Introduction

This chapter is based on the presentation and analysis of an exploratory street survey carried out in the city of Chimoio, northern Mozambique, in April 2003. The original purpose of the survey was to gather basic data on the perception of local communities on security, police performance, and the presence of firearms in Mozambique.

As a pilot study, however, the specific purpose of this research exercise was to pilot the consistency to local realities of the survey, which was developed by the Institute for Security Studies and which has been carried out in other African countries, namely in several regions in South Africa and nationally in Tanzania.

There are several advantages in using surveys previously designed and tested– they have been used and the data analysed and they enable the gathering of similar data, enabling comparative analysis. The danger of using these surveys can be their inadequacy to different socio-economic contexts. It is therefore important to test these surveys with a sample of the target population and adapt them accordingly before starting the actual broader research.

Background

A particular set of circumstances including the proximity to the border with Zimbabwe, its placement along a transport corridor, being in a politically polarized region with a relatively small population made Chimoio the selected location for an exploratory study on security issues in Mozambique.

The province of Manica is crossed by the Beira corridor, which is the main rail and road connection linking the Indian Ocean and the east with the interior of southern and central Africa. It is a fertile agricultural region, producing among others maize, bananas, citrus, and cotton and home to a range of industries, including textiles, saw mills, processing of cotton and sisal. Relative to the rest of Mozambique, Manica can be considered a prosperous province.

The economic potential of the province and the transport corridor that crosses it turned this region into a region of strategic importance during both conflicts. The province was stage to violent clashes between Frelimo and Zimbabwe African National Union – Patriotic Front (ZANU-PF) fighters and the then - Rhodesian and Portuguese armies. Later, the contenders were to be Renamo fighters against soldiers from both Mozambique and independent Zimbabwe.

During the colonial administration, income to Mozambique was provided by labour and transport. If defending the Beira corridor was paramount for the colonial government, Frelimo could not afford not to attack it. The presence of a large settler community in the province was an added bonus. Attacking the region meant not only economic disruption for both the colonial administration and the Rhodesian regime and the infliction of some damage; it showed also the presence of Frelimo soldiers in an area void of external allies. The psychological impact on the settlers was tremendous.

However, given the lack of back up support in the area, Frelimo could not count on their political activists to persuade the population on the justice of their cause. Thus the population was ‘compelled’ to cooperate with Frelimo, only to be at the receiving end of the Portuguese fury after each Frelimo incursion.

The importance of this corridor was obvious again when, after independence, President Machel threatened to close the Beira corridor to the Rhodesian regime and decided to openly support Zimbabwean ZANU-PF fighters. When dissent with the Mozambican regime started to grow, Rhodesia was quick to provide military and logistic support, which ignited a conflict that would last 16 years. In a kind of poetic justice, from the 1980s on the Beira corridor was being protected by those who tried to disrupt it previously – the then-ZANU fighters, now as Zimbabwean soldiers.

After the independence of Zimbabwe in 1979, Renamo, now without an ally in the region, experienced the same problems Frelimo had experienced
during the anti-colonial war and resorted to the same tactics – violence on the population. As a former Renamo soldier told an ISS researcher: “We had to be violent because we had to show the population that the government was unable to protect them, otherwise the population would not believe in us and would not support us.”1

This combination of diverse warring parties – Rhodesian, Portuguese, Mozambican and Zimbabwean soldiers, Frelimo, ZANU-PF and Renamo fighters – in a continuum of nearly three decades of conflict with a high level of violence singles this province out.

Chimoio, being the provincial capital, is a regional administrative, commercial, and transportation centre. Situated along the Beira corridor, Chimoio became central in both conflicts as attested by the huge army barracks still standing in the city.

The social disruption caused by conflict is expressed in the large numbers of refugees both fleeing into neighbouring countries and flocking from rural to urban areas. Chimoio was not an exception to this scenario and the influx of refugees from rural areas brought the city’s population to 105,818 in 1991.2 Five years after the peace process, the census data of 1997 places the population of Chimoio at 171,056.3

Decades of violent conflict in Mozambique – from the 1960s up to the General Peace Agreement (GPA) in 1992 – facilitated the proliferation of small arms as well as their availability throughout the country.

The little research carried out in Mozambique regarding security issues has focused mainly on Maputo city – the capital of the country. If on one hand it is true that Maputo has higher levels of crime, research based only on Maputo may not be representative of the rest of the country. In post-conflict countries, such as Mozambique, many capital cities tend to grow disproportionately to the rest of the country both in urban and economic terms. The high presence of a relatively large – at least larger than in the rest of the country – expatriate community facilitates the expansion of a service economy catering for it at inflated prices, which can distort the economy. Greater concentration of wealth attracts greater concentration of crime too. Illegal arms caches though are more likely to exist in more rural and less populated areas.

Since the perception has been that urban crime both in South Africa and Mozambique is being committed with illegal small arms from war caches, choosing a location where the presence of such caches is likely to be present seemed to make sense. The historical background of the region only made the choice of Chimoio more obvious.

Methodology

The research team decided to carry out the pilot test with population less used to surveys and also here Chimoio seemed to have an advantage over Maputo.

