No matter how many creative disarmament initiatives may be developed, none will be successful if equal measures to develop an institutional culture guided by a common national vision within the security sector are not taken.

The concept of security sector used in this monograph is based on the definition provided by Nicole Ball, who identifies the key actors in the security sectors as:

• "Defense and intelligence bodies: armed forces; paramilitary forces; coast guards; militias and intelligence services;
  
  

• Criminal justice organizations: police, judiciary and correctional services;
  
  

• Security-sector management and oversight bodies: legislatures and legislative committees; ministries of defense, internal affairs, justice, foreign affairs; office of the president; and financial management bodies (ministries of finance, budget offices, auditor’s general’s offices); other oversight bodies such as human rights ombudsman, police commissions;
  
  

• Non-core security institutions: customs and other uniformed bodies”1

The violent history of Mozambique has created a culture of violence, aggression and impunity within the security sector, which still pervades security institutions and which needs to be dispelled. Colonialism came to Mozambique with the contours of the fascist dictatorship in mainland Portugal and the security sector was used to cement the regime and destroy any type of eventual opposition. Police, for instance, were used to imposing public order by whichever means deemed necessary. Thus, security forces became the public face of an oppressive regime. The boundaries between army, police and secret service were blurred enough to create the perception that none of these institutions would have security concerns other than the ones connected to the maintenance of the state. Furthermore, this perception, coupled with constant harassment, alienated the population from giving any eventual support to these institutions. For Mozambicans in colonial times the sight of soldiers or policemen could only be an omen of bad news.

In mainland Portugal this image was dispelled when the security forces overthrew the oppressive fascist regime, however in Mozambique the picture was different. Mozambicans saw the colonial administration replaced by a party that was supported by the majority of the population while deeply rooted in military culture. Frelimo became a highly militarized political party partly by force of circumstances and partly by choice.

Once independence was achieved, on 25 September 1975, there was no opportunity for either demilitarization of the Frelimo party, or for a re- structuring of the security sector. Surrounded by the hostile regimes of South Africa and Rhodesia, Mozambique felt justifiably threatened and hence adopted the need for strong armed forces able to defend the country against foreign intervention. President Machel envisioned a professional army but the period of peace was too short to allow for its development. Soon after independence, the then-Rhodesian regime, tapping on internal dissent to Marxist policies and angry at Mozambique’s support of Mugabe, established and supported the Renamo movement in Mozambique. Civil war soon ravaged the country. As such, the culture of aggression instilled in the guerrilla forces during the struggle for independence was never questioned, but rather encouraged. It also meant that the military wing of the party was strengthening its position instead of assuming a subordinate role to the political leadership. The Ministry of Interior and the military were perhaps the most powerful institutions during the civil war and were accountable only to the President; budgets and recruitment, for instance, were never questioned and the activities of the security sector were conducted with little oversight.

The Armed Forces

“What the UN never understood is that you do not dismantle an army; you reform an army. In our case, our army was dismantled.”2 This seems to be the prevailing sentiment among armed forces of Mozambique (FADM) officers.3 But this resentment against ONUMOZ may be partially misdirected. The armed forces themselves and the government of the time contributed equally to the current situation of the military. In fact, instead of contributing towards an effective demobilization of the armed forces and their reform, the government simply shifted great numbers of soldiers to the police force
(outside the ONUMOZ mandate) and let the armed forces, now perceived as less manageable since members of Renamo were being incorporated into the new army, wither.

But the military, by failing to put their own house in order, must also share the blame for inaction. The time may not have been the most suitable for a reform of the security sector, as the militaries of both sides represented the greatest threat to the peace process. Frelimo officers had everything to lose. Salomons provides a very accurate picture of the status within the armed forces:

