of the aggregate sample

by daily wage and occupation in figure 39. The figure excludes variable outliers by reporting the


range of wages received by the middle 80th percentile of ex-combatants. Notably, typical

occupations in Liberia do not exceed 400 Liberia Dollars (LD), or approximately US$7.25/day.

Disturbingly is the daily wage secured by teachers – a meager US$1.50 /day.

Figure 39: Wage and Occupation of the Sample Population of Ex-combatants

0 200 400 600 800

Large Scale Trader/Business

Other

Petty Trader

Professional

Unskilled

Trades/Skilled

Rubber Tapper

Farmer

Home Duties

Student

Teacher

NB: Approximate exchange rate 1 US$ = 55 LD

10th and 90th percentile range in Liberian Dollars

Occupations and Daily Wages

10th percentile 90th percentile

Determining employment status in a post-conflict society with little formal economic capacity can

be problematic for investigators. A new construct, however, for deciphering employment status

was introduced in order to better understand the employment opportunities and challenges. The

construct for employment was formulated in coordination with the Bureau of Statistics,

Government of Liberia. A series of six questions based upon a 1999 Government of Liberia

demographic survey31 was designed to assess the respondents according to a particular

employment status. Ten possible categories of employment are used: employer, employee, self-

employed, unemployed, home duties, student, retired, not looking for work, don’t know/other, and

sick/disabled. This construct differs from the Humphreys and Weinstein’s Sierra Leone study as

the latter focuses on distinguishing employment status through asking what occupation the

respondent is in. In the Humphreys and Weinstein variable, a response of “unemployed”

distinguishes the respondents from the various “employed” occupations of farmer, teacher, etc.

With unemployment estimated at 80% in the formal sector,32 a construct that provides more

variation on employment status is critical to analysis.


Figure 40 illustrates the employment status findings by DDRR participation status and by locality.

The aggregate findings for employment status are summarized as follows:

• Formal sector employment (26%)

• Self-employed (31%)

• Unemployed (25%)

• Student (15%)

• Inactive – home duties, sick/disabled, retired (3%)

In the findings that correspond to DDRR participation, those who completed a course of training

show the highest rate of entry into the formal employment sector at 41%, followed by non-

participants at 32%. Non-participants and those who have not yet enrolled in a reintegration

training program are the most prone to unemployment at 29% and 28%, respectively. When

examination of employment status by locality is taken into account, the significance of self-

employed ventures becomes apparent outside Monrovia. Ex-combatants outside Monrovia

reported a status of self-employed 42% of the time, compared with 18% reported in the capital

region. Notably, unemployment is more than 10% less when former combatants choose to work

outside Monrovia.

Figure 40: Employment Status, by Category of Program Participant and Locality

21.62%

41.74%

20.42%

12.01%

22.05%

17.72%32.28%

18.5%

Not Monrovia Greater Monrovia

Employer Employee Self-employed

Unemployed Home Duties Student

Sick/disabled

NB: 590


Excombatant Employment Status

Over 80% of the ex-combatants report a daily wage of less than US$5/day. Almost half of the

sample reported a wage of US$2/day. Figure 41 disaggregates the intervention categories and

illustrates their self-reported daily wage. Reintegration training completers and non-participants

appear to be securing the best wages, although the non-participants also show one of the highest

rates of those below US$1/day. Those registrants who have not enrolled in a course of training to

date appear to be worst off of all.


Figure 41: Daily Income Level, by Category of Program Participant

6.977%

29.07%

43.02%

20.93% 15.29%

26.47%

40.59%

17.65%

17.74%

36.23%29.06%

16.98% 14.49%

21.74%

37.68%

26.09%



Below US$1/day Below US$2/day

US$2 to $5/day Above US$5/day

NB: 590


Excombatant Daily Income

Figure 42 looks to the indicator of home ownership as a facet of the economic reintegration

dimension. A modicum of economic independence should be perceived from those former

combatants able to secure living accommodations with their own resources – renting or

ownership. With this in mind, those who have completed reintegration training appear to do well

with this indicator as over 85% have paid their way in some capacity. Non-participants are the

least capable category in securing their own living accommodations, with over 42% living with

friends or family for free.

Figure 42: Home Ownership, by Category of Program Participant

25.88%

60%

12.94% 13.02%

47.34%

34.91%

39%

30.5%

27.41%

14.93%

43.28%

35.82%



Owner Renter Family(free)

Friend(free) No home

NB: 580


Home Ownership

The final determinant of economic reintegration is presented in figure 43 – education levels. In the

aggregate, only 18% of the sample reported that they had never received any formal education.

