Reintegration of Combatants: Were the

Right Lessons Learned in Mozambique?

JAREMEY McMULLIN

Efforts to reintegrate combatants following Mozambique’s civil war concentrated

exclusively on avoiding a return to violent conflict. Though conflict has not
resumed, two challenges to long-term security remain: first, involvement among
certain combatants in organized criminal activity; second, political instability
from the continuing politicization of reintegration issues. Mozambique’s reinte-
gration programme, in aiming only to avoid a return to war, failed to address
these two issues. This has hurt Mozambique and has repercussions for southern
Africa and the international community.

A standard worst-case scenario is often invoked to justify the critical

need to reintegrate ex-combatants into civilian life following armed con-
flict within states. Large numbers of combatants on the various sides of
a conflict are demobilized after a peace agreement. In the absence of
proper care for those soldiers after their demobilization – programmes
to help them survive, find employment and adjust to life as civilians –
they become disgruntled with peace and use their weapons and skills
to re-ignite conflict. Renewed violence initially takes the form of
public disruption and rioting, and then escalates into a return to all-
out civil war.
Tragically, this devastation is not merely hypothetical: the above
model has its roots in experience. In Angola the lack of alternative
employment for troops from the government’s Popular Movement for
the Liberation of Angola (MPLA) and the Union for the Total Indepen-
dence of Angola (UNITA), and indeed of any significant reintegration
efforts from the second UN Angolan Verification Mission (UNAVEM
II) deployed in 1991, left each side still mobilized and well-armed at elec-
tions on 29 – 30 September 1992. Once UNITA denounced the election
results, the slide back into civil war was rapid and bloody, exacting a
devastating human and economic toll.1 Reintegration programmes have
since been designed specifically to avoid this scenario. But such a template
International Peacekeeping, Vol.11, No.4, Winter 2004, pp.625–643
ISSN 1353-3312 print=1743-906X online
DOI:10.1080/1353331042000248704 # 2004 Taylor & Francis Ltd.
626 INTERNATIONAL PEACEKEEPING

can overemphasize the immediate, post-election instability caused by

demobilized combatants. Implementing demobilization and reintegration
programmes (often called DRPs) with an eye solely to avoiding repetition
of the Angolan experience of 1992 obscures the long-term security threat
that certain combatants pose for a state after conflict.
‘Angola Anxiety’ led the UN and observer governments (the United
States, France, Portugal, and the UK) to put a high premium on the reinte-
gration of government and opposition forces in Mozambique during
ceasefire negotiations. The fear of combatants disrupting peace prior
to, or immediately after, elections is reflected in the prominence given
to reintegration support in the General Peace Agreement for Mozambi-
que (GPA) concluded by the RENAMO opposition and the ruling
FRELIMO government in Rome on 4 October 1992. Conflict between
the two groups had destroyed the country during 17 years of civil war.
The demobilization and reintegration programme (DRP) implemented
in Mozambique was the most comprehensive ever attempted at the
time, aiming to ease the combatant-to-civilian transition of about
100,000 fighters. The UN Secretariat and international aid agencies con-
tinue to judge Mozambique’s as one of the most successful war-to-peace
transitions. So do many of those who write about such transitions.2 The
successes of the peace operation are not easily dismissed. Armed hostili-
ties between RENAMO and FRELIMO did not resume after elections
were held in October 1994, after which the UN forces withdrew. Peace
has held in the ten years since. These successes are remarkable because
they came despite delays in the deployment of the UN force, delays in
the demobilization of each side’s troops, an escalation of violence in
assembly areas where troops were eventually cantoned, and riots by
demobilized combatants. At first glance, then, the reintegration issue in
Mozambique seems resolved.
A closer examination, however, suggests limits to the programme’s
success. Stability, though less precarious in 2004 than a decade before,
is still threatened by two issues in particular. First, there is entrenched
involvement in organized crime among certain former combatants
(including trafficking of drugs and arms). Second, the highly politicized
nature of reintegration issues has fuelled further distrust and animosity
in an already highly-charged political atmosphere. These issues have
impeded Mozambique’s development and eroded its stability. They
should prompt re-conception of the lessons learned from reintegration
efforts. Mozambique’s DRP did not address broader threats because it
focused on one aspect of security alone: the avoidance of the worst-case
scenario. If the international community continues to implement reinte-
gration programmes around a model that focuses exclusively on avoiding
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 627
a return to war, peace consolidation as it relates to combatants will be
incomplete.
Mozambique’s experience with ex-combatants demonstrates the
need to re-conceive reintegration to address long-term security threats.
This article provides an overview of the reintegration programme in
Mozambique. It then evaluates the continuing threats to Mozambique
posed by combatants. Third, it will be argued that reintegration
goals, tactics, and beneficiaries need to be re-conceived, and I also
look at obstacles to implementing these changes. Finally, the article
concludes that an exclusive focus on avoiding the worst-case scenario
allowed longer-term security threats to fester in Mozambique.
Moreover, traditionally lower-order security threats (such as drug
trafficking) are increasingly interconnected with higher-order threats
(such as terrorism).