Random sampling

Due to time and logistic constraints the research team decided to carry out a random street survey in the main markets of Chimoio. The survey team approached people randomly in the market – both stall owners and shoppers – of both genders and with as diversified ages as possible. People were asked questions regarding security issues in their residence areas, provided they lived within the city limits.

The advantage of such random sampling is that it is possible to interview a wide range of people, without running the risk of carrying out what is often called a ‘housewife survey’, as is sometimes the case of house surveys.4 As people are being questioned outside their home environment, they are more likely to talk openly as they know they are not being observed by neighbours or relatives. On the other hand, people may be in a hurry to get home and thus be less available to be interviewed, or may also feel intimidated by talking in a public place. To overcome this constraint, people were invited to join the interviewer in a public cafe, replying to the questionnaire over a soft drink.

Limitations of the sample

As the interviews were conducted during the day the risk of targeting only a certain type of population, for instance, only the jobless or students, was
minimized by asking security questions regarding not only the respondent, but also including his/her family. As ‘family’ the interviewers defined ‘everybody living under your roof’.

The general reception to the research team was positive. People were cooperative, took their time and some people even waited for their turn to be interviewed. The research team was able to collect surveys from 34 respondents. Although this sample is enough for a pilot study, any quantitative analysis of such a small sample has to be regarded with caution and should not be considered as representative of the population.

It is legitimate to say for instance that “3 out 10 people interviewed in Chimoio have experienced a certain type of crime” but extrapolations such as “one third of the population of Chimoio have experienced the same type of crime” cannot be inferred. Consistencies in the surveys though enable us to draw a picture of the security concerns of the communities and point directions for further studies. Analysis of pilot studies also enable researchers to correct procedures and/or questions for further surveys.

Survey procedures

People were told the purpose of the study, the confidentiality agreement and also that they could stop the interview at any time. All but two respondents completed the questionnaire. The actual answers of these two respondents are included in this report. From the point where they decided to stop the interview, they are included in the ‘no reply’. A third respondent started his interview. By Section 4 he decided to stop and destroyed the questionnaire.

In spite of the good reception to the team, it is obvious from the analysis of the surveys that people are not comfortable replying to questions about firearms and sexual assaults. This issue will be discussed below.

As the survey sample was small, researchers complemented the survey with qualitative data. Interviewers were trained to conduct the interview as a
structured interview and to take notes of remarks made by the respondents to the questions.

The survey questions were read aloud to the respondent and the reply was written down by the interviewer. The decision to proceed so is that the
census of 1997 recorded an illiteracy rate of 57.7% in Manica Province. Many people are reluctant to admitt to illiteracy and would probably refuse to reply to the survey without stating the reason. This could create a wrong perception by the research team as reluctance to participate would not be understood and could be perceived as being related to the sensitivity of the questions. On the other hand, previous research experience in Mozambique has shown that people prefer to reply orally rather than write down their own replies.

The survey included five sections:

• Section 1: individual data on the respondent
  

• Section 2: crime trends and perceptions of security
  

• Section 3: security within neighbourhoods
  

• Section 4: attitudes towards firearms
  

• Section 5: perceptions on security and firearms

Language

The survey questionnaire (Annex 1) was first translated into Portuguese. It was then discussed with Mozambican nationals to test the pertinence of each question and to adapt any question that needed so.

In Chimoio, the survey questionnaire was then translated into the local
languages of Cisena, Cindau, and Shona by the three locally hired interviewers, who also led the interviews. The translation was done orally by the interviewers, so that the concept of words without translation into local languages would have a single meaning, previously clarified and commonly agreed upon.

Section 1: Individual data

The total sample consisted of 34 surveys from 18 male and 16 female respondents. Twenty of the interviews were made in public places, such as cafes; eight interviews were done at the home of the respondent; and six of the interviews were done at the working place of the respondents. The ages of the respondents range from 16 to 70 years of age, giving an average age of 27 years for the sample.

Questions on security were posed in relation to the home area, even for those respondents who had market or street stalls and small businesses in the area of the interview.

All respondents attended or were still attending school and yet most of them are busy in the informal sector. The formal education of the sample goes from completion of grade one to pre-university studies (grade 12), one female was a nurse/birth attendant, another woman an accountant, and two of the respondents were electricians. Almost half of the respondents (15) have completed schooling between grades 9 and 12.

For analysis purposes, the activities of respondents were divided into five categories: public sector, private sector, informal sector, agriculture, and other:

• Public sector, meaning formal employment with a government institution (three respondents)
  
  

• Private sector, defined as legally licensed shops or businesses (eight respondents)
  
  

• Informal sector, meaning informal economic activity such as street
  vending, market stalls and street stalls (14 respondents)
  
  

• Agriculture –nearly all respondents practised some kind of agriculture, but mainly for home consumption. None of the respondents had
  agriculture as a main activity.
  
  

• Other – students, housewives and other employment (nine respondents)

Section 2: Perceptions on Crime and Security

This section includes information on types of crimes and victimization during the past 10 years. The time frame of the questions is used as a tool to encourage accuracy in dating a crime and is generally not used in the analysis. Time is usually referenced in relation to a public holiday or event. In this case the research team used the general elections of 1994 for the time reference.