Those ex-combatants who had registered but had not enrolled in a reintegration training program

appear to be the least educated, with more than 60% of the sub-sample able to claim only an


elementary-level education at best. Notably, the non-participants include a portion of fairly

educated members (high school-level equivalent or higher) – 26%.

Figure 43: Educational Levels, by Category of Participation

0 20 40 60 80 100percent

Non-participant

Disarmed/Demobilized

Reintegration Participant

Completed Reintegration Training

NB: 589 respondents


Excombatant Educational Levels

No Education Elementary Junior High

High School Some University Vocational

3.4.3 Political Expressions and Reintegration

The political activism of the former combatants as shown in figure 44 is in stark contrast to the

democratic involvement exhibited at the onset of hostilities in 1989 (see figure 3), where over

86% claimed no party affiliation – as of February/March 2006, all but a little more than 2% of the

sample claimed a party affiliation. Of the 30 political parties that registered in the October 2005

elections, 18 were old and 12 were new.33 The UP is led by Liberia’s current executive and

winner of the presidential election, Ellen Johnson-Sirleaf, while the CDC was represented George

Weah’s in the two-person November runoff elections. The Liberty Party (LP) placed third in the

2005 elections with Charles Brumskine serving as its candidate. The National Patriotic Party

(NPP) is former Liberian president Charles Taylor’s old party.


Figure 44: Former Combatant Political Party Affiliations (February/March 2006)

36.56%45.7%

37.3%

54.76%

57.89%

30.26% 27.33%

52.67%

GOL/Militia LURD

MODEL NPFL

Not answered CDC LP

NPP UP None

Other

NB: 590 respondents

by faction

Ex-combatant Current Political Affiliations

The organizational linkages that bound members to their commanders are the second facet of

political reintegration that will be explored by this study. Figure 45 illustrates the degree to which

the ex-combatants in the sample have disassociated themselves from their former commanders.

The question asked how often they had contact with their former commanders for financial or

employment assistance. Significantly, across categories of intervention as well as across former

warring factions, greater than three-quarters of all respondents reported that they never had

contact with their former commanders. Disturbing in the findings, however, is that 8% of ex-

combatants who have completed their reintegration training are in contact “often” with their former

commanders.

Figure 45: Breaking Ties with Faction Leaders

11.96%

82.61%

5.556%

89.68%

18.42%

76.32%

11.33%

81.33%

GOL/Militia LURD

MODEL NPFL

Often Sometimes

Seldom Never

NB: 536 respondents

for financial or employment assistance

Ex-combatant Contact with Former Commander

13.95%

75.58%

14.2%

80.47%

8.333%

88.26%

13.24%

75%



Often Sometimes

Seldom Never

NB: 587 respondents

for financial or employment assistance



3.4.4 Psychological Reintegration

The psychological dimension proved to be a difficult component to address. A construct for “self-

esteem” was employed in this study in an exploratory capacity. The Rosenberg Self-Esteem

Scale (SES), a standard ten-item, four-point Likert scale, was administered to measure global

self-esteem. The resulting score from the ten-item construct provides a quantifiable measure. The

Rosenberg SES “is a unidimensional scale designed to measure only perceptions of global self-

esteem.”34 Originally designed in 1965 for assessing the construct in adolescents, it has been

widely used and accepted for studying adults as well.35 While the instrument has been widely

employed internationally, there are no known uses of it in Liberia. In a report to the USAID that

discusses the psychosocial well-being of youth in the African context, the Rosenberg SES is

highlighted as a very promising psychometric tool.36 The construct has proved to be useful in

other reintegration instigations.37

Unfortunately, the results obtained from the survey were determined to be unreliable as a whole,

but did indeed appear to show some promise when the cases were filtered by enumerator

identification. Overall, a statistical reliability analysis (Cronbach Alpha) yielded a rating for the

construct that was below the threshold usually accepted in the social sciences, and therefore the

decision was made not to advance the findings. Although a tangible finding cannot be presented

here, it is important to share that one of the enumerators appears to have accurately captured the

intent of the construct as his cases, when analyzed for reliability, met the threshold for

acceptance. This particular enumerator received extra training on the delivery of the construct,

and it is believed that insufficient training of the enumerator staff played a significant role in the

inability of the construct to yield reliable findings. More work needs to be done on evaluating the

psychological dimension, but the one small sub-sample (95 cases) indicates some promising

exploration; the self-esteem construct appears to show a strong correlation with numerous

reintegration dimensions.