Mozambique’s Reintegration Experience

Security Council resolution 797 established the UN Mission in
Mozambique (ONUMOZ) on 16 December 1992. ONUMOZ had four
main components: political, military, humanitarian and electoral. Con-
ceived under the gloomy shadow of failure in Angola, ONUMOZ’s
mandate stipulated the critical need to link the holding of elections in a
post-conflict Mozambique with the complete implementation of all mili-
tary aspects of the General Peace Agreement (GPA). These were: the
monitoring and verification of the ceasefire; the separation and concen-
tration of RENAMO and FAM (government) forces and their subsequent
demobilization and reintegration; the collection, storage, and destruction
of weapons; the withdrawal of foreign forces (Malawian and Zimbab-
wean contingents) from Mozambican territory; the provision of security
for key transport corridors; the formation of a single, unified army (the
Mozambique Defence Force, or FADM); and the disbanding of private
and irregular armed groups.3
The GPA specifically called for the economic and social reintegration
of demobilized soldiers, but was vague on details. The precise nature of
the DRP was left to the UN, to donors, and other state and non-govern-
mental agencies, under the direction of the UN Office for Humanitarian
Coordination (UNOHAC). A fierce debate on programme design fol-
lowed, which one participant labelled ‘acrimonious and chaotic’.4
There was fundamental disagreement over the concept of reintegration.
For some reintegration should have been limited to removing former
fighters as an immediate threat to peace: reintegration would be achieved
when combatants became ex-combatants. For others, reintegration
628 INTERNATIONAL PEACEKEEPING

should have entailed finding jobs for the demobilized through training
and credit projects.
The former view essentially triumphed, with most resources
(US$35 million of the $60 million total cost of the DRP) supporting a
cash compensation programme, the Reintegration Support Scheme
(RSS). The RSS supported the minimalist goal of providing financial
support to combatants over a fixed period of time, enough to ‘pay
them and scatter them’ over a relatively short period to remove them
from the conflict equation.5 The push for employment assistance did,
however, find some expression in three additional programmes targeting
ex-combatants: the Information and Referral Service (IRS), the Occu-
pational Skills and Development Programme (OSD), and the Provincial
Fund (PF). These, too, were limited in scope, and each had ended by
1997.
A prerequisite for reintegration was the successful demobilization of
RENAMO and FAM forces. The GPA mandated that demobilization
be complete by April 1993, with elections to be held in October of the
same year.6 Troops did not begin to arrive in assembly areas (AAs),
however, until November 1993, and demobilization from the AAs
began in early 1994 and finished in late August of the same year,
leaving about two months before elections were held on 27 –28
October 1994.
Two factors caused the delays. First, the slow deployment of
ONUMOZ forces significantly stalled cantonment of troops, as
RENAMO in particular did not want to proceed without a strong UN
security presence. Second, tactical manoeuvring and a lack of trust
between parties meant that neither side was willing to cede military
advantage and positions – these tactics stalled the identification of
AAs and led to bitter disputes about the numbers of troops on each
side to be cantoned and demobilized. Problems within the AAs them-
selves exacerbated the situation. Soldiers stayed in the AAs longer
than planned, and, particularly for government soldiers, whose salaries
had not been paid and whose demands for increased salaries had been
unmet, the lengthy period of time spent in the camps became unbearable.
Riots and mutinies escalated: six violent incidents were reported in
January 1994, 13 in March, and 36 in May.7 In RENAMO camps
there was also rioting (12, 21, and 31 incidents for the same
months),8 where dissatisfaction was linked to the physical conditions
in the AAs, as the camps were overcrowded and there were serious
food shortages.
Demobilization delays had a significant impact on reintegration: they
increased disgruntlement among both parties’ combatant populations;
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 629
they decreased the willingness of combatants to volunteer for the new
army; and they decreased combatant trust in the institutions and agencies
that were to implement reintegration.
It was in this uncertain context that the reintegration programme began
in early 1994. The RSS was the boldest of the components in size, scale and
duration. Payments to demobilized soldiers varied according to their rank
and were drawn out over a long period of 24 months from demobilization,
with six months provided for by the government and 18 months funded
through donors and managed by the UN Development Programme
(UNDP). The decision to draw payments out over a relatively long
period (unprecedented for reintegration programmes) was meant to
ensure guaranteed income to combatants during their most vulnerable
transition period. The money also helped combatants to win acceptance
of resettlement in their communities, because it provided a steady
source of spending in those communities. The lure of the monthly pay-
ments was also meant to encourage any soldiers not yet accounted for to
come forward, register and demobilize in order to claim the benefits.9
There were some problems with implementation of the RSS, especially
delays in distribution and combatant confusion over procedure. Sometimes
combatants would abuse branch officials of the Banco Popular de Desen-
volvimento (the Mozambican bank that distributed payments), and given
that sometimes upwards of 6,000 soldiers were collecting cheques from
one branch, there was a danger of their anger spilling into violence.10
But discontent was never unmanageable, so ‘pay and scatter’ succeeded
at paying and scattering. Additionally, the IRS ironed out many of the pro-
cedural problems. The service also offered general advice and counselling,
but it never functioned as a job referral service, as was initially planned and
as many combatants expected.
The success of the RSS did not rub off onto the employment-oriented
components of the DRP. The first of these, the OSD, run by the Inter-
national Labour Organization (ILO) developed an employment-training
curriculum featuring 49 courses geared toward skilled and semi-skilled
employment sectors.11 Only a fraction of combatants participated in
OSD courses, and the ILO never performed a market survey to assess
which skills were most in demand. Training programmes were legendary
in Mozambique more for their dark comedy than for their success. For
example, some combatants trained as electricians in villages without elec-
tricity.12 Roughly 70 per cent (or about 6,000) of the programme’s trai-
nees secured employment, but it is unlikely that many of these retained
their jobs beyond six months. If anything, combatant experience with
the OSD may have made the situation worse, raising expectations
beyond what the market could offer.
630 INTERNATIONAL PEACEKEEPING