“... While the leadership in both parties was committed to peace, not all their followers were. The big losers in the process, on both sides, were the military. They had not played much of a role in the peace negotiations; in fact, one could argue that the Frelimo civilian leadership, in framing the scope of the peace accord, had managed not only to eliminate Renamo’s threat but also the crushing burden of its own military... This was the end of a lucrative industry. The Government handed over some $240 million each year to the military in one large lump sum, and the military were not accountable to anyone for the way this money was spent. When we saw the dilapidated barracks, the equipment rusting away in the rain, the lack of preparedness of the Mozambican troops, we wondered where this money had gone – certainly not into any military infrastructure. When we heard the soldiers’ laments about the many months they had not been paid, we realized the money had not gone into payroll either. The Ministry of Finance confirmed that the Army had never submitted a payroll, and that the estimates of the number of soldiers under arms had never been audited.... Clearly, the demilitarization of Mozambique was going to cramp the style of some of the military leaders....”4

On their side, Renamo officers, with generally low literacy rates and few skills other than those pertaining to warfare, feared civilian life and felt uncertain of their role in the new armed forces. Their status during the war was high and their positions and opinions were respected, even if only out of fear. Survival for them and their families was guaranteed as long as the rule of ‘might is right’ prevailed. It is suspected that many had strong ties to smuggling routes and traffickers and thus controlled businesses providing extra income. The General Peace Agreement (GPA) brought an end to this accomodating situation and made them face the possibility of a dubious future in civilian life. For both Renamo and Frelimo, ONUMOZ represented a great opportunity to curtail the power of a sector they may have perceived as too powerful. For the government of Mozambique (GoM), the army had long been a financial burden; for Renamo, the transition from armed movement to political party meant that more power had to be ascribed to its political wing. ONUMOZ offered both parties an opportunity to dismantle their militaries with little fuss and the added bonus of shifting accountability to foreign institutions.

Demobilization and the new armed forces

In spite of this military-political transition, the distrust between the two warring parties was still too recent to allow complete demobilization of both sides and naturally they wished to keep some kind of military leverage. Frelimo could count on the militias it had created during the war and also on the soldiers who were transferred from the army into the police; Renamo excluded some batallions from the demobilization process and had them stationed in remote areas. The suspicion that such batallions exist, persists today – rumours of Renamo soldiers in the area of Maringue abound; during the survey carried out in Chimoio, the research team was often told of a group of two female batallions still stationed around Inhaminga.5 The result created many of the problems the security sector still faces today – disgruntled armed forces with little capacity to protect the territory, albeit overarmed, and a police force overstaffed with men untrained for police service.

The GPA, in Protocol IV, provided for the formation of new national armed forces manned with 30,000 soldiers coming from both sides (24,000 for the army, 4,000 for the air force, and 2,000 for the navy). This target has never been achieved in spite of the new military law that includes mandatory conscription for every Mozambican citizen at 18 years.

Military leaders interviewed during the fieldwork complain that benefits given by ONUMOZ to soldiers being demobilized were better than anything the army could offer and thus most soldiers chose to demobilize rather than join the new army. Research conducted for this monograph on demobilization in Mozambique suggests other motives as well, for example the negative image of the armed forces among the population and the fact that few were given the choice to join the new FADM.

During interviews with demobilized soldiers, many stated that it had been their leaders who had selected those who should be integrated into the new army;6 only a few were given the choice and in the sample all those asked declined the offer. None expressed the wish to return to the military service and most extended this rebuttal to their children, stating that they would not like to see their children doing military service. Most seem to consider the period they spent in the military as a disruption of their normal lives. They also admit that if they were with the military now their lives would probably be easier from an economic point of view, but this perceived advantage does not seem motivation enough to join the military.

Whatever the reasons, the reality is that when ONUMOZ left, what had once been a powerful army in the region was reduced to about 9,000 sergeants and officers and about 3,000 foot soldiers, most of them too old for military life; dilapidated barracks and warehouses; fleets of airplanes and boats unable to move from the places where they were stationed; and huge stockpiles of small arms and light weapons (SALW) with inefficient control mechanisms. These material conditions were complemented with the perception of widely disseminated corruption among the ranks and a public image that instilled fear in the population. Without any meaningful support from donors for the sector, and unable to attract young people into the military, the GoM decided it had little choice but to reintroduce mandatory conscription.