With regard to gender differences in post-conflict reintegration, a few perspectives are

disaggregated in an effort to uncover any disparities in the process. A brief presentation of

findings in figures 46-48 includes analysis on three of the dimensions covered - social, economic,

and political. The social determinant of “acceptance” is addressed in figure 46. For completeness,


both the “self-reported” and the “perceived” findings are included for consideration. Variation only

appears to surface in how males and females perceive acceptance. Notably, females are

interpreting their community’s interactions with the population of ex-combatants as more

conciliatory than their male counterparts, with over 77% reporting “acceptance” versus 63% in the

male population. Figure 47 contrasts the educational levels of the male and female ex-

combatants in the sample. Strikingly, females reported not having ever received formal education

at a rate twice that of the males. Additionally, only 28% of the females in the sample had an

education above elementary school compared with the males at 57%. Finally, figure 48

addresses the status of ex-combatants in their ability to “break ties” with their former

commanders. The illustration doesn’t show much variation with regard to this variable with the

exception that not a single female respondent reported that she kept in touch “often” with her

former commander.

Figure 46: Gender - Social (Acceptance)

13.01%

23.67%

63.33%

5.882%

16.81%

77.31%

Male Female

With fear Watchful/Distrustful With Acceptance

NB: 588

How ex-combatants perceive community acceptance

Ex-combatant Acceptance - Perceived

Figure 47: Gender - Economic (Education)

15.32%

27.66%57.02%

30.25%

42.02%

27.73%

Male Female

None Elementary Junior High or More

NB: 589

by Gender

Ex-combatant Educational Levels

2.772%

94.03%

1.681%

94.12%

Male Female

Yes, big problems Some problems No problems

NB: 588

Degree that ex-combatants report community acceptance

Ex-combatant Acceptance - Self-reported


Figure 48: Gender - Political (Breaking Ties)

3.419%10.9%

3.632%

82.05%

13.45%

1.681%

84.87%

Male Female

Often Sometimes

Seldom Never

NB: 587

by Gender


3.4.6 Key Patterns

Evaluating post-conflict ex-combatant reintegration is a quite complex endeavor. Four dimensions

of reintegration – social, economic, political, and psychological – have been examined in this

section and have yielded some interesting patterns. With regard to the social dimension,

understanding “acceptance” continues to be challenge, but the construct introduced in this study

which asks for the respondents’ “perceived” view of community acceptance shows promise in

obtaining a degree of variance. In general, both variables illustrated findings that showed that

reconciliation is on the right track in Liberian communities. Resettlement also proved to surface

some interesting patterns by faction. The former Taylor fighters were the least likely to return to

their home communities as of the time of the survey – 43% had not yet returned home. The

LURD reported a rate of almost 63%, while fully 75% of the MODEL fighters had returned to their

home communities – the highest return rate of any faction.

Significant insight has been gained through the incorporation of the construct developed to

determine employment status. The differences in employment status patterns in the urban

Monrovian area vice the rest of the country illustrate the need to plan for varied economic

opportunities. The primacy of self-employed ventures in the lives of former combatants also

became apparent. Politically, across factions and categories of intervention, the former

combatants are actively involved in the politics of their country. They also appear to be breaking

their ties to their former factions with great consistency. Unfortunately, tangible results that

illustrate a linkage between “self-esteem” and reintegration were not achievable in the report, but

the exploratory findings with this construct appear quite promising.


3.5 Conflict and Conflict Resolution

A number of questions were asked of the ex-combatants that addressed sources of conflict within

their respective communities as well as the mechanisms available for resolution. Important to

note within Liberia’s governance structure is a traditional leadership hierarchy that serves as the

mitigator of conflict within most of the communities outside Monrovia and the larger towns. The

general model for conflict resolution is to begin by settling disputes through family; should that fail,

a number of elders may be called upon to resolve the situation. The system of “chiefs” then

provides disputants with another layer of appeal (village, town, Paramount). This traditional

system parallels the police and the courts system.

3.5.1 Sources of Conflict

The survey asked the ex-combatants to evaluate numerous sources of conflict by occurrence

within their communities. Figure 49 provides a snapshot of all sources of conflict and their scaled

scores as pertaining to frequency – the higher the average score, the higher the frequency of

occurrence. The top four reported sources of conflict within Liberian communities were 1) bad

Leadership; 2) crime/lawlessness/lack of prisons 3) land; and 4) unemployment, idleness, youth

dissatisfaction. The sources of conflict reported to occur less frequently include 1) IDP/refugee

return, 2) firearms possession, 3) forced labor, and 4) ex-combatant violence. Significantly, the

indicator of “firearms possession” is also corroborated by a question asked about arms presence

within the community. Only 2.2% of the respondents believed there was an issue with small arms

in their communities. The strength of this finding should not be underestimated because a parallel

study which collected surveys from non-combatants who resided in the same geographic cluster

also indicated that small arms were not a problem at a rate of 2.5%.