The second employment-centred component was the PF, a pro-

gramme that intended to address fears that, once the RSS subsidies
ended, there would be no employment opportunities for combatants.
The International Organization of Migration (IOM) and the German
aid agency, Gesellschaft für Technische Zusammemarbeit (GTZ) coordi-
nated micro-enterprise and income-generation activities and offered grants
and technical assistance to businesses and other enterprises willing to hire
or train combatants. The World Bank provided additional funds in 1995
to strengthen the efforts of GTZ and IOM.13 The PF benefited a small
percentage of the total combatant population. Combatants were placed in
arbitrary cooperative groups with little market incentive to succeed. Once
grants ran out, many employers laid off ex-combatants who they had
hired as part of the scheme. Other employers were not employers at all:
they filled out impressive applications only to take the money and run.
Employment programmes highlight a central problem with reinte-
gration: post-conflict states with impoverished economies offer little to
reintegrate into. Mozambique, where only a tenth of the population
had formal employment, was no exception, giving rise to the quip: ‘The
government told us, “Now you are all equally poor. You have been
reintegrated back into basic poverty”’.14 Despite problems in securing
employment, combatants did not return to arms to protest about their
poverty, which suggests that ‘pay and scatter’ indeed averted the worst-
case scenario. Thus, while the overall results of different reintegration
components were mixed, programme failures did not compromise the
central goal: ‘do not repeat Angola’. In post-conflict Mozambique,
however, other threats emanating from problems with reintegration
have come to the fore. These threats have developed outside the frame-
work of the DRP.

Nature of the Peace: Problems

Initially, concern about a return to civil war trumped consideration of
other security challenges. Based on assumptions about how to avoid a
return to war, analysis focused on criminality and violence among the
general combatant population.15 After elections, the IOM catalogued a
number of incidents of insecurity or criminality with confirmed or sus-
pected involvement of demobilized soldiers, former militiamen or other
paramilitary groups.16 The worst violence in the country since the end
of the war ocurred on 9 November 2000 in Montepuez, Cabo Delgado
province, when armed RENAMO men killed ten people in an attack in
the town centre. The police arrested scores of RENAMO men linked to
the attack, and almost 100 later suffocated to death in an overcrowded
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 631
cell.17 Tensions ran high in the community again during municipal elec-
tions in November 2003. Elsewhere during these same elections, there
were allegations and evidence of police harassment of RENAMO suppor-
ters and candidates.18 Although return to civil war seems unlikely,
relations between the former warring parties are hardly serene.
It is, of course, a logical strategy to prioritize avoidance of a return to
conflict in reintegration programmes. Other challenges, however, include
organized criminal activity and the continuing politicization of reinte-
gration: each endangers the stable functioning of the democratic political
process.
Organized criminal activity in Mozambique takes several forms: drug
trafficking, arms trading, money laundering, contract assassinations and
the smuggling of stolen vehicles, human organs, and items illegally
imported into the country.19 Each of these fields of activity depends on
the subversion of the state’s criminal justice system, the buying off of
well-placed officials through bribery and coercion, and the use of vio-
lence. The impact of such activity on state security is acute: ‘Powerful
criminal networks can almost be seen as having created a parallel
power base from which to challenge the structures and capacity of the
state’, leading many within Mozambique to wonder whether it has
become a ‘gangster state’ or ‘an impracticable country’.20 The idea of
failed reintegration sparking a return to war relies on the premise that
combatants will turn to criminal activity and violence in pressing their
economic grievances and that this will spiral into full-scale combat.
The pattern of criminality in Mozambique, however, suggests that this
is not the case, at least for all combatants. Peasants made up the vast
majority of the two armies and lack the connections or wherewithal to
transform themselves from foot soldier to crime boss, nor is there evi-
dence that this has happened. The ex-combatant – criminal nexus is
more apparent, however, among middle and high ranking officers, who
have the stature and connections to be caught up in such activities.21
Their involvement is viewed as particularly significant in the trafficking
of drugs and arms.
Criminals in Colombia, Chile, Spain or elsewhere in Europe direct
most transnational drug smuggling and use Mozambique as a transit
point only. Within Mozambique, much of this is conducted by Nigerian
criminal networks (cocaine) or Pakistanis and Mozambican citizens
belonging to the local Pakistani community (hashish and methaqualone).
These groups rely on the protection of senior members of Mozambique’s
political establishment, including former combatants occupying positions
within the military and police force.22 Officers living in Maputo’s military
quarter were implicated in the trade of drugs for local consumption in
632 INTERNATIONAL PEACEKEEPING