Conscription

Mandatory conscription however is not the best tool to professionalize armed forces. It is estimated that each year 15 per cent of the selected conscripts fail to come forward and present themselves, and approximately 90 per cent of the recruits who do enroll choose to leave the army after the two years of mandatory service. Those who decide to stay are not always the most skilled.7 Most military leaders interviewed by the research team expressed regret at this situation – they know that conscription is resulting from the need to staff an ‘army of generals’ and to man nearly abandoned barracks rather than forming any part of a strategy for reform.

The recourse to mandatory conscription has been controversial – the opposition in parliament questioned the rationality of such an exercise given the scarce financial resources of the Mozambican state. Young Mozambicans were not happy with this measure either and seem to have found ways of evading it. According to the law, every Mozambican citizen must register for military service the year s/he turns 18. The army then puts them through medical tests, selects from those deemed fit and trains them for two years. However, most Mozambicans upon reaching 18 simply do not register and there is no legal mechanism to make them do so—legal sanctions were foreseen only against those who register and then fail to show up when called to join the FADM.

Military registration started with an ‘extradordinary period’ of two months (August and September) in 1998. It was expected that over a million Mozambicans would register but one week before the final date only about 51,634 of the expected number had registered; at the end of the exercise only 140,000 from the expected million had registered. The government had foreseen the inclusion of 3,000 conscripts but ended up with only 1,000, all males.8 This pattern has been repeated each year. In 2003, from the estimated 424,000 Mozambicans turning 18, only 21,000 youth registered.9

Military leaders are aware of the important contributions the security forces can make to the building of national cohesion and identity. “Army structure, with its egalitarian principles, is ideal for reconciliation among fighters,” stated a former Renamo general, currently with the FADM.10 They see the army as an institution able to bring some structure to a society whose fabric has been disrupted by civil war; as an institution with the potential to provide capacity and to integrate young people, who otherwise will have little or no education/training opportunities; as an institution worthy of respect rather than neglect. And they are right in their assessment of what the FADM can be and represent.

Currently the FADM have approximately 15,000 men and women, 9,000 of whom are officers.11 To assist in providing food for these troops, the FADM have recently started a farming programme in some areas. This programme also intends to build capacity among the soldiers. Members of the armed forces will spend 40 per cent of their time with farming and livestock activities and 60 per cent with military activities.12 The FADM are looking for donor support for these activities. However unusual, such intentions are commendable and show a tremendous will within the FADM to overcome the difficulties they face. This initiative has raised some concerns among civil society regarding the legitimacy of using soldiers as labour and may not find sympathy among the international community for similar reasons but does need to be recognized for its approach.

The dilapidated barracks and warehouses are a constant cause of concern, as the explosion of an ammunition warehouse in Beira, which was hit by lightening, testifies.13 These facilities, which once stood alone, are now surrounded by populations and represent a risk their design does not accommodate. But these barracks and warehouses also represent assets that can be used to different purposes and generate some income – they can be sold or rented out to other institutions or private parties, for instance. The amount of infrastructure needed by the FADM will depend on the nature of the FADM itself and as such decisions on the infrastructure – how many, which type and where – should be part of any strategy for security sector reform.

Stockpiles

Regarding armaments, it was not possible to obtain exact figures for the weapons under FADM control. However the central question should not concern numbers of weapons but rather how equipment is recorded, stored and controlled. Interviewees seemed to agree that the FADM are overarmed, in the sense that the firearms handed over by ONUMOZ alone can arm each soldier 10 times over, but no one seems to know exactly to what extent the FADM is overarmed. That is, in quantitative terms the numbers are high but there has been no assessment on the quality of the equipment currently under the supervision of the armed forces. There is a registration system for every weapon or piece of equipment in possession of the FADM but it is a manual system of simple entries in a logbook, viewed as ineffective and unreliable. During the workshop undertaken during the fieldwork, participants of the FADM acknowledged the need to take an inventory of both stockpiles and storage facilities, recognizing that such inventories are integral to a broader strategy for reform.