Figure 49: Sources of Conflict in the Community

0 .5 1 1.5 2 2.5Frequency of Potential Conflict Source

rated from 'Never=0' to 'Often=3'

Sources of Conflict within Communities

IDP/Refugees Firearms Possession Forced Labor

Ex-com Violence Ethnic/Tribal Illegal Squatting

Big-man Politics Commercial Sex Poverty

Polygamy Idleness Land

Crime Corruption


3.5.2 Mechanisms for Conflict Resolution

As noted above, Liberians rely upon many different mechanisms for conflict resolution. The

traditional venues for resolution (especially in the more rural areas) include family-elder-chief,

while the police and the courts are also available to the general public but in a limited capacity in

some cases. In the aggregate, 88% of all conflict was reportedly resolved at one of the traditional

levels. There was a definite distinction between what was reported in Monrovia vice the rest of

Liberia, however. The “chiefs” were looked to settle disputes 68% of the time outside Monrovia

compared with 2% in the metropolitan region. In place of the “chiefs” in Monrovia, “elders”

assumed the lead role at 52%. Utilization of the courts system at the time of the survey stood at a

meager 2%. Figure 50 illustrates the distribution by category of DDRR participation. While there

seems to be consistency across categories, only one – those who had registered but had yet to

enroll in training – shows any real deviation, and it is their dependence upon the “chiefs” to

resolve issues. Figure 51 presents the findings on the impartiality of the conflict resolution

mechanism. Overwhelmingly, great news is presented – over 96% of the former fighters indicated

that their home community’s mechanism for conflict resolution was either “mostly fair” or “almost

always fair.” Interestingly, the ex-combatants who had completed their reintegration training had

the highest percentage of respondents who appeared to be disgruntled with the process, with

over 8% citing that the mechanisms for conflict resolution were “mostly unfair.”

Figure 50: Judges for Community Conflict Resolution, by Category of Program Participant

19.77%

40.7%

17.44%

17.44%15.29%

42.35%

27.65%

14.12%

6.464%

31.18%

55.89%

5.323% 11.76%

39.71%33.82%

13.24%



Family Elders/Zoe Chiefs

Police Courts

NB: 587

as reported by ex-combatants

Level at which Conflict is Peacefully Resolved


Figure 51: Impartiality of Judges of Conflict Resolution, by Category of Program Participant

35.29%

56.47%

8.235%

38.46%

58.58%

2.367%

26.34%

71.37%

1.145%

31.34%

67.16%



Almost always fair Mostly fair

Mostly unfair Almost always unfair

NB: 583

as reported by ex-combatants

Fairness of Mechanism for Conflict Resolution


A few differences surface with respect to gender in the examination of conflict and its mitigation in

Liberian communities. A disaggregation by gender of the frequency of sources of conflict yielded

little variation, indicating that a generally common perception of conflict is present across gender.

Female respondents also demonstrated a variance in the location of the venues chosen for

resolution. Primarily, females responded that the “chiefs” resolve conflict for them in their

communities at 52% versus 36% for their male counterparts. With regard to the police and courts,

females cited these venues only 5% of the time, while males indicated them 14% of the time.

Finally, figure 52 presents the findings on the perception of impartiality when gender is

disaggregated. While a quarter of the female sample gives its highest endorsement of the

impartiality of Liberian “judging” systems, the male sample puts a full third of its support behind

this testament. Interestingly, not a single female respondent in the survey indicated that the

mechanisms for conflict resolution were “mostly unfair.”

Figure 52: Gender – Perception of Impartiality

33.62%

62.72%

24.37%

73.95%

Male Female

Almost always fair Mostly fair

Mostly unfair Almost always unfair

NB: 587

by gender

Fairness of Mechanism for Conflict Resolution


3.5.4 Key Patterns

Interestingly, with regard to the sources of conflict within Liberian communities, the potential

catalysts commonly associated with post-conflict reintegration were reportedly occurring at a

lesser rate than those sources associated with security and legal issues. While the reported

usage of the police as a mechanism for resolving conflict was reported to be minimal, the former

fighters as a group fully endorsed the impartiality of the “traditional” judges – family, elders, and

chiefs. Significantly, the issue of small arms presence within communities has been reported to

be non-existent by both the population of former fighters and a sample of non-combatants drawn

from the same geographic clusters.