April 2001.23 Trafficking networks also rely on some of Mozambique’s

banks, foreign-exchange bureaux and casinos to launder money, some
of which have been investigated since September 2001 for having links
with the Al Qaeda terrorist network.24
Additional evidence suggests that soldiers are implicated in the drugs
trade in other ways. Surveys of Mozambican military personnel in 1997
show that 14 per cent believed that demobilized soldiers were responsible
for transporting drugs to urban markets in Maputo, and 17 per cent
believed that other military personnel were responsible. Most (41 per
cent) attributed this involvement to ‘trafficking networks’, but 29 per
cent also held former and present military personnel responsible for the
Mozambique networks. Eleven respondents acknowledged being
involved in selling illicit drugs. Those interviewed identified unemploy-
ment and problems with reintegration as major causes of combatant
involvement in drug trafficking.25 Drug trafficking poses serious threats
to Mozambique and the region. Mozambique is the southern African
transit point for South Asian hashish, South Asian heroin, and South
American cocaine destined for South African and European markets,
and it is a significant producer of cannabis for local trade and methaqua-
lone for export to South Africa.26
The abundance of arms in the country also threatens regional security.
ONUMOZ collected just over 200,000 weapons during and after demo-
bilization: none of these was destroyed. Originally, the UN envisaged
sending all collected weapons from AAs to regional depots. The govern-
ment opposed the transfer as differences in the pace of cantonment and
demobilization meant that, initially, more FAM troops had demobilized
than RENAMO troops, and the government argued that transfer of
weaponry away from the AAs would leave FAM troops vulnerable
should RENAMO break the ceasefire. Destruction of weaponry, there-
fore, was not politically viable.27 Still, it is puzzling why the UN did
not advocate weapons destruction once the demobilization of the two
sides had evened out. Some commentators suggest that once it became
clear that the two sides were unlikely to return to war, there was no
motivation within ONUMOZ to give disarmament a high priority and
that the mission shifted its emphasis from military-security matters to
electoral-political ones.28 It was clearly a missed opportunity to destroy
over 200,000 arms, which eventually found their way into the hands of
criminals and which continue to be peddled to willing buyers.
Additionally, the weapons collected represent a fraction of the total.
Even the most conservative estimates place the number of AK47s
imported to the country during the war at between 500,000 and one
million.29 There was no attempt to locate and destroy the arms caches
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 633
of the two sides during the post-conflict transition. Widespread illicit
trading of weapons occurred after combatants dumped weapons at
caches (later to be looted by profiteers) or onto the market.
As thousands of guns and other weapons flooded into the hands of
South African criminals, the South African government labelled the flow
a major threat to regional security. This led to some progress on the
issue when, in January 1995, President Nelson Mandela and President
Joaquim Chissano of Mozambique signed an agreement to increase
cross-border police cooperation to fight the spread of illegal weaponry.
This cooperation produced a series of joint operations, codenamed
Rachel, that succeeded in locating and destroying thousands of
weapons within Mozambique. Thousands, but not hundreds of thou-
sands, and in large part, the damage had been done. Although various
buy-back programmes and ‘tools for arms’ initiatives have tried to
reverse that damage, arms in circulation and arms still buried in the
ground present a significant threat to state and regional stability. South
African intervention on the arms issue highlights the general weakness
of the Mozambican state to confront these issues on its own.
The second long-term threat to stability in Mozambique is the
continuing politicization of reintegration issues not resolved with the
DRP. One cause of this is the exclusion of certain groups from reinte-
gration benefits. The first such group, paramilitary fighters and officially
sanctioned government militias, plays a particularly destabilizing role.
During the ONUMOZ mission, paramilitary groups, jealous of RSS
benefits afforded to demobilized RENAMO and FRELIMO soldiers, pro-
tested and demanded comparable treatment. Sometimes this protest took
violent forms.30 Although such groups lack the power and resources to
return the country to civil war, their alienation from the formal reinte-
gration process and their dissatisfaction with their economic and social
lot has kept reintegration politically contentious. This has contributed
to instability within local communities, and rival groups have clashed,
sometimes violently, during election periods.
RENAMO soldiers have also complained of exclusion from full
reintegration benefits because they are not eligible for pensions.31 The
government argued that RENAMO fighters did not deserve pensions
because they did not have pension allowances deducted from their sal-
aries as did FAM soldiers. RENAMO countered that its fighters had
received no salaries during the war, so there was nothing from which
to make deductions. RENAMO proposed extending pension benefits to
its soldiers, appealing to national reconciliation, but the FRELIMO
majority in parliament voted against it. RENAMO leaders believe that
FRELIMO used the pension issue for political gain to make RENAMO
634 INTERNATIONAL PEACEKEEPING

seem ineffective in providing for its own supporters. RENAMO asked the
international community to support its position, but outsiders refused to
intervene on grounds that it was a purely domestic issue. RENAMO has
charged that FRELIMO uses heavy-handed tactics to undermine the
opposition, which in turn weakens transparency and accountability and
encourages corruption.32 FRELIMO counters that RENAMO is an inef-
fective opposition. Some politicians incorporated the pension issue in
their municipal election platforms in November 2003, and others
planned to use it to secure votes in the 2004 national elections. Part of
this is democracy at work: finding issues that mobilize voters and creating
reintegration policies that better the lives of citizens. The problem,
however, is that keeping a contentious issue alive further polarizes an
already fragile political divide and revives in ex-combatants a collective
grievance that could prove dangerous.
High unemployment rates among former combatants do not help
matters. Despite relatively quick economic growth, few Mozambicans
see tangible improvements in their quality of life, leading many to ask
where the spoils of increased growth are going.33 A widespread percep-
tion is that the spoils of any growth go to party supporters and increase
opportunities for corruption and organized crime.34
Within RENAMO, there are serious divisions and a loss of confidence
in the party’s leadership.35 One RENAMO ex-general comments that:
‘All of us generals who fought during the war expected better. If we do
not get more we will go back to war’.36 The extent to which such
threats could or would be implemented is debatable, but such anger
demonstrates that the parties are not separated by differences of policy
or ideology but by outright animosity.
For the government, the issue is resolved: reintegration is a problem of
the past. Part of this is understandable. A state trying to rise from the
ashes of devastating conflict wants to consolidate its authority. Treating
the reintegration issue as resolved is one way to reinforce that authority.
But this strategy belies a crisis of political representation in Mozambique
and can lead the state to ignore serious warning signs. A high level of
organized criminal activity combined with widespread disenchantment
with government is not a recipe for stability.