The Police

If the armed forces present a dismal picture, the police forces, albeit having more support from both government and donors, seem to suffer from an equally damaged image while still struggling with the legacy of the peace process. Although the GPA called for a new army combining men from both sides, such a clause was not included for the police. Moreover, the GPA also stipulated that both armies should be disarmed and demobilized but policing activities should continue under the scope of the PRM – Polícia da Republica de Moçambique (Mozambican Police Service).

The CIVPOL mission

The police thus represented the ideal tool for Frelimo to retain a kind of military leverage, since it could not count any longer on the support of a Frelimo army. Much has been written on the difficulties surrounding the United Nations Civilian Police (UNCIVPOL) mission in Mozambique, whose mandate was “... to monitor the conduct of the indigenous police force during this delicate time in Mozambique’s history”.14 Both the UN and Renamo desired a large UN police presence to guarantee public security whereas Frelimo saw such a presence as an incursion into national sovereignty – public security being the competence of a national government. Frelimo may have been right on this point, but there was another reason to restrict CIVPOL – the transfer of military personnel to the national police had started in 1990, with the peace negotiations,15 and would continue well into the CIVPOL mission in Mozambique. According to Woods:

“Over time, it became increasingly clear to CIVPOL officers that government military troops and equipment were being transferred to the police, especially to the Presidential Guard... CIVPOL officers were sometimes denied access to site where military personnel and equipment were suspected of being converted to police use. In one case, a visiting Security Council delegation was inadvertently shown an undeclared police training camp where police recruits were being trained in machine-gun use. Later, CIVPOL officers were denied access to the camp when they asked to verify the report.”16

The government of the time, motivated perhaps more by political survival rather than by concerns of public security, delayed as much as possible the deployment of the UN civilian police, whose first personnel arrived only in September 1993. Given the lack of cooperation within the PRM and the low capacity of many members of the CIVPOL contingent, when the CIVPOL left in 1994 little had been done to improve the capacity of the PRM or to improve the crime situation. As Martinho Chachiua states “ ... despite the internal security environment, characterized by criminal violence, the need for political survival kept security policies apart from security needs.”17

Transfers from the military to the police

Whatever the circumstances, the reality is that when ONUMOZ left, the GoM had a police force that was more of a partisan body than a modernized institution; overstaffed with members who had military rather than police training; with little regard, or knowledge, of human rights and internationally accepted procedures; unable to stop crime; a police force that the population perceived as a foe rather than as a friend, awash with corruption and operating with impunity. Nevertheless, the transfer of military personnel to the police continued – although no new recruits were hired between 1994 and 1998, the police grew from 18,047 to 21,666.18 This transfer meant that the problems within the police forces, such as low levels of education and inadequate training, were being compounded rather than addressed.

Another outcome of transferring military personnel to the police is that the police budget is spent on salaries for officers that may have never wanted to be police, thereby diverting funds that could have been allocated to equipment and modernization of the police force. It also meant that no new staff could be recruited, as there was no budget to pay additional salaries.

However since 1997 the PRM has been included in several bilateral assistance projects, which are having a visible impact; emphasized by several of the interviewees. Yet not all the problems from the past are solvable with assistance; they require political will translated into practical measures. The PRM has started recruiting new officers in 2003, which is a good sign. These new recruits will be trained in the Police Academy, established with donor funds in 1999.19 The PRM are currently about 20,000 strong giving a ratio of one policeperson per 1,089 inhabitants, which is low for the size of the country (overstaffing in the police forces refers not to the amount but rather to the composition of the police staff).20 However recruiting alone will not be enough and will need to be coupled with a restructuring of resources, including equipment and personnel.

