Introduction

The firearm indicators measured in the survey aimed to obtain information on the extent and nature of firearm penetration within Tanzania. The extent to which the community is exposed to firearms, as well as is dependent on firearms, is critical to an understanding both of the reasons for concern regarding the proliferation of firearms and of how to design programmes to redress firearm dependency.

Firearm ownership is one of the most important firearm penetration indicators. It is often difficult to obtain information about the direct level of firearm ownership, as interviewees are often reluctant to be fully open on the topic of firearm possession and use. Given that acquisition and/or possession may not be legal in some cases this reluctance is entirely understandable. In order to overcome this, questions of associated firearm ownership are utilised, which avoid the direct and highly sensitive question of personal ownership but instead survey connected areas from which deductions about the nature and extent of actual firearm ownership can be made.

Measuring other perceptions of the respondent towards firearm proliferation is also important in order to gauge how the community will react to measures to curb proliferation. If the community is of the opinion that firearms are problematic, then a high level of co-operation could be expected towards government- and community-based initiatives to explore measures to curb firearm proliferation.

Analysis

The indicators of firearm penetration point to a country in which the prevalence of firearm possession is increasing but at a rate that does not as yet constitute a crisis. Overall levels of the use of firearms in crime, exposure to violent conflict involving a gun and the frequency of hearing gunshots were relatively low, although certain regions did appear to be more afflicted than others; notably Kigoma, Mwanza and Morogoro. When perceptions of the change in the number of firearms are considered, Kigoma and Morogoro appear to be the regions in which firearm penetration is greatest. These regions are those to which crime prevention strategies might be usefully targeted as a means of tackling the proliferation of small arms. In Kigoma the main determinants of this relatively high level of penetration appear to be its geographical location, bordering the conflict zones of Burundi and DRC, and the resultant destabilising influx of refugees from these regions.

The region of Kilimanjaro is where firearm proliferation is perhaps lowest. This is evidenced by the low incidence of residents having friends or family members who possess a firearm and of residents having access to firearms. This is emphasised further by low levels of willingness to possess a gun. Given these findings, Kilimanjaro is a region that deserves further study to ascertain why firearms are less of a problem there than elsewhere. One obvious factor, already mentioned above, is the significance of this region economically as a tourist destination and the importance placed on policing to ensure the safety of tourists, as a result. One can thus deduce that improved levels of policing both quantitatively and qualitatively is likely to have an impact elsewhere in Tanzania.

However, identifying clear trends across indicators of ownership proves difficult. Morogoro had the highest indicated level of actual possession and showed the highest willingness to possess a firearm, suggesting that the prevalence of guns in this region is great and moreover, that the situation is unlikely to improve in the near future. Morogoro, a region where crime is known to have burgeoned recently, should thus be prioritised by agencies, both governmental and non-governmental, that are to address issues of firearm control. While less immediately obvious, Kigoma also appears to have a notable firearms problem, particularly with regard to illegal weapons. This is illustrated in the respondents' answers by having the highest level of access to firearms and a higher than average willingness among its residents to own a weapon, which suggest guns are relatively widely dispersed. When this is considered alongside the comparatively very low number of respondents in this region indicating that they had friends or family members possessing a firearm and the relatively low levels of actual firearm possession, it is possible to conclude that ownership is relatively high but that respondents were, naturally, unwilling to implicate themselves or close associates directly.

That Kigoma borders countries in conflict, in which rebel factions—by their very nature illegal possessors of arms—are fighting, gives weight to this deduction. Given Kigoma's situation the responses of residents in Kagera, also bordering conflict zones, are of particular interest. Levels of actual possession were relatively low (only Zanzibar had lower levels) and the willingness to possess a firearm equally so (only respondents from Kilimanjaro indicated a lower willingness), yet only three regions had a higher level of access. Drawing concrete conclusions about the situation in Kagera from these results is not easy but a possible scenario can be reasoned. It may be that, as in Kigoma and for similar reasons, there are a large number of illegal firearms (indicated by high acknowledged access yet low acknowledged actual possession and low levels of licensing), yet the situation may be more stable in Kagera, which would explain the lower willingness to possess. Morogoro and Kigoma appear in the most urgent need of attention, while Kagera, but also Mwanza and Arusha, seem to have significant levels of firearm possession.