The importance of definitions

Crimes often have a legal definition, which is not always consistent with the public idea of the crime. Concepts such as theft, assault, or robbery are used loosely and deprived of their legal meaning. This becomes particularly important when the surveys have to be translated into another language and social context.

The survey used in this research was translated first from English into Portuguese and then from Portuguese into local languages. In Portuguese, for instance, the common word used for ‘car hijack’ and ‘car theft’ is the same and does not exist in some of the local languages. To overcome this constraint, the research team agreed on common definitions for the crimes listed in the survey.

Thus, the following definitions were used to describe types of crimes mentioned in the survey:

• 'home burglary’ was defined as ‘when thieves come, or try to come, into the house independently whether you and/or your family are inside or not’
  
  

• ‘stock theft’ was defined as ‘when someone steals animals, such as sheep, cows, goats, chicken, but not dogs or cats, belonging to you’
  
  

• ‘crop theft’ was defined as ‘when crops get stolen from your farm, be it from the barn or from the field’
  
  

• ‘car hijack’ was defined as ‘when your car is stolen while you are driving it or when you are parked and siting in it; when you get pushed out and the person runs away with the car’ (long but necessary)
  

• ‘car theft’ was defined as ‘when your car or parts of your car disappear while you are not present. For instance, during the night, or while parked somewhere’
  
  

• ‘deliberate damage’ was defined as ‘when things belonging to you have been damaged on purpose for no apparent reason’
  
  

• ‘rural equipment’ was defined as ‘not only tractors or mechanized implements, but also hand tools such as hoes, machetes, etc.’
  
  

• ‘violent assault’ was defined as ‘if you were ever beaten up’
  
  

• ‘robbery’ was defined as ‘when you are walking down the street or riding the bus and someone approaches you and threatens you unless you give something; or when you get home and you realize that your wallet is missing’
  
  

• ‘murder’ was defined as ‘when someone was killed by another person on purpose and not by accident’
  
  

• ‘sexual assault’ was defined as ‘not only violent rape, but also when a person has to submit to get the marks at school, for example’

Potential constraints of definitions

Definitions are very important in surveys as they can represent serious constraints during the analysis. In a broader survey, the definition of ‘robbery’ used here, for instance, would prevent an indicator of violent crime in Chimoio.

Traditionally robbery is defined as ‘theft against physical force’, that is, robbery is considered only when the theft occurs with either violence or threat of thereof. The definition worked out by the research team though, includes pick pocketing in this category. Were this a broader survey, the results could not be used for comparative analysis with surveys carried out in other countries, due to the different definition.

Whereas in an exploratory survey, given the diminutive size of the sample and the different purpose of the research, this may not pose a problem in terms of analysis, larger surveys have to have carefully drafted definitions. For the purpose of this study these questions were used rather as a test of people’s willingness to reply to this type of question and also to test the relevance of these types of crime in the local context.

Credibility of the respondents

Surveys depend on the willingness of people in giving their time and also on the reliability of the respondents. There is always the risk that one of the respondents may not be telling the truth. To overcome this obstacle in bigger samples, the surveys showing replies very different from the median are excluded from the analysis, so that distortions do not occur.

The research team experienced some doubts when analysing the surveys of two male respondents who reported together:

• 3 home burglaries
  

• 11 crop theft
  

• 2 assaults
  

• 2 murders
  

• 1 rape
  

As this was an exploratory study no survey was excluded but numbers have to be considered very carefully.

The importance of classification for the analysis

There is another equally important classification for types of crimes, which was not used in this survey: violent crimes and economic crimes. Violent crimes are those where there could have been weapons, such as car hijacking, assault, murder and robbery. This distinction is important as it can represent an indicator of the level of violence.

Observations made during the field work and remarks noted by the researchers created the impression among researchers that violence does not seem to be a common pattern in the criminality of Chimoio. The team had the same impression regarding the use of firearms in crime. People still seem to associate firearms far more to war than to crime. One of the reasons many people stated for not having a gun is that it would not make sense to have one in a country at peace. The association of firearms to crime though is latent in people’s minds, as many of them state that if they had a gun they think they would be tempted to use it to solve financial problems, and also that firearms bring instability into the communities.

Questions on sexual assaults

The same disparity can be found regarding the question on sexual assault. Surveys carried out in other places note that it is extremely difficult to get replies to such questions. In a recent survey carried out in central Johannesburg by the ISS,5 questions related to sexual crimes were excluded from the survey, due to the reluctance of people in replying to them. The research team in Chimoio faced the same challenge.

While adapting the survey, the research team anticipated this constraint but decided to keep the question on sexual assualt because the purpose was to test people’s willingness to participate in the survey. Thus, the question remained but the definition of sexual assault was broadened to include sexual favours. After the first day of interviews, the research team decided to modify the question and ask if the respondent knew of anyone in his/her community (and not a family member) who had been raped.

The survey recorded two sexual assaults and one attempted sexual assault. Both victims were relatives of the respondents. Firearms were used in one sexual assault.

To have two actual sexual assaults and one attempt in a sample of 34 seems to suggest that people in Mozambique would be willing to discuss this type of crime. However and again to stress the danger of extrapolating factual data from such small samples, qualitative field observation contradicts this. It was obvious that people felt uncomfortable with the question – they would shift in their seats or immediately say no without pausing.