In terms of equipment, the PRM seems to lack many of the things that could turn it into a well functioning institution: the criminal labs have a lower productivity rate than in the 1980s; there are not enough laboratories; of the 321 vehicles nationwide owned by the PRM only 200 are operational.21 In terms of weapons it was not possible to obtain stockpile figures, but the Stragegic Plan of the PRM states: “The PRM has enough armaments for the current amount of personnel, however, the characteristics of the armaments are not adequate to the mission of maintaining law and order.”22 The document notes further that the PRM lacks equipment traditionally used by the police, including batons, handcuffs, and whistles. This seems to imply that in spite of donor support to the PRM, the equipment at its disposal is of a military rather than policing nature. Each PRM station is supposed to have a registry for weapons, both the station’s equipment and privately owned firearms, but this registry is a handwritten ledger book. These records should be periodically sent to the Central Command in Maputo, but no interviewee could specify exactly how often this occurs. The PRM and the Ministry of Interior are the main actors in the control of SALW – they issue firearm licenses and are responsible for overseeing and updating the records. The degree to which this can happen under current resource constraints, however, is unclear.23

National legislation on firearms

National regulation on SALW is important as most illicit weapons were first diverted from legal sources. As such, national laws managing the possession, manufacture and use of firearms must be part of any strategy aimed at reducing the availability of firearms within a country. The Mozambican law on the licensing of firearms dates from pre-independence and was written for another era: it is vague, outdated and bureacratic to enforce.

The legislation regulating firearm ownership – Law 1/73, issued in January 1973 covers firearms and ammunition. It is an elaborate and complicated decree that includes provisions covering knives, daggers and other artifacts. The law sets out classifications for different types of weapons. These include:

• personal defence
  

• hunting
  

• sport
  

• ornamentation
  

• arms other than firearms (knives, daggers, traditional arms), and
  

• war equipment.

According to the legislation, firearms with characteristics other than those defined in the decree are to be classified according to the discretion of the General Command of the Police.

In addition, the law defines:

• rules for import and export of firearms by individuals;
  
  

• procedures for the establishment and running of commercial firearm dealerships;
  
  

• yearly limits for the ammunition sold to individuals;
  
  

• possession limits for individual citizens (up to three firearms: one for self-defence and two from the other categories (hunting, sport) with the equivalent ammunition for one year);
  
  

• rules to register and to control firearms in possession of individual
  citizens and of legal traders;
  
  

• that the President, the Prime Minister, Ministers, Secretaries and Under-Secretaries of State, General Directors and Vice-directors in Ministries, Senior Inspectors, members of Parliament, General and Provincial secretaries, judicial magistrates or magistrates from the Public Attorney, district governors, army officers still in service, as reservists or retired, and staff with the General Directorate of Security while on duty, can own (and use) whichever firearm they wish, without registering them and without aplying for a license (chapter IV; Section I; Article 55);
  
  

• two types of licenses – one solely for possession and one for possession, use and transport of firearms as well as the procedures to apply for such licenses;
  
  

• rules for building and maintaining warehouses for firearms and ammunition, as well as amounts allowed for storage;
  
  

• rules for the establishment of ammunition workshops;
  
  

• rules for transport of firearms and ammunition;
  
  

• how control should be undertaken by the government bodies;
  
  

• sanctions for non-compliance;
  
  

• amounts to be paid by type of license; and
  
  

• responsibilities of the police in keeping records and in controlling compliance with the law.

The decree also contains models for the several types of forms required and for the licenses being issued.

If the applicant’s request to buy a weapon is granted, each year the licensed owner is supposed to report to the police station of the residence area, in order to either confirm or update residential information. Each license must be renewed every two years. In the case of the death of the licensed owner, the survivors must report his/her death to the police station of the residence area and hand in the weapon to the PRM. Records are supposed to be updated with each of these procedures.

The research team was told on more than one occasion that firearms were handed to members of parliament and to members of the government
without making an application; however, this perception was refuted by one officer with the Ministry of the Interior, who stated that “Members of Parliament or of the government also have to apply for firearms.”24 And have they ever refused such applications? “Not so far, but not everybody who is entitled to have a firearm has applied for one.”25 And another officer, present during the interview, added:

“Listen, we try to convince these people they do not need a firearm. In fact, we try to convince anyone applying for a firearm that they do not need one. We do not like the idea of having firearms around. After all, we are the ones directly hit when crimes with guns go wrong. It is in our interest to know what is out there and who has it.”26

Applications for firearms licenses are analysed on a case-by-case basis, motivations are questioned and officers with the Ministry of Interior were visibly reluctant to issue licenses, as they themselves admitted to the research team. Currently, the Ministry of Interior has about 7,000 registered legal owners of firearms.27

The officers interviewed in the Ministry of Interior and in the police forces seemed to be well aware of the difficulties in their jobs. They are committed officers doing their best with the existing conditions, they are well aware of the steps that should be taken and of the measures that should be adopted. They show some embarassment when talking about the shortcomings in their departments, which shows a high degree of devotion; and they discouraged any political interference in their jobs.