Re-conceiving Reintegration
The UN and international aid agencies have yet to make connections
between the DRP of the past and security challenges of the present,
suggesting a gap between the political and development responsibilities
of the UN. The main lesson that UNDP has drawn from Mozambique’s
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 635
reintegration experience is that money and time matter: the RSS success-
fully bought peace during and after the elections of 1994 by giving war-
weary combatants the financial means and the time needed to make
nominal adjustments to civilian life.37 A second lesson is that peace in
Mozambique worked because combatants were willing partners in their
own demobilization. The military (and civilian) population was tired of
war and sceptical of the gains its continuation or re-ignition could produce.
But reintegration cannot be viewed solely in terms of the RSS and a
non-return to war. Mozambique’s trajectory after the GPA suggests
new ways of conceiving reintegration and highlights the difficulties that
the UN and others had in assisting Mozambique with reintegration.
The two sets of lessons are linked, as a re-conception of reintegration
approaches can assist international and state organizations in their
implementation of actual programmes. These lessons can then be used
to ask whether additional steps can be taken to make reintegration
more complete, to counteract long-term security challenges and
enhance long-term development prospects.
To ‘re-conceive’ reintegration in the design of programmes requires
questioning assumptions concerning reintegration goals, tactics, and
targeted beneficiaries. Re-conceiving the goals of reintegration involves
revisiting the debate about whether programmes encourage combatant
demands for special group status. According to the ‘pay and scatter’ stan-
dard, many in the international community judged the lack of widespread
and sustained violence by combatants in Mozambique, and the shift in
resource allocation from short-term emergency aid to long-term develop-
ment assistance, as demonstrating that combatants were becoming
civilianized.
There is evidence that this is true. Combatants feel accepted by host
communities. The ‘combatant’ label is not their primary identifier. But,
they also believe in the value of organizing around interests that they
share and in forming associations to advance those interests. In response
to the idea that ex-combatant organizations are potentially menacing,
they disagree and say that they made unique sacrifices and deserve
special treatment from the government.38 In some ways, the formation
of interest groups is natural to democratic politics. Veterans of many
wars in many states continue to demand special treatment long after a
war has ended, whether through economic assistance or symbolic recog-
nition. Why should the international community expect combatants of
civil wars to behave differently?
If reintegration is depoliticized then it is unproblematic when comba-
tants group together or desire special group status (similar to veterans
groups elsewhere). But, if reintegration remains highly politicized, as in
636 INTERNATIONAL PEACEKEEPING

Mozambique, then group identification presents problems. It is less pro-

ductive to centre reintegration design on whether or not combatants
should be treated as a special group. Some group concessions, like cash
compensation programmes, are always necessary to win support of comba-
tants as key conflict actors, and combatants are naturally prone to group
identification. Instead, the international community could apply political
pressure during the peace process to grant equal access to reintegration
benefits in order to depoliticize the issue. This does not mean giving com-
batants excessive concessions that could alienate civilian populations by
creating resentment; it implies emphasizing actions that can make reinte-
gration less politically volatile between competing groups.
The tactics of reintegration also need to be rethought. This involves
using a combination of ‘carrots and sticks’ to link reintegration to the
security requirements of the peace agreement. Reintegration is not just
a process of granting benefits but of expecting compliance as well.
In the light of the Angolan experience, the programme in Mozambique
neglected the ‘sticks’ and placed excessive reliance on the ‘carrots’, as
the failure to link reintegration to disarmament demonstrates.
A strong UN or regional role in disarmament efforts was needed to
prevent the flow of arms from combatant to black market outlets during
the reintegration process. Such a role was feasible for ONUMOZ during
its mission, but it was not pursued. Under its mandate the UN could
have taken additional action by destroying seized weapons or seeking
out undeclared or illegal arms caches. It could have empowered military
personnel to seize weapons transported in vehicles across borders;
instead, ONUMOZ troops could only watch and note license plate
numbers as scores of weapons flowed from Mozambique to other areas
in the region.39 The UN took a narrow view about the duties of the
armed ONUMOZ contingent, both to avoid confrontation between the
parties and because it considered disarmament to be a function of those
administering the AAs and not the troops stationed along security corri-
dors. Using a combination of carrots (in the form of buy-back pro-
grammes) and sticks (disqualification from reintegration benefits for
combatants caught selling or transporting arms illicitly, prosecution, or
more aggressive reporting of such transactions as violations of the cease-
fire), the UN could have averted some of the problems now associated
with disarmament.
Finally, re-conceiving reintegration requires changing the way benefits
are distributed, by focusing on the combatants most likely to disrupt the
peace: middle and high ranking officers. Combatants are not all the same.
They have different skills, experiences and expectations. Programmes
must therefore disaggregate ex-combatants, addressing the needs of
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 637
different groups of demobilized soldiers. Among ex-combatants in
Mozambique, the higher-ranking soldiers have tended to fall into two
extremes: those who reaped huge economic rewards through connec-
tions, where giving in to criminality is ever present, and those who com-
manded large battalions only to find themselves now living in poverty.40
To reach a more satisfactory outcome in Mozambique, the UN needed
to actively counteract elite ex-combatant involvement in criminal and
illicit activity. To counteract the economic lure of illegal activity,
implementing agents could sweeten the deal on offer to those combatants
most likely to become spoilers. This approach was proposed by
Aldo Ajello, the UN Secretary-General’s Special Representative in
Mozambique, shortly after his arrival in the country in 1992. He
suggested that senior officers and generals could be given shares in
newly-privatized industries to buy their participation and reduce any
incentive to become involved in illicit pursuits. The donors did not
respond to the idea, however, both because they did not know how to
make it operational and because they disliked the prospect of giving the
most brutal of the soldiers the most generous packages.41 The problem
is that much of Ajello’s plan came to pass anyway, but without any over-
sight, through cronyism and, sometimes, corruption.
If buy-off programmes are not palatable, then training or micro-credit
programmes could be targeted directly at this class of soldier (where they
may be more likely to succeed, given different levels of education and
expertise, than in the selective targeting among all combatants). Direct
negotiation with senior officers and generals could have ameliorated the
problem as well. Finally, reintegration efforts should not rule out negative
incentives to deal with criminal activity during and after the transition,
including disqualification from benefits or prosecution. These blows
could also be softened through conditional amnesties or rehabilitation
programmes.
DRPs are designed on the assumption that combatants want to be
reintegrated following a conflict. To consider that certain incentives
exist for some combatants not to reintegrate requires changing a
number of assumptions held in the international aid community, not
least of which is a shift from focusing on a passive programme of
development stimuli to a more active one of combating involvement in
criminal or illicit activity within certain segments of demobilized soldiers.