During a recent interview with a Mozambican psychologist6 he was asked about rape and how communities perceived it. According to him, rape seems to be something that most women expect to experience at some time or other during their lives; they do not like it but came to accept it as a fact of life, as an occupational hazard while farming or collecting wood or fetching water. It must be said again that this reflects a personal perception rather than a research finding. Research on sexual crimes although difficult, is important and it is a field that is unexplored in Mozambique. Given the government’s efforts to curb domestic violence, research in this field could be useful to inform policies, correct guidelines and design awareness campaigns. As this is a very sensitive topic, research on sexual crimes is usually carried out by experienced psychologists and social workers, who know how to deal with the inflicted trauma.

Crime in Chimoio: Survey results

Chimoio is said to have one of the lowest crime rates in Mozambique for the past five years. According to the PRM in Chimoio, the crime rate in Manica Province is 15 per cent. Although Chimoio has a higher crime level than other urban centres in the province it is still among the safest in Mozambique. And yet, in reviewing the survey results, there appears to be a high number of incidents reported for such a small sample. Half of the respondents reported home burglaries. As respondents in Chimoio did not live in apartment buildings, stealing a bicycle from the house yard, for instance, was counted as home burglary as criminals had to enter the physical premises of the house. Home burglaries and robberies (as defined in the survey) seem so prevalent that people classify them as “being normal”. Shoulders were often shrugged and many respondents replied “of course” as if there were no alternative (Figure 1).

Figure 1 - Types of crime



The (under)reporting of crime Underreporting of crime is common to many countries and societies. Reported crimes are usually those that cannot be concealed or dealt with otherwise, such as murder, where legal proceedings are needed in order to dispose of the body; those that can be covered by insurance companies, such as car theft or home burglaries; and those too valuable not to report, such as robberies of particular expensive items like jewellery. Petty theft and pilfering are seldom reported to the police.

Future research could profit from a deeper understanding of this problem. This would require a survey with better-defined concepts of crime, with crimes being duly classified in household or individual, violent or economic. This section should also be complemented with detailed questions on the crime reported and on the proceedings, so as to give some insight on areas where police / judiciary services could be improved.

In the reporting of crime, respondents indicated that firearms had been used in primarily in incidents of murder and car hijacking (see Figure 2). Generally, the use of a firearm in crime is very low.

Figure 2 – Use of firearms in crime, 2002



Figure 3 – Types of crime in 2002, by percentage



As indicated in the survey results, burglary is now the most common crime among respondents in Chimoio, followed by robbery, crop and stock theft (Figure 3).

Security

The purpose of section three of the survey was to assess how safe people perceive their communities to be and how people rate police performance. If a neighbourhood is perceived as unsafe it is unlikely that individuals will invest in the area; perception of poor police performance may be decisive when voting.

This is also the reason why it is important for the police to know the dimension of petty crime that is not being reported. A person may not report to the police thefts of small amounts but if these occur repeatedly it is unlikely for the person to feel ‘safe’ in a neighbourhood.

Official crime statistics cannot include non-reported incidents. Chimoio is one of the Mozambican cities with the lowest reported crime rate and yet respondents to this sample expressed concerns with security. The city of Chimoio was often referred to the research team as a role model for other Mozambican cities and is currently a pilot site for a community policing project. Police efforts to control crime in Chimoio were often referred to and acknowledged by the respondents. If, as this study suggests, crime in Chimoio is not being reported to the police due to its nature, these efforts are probably not being maximized.

Urban and rural population

Although Chimoio is an urban centre, many of its inhabitants have small family farms on the outskirts of town. They do not live there, but they attend the fields regularly and most of the farm stands have a small hut where the owner can stay overnight if necessary. As such, questions were also asked regarding how safe they feel while working at their farm stands. Figure 4 summarizes the degree to which respondents worry about crime.

Figure 4 – Individual concerns about crime



Within the sample, slightly less than half (45%) are occasionally worried about crime. However 52% of respondents are ‘always’ or ‘often’ worried about crime in Chimoio.

Figure 5 - Feeling of security at home and in countryside



Figure 5 presents feelings of safety among respondents at home and in the countryside. More respondents felt ‘more or less’ safe at home during the day, while respondents generally felt ‘very insecure’ at night, whether at home or in the countryside.

The perception of the respondents is that most crime happens at night. They referred to the lack of streetlights and lack of night patrols by police as the main factors affecting security in their communities.

Concerns with crime and gender

Crime seems to be an issue that is ever present in respondent’s minds. Five female respondents said they worry ‘always’ about crime and another six female respondents said they worry ‘often’ about crime. From a sample of 16 women, 11 reported high concerns with the issue.

This could mean that women may feel physically more vulnerable and thus fear crime more than men or they could see themselves as preferential crime targets. But, it could also mean that women perceive that they have fewer rights in the administration of law. In an informal conversation with a police officer from the Domestic Violence Unit in Beira, it was stated that the biggest constraint the unit faced was to convince women to come forward and report violence exerted upon them. Rape victims may decide not to report the incident, as the reporting could bestow a social stigma upon them. All other respondents worry to a certain degree with the level of crime.