3.6 The Impact of the DDRR Program

The Integrated Disarmament, Demobilization, and Reintegration Standards (IDDRS) suggests

that program evaluations can assess a variety of aspects – relevance, effectiveness, efficiency,

impact, and sustainability.38 This study has made a concerted effort to understand the “impact” of

the program upon the lives of former combatants. The study is inclusive of both types of Liberian

ex-combatants – those who registered with the NCDDRR and those who never formally disarmed

and opted to reintegrate on their own. That said, evaluating the impact of the DDR program is

quite a challenging and complex endeavor. A methodical approach to the presentation of findings

has been presented: 1) characterize the beneficiaries or ex-combatants; 2) review perspectives

on the program and participation levels; 3) examine outcomes based upon four dimensions of

reintegration; 4) document the mechanisms for resolving conflict within communities and gauge

ex-combatant perceptions of impartiality.

3.6.1 Socio-Economic Reintegration

In an effort to operationalize a variable that attempts to account for the many different dimensions

of reintegration into a single indicator, the respondents were asked to indicate how their situation

in February/March of 2006 compared with that of the previous year (a point near the end of the

formal disarmament period). It is believed that a variable such as this will serve well in the

capacity of a criterion (or dependent) variable that will help to make sense of many detailed

“predictor” or independent variables. This socio-economic criterion variable is derived from a one-

item, five-point Likert scale question that relies upon the assumption that the intervals between

the nominal Likert categories are equal. Figure 53 presents a graphical representation of findings

as they correspond to the categories of DDR intervention. The good news is that over 80% of the


respondents indicated an increase in their “overall situation” since the completion of the formal

disarmament period. Notably, those respondents who had completed a course of reintegration

training indicated the largest percentage of improvement in their situation at 89% with 45% or

almost half of the sub-sample indicating that their situation was “much better.”

Figure 53: Overall Situation Compared with One Year Ago, by Category of Program Participant

45.12%

43.9%

7.317%30.43%

52.8%

13.04%

24.81%

52.67%

19.47% 29.41%

44.12%

22.06%



Much Better Better Same

Worse Much Worse

NB: 573


Socio-Economic Status Since January 2005

3.6.2 Assessment by Intervention Category

This section provides a summary “snapshot” that incorporates the preceding discussions on

reintegration as well as the findings of a more rigorous assessment of the data through a higher

order statistical process. The quantitative assessments regarding the impact of the DDR program

upon the lives of the sample of ex-combatants are based upon an ordinary least squares

regression (not provided in this report) of reintegration variables that controls for many of the

social, economic, and political predictors presented throughout this paper. Table 6 provides a

matrix of the variables tested against the categories of DDRR program participation and table 7

presents the results of difference of means testing between program non-participants and two

categories of program participants – program completers and those yet to enroll in a training

program. The empirical results provide an illustration of the difference between categories of

intervention class. They not only demonstrate that each class exhibits unique reintegration

characteristics, but that enrollment into a DDR program does not result in a decidedly linear

benefits progression. Importantly, as table 7 demonstrates, those who disarmed and demobilized

and have yet to enroll in a training program have not fared as well as those who chose not to

register at all (or any other category for that matter).


Table 6: Reintegration Outcomes by Participation Category (Mean Score Comparison Matrix)

(1) (2) (3) (4)

DD only

Enrolled

Complete Non-part.

N Mean SD Min Max (mean) (mean) (mean) (mean)

Daily Income (fig. 41)

590 1.58 0.96 0 3 1.452 1.605 1.779 1.753

Home Owner (fig. 42)

580 0.27 0.44 0 1 0.389 0.130 0.258 0.149

Education Level (fig. 43)

589 1.33 0.77 0 2 1.045 1.576 1.581 1.478

Idleness/Unemployment (fig. 40)

503 0.32 0.47 0 1 0.339 0.278 0.285 0.350

Perceived Acceptance (fig. 33)

588 0.66 0.47 0 1 0.738 0.652 0.616 0.441

Reported Acceptance (fig. 32)

588 0.96 0.18 0 1 0.971 0.964 0.906 0.948

Community Participation (fig. 38)

579 1.98 0.99 0 3 1.988 2.000 2.000 1.880

Home Resettlement (fig 35)

588 0.58 .49 0 1 0.666 0.482 0.593 0.441

Impartiality (fig. 51)

583 2.28 0.55 0 3 2.229 2.349 2.270 2.283

Breaking Ties (fig. 45)

587 2.66 0.79 0 3 2.784 2.603 2.453 2.558

Overall Situation (fig. 53)

573 3.06 0.78 0 4 2.988 3.099 3.304 2.985

Registered ex-combatants yet to receive training benefits

There is a major risk of leaving behind a very vulnerable grouping of NCDDRR registered ex-

combatants – those who have disarmed and demobilized but have yet to receive training. This

category of former fighters is the least educated, most agriculturally oriented, and poorest of the

four classes under investigation. Most important, they have been shown to be the least

reintegrated of all categories under investigation when examined through the lens of the four

proposed dimensions of reintegration.