Obstacles to Re-Conceptualization
Even if DRPs are re-conceived for reintegration goals, activities, and
distribution of benefits, there is no guarantee that such changes can be
638 INTERNATIONAL PEACEKEEPING

successfully implemented. First, aggressively enforcing the rule of law and

cracking down on organized criminal activity requires a strong police
force. ONUMOZ had difficulty supporting Mozambique’s existing
police force. The GPA charged the FADM with cooperating with the
country’s Police Command to protect civilian inhabitants against crime
and violence of all kinds. But the FADM is short-staffed, having attracted
only about a third of the 30,000 soldiers envisaged in the Agreement,
leaving the country’s new military in no position to fulfill the police-
assistance role given to it by the GPA. This led the UN to consider
more policing. Although, in 1994, ONUMOZ eventually added a civilian
police (CIVPOL) component and the Security Council authorized a force
of up to 1,114, the Council also made it clear that the creation of CIVPOL
should accompany a reduction in the military component in order to keep
costs the same.42 Total manpower was not increased. Additionally,
CIVPOL’s activities were passive, limited to monitoring police neutrality
in the provinces, and never intended to contribute robustly to the conso-
lidation of the rule of law. The lack of a strong police presence allowed
organized criminal activity to flourish without significant challenge.
Second, the UN’s duty to avoid interference in ‘domestic’ matters
within states has made it difficult for the organization to depoliticize
reintegration in a highly-charged political atmosphere. Incentives,
especially in the form of money, are almost always welcomed by post-
conflict governments, but punitive or enforcement measures are out of
the question unless there is consent. Reluctance to intervene on issues
typically considered to be part of a state’s sovereign control, such as
domestic criminal activity and corruption, meant that the UN main-
tained only a consultative role through its agencies on these issues.43
This is not an atypical stance for the UN, but inaction led to an atmos-
phere of impunity, and compounded the weakness of the state to resolve
law and order issues during the crucial transition period. A continuing
failure to resolve such issues threatens to compound a cycle of violence
that could threaten stability on a larger scale. The costs of inaction,
then, outweigh whatever political goodwill is bought through a
general deference to the disputants, and suggests that the UN could
deal more aggressively with parties to civil conflicts in negotiating the
post-conflict security of the state.
Finally, basic organizational and operational limitations of the UN
lead to a prioritization of emergency measures over long-term security
concerns. Problems with troop deployment and limited funding, commit-
ment and political will on the part of troop-deploying countries, and
coordination of humanitarian efforts during operations with numerous
UN agencies and NGO and state partners were obstacles not unique to
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 639
Mozambique in the 1990s. These limitations cause the UN to emphasize
exit strategies and to dedicate resources to short-term efforts designed to
guide a post-conflict state to its first elections and no further. Host
countries are concerned about the UN outstaying its welcome. But the
focus on security immediately before and immediately after elections
can ignore long-term threats to state stability, particularly by individuals
who come to see themselves as losers in the distribution of post-conflict
spoils. Reintegration in Mozambique aimed to buy peace in the short
term and expected that local reconciliation and broader development
processes would stabilize that peace in the long term. It is not clear that
this stabilization occurred. UN strategy treated the 1994 election as an
end in itself, and withdrawal shortly thereafter meant that a long-term
investment in the security and development of the state was rhetorical.44
The price of an exclusively short-term approach may well be a subsequent
breakdown in law and order among elite segments of the ex-combatant
population and, if dissatisfaction remains unchecked, the possibility of
political conflict among a broader segment of the population later.

Conclusions
The UN Secretary-General called ONUMOZ ‘a major success story in
UN peacemaking, peacekeeping and humanitarian and electoral assist-
ance’.45 Certainly, in the wake of the Angolan disaster in the early
1990s, the intervention in Mozambique stands out as comparatively
more successful. But in the reintegration context, more could have been
done to anticipate long-term security challenges. For some states, the
threats emanating from criminality constitute mere ‘law and order’
concerns. But for Mozambique, as for many developing countries, they
threaten the stability of the state itself. The effects of a breakdown in
the rule of law are more acute in a state where the legitimacy and
authority of governmental institutions are already tenuous. Domestic pol-
itical conflict regarding a highly-charged issue like reintegration is often
interpreted by the ruling party not as spirited debate or criticism but as
a threat to the survival and cohesion of the nation. Combined with
persistent unemployment and the ready availability of arms, allowing a
breakdown in the rule of law could be merely a recipe for postponing
the worst-case scenario. The cross-border manifestations of such a break-
down threaten regional security, as has been the case with Mozambique
and its neighbour, South Africa.
The fundamental challenge in Mozambique was to create a durable
security environment. ‘Pay and scatter’ took a least common denominator
approach to development, which was logical given the limitations of
640 INTERNATIONAL PEACEKEEPING