Geographical dissemination of crime

The impression from the sample is that crime in Chimoio is not restricted to any particular area or neighbourhood, rather affecting the whole city. Crime statistics provided by the police in Chimoio for the four police stations in Chimoio seem to confirm this (Table 1).



Police stations 1 and 2 cover each nine neighbourhoods; police stations 3 and 4 cover seven and eight neighbourhoods respectively.

In terms of neighbourhoods with higher crime levels, Police station 1 covers two and the other stations each cover one neighbourhood with higher levels of crime. These figures, provided by the Central Command of the Police in Chimoio, suggest that crime is relatively widespread rather than concentrated in certain neighbourhoods.

Organized versus economic crime

In a recent focus group discussion carried out with youth in Maputo, the participants made a clear distinction between what they called two types of crimes: organized crime – violent, using firearms, and involving people of financial means; and what was referred to as ‘the crime of ignorance’, described as petty crime committed by people who see no other way forward in their lives. It may involve the occasional use of firearms but the motivation of the criminal is the economic stress of poverty. A similar distinction was equally made by an officer in the Ministry of Interior in Mozambique:

“In Maputo there are two types of crime – organized crime and economic crime. They tend to use the same types of guns, only some organized crime use other tactics, mainly when the criminals are former soldiers (like the murderer of Siba Siba7). The former soldiers of Casa Militar, who guarantee security to the government, are known to organized crime bosses.”

The attitudes of respondents to changes in crime and violence are shown in Figure 6. The majority of respondents felt that both crime and violence are on the increase, although more noticeably in terms of violence.

Figure 6 – Changes in crime and violence

The fact that 17 and 14 respondents respectively believe that the levels of violence and crime in their areas has increased is quite remarkable for a
country which experienced three decades of violent conflict.

This could be due to the sample, to the lack of definition of violence in the question, and/or to the fact that Mozambicans often state that ‘before the war there was no crime’. What seems to be apparent from field observation is that the respondents made a differentiation between violence that occurred during the war and criminal violence. The research team had no access to crime data and as such cannot confirm the accuracy of this statement, but an officer in the Ministry of Interior in Mozambique told the research team that “criminals used the war scenario to commit crimes, but the problem was not general and was ‘diluted’ due to the war”.

Policing in the communities

Policing in Chimoio does not seem to be a problem overall. More than half of the respondents replied they see police on duty at least once a day and a total of 28 respondents see police patrolling their communities once a day to once a week (see Figure 7).

Figure 7 – Presence of police in area



However most respondents complained about the quality of patrols, claiming that it is irregular and generally nonexistent at night.

Respondents felt that police efforts to control crime are ‘average’ to ‘bad’, and police performance is rated between ‘the same’ and ‘worse’ (Figure 8).

Figure 8 – Perceptions of police efforts



Almost equal numbers of respondents stated that police performance had improved in the past few years (n=9) as stated that it had worsened (n=7) (Figure 9).

Figure 9 – Perceptions of police performance



Police performance

The reasons presented by respondents for the change or lack of change in police performance were similar among respondents and were used both to justify the improving and the worsening of police performance. Selected reasons given by respondents follow, each respondent was asked to give up to three reasons:

• salaries for policemen (n=8) – low salaries viewed as a reason for poor performance and an increase in salaries explained improvements;
  

• lack of or inefficient patrolling of the community (n=7);
  

• presence (or the lack of) street lights to security in communities (n=6);
  

• security improved in neighbourhoods due to better police performance (n=6);
  

• cooperation of policemen with criminals deteriorates the security in their area (n=5);
  

• cooperation, or lack of, with the community was also referred to as a reason of change in the public security (n=5);
  

• corruption was identified as a deteriorating factor in public security (n=5);
  

• poverty level of police officers (n=3);
  

• other reasons included: faster reaction by the police; reduction of crime in the area; lack of police ethics; awareness campaigns; good patrolling; vigilantes; honesty of some policemen; bad training and lack of resources for the police.

Factors influencing police performance

The consistency detected in the replies of the sample seems to identify three main factors that affect police performance: economic reasons, corruption and cooperation with the community. Thus, the replies of the sample were clustered into three categories: cooperation with the community, including all the replies that mention cooperation, or lack of, with the community as an important reason for change; economic reasons, including replies mentioning salaries and levels of poverty; and corruption, which includes the replies mentioning cooperation with criminals, lack of ethics, and bribery.

Out of 60 suggestions, more than half (33) fall under these three clusters:



Respondents seemed to perceive these three problems as key factors in public security: living conditions need to be improved; police and community have to cooperate; and corruption has to stop. Respondents seemed to relate corruption in the police force to poverty.

Security in the community

One of the major constraints in community-based projects to control crime and firearms is the difficulty in encouraging people to report on their neighbours. If popular attitude towards corrupt officials is one of understanding and empathy, the denouncing of corrupt officials can become very hard. Respondents do not like corruption and complain about it but it seems that they have largely come to accept it as largely inevitable.