Private security companies

This was particularly obvious when the team tried to discuss the control of armament in possession of private security companies. As one police officer told us: “Don’t ask me anything on private security companies because I do not want to go into that. Too many important people are involved.”28 The law on security companies was approved in 1990 at a time when crime rates were increasing in urban areas and, with the war still going on, the security forces were unable to cope. There are currently 31 private security companies registered in Mozambique.

The law in itself is quite simple, although with ample room for interpretation. In essence, it establishes the bureaucratic steps to start a company and sets minimum requirements for hiring staff and oversight by the Ministry of Interior. It excludes private security companies from any criminal investigation and from using systems or methods potentially harmful to citizens.

The decree permits private security companies to protect property and individuals and undertake surveillance activities. Private security companies may also “manufacture and market equipment and other goods concerning private security, in accordance with the Ministry of Industry, Energy and Commerce after consultation with the Ministry of Interior.”29 The law also defines the forms of security (guard, garrison, and patrol) and establishes the rules for licensing and documents to be assembled by the applicant, giving scope to the Ministry of Interior and/or to the provincial governors to require additional information. It sets deadlines for approval and establishes the minimum amounts for financial guarantees. It also states that the licenses cannot be sold or passed over to third parties and establishes fines for those not licensed.

In terms of who may be hired as guards, the law specifies that they must be:

• national citizens with full political rights;
  

• over 25 years old;
  

• able bodied and approved by a team of medical doctors;
  

• without criminal or police records;
  

• and to have completed at least grade seven, or equivalent training (not specified).

Companies hiring staff without these requirements are subject to a fine between 200,000 and 600,000 meticas (equivalent to USD 10-30), but it is not clear if this fine is one-off or if staff in breach of these requirements must be fired. This is an important point, as from the security guards questioned by the research team, most were demobilized soldiers and none had completed grade seven.30 Regarding the training of the security guards, the law stipulates that it is left to the companies to train their own guards, at their discretion. Further articles establish the obligations and duties of private security companies and fines for non-compliance. The law also defines the equipment that private security guards may use, including firearms, dogs, vehicles, uniforms, and identification tags.

Regarding firearms allowed to be in possession of private security guards the law states:31

“1. Private security guards can only possess defence guns when on
duty as body-guards, protecting banks or cash in transit.

2. In terms of this Regulation, defence guns are:

a) semi-automatic pistols with caliber up to 7.65 mm, whose barrel is no longer than 7.5 cm;

b) revolvers with a calibre of less than 9 mm, whose barrel is no longer than 10 cm;

c) semi-automatic rifles of a caliber up to 7.65 mm”32

This article is ambiguous in its coverage of automatic weapons as they are not explicitly excluded. In addition, private security guards were seen with machine guns and carrying weapons on occasions other than the ones mentioned in the above reference.

Article 33 of the law establishes that the police should monitor and survey the activities of private security companies and companies that do not compy with inspections can be fined. The weapons used by private security guards are controlled by the hiring company. Each company undergoes a monthly inspection of its stockpiles by two officers from the PRM.

Some of the interviewed officers provided insight on what the PRM sees as major gaps in the current legislation. According to them, the law should be far more specific on the criteria needed to start a security company and also on the criteria for hired staff. It should establish a minimum salary for armed security guards. The average monthly salary for armed security guards is approximately MZM 800,000 (USD 33), which leaves them vulnerable to bribery or criminal involvement. The law should also regulate the number of hours on duty for armed security guards.