The indicators concerned with the need for controls point to a country in which there is general acceptance of the need to control firearms. These indicators help to enhance the overall picture of firearm proliferation and in particular point towards the likely willingness of regions to accept, co-operate with and welcome initiatives to control and reduce firearm proliferation. They can also help to illuminate which are the most affected, perhaps not in absolute terms, but in the equally important sphere of the impact on the quality of life. Encouragingly, there appears to be very widespread recognition of the need to control firearms. This suggests a recognition of the desirability of improved arms control. One may also infer, however, that this may be the result of personal experience of the deleterious impact of guns and/or from an awareness through the media that the proliferation of small arms is a growing issue of more general concern within Tanzania, and perhaps also in the wider world.

Kigoma, Pwani and Kagera are the regions where the impact of firearms is greatest, while Mbeya and Arusha appear to be the least affected. For the reasons given above the serious concern about the situation in Kigoma, and similarly in Kagera, is unsurprising. What is interesting about Kagera is that firearm injuries are decreasing, while the risk of injury, the level of worry and the level of impact are high. This suggests that Kagera has suffered greatly in the past and continues to be badly afflicted but that the situation has, in recent times, improved slightly. Although Pwani did not stand out as one of the regions of greatest concern in terms of firearm penetration or ownership, there are indications (in terms of a general decline in safety and the relatively high incidence of friends and family possessing a firearm) that should be monitored in this region.

Indicators of firearm penetration

Frequency of use of firearms in crime

The respondents were asked to indicate how frequently firearms are used to commit crimes in their community. Overall, firearms were infrequently used to commit crimes. This could be strongly related to the fact that violent crime in Tanzania is relatively low. Kigoma, Pwani, Arusha, Morogoro and Mwanza were those regions with relatively more incidences of firearm usage in crime as compared to the other regions surveyed. However, in their own right the use of firearms in crime occurred infrequently (as indicated by a mean greater than 3.0 and less than 4.0, that is, somewhere between 'sometimes' and 'seldom').

Table 26: Frequency of use of firearms in crime (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
All the time	0	1	3	6	8	0	0	1	6	2	5	0
Often	0	6	10	11	6	3	6	6	10	2	2	0
Sometimes	33	8	7	24	7	5	39	22	10	3	10	1
Seldom	61	43	38	55	35	55	31	51	59	34	58	16
Never	6	42	42	4	44	37	24	20	15	59	25	83
Mean	3.73	4.21	4.08	3.38	4.00	4.27	3.73	3.83	3.66	4.46	3.97	4.83
n =	79	402	238	238	237	240	219	367	159	217	233	163
Key for mean score: 1 = Most frequent (All the time) < > 5 = Least frequent (Never)

Frequency of being exposed to violent conflict involving a firearm

Across all the regions, the mean score indicating the frequency of exposure to violent conflict involving a firearm was above 4.0. This showed that most respondents had 'seldom' or 'never' been exposed to violent conflict involving a firearm. Respondents from Arusha, Morogoro and Mwanza have been exposed to marginally more violent conflict than respondents from other regions, although the level of exposure still remained low.

Sensitivity to the experiences of the people of Arusha, Morogoro and Mwanza needs to be shown when developing a strategy to reduce firearm proliferation. The respondents from these provinces would in all likelihood feel more insecure, and thus be less willing to engage in discussions or processes linked to micro-disarmament. Consequently, linking any disarmament initiatives to conflict prevention and resolution is likely to enhance both their acceptability and their long-term success.

Table 27: Frequency of being exposed to violent conflict involving a firearm

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Mean	4.23	4.63	4.47	4.82	4.38	4.73	4.47	4.23	4.81	4.78	4.81	4.61
n =	79	387	238	203	237	227	220	344	160	218	238	159
Key for mean score: 1 = Most frequent (All the time) < > 5 = Least frequent (Never)

Frequency of hearing gunshots

The regions of Arusha, Mwanza and Kigoma heard relatively more gunshots than the other regions, although hearing gunshots was still a relatively infrequent occurrence. In Zanzibar, the incidence of hearing the discharge of firearms was very low with only 19% seldom, and 80% never, hearing gunshots.