Ex-combatants currently enrolled in a course of reintegration training

The situation of this category of ex-combatants is only slightly better than that of those who chose

not to participate in the DDRR. The category of combatants is challenged by their economic

situation (42% live on less than US$2/day) and their physical separation from their home

communities (52%). The reason for the separation was unclear in the study but may be related to

limitations regarding training program availability. The question still remains, though, will the

Liberian economy have the capacity to capitalize on this group’s newly acquired education and

training?


Ex-combatants who have completed a course of reintegration training

While this group was found to be the most capable of reintegrating into their respective

communities, the international community must not forget them. Many of the former fighters are

moving on economically, but some are still having a hard time reintegrating socially within their

respective communities. The comparatively lower self-reported acceptance rates, lack of faith in

the communities’ judges of conflict resolution, and large proportion of these former combatants

spending their free time on their own indicate that much more progress is needed.

Non-NCDDRR-registered ex-combatants (program non-participants)

While this group appears to be reintegrating to some degree, there are a couple of significant

areas of concern. Members of this group are the least likely to resettle their home communities

and have also indicated an overwhelming concern about being identified as ex-combatants. The

survey found a DDRR non-participation rate of 12%, which was nearly identical to the finding in

the Sierra Leone case (13%). The significance of the convergence between the two country

cases is quite important to policy makers in itself as it may now be able to be utilized as a

planning factor for future cases. Non-participation is really a non-measurable action, but

convergence should now enable a bit more confidence in the planning stages.

Table 7: Impact of DDR Program Participation (Compared against program non-participants)

Program Completion DD only (not training) (3)minus(4) (1)minus(4) (mean difference) (mean difference)

Daily Income (fig. 41)

0.03 (0.17)

-0.30 (-2.27)**

Home Owner (fig. 42)

0.11 (1.65)*

0.24 (3.78)***

Education Level (fig. 43)

0.10 (0.97)

-0.43 (-3.78)***

Idleness/Unemployment (fig. 40)

-0.06 (-0.80)

-0.01 (-0.159)

Perceived Acceptance (fig. 33)

0.18 (2.18)**

0.30 (4.83)***

Reported Acceptance (fig. 32)

-0.04 (-1.06)

0.23 (1.08)

Community Participation (fig. 38)

0.12 (0.73)

0.12 (0.82)

Home Resettlement (fig 35)

0.15 (1.88)*

0.23 (3.47)***

Impartiality (fig. 51)

-0.01 (-0.14)

-0.05 (-0.76)

Breaking Ties (fig. 45)

-0.11 (-0.70)

0.23 (2.48)**

Overall Situation (fig. 53)

0.32 (2.44)**

0.00 (0.03)

Note: *significant at 10%, **significant at 5%, ***significant at 1%



Significant to the findings of this paper is that no major differences between genders were

uncovered with regard to the impact of the program upon the lives of Liberian ex-combatants.

Figure 54 separates the socio-economic variable by gender for illustration. The findings actually

indicate that the female population believes that its situation has improved at a slightly greater

rate than that of males. The regression model utilized to quantify the evaluation, which also

accounts for gender, did not uncover any evidence of differences between males and females.

Figure 54: Gender – Overall Socio-Economic Situation

30.13%

48.91%

17.25%

3.493%

28.7%

56.52%

12.17%2.609%

Male Female

Much Better Better Same

Worse Much Worse

NB: 573 (473 males, 115 females)

by gender

Self-Reported Socio-Economic Status Improvement


4 CONCLUSIONS

The purpose of this report is to present the initial findings of the study and lay the groundwork for

more detailed and topical analytical work. An evaluation of the impact of reintegration and

reconciliation activities in Liberia has been discussed though a methodical approach that has

dissected the lives of the country’s civil war combatants.