Communities who perceive themselves as insecure tend to assume responsibility for their own security, not always in the best of ways. Mozambique is not alien to popular justice – around 1994 in Chipamanine, the biggest market in Maputo, if the population caught a thief he would be executed on the spot. The research team got the clear impression that there may be some vigilante-type movements in the neighbourhoods of Chimoio, primarily through references to ‘popular defence’.

Youth and demobilized soldiers

Reports written by organizations involved in the reintegration of former combatants hint that in fact, and contradicting common wisdom, ex-combatants are not more prone to crime than any other group.8

However, an interview with an officer in the Ministry of Interior indicated that most criminals are middle aged, demobilized soldiers and unemployed. According to this officer, demobilized soldiers prefer to say that they are unemployed rather than former soldiers or former Casa Militar.
The streets of Mozambican cities are crowded with young people trying to fend for themselves. Most of these people have some degree of formal
education but find it difficult to enter the job market. Youth is traditionally looked upon suspiciously for varied reasons.

Community protection

Regarding the question on what are respondents doing to protect the community against crime, responses varied between individual response (installing burglar bars or buying a dog) to more community-focused approaches, including attending community police meetings. Table 3
summarizes the replies to this question.



Private initiatives to prevent crime seem to prevail. Even those respondents who replied ‘nothing’ justified the answer with lack of financial resources to install an alarm or burglar bars.

This could mean that people are not relying on authorities to prevent crime in their areas. This can also be due to the nature of crime being committed in the community. As stated above, it may be crime that is often unreported.

Community alternatives to public security

According to information from the police in Chimoio, there have been no reports of vigilantism or self defence units in the area. Police stated that
communities have been contributing in a positive way to preventing crime and welcome this contribution.

According to the survey respondents, there is also a will among the population to cooperate with the police. The police in Chimoio have been organizing talks and awareness campaigns in the communities, which were mentioned by respondents.

An interesting observation is that respondents seem more prone to denounce criminals to the community authorities (defined as the community informal leader) first; to the local authorities (defined as the secretary of the neighbourhood or government representative) second and finally to the police.

This could mean several things and may deserve further research. It could be that communities prefer to respond themselves to petty crime and only go to the police in certain instances or signal a lack of trust in the institution.

Cooperation with the community

Chimoio is currently testing a community policing pilot project in the neighbourhood of Fepom. None of the respondents of the sample came from this part of town. However according to the police in Chimoio, this project is being implemented with the support of community leaders.

The last question of this section asked respondents what they think the government of Mozambique could do to improve the security in their communities. It was an open question however most respondents refer to providing street lights, creation of employment, better salaries, improving police resources, and better patrolling as actions the government could take to improve public security. Again it seems that respondents make a clear link between structural problems and crime. It also seems that respondents respect the police as an institution, as most replies emphasized stronger cooperation with the communities and many ask for stronger ties and more awareness campaigns.

Attitudes towards Firearms

The purpose of the fourth section of the survey was to research popular attitudes towards firearms; whether firearms are important in the local culture and also how they are being used. Knowing popular attitudes towards firearms enables governments to draft policies and to design specific and targeted awareness campaigns.

In the case of this study, the main purpose was to verify how willing people would be to answering questions related to firearms.

During previous meetings with organizations collecting firearms, such as TAE, the research team was told that this was a sensitive topic. It was no different in Chimoio. It was obvious that people were uncomfortable replying to questions about weapons. Respondents had to be encouraged to reply with answers other than ‘I don’t know’.

Although respondents were reluctant in replying to questions such as “If you need a gun do you have easy access to one” (question 4.7), they were less reluctant when questioned regarding their own attitude towards firearms. Respondents did not hesitate in saying why they would like, or not, to have a gun.

Question 4.12 “What would you recommend to reduce weapon availability in your community” presented two clear responses: give police more power and promote more weapons collection programmes. Respondents were then asked whether they agreed or not with any of the alternatives and also if they would like to add any other type of initiative. Almost all respondents added their own proposals to control weapons. It seems that reluctance on this subject is linked to the nature of the questions and can be overcome, so further research on this issue should not be deterred from including these questions.

During this section, respondents would often look around to make sure no one was listening to them before they would venture a reply.

The first set of questions inquires on how often weapons are used in crime in the community, how often the respondent hears gunshots in his community, and if compared to previous years the availability of weapons has changed. The majority of the respondents (n=21) says that firearms are used ‘sometimes’ and ‘seldomly’ to commit crimes in their communities, and half of the sample (n=17) replied that firearms were used ‘seldom’ and ‘never’ (see Figure 10). This seems to confirm the pattern of firearms use noted earlier the survey, where most crime incidents did not involve the use of firearms.

Figure 10 – Frequency of firearms used in crime



The same can be said regarding the next question on the frequency of gunshots in the community. The majority (n=27) reported hearing gunshots ‘sometimes’ (n=18) and ‘seldom’ (n=9) whereas 13 respondents fall reported ‘seldom’ and ‘never’ (n=4). These figures are also consistent with the impression the team had on the field that firearms may not yet represent a problem in crime in the area.

Figure 11 – Availability of firearms



A clear majority of respondents (n=21) stated that the availability of firearms has decreased (Figure 11). However, approximately one-third of the sample saw no change or an increase in weapons.