Ultimately, the law should demand proof that the guards know how to handle the weapons they are using. This was identified as a breach in the current legislation both for firearm licensing of individuals and for private security companies. Recently clay pigeon shooting-ranges have opened around Maputo. These premises function without regulation, as there is no Mozambican legislation that provides for this type of activity. The Mozambican police are not against such facilities, on the contrary, they seem to agree that such premises could be useful as learning facilities, should the law be updated. However, they expressed concern at the legal gap in which they operate currently.

Efficient law enforcement needs more than a well-structured and professional police force – it needs also a judiciary system able to complement police activities and able to enforce laws and sanctions. This is not the case currently in Mozambique, where the judiciary system lacks human resources with capacity, survives in degrading infrastructures and is vulnerable to corruption. In spite of the great improvements to the judiciary and the ongoing donor support much needs still to be done.

Notes

1. N Bale, “Security sector reform and good governance in developing countries’, <http://payson.tulane.edu/seminars/Security_sector_files/outline.htm> (October 2003).
   
   

2. Personal interview, September 2003.
   
   

3. FADM – Forças Armadas de Defesa de Moçambique (Mozambican Armed Forces) are the post-GPA army, incorporating fighters from both sides of the civil war.
   
   

4. Salomons, ONUMOZ: The United Nations in Mozambique, 2000, <http:// www.intlmgt.com/publicmanagement/mozambique.htm> (May 2003).
   
   

5. Maringue and Inhaming were Renamo strongholds in the northern Province of Sofala in Mozambique.
   
   

6. This can also be attributed to the fact that most interviewees had been recruited under age.
   
   

7. Personal interview, September 2003.
   
   

8. AIM report No. 143, September 1998, <http://www.poptel.org.uk/mozambique-news/newsletter/aim143.html> and No. 149, January 1999, <http:// www.poptel.org.uk/mozambique-news/newsletter/aim149.html> (February 2003).
   
   

9. Panafrican News Agency Daily Newswire, 8 April 2003, “Young Mozambicans shun military service”.
   
   

10. Personal interview November 2002.
    
    

11. Personal interview October 2003.
    
    

12. AIM, October 2003, posted on <http://allafrica.com/stories/printable/2003101 00566.html> (October 2003)
    
    

13. During the workshop this was a constant concern expressed by the members of the armed forces attending. They mentioned specifically this incident in Beira in November 2002. A live account of it can be found on <http://www.decaturdaily.com/decaturdaily/news/021101/missionary.shtml>. According to this source, “the exploding munitions caused thousands to flee, killed six, injured at least 50 others and destroyed fifty houses within a three-mile radius of the depot.”
    
    

14. J L Woods, Mozambique: The CIVPOL operation, <http://www.ndu.edu/ inss/books/Books%20-%201998/Policing%20the%20New%20Disorder%20-%20May%2098/chapter5.html> (October 2003).
    
    

15. M Chachiua, Internal security in Mozambique: Concerns versus policies, African Security Review 9 (1), 2000, <http://www.issafrica.org/Pubs/ASR/9No1/%20Security Mozambique.html>, (October 2003).
    
    

16. J L Woods, op.cit.
    
    

17. M Chachiua, ibid.
    
    

18. M Chachiua, ibid.
    
    

19. Ministério do Interior, Plano Estratégico da Polícia da República de Moçambique – PEPRM, May 2003.
    
    

20. Ibid. The “normal” ratio would be 1/350 to 1/450 inhabitants.
    
    

21. Ibid.
    
    

22. Op cit., paragraph 5.2.1.4.
    
    

23. The research team spent one afternoon in a department of the criminal police, where three officers were sharing one pen, having agreed on shifts to use it. This situation did not seem abnormal to them.
    
    

24. Personal interview in September 2003.
    
    

25. Personal interview in September 2003.
    
    

26. Personal interview in September 2003.
    
    

27. Personal interview in September 2003.
    
    

28. Personal interview in August 2003 reconfirmed by another officer in September 2003.
    
    

29. Chapter 1, Article 2, paragraph 2.
    
    

30. During the fieldwork from April to September 2003 the research team would ask each security guard we came across how old they were, they been demobilized and had they completed grade 7.
    
    

31. Author’s own translation.
    
    

32. Decree no. 26/90, Chapter VII, Article 20.