This study has uncovered a modicum of empirical evidence that the DDRR program in Liberia

has indeed enabled a much better life for those ex-combatants who have completed their

program of training when compared with those former fighters who chose not to register,

preferring to reintegrate on their own. In almost every sub-facet of the dimensions of reintegration

(social, economic, and political), the DDRR program completers were decisively more advanced

in their efforts to reintegrate than those combatants that chose self-reintegration or any other

category, for that matter. One important point must be clear with regard to the national DDRR

program. The program was designed not to increase the economic capacity of the country—i.e.,

job creation—but to enable former fighters to reintegrate with the assistance of reinsertion

benefits and training and education. The international community needs to fight the perception

that the DDRR is a “jobs” program when, in fact, the capacity for the Liberian economy to absorb

many of the training graduates into formal sector employment is quite limited.

There is strong evidence that indicates that the presence of small arms has ceased to be an

immediate threat to the peace and stability of Liberian communities, but other sources of violent

conflict are still very prevalent. The socio-economic-political environment is in the process of

healing itself in this post-conflict country, but a serious commitment on behalf of the international

community is, obviously, still very much necessary to ensure a full recovery and to keep Liberia

away from the edge of chaos.


5 ACRONYMS

AFL Armed Forces of Liberia

CAFF Children Associated with Fighting Forces

CBR Community Based Recovery

CESD Community Empowerment & Skills Development Inc

CPA Comprehensive Peace Agreement

CWC Child Welfare Committee

DEX Direct Execution

DFID Department for International Development (UK)

DDC District Development Committee

DDR Disarmament, Demobilization and Reintegration

DDRRP Disarmament, Demobilization, Rehabilitation and Reintegration

Program

DSRSG Deputy Special Representative of the Secretary General

EC European Commission

GOL Government of Liberia

JIU Joint Implementation Unit

LNP Liberia National Police

LURD Liberians United for Reconciliation and Democracy

MODEL Movement for Democracy in Liberia

MIS Management Information System

M&E Monitoring and Evaluation

NCDDRR National Commission for Disarmament, Demobilization,

Rehabilitation and Reintegration

NGOs Non-Governmental Organizations

NTGL National Transitional Government of Liberia

RBM Results Based Management

RRR Reintegration, Rehabilitation and Recovery

SF Strategic Framework

TSA Transitional Safety Allowance

SRSG Special Representative of the Secretary General (United Nations)

UNDP United Nations Development Program

UNMIL United Nations Mission in Liberia (established by Security Council

Resolution 1509 of 19 September 2003)

USAID United States Agency for International Development


Figure 55: Map of Liberia


Endnotes:

3 A downloadable version of the study “What the Fighters Say: A Survey of Ex-combatants in Sierra

Leone” can be found at URL:< http://www.earthinstitute.columbia.edu/cgsd/documents/ humphreys_combatantsurvey.pdf>.

4 United Nations Development Program (UNDP) Liberia, to serve as the principal investigator for

the nation-wide "UNDP Liberia Community Reintegration and Reconciliation Study," 2005-2006. URL:<http://www.lr.undp.org/Ex-combatants%20Nationwide%20Survey%20-%20Key%20Findings.pdf>

5 Jeremy I. Levitt, The Evolution of Deadly Conflict in Liberia (Durham, NC: Carolina Academic

Press, 2005), 4. Hereafter known as The Evolution of Deadly Conflict in Liberia. 6 The Evolution of Deadly Conflict in Liberia, 195.

7 Ibid., 201.

8 Ibid., 216.

9 Ibid., 223.

10

UNOSAA (UN Office of Special Advisor on Africa). 2005. Disarmament, Demobilization, Reintegration (DDR) and Stability in Africa. Conference report Freetown, Sierra Leone. New York: UN. June.

11

LNP (Liberia National Police). 2006. Annual Crime Report and Statistics for 2005. October. 12

IPA, (International Peace Academy). 2006. “Counting What Counts: Ten Steps toward Increasing the Relevance of Empirical Research in the UN system,” Meeting Note, February.

13

Koeb, Eleonora. 2006. Summary Report on Twelve Focus Group Discussions held in Montserrado, Bomi, Grand Gedeh and Sinoe Counties (draft), report for the UNDP Liberia.

14

Kalyvas, Stathis N. 2005. “Warfare in Civil Wars,” in Rethinking the Nature of War, Isabelle Duyvesteyn and Jan Angstrom (eds.). Abingdton: Frank Cass.

15 The actual number of ex-combatants in the Liberian conflict will never be known for certain.

Although the NCDDRR registered over 91,932 adult former fighters, many others will not have disarmed for a variety of reasons and will remain anonymous to any database. In order to quantify the sampling frame, a “best-guess” rate of 13% (or 13,737) was used to determine the number of DDRR program non-participants. The rate corresponds to the finding on non-participation in the Sierra Leone study. Humphreys and Weinstein (2004), 30.