targeted employment schemes. But it also took a least common denomi-

nator approach to security, where reintegration meant buying off success
in the short term. Failure to encourage the full participation of the former
military elites in the project of peace and democracy allowed corruption
and criminality to go unchecked, which in turn creates an atmosphere
where more dangerous security threats can develop. Failure to guarantee
equal treatment of combatants in the ten years since the end of war has
soured relations between and within the two main parties.
For the UN and its partners, a preference for reintegration in the short
term rather than investing in long-term solutions also calls into question
their credibility as peacebuilders. More states could choose reconstruc-
tion without UN involvement (as Angola decided in 2003 to manage
reintegration of about 105,000 UNITA troops in the wake of their
leader, Jonas Savimbi’s, death). Or coalitions intervening in other states
can dictate the terms of reconstruction, demobilization and reintegration
(as in Afghanistan and Iraq). These are not necessarily desirable
approaches, as the UN tends to have more operational expertise and
more experience accessing and mobilizing donor funds than war-torn
societies or external forces. The UN tendency to withdraw after post-
conflict elections signals a frail commitment to the long-term security
of states that lack the means to consolidate the rule of law on their
own. It calls into question the extent to which elections alone should
justify a full transfer of sovereignty to newly-reconstituted states.
For security, a failure of reintegration programmes to recognize the
importance of threats other than all-out war comes at a time when tra-
ditionally lower-order security threats (such as drug trafficking) are
increasingly interconnected with higher-order threats (such as terrorism).
The nexus between the drug trade and terrorism in the Tri-Border region
of South America, for example, has been a cause of concern for the United
States. Will alleged links between criminal networks in Mozambique and
their external supporters in Pakistan or Colombia exacerbate the syner-
gistic trend? It is increasingly clear that it is a mistake to marginalize
security problems short of immediate resumption of civil war. Mozambi-
que’s experience with reintegration demonstrates that a prolonged and
generous subsidy can placate and disperse the majority of demobilized
soldiers, but is less capable of preventing certain combatants from threa-
tening stability in the state and region, or of diluting the ability of political
leaders to use incomplete reintegration to stir up old grievances.
That combatants did not return Mozambique to war may be gratify-
ing. But unless reintegration can encompass other security challenges that
certain ex-combatants pose, then states and international organizations
are left with an outdated template. In this sense the Angola of today is
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 641
perhaps more instructive than the Angola of 1992. UNITA representa-
tives would not plunge the country into civil war again, at least not as
UNITA. But they warn that there will be ‘banditry, more organized crim-
inal activity, ambushes and similar violence if [reintegration] needs are
not addressed’.46 Call it ‘Angola Anxiety Redux’.

NOTES

1. The Secretary-General reported that, within three weeks of the elections in 1992,
only 41 per cent of government and UNITA troops had been demobilized, ‘Report of
the Secretary-General on the United Nations Angola Verification Mission II’, UN
doc. S/24245, 7 July 1992, p.10.
2. See, for example, Kees Kingma (ed.), Demobilization in Sub-Saharan Africa: The
Development and Security Impacts, New York: St. Martin’s Press, 2000.
3. ‘Report of the Secretary-General of the United Nations Operation in Mozambique
(ONUMOZ)’, UN doc. S/24892, 3 Dec. 1992, p.5.
4. Author’s interview with Timothy W. Born, Team Leader, Private Sector Enabling
Environment, USAID Mission to Mozambique (formerly chief coordinator for
USAID’s involvement in demobilization and reintegration activities), Maputo, 18
Sept. 2003.
5. Ibid.
6. After a visit from Secretary-General Boutros Boutros-Ghali on 22 October 1993, the
government and RENAMO agreed to reschedule the implementation of the GPA,
and consequently the date of elections. The UN saw this as a necessary step to avoid
elections taking place in a context where both parties were still mobilized and
armed, and therefore to avoid the result that occurred in Angola.
7. Joao Paulo Borges Coelho and Alex Vines, ‘Pilot Study on Demobilization and Re-
integration of Ex-Combatants in Mozambique’, Oxford: Refugee Studies Programme,
1994, p.16.
8. Ibid.
9. Eric Berman, Managing Arms in Peace Processes: Mozambique, Geneva: United
Nations Institute for Disarmament Research, 1996, p.82.
10. Authors’ interview with Gareth Clifton, former Northern Region Coordinator for the
Reintegration Support Scheme, Maputo, 19 Sept. 2003.
11. Chris Alden, ‘Making Old Soldiers Fade Away: Lessons from the Reintegration of
Demobilized Soldiers in Mozambique’, Security Dialogue, Vol.33, No.3, 2002, p.345.
12. Author’s interviews with 15 ex-combatants from RENAMO and FRELIMO in urban
and rural districts, Maputo and Moamba, 23–26 Sept. 2003.
13. World Bank, ‘War-to-Peace Transition in Mozambique: The Provincial Reintegration
Fund’, Findings (Africa Region), No. 90, July 1997, accessed at www.worldbank.
org/afr/findings/english/find90.htm).
14. Interviews with ex-combatants (see n.12 above). For a detailed account of the economic
devastation caused by Portuguese colonialism, the Cold War, Renamo destabilization
tactics, and structural adjustment programmes imposed by the International Monetary
Fund, see Susan Willett, ‘Ostriches, Wise Old Elephants and Economic Reconstruction
in Mozambique’, International Peacekeeping, Vol.2, No.1, 1995, pp.34–55.
15. Willett, for example, warned about the potential for warlordism resulting from the col-
lapse of social and economic order, ibid., p.48.
16. International Organization for Migration, ‘After One Year: What is the Status of
Reintegration in Mozambique?’ Maputo: Information and Referral Service/Provincial
Fund for Demobilized Soldiers, May 1996.
17. Joe Hanlon, ‘Local Election Email Special Issue 1’, Mozambique Political Process
Bulletin, Maputo: European Parliamentarians for Africa, 16 Nov. 2003, accessed at
642 INTERNATIONAL PEACEKEEPING