Respondents were asked what they viewed as the main reason for the change in weapons availability. A summary of responses is presented in Figure 12. It should be noted that the responses could refer to either an increase or decrease in the number of available weapons and respondents could give more than one answer.

Figure 12 – Main reason for weapons availability



War and the availability of firearms

The majority of the respondents who recorded a reduction in the availability of weapons seemed to attribute the reduction in weapons mainly to war-
related events: arms collection programmes, demobilization, end of the war and awareness campaigns. Eight respondents make a link between firearms and crime and attribute the reduction to police efforts, with and without the support of the community.

The respondents who think that the availability of weapons has increased (n=8) attribute the change to the fact that the collected weapons are in the wrong hands and/or the police are unable to control the situation. Enquiries as to what respondents meant by the term ‘wrong hands’ they were quite blunt and explained to the team that army and police officers were renting to criminals either their own weapons or weapons collected after the war. One of the respondents attributed the decrease in weapons to the fact that “many policemen were fired because they had sold or rented out their guns to criminals”.

The origin of firearms

The original idea of the research team was to verify if remains of war still in arms caches were perceived as being used in crime. According to eight of the respondents it is rather weapons already collected that are being used in crime. According to an officer of the Ministry of Interior in Mozambique, weapons used currently in urban crime are either being rented out by legal owners or have been hidden in arms caches, which seems to concur with the statements of the respondents.

Access to firearms

The next questions were more personal and tried to establish how respondents relate to firearms as individuals. Respondents are asked if s/he knows any friend or relative who owns a firearm; if that person has a gun license; if the respondent has access to a firearm should it be necessary; and if the respondent would like to possess a gun. The questions were complemented with an open ended question that asked respondents to state the reason why s/he would like, or not, to own a gun.

Most respondents stated that they did not have friends or relatives who were gun owners. No respondent seemed comfortable replying to this set of
questions. The majority of respondents said they did not have access to a firearm and would not to be willing to have one.

Only two respondents said they have easy access to a firearm, which is interesting because both stated that they did not know anyone with a gun.

This seems to suggest that there is not a culture of weapons in Mozambique. Respondents did not seem to think of weapons as something one should have, but rather as something that is not quite honourable or right. It seems that there is an association between weapons, war and social disturbance. The team was surprised that no respondent made the association of firearms to hunting, as had been expected.

Reasons for owning a firearms Those respondents stating that they would like to possess a firearm justify the wish with security reasons. The reasons given by respondents who did not want a firearm fit into three clusters:

• having a weapon in a time of peace does not make sense
  

• afraid of having and handling a firearm
  

• fear they might use the gun
  

Fourteen respondents attributed their wish of not having a gun to either considering that guns bring violence into the community or because they are afraid they may be tempted to use the gun.

It seems that respondents in this sample feel that weapons can have an impact on one’s behaviour and sense of self. Some respondents said that weapons “bring disobedience” and they were asked to explain what they meant. According to them, if you have a weapon you feel empowered to do more of what you want and less of what you have to. It is easier to disobey the norm.

Another respondent replied with a parable about lions and lambs that you cannot change nature of things and guns are meant to kill.

Controlling firearms

The next set of questions focus on gun control and inquires respondents how they feel about gun control measures in his/her community. The first two questions ask the respondent’s opinion on whether gun control should be improved and if respondents would be willing to encourage people in handing over their guns, should security in their communities improve (Table 4).

Communities and disarmament efforts

Overall respondents expressed a willingness to support efforts to disarm communities. A majority of respondents thought that gun control needs to be improved and were willing to contribute to disarmament efforts. If future research confirms this trend, the government of Mozambique seems to be in a very comfortable position to strengthen gun control efforts.

Respondents were given two recommendations on improving controls on firearms in their community. Respondents were asked which they would
prefer and also if they would like to add additional recommendations. Table 5 presents the replies.



Most respondents support giving the police more power to control illegal weapons. The other suggestions made by respondents are in line with previously given answers. These include more coordination between the police and communities and promoting greater awareness. Respondents also refered to the structural problems of poverty and also the need for improvements in police working conditions.

Respondents were asked their opinion on the sources for illegal weapons both within Mozambique and externally. Most respondents (n=20) stated that the weapons in the community were remnants of the war. However these are not necessarily weapons that remain in arms caches. In some instances, the respondents specifically identified weapons that were seized during the ONUMOZ operation but were not destroyed at the time.

Perceptions on security and firearms

The final section of the survey explored the link between firearms and security and asked respondents questions on crime, firearms and weapons collection programmes.

Respondents were asked to identify what they believed to be some of the motivations for crime in Chimoio. Their replies are summarized in Table 6. Respondents were able to indicate more than one reason.



Most respondents attribute crime to structural problems (unemployment, poverty and difficult living conditions).

Respondents were then asked a series of questions related to firearms that are most common in their area. The purpose of these questions was to establish a pattern, if any, of the presence of weapons in each community and identify a possible link to the crime situation.

Figure 13 – Most common firearms



Respondents stated that there were more automatic weapons than hand guns in the community (Figure 13). No respondent mentioned shotguns or rifles, even though Manica Province is a region where hunting was common before the war. The majority of respondents (75 per cent) stated that the presence of firearms worsens crime in the community (Figure 14).