16

STATA release 9 statistical software was utilized for all of this paper’s findings and analysis. 17

JIU (2004). NCDDRR Statistical Summary Report. August, page 3. 18

Dr. Macartan Humphreys provided a coded Microsoft Excel spreadsheet (used in determining the location and number of clusters in the Sierra Leone study) that allowed districts to be selected randomly based upon weights.

19

NCDDRR ex-combatant reintegration intake form FORM-DDRR003A indicates the former combatant’s preferred choice of location resettlement.


20

This mapping project was completed in 2004. The data lists all villages with a unique “P-code” (Place Code) that can subsequently be linked to many other sources of information. The database currently associates the following information as well: county, district , number of houses, percentage of displacement, percentage that have returned, presence and functionality of schools and health facilities, accessibility by car in wet and dry seasons. The database can be accessed at: URL:<http://www.humanitarianinfo.org/liberia/ mapcentre/ reference/village_ws_maps/index.asp>.

21

Humphreys and Weinstein (2004), What the Fighters Say: A Survey of Ex-combatants in Sierra Leone, page 10.

22

Information access through the on-line UNHCR report “Refugees by the Numbers, 2006 Edition,” URL:<http://www.unhcr.org/basics/BASICS/3b028097c.html#Refugees>

23

Draft Interim Secretariat (2003), Liberian Disarmament, Demobilization, Rehabilitation, and Reintegration Programme: Strategy and Implementation Framework. Monrovia. October.

24

NTGL (National Transitional Government of Liberia) (2004), United Nations, World Bank: Joint Needs Assessment, February, 41. URL:<www.un.org/Depts/dpko/missions/unmil/Needs_Assessment_for_Liberia.pdf>

25

Fortnightly Statistical Monitoring Report (22 August 2004) produced by Monitoring and Evaluation Unit (M&E) Unit at the JIU.

26

NCDDRR status reports can be viewed at URL:<http://www.lr.undp.org/ddrr.htm> 27

“Forced labor” is a most euphemistic term. A more correct terminology would be “slavery” or “human trafficking.” Taking civilians as “spoils of war” in Africa has been most common throughout the history of the continent.

28

UNDP (2000), Sharing New Ground in Post-Conflict Situations, 55, January, URL:< www.undp.org/eo/documents/postconflict_march2000.pdf>. 29

Kees Kingma, “Demobilization, Reintegration, and Peacebuilding in Africa,” International Peacekeeping (Frank Cass) 9, issue 2 (Summer 2002), 183.

30

Mats R. Berdal, Disarmament and Demobilization after Civil Wars: Arms, Soldiers, and the Termination of Conflict, Adelphi Paper No. 303 (Oxford: Oxford University Press), 14.

31

Government of Liberia, “1999 Liberia Demographic and Health Survey Field Manual,” 9. 32

US Department of State, “Liberia,” Background Notes, January 2006, URL:<http://www. state.gov/r/pa/ei/bgn/6618.htm>, accessed 14 February 2006.

33

UNDP “Facts and Figures,” URL:<http://www.lr.undp.org/facts.htm>, accessed 3 July 2006. 34

Barbara M. Byrne, Measuring Self-Concept across the Lifespan: Methodology and Instrumentation for Research and Practice (Washington DC: American Psychological Association, 1996), 141. Hereafter known as Measuring Self-Concept.

35

Measuring Self-Concept, 189. 36

Sweta Shah, Georgina Graidage, and Josephine Valencia, “Youth on the Streets: The Importance of Social Interactions on Psychological Well-being in an African Context,” Report for USAID, May 2005, URL:< pdf.dec.org/pdf_docs/PNADD170.pdf>, accessed 16 July 2006.


37

The study presented by Hee-Young Song addressing the social reintegration of persons with spinal cord injury proved to be valuable not only in providing construct recommendations but for methodological design as well. The study implemented “structural equation modeling to examine the direct and indirect effects of self-esteem, social barriers, physical function, family support, informational support, perceived stress, emotion-focused coping, and problem-focused coping on social reintegration.” The relevance of self-esteem to reintegration is of paramount importance, and the model warrants inclusion of this construct in ex-combatant models for further investigation. For more, readers should see: Hee-Young Song, “Modeling Social Reintegration in Persons with Spinal Cord Injury,” Disability and Rehabilitation 27, no.3 (2005), 131-141.

38

United Nations (2006). Operational Guide: To the Integrated Disarmament, Demobilization and Reintegration Standards., 114.

what the fighters say - ropv~what_the...what the fighters say: a survey of ex-combatants in liberia...

Documents