www.jiscmail.ac.uk/cgi-bin/wa.exe?A2 ¼ ind0311&L ¼ Mozambique-study-group&

D ¼ 1&T ¼ O&F ¼ &S ¼ &P ¼ 414).
18. Ibid. The municipal elections were hardly serene: Renamo and Frelimo contested results
in multiple municipalities, and the Constitutional Council, although validating the ulti-
mate results, harshly criticized the National Elections Commission for committing
elementary errors and for violating ‘the principles of security, stability, confidence and
transparency that should guide the electoral process.’ See Joe Hanlon, ‘Constitutional
Council Validates Local Elections’, AIM Report No. 268, Maputo: Mozambique
News Agency, 19 Jan. 2004, accessed at www.jiscmail.ac.uk/cgi-bin/wa.exe?A2 ¼
ind0311&L ¼ Mozambique-study-group&D ¼ 1&T ¼ O&F ¼ &S ¼ &P ¼ 182).
19. For a detailed investigation into the dynamics of each of these activities see Peter
Gastrow and Marcelo Mosse, ‘Organised Crime, Corruption and Governance in the
SADC Region’, presented at the Institute for Security Studies Regional Seminar,
Pretoria, 18–19 Apr. 2002.
20. Ibid., pp.10, 17.
21. Perceptions of these connections are rampant in Mozambique. Sometimes allegations
are overblown, and direct evidence is not always publicly available. Nevertheless,
interviews with journalists, researchers and community leaders conducted in
Maputo, 15–26 September 2003 confirm that some links exist. Additionally, inter-
views with former government officers conducted by other researchers have produced
the same conclusion (see Alden, n.11 above, p.350). An atmosphere of fear developed
in the wake of the murder of Carlos Cardoso, a journalist assassinated in 2001 while
investigating the disappearance of US$14 million from the Commercial Bank of
Mozambique. Before his death, Cardoso made similar claims in articles and interviews
about the complicity of high-ranking government officials (including military
personnel) in criminal activity.
22. Gastrow and Mosse (see n.19 above), pp.5–8.
23. Ibid., p.5.
24. Ibid., p.9.
25. UN Office of Drugs and Crime, The Drug Nexus in Africa, Vienna: UNODC, Mar.
1999, p.101.
26. United States Central Intelligence Agency, World Factbook 2002: Mozambique, 1 Aug.
2003, accessed at www.cia.gov/cia/publications/factbook/geos/mz.html).
27. Berman (see n.9 above), pp.46,48,87–88; Martinho Chachiua, ‘The Status of Arms
Flows in Mozambique’, in Tandeka Nkiwane, Martinho Chachiua, and Sarah Meek
(eds.), Weapons Flows in Zimbabwe, Mozambique, and Swaziland, Pretoria: Institute
for Security Studies, monograph 34, Jan. 1999, p.26.
28. Martinho Chachiua and Mark Malan, ‘Anomalies and Acquiescence: The Mozambi-
can Peace Process Revisited’, African Security Review, Vol.7, No.4, 1998, accessed
at www.iss.co.za/Pubs/ASR/&No4/Anomalies.htm).
29. Christopher Smith, ‘Areas of Major Concentration in the Use and Traffic of Small
Arms’, in Jayantha Dhanapala et al. (eds), Small Arms Control: Old Weapons, New
Issues, Aldershot: Ashgate, 1999, p.107; Graduate Institute of International Studies
(Geneva), Small Arms Survey 2001: Profiling the Problem, Oxford: Oxford University
Press, 2001, p.64.
30. Berman (n.9 above), pp.76–7.
31. Disabled RENAMO veterans are entitled to pension benefits under a separate scheme,
but the bureaucratic process involved is too complicated for many disabled veterans,
and not all disabled RENAMO fighters know that they are eligible for these benefits.
32. Author’s interview with Raul Domingos, founder and president of the Democratic
Institute for Peace and Development (IPADE) and former unofficial ‘second in
command’ of RENAMO and chief RENAMO negotiator during the Rome talks,
Maputo, 19 Sept. 2003.
33. Brian Slattery, ‘Development without Equality: An Interview with Raul Domingos’,
Journal of International Affairs, Vol.57, No.1, 2003, pp.129–34.
REINTEGRATION OF COMBATANTS IN MOZAMBIQUE 643
34. Interview with Domingos (see n.32 above). Domingos makes the same point in Slattery,
ibid.
35. RENAMO performed much poorer than expected in the municipal elections; and, after
having been expelled from the party earlier, Raul Domingos formed a third party to
challenge RENAMO candidates.
36. Author’s interview with General Vareia Mange (‘Chimuanza’), Costa do Sol, Maputo
Province, 15 Sept. 2003.
37. UN Development Programme, Report on the Reintegration of Demobilised Soldiers in
Mozambique 1992–1996, Maputo: UNDP, 1997, p.27.
38. All combatants interviewed by the author shared this view (n.12 above).
39. Dennis C. Jett, Why Peacekeeping Fails, Basingstoke: Palgrave, 2001, p.84.
40. RENAMO officers, such as General Vareia, have tended to fall into the latter category,
unless they joined the FADM.
41. Interview with Born (see n.4 above).
42. United Nations Security Council, Resolution 898, UN doc. S/RES/898 (1994), 23
Feb. 1994.
43. The lack of a more active police role was also due to resource constraints, a desire to
show trust in the government, and the lack of a precedent for more involved police
intervention.
44. Willett (see n.14 above), pp.48–9.
45. The United Nations and Mozambique 1992–1995, UN Blue Book Series, Vol.V,
New York: UN Department of Public Information, 1995, p.3.
46. International Crisis Group, ‘Dealing with Savimbi’s Ghost: The Security and Humani-
tarian Challenges in Angola’, Luanda and Brussels: ICG, Africa Report No. 58, 26 Feb.
2003, p.6.