Figure 14 – Do firearms negatively affect crime in the community?




The two questions on how frequently firearms are being used compared to before and after the elections, respondents seem to be consistent in a less frequent use of firearms (Figure 15).

Figure 15 – Change in availability of firearms

Post-war disarmament efforts

Respondents were asked where they thought weapons were turned in after the GPA (Figure 16). The majority of the respondents said that the firearms handed over by militia were delivered to the military barracks, apparently the appropriate procedure.

Figure 16 – Weapons collection points




None of the respondents had participated in firearms collection efforts although thirteen of the respondents were old enough to have been either soldiers or militia. The survey did not include a direct question on whether the respondent had or not been part of a militia group. No respondent admitted having belonged to a militia group; on the contrary, some would state they had not been in the militia.

Respondents were asked why they thought some militia did not want to hand over their firearms after the war. Eleven respondents did not reply to the
question or said they did not know and one refused to reply to further questions. The most frequent reason attributed to some militia to keep their firearms were fear of renewed war, for protection, and not knowing where to deliver them.

Respondents were visibly uncomfortable with the subsequent questions on arms caches and many respondents would take care in stressing that they did not know of any cache around their community; or that they had never found one. The majority of the sample denied both the presence of arms caches in their communities (n=24) and participation in finding them (n=19). Some of the respondents had heard of both Operation Rachel and the CCM TAE programme.

Regarding the question about how communities identify arms caches, nearly all respondents replied the same: arms caches were found when refugees started to resettle and had to open farm land, or while gathering construction materials for their houses.

Given the existence of two successful firearms collecting programmes in Mozambique – Operation Rachel and TAE, the survey included questions on both these projects. Although most respondents stated that their community had not cooperated with either initiative, seventeen respondents would provide information on arms caches to TAE, while 18 would provide information to the police.

Lessons learned and recommendations for further research

The most important finding of this survey was the commitment of the respondents to completing the survey, even when the questions started to become uncomfortable. People showed reluctance in replying to some questions and it is doubtful that they have been candid with some of the replies, but this should not deter further research exercises.

The main finding of this survey is that this type of research is possible in Mozambique – people understand the questions and, in spite of some reluctance, people do reply and participate. Many respondents at the end of the survey told the interviewers how much they had liked to be interviewed and would the team please pass the concerns of the population to those who could respond.

However this chapter also raises several issues for further research and suggests some recommendations.

Research in Mozambique should preferably be extended to at least one province outside Maputo province. As stated earlier, research based only on the capital city may present trends and patterns not common to the rest of the county. One of the main purposes of research is to provide institutions with information that can help draft and implement policies and strategies. As such, data gathered in Maputo should be cross-checked with data gathered at provincial level.

There is a need to know exactly which type of crime is being committed in these communities and the extent of under-reporting. Statistics which do not include these two dimensions of the crime problem may be presenting a distorted image of the situation. Policies and strategies based on this distorted image may become difficult to implement or not ripe the expected benefits. This can lead to further frustration both in the population that does not experience improvements, and in the police force, who do not see their efforts rewarded, no matter what they do.

This report suggests that there may be a link between structural problems and crime in Mozambique. Understanding the motivations of crime is important to draft policies and to allocate resources. It may also point to situations that require a more global approach, involving resources and strategies from more than one ministry. Crime and security should not be dissociated from the existing social and economic environment and crime control measures can also come from quarters other than the police or the Ministry of Interior.

Some circumstances seem to be particular to Mozambique and should be taken into consideration when designing further research projects. When defining the sample population, care should be given to the fact that some Mozambican households are headed by children or youth. Questions including comparisons should include a reference in either size, quantity or time. Concepts such as crime types, violence and crime should be carefully defined, taking into consideration the decades of historical violence.

Another perception that emerged in this survey is the suspicion of corruption within the police. This is not new and the Ministry of Interior is aware of the perceptions. Many efforts have been made in Mozambique to curb corruption and many institutions show visible signs of improvement.

None of these problems is new in Mozambique neither are they unique to the country. Governments all over the world have faced or are still facing
similar problems.

More than anything, this small survey illustrated how research can provide information that can be formulated into policies and approaches by governments. The Ministry of Interior recognizes the importance of research and it can benefit from engaging with others in additional research on research proposals onfields of particular relevance for the Ministry to aid it in its work.

Notes

1. Personal interview.
   
   

2. Source: website of INE.
   
   

3. Source: website of INE.
   
   

4. House surveys carried out during office hours, for instance, run the risk of interviewing mostly housewives, as they are the most likely to be home during normal business hours.
   
   

5. T Legget, Rainbow Tenement, Crime and Policing in Inner Johannesburg, ISS, Monograph 78, 2003.
   
   

6. Personal interview on 11.2.2003 in Maputo.
   
   

7. The firearm used to murder Carlos Cardoso has never been found and it is suspected it may belong to the police stockpiles.
   
   

8. See the most recent report published by PROPAZ, a Mozambican NGO that deals with the social re-integration of former combatants. Alex Vines, in op.ci also refers to the fact that demobilized soldiers are not more involved in crime than other social groups. Interviews with many Mozambicans expressed the same view.