Change in the number of firearms

The region with the most significant increase in the number of firearms compared to previous years was Kigoma. Overall, 80% of the Kigoma respondents indicated that the number of firearms had increased. The reason for the significant increase in the number of firearms in Kigoma may be attributed to the decades long destabilisation caused by the spill over from conflicts in Burundi, Rwanda and the Democratic Republic of the Congo. This spill over included rebel groups training in this region of Tanzania, as well as the victims of conflict seeking refuge in Tanzania.

Table 28: Frequency of hearing gunshots

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Mean	3.95	4.12	4.02	3.61	4.47	4.48	4.07	3.89	4.27	4.67	4.03	4.80
n =	79	401	238	238	238	239	227	372	160	220	239	162
Key for mean score: 1 = Most frequent (All the time) < > 5 = Least frequent (Never)

In addition to this, frequent raids are carried out across Lake Tanganyika by rebel groups seeking supplies for their soldiers. Having stated this, one should also note that Kagera seems to be afflicted by some or all of these destabilising factors, yet there does not appear to be the same level of increase in the number of firearms. Possible explanation for this lower level of increase in the number of firearms may be that the instability in Kagera was very bad, and although still by no means particularly good now, is relatively speaking getting better. This would explain why the number of firearms is decreasing and there is a general sense that Kagera is safer than it was but also why there is still a considerable concern about being shot and about the general impact of firearms, as identified elsewhere in the survey.

Table 29: Change in the number of firearms (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Increase	51	54	53	80	39	25	72	43	45	12	64	34
No change	24	30	21	9	27	30	12	40	35	59	26	55
Decrease	25	16	26	11	34	45	16	17	20	29	10	11
Mean	2.71	2.60	2.67	2.16	2.74	3.23	2.12	2.58	2.73	3.25	2.37	2.74
n =	79	334	224	224	215	193	122	295	140	200	163	73
Key for mean score: 1 = Maximum increas < > 5 = Maximum decrease

Other regions that experienced comparatively significant increases in the number of firearms were Morogoro (72% of the Morogoro respondents indicated an increase in firearms) and Tanga (with 64% of the respondents reporting an increase). Training camps in Morogoro used by the Umkontho we Sizwe combatants in the 1960-1980s are now housing Burundian refugees The significant quantity of firearms present in Morogoro stems both from the legacy of the resistance fighters based in Tanzania— and the large quantities of weapons that they left behind on their return to South Africa—and from the fresh influx of weapons that has accompanied the arrival of many of the Burundian refugees. The increase in the number of firearms in Tanga, unlike in Kigoma and Morogoro, can be attributed to issues relating to both armed conflict and criminal activity. With the expansion of the mining industry in Arusha and Tanga, the police have been faced with an increase in crime, which is specifically linked to the mining of tanzanite and other semi-precious stones. Mbeya was the only region where a significant number of respondents (46%) felt that the number of firearms had actually decreased. On balance, although a clear trend across the whole country is difficult to identify, the number of firearms appears to be increasing, although not at an overly alarming rate.

Indicators of firearm ownership

The Tanzanian police award firearm licences to civilians for three main reasons, namely: security; hunting; and the protection of crops from wild pigs and monkeys. Firearms are registered at the Firearm Registry of the Tanzanian Police in Dar es Salaam. Firearm licences are approved in a hierarchical manner, whereby recommendations for approval are transferred from the district, to the regional and finally to the national level. This process is to ensure that only those people fit to possess a firearm are awarded a licence. The main limitation faced by the Firearm Registry is the lack of computers and a suitable computer network which hinders staff in keeping records up to date. The need for computer networking to link regional offices to the head office was seen as a very important mechanism to overcome existing delays in the submission of data. Electronic or technical communication was seen as the most effective mechanism in transmitting reliable information to the department1 to enable the Firearm Registry to keep and maintain up-to-date records.

No assessment was made of what proportion of the Tanzanian population licensed their firearms. However, given the large rural population and the hierarchical approval process, it may be assumed that a fair proportion of the population did register their firearms. There are a number of possible reasons for this, including the ease with which prospective licensees can submit applications and the relatively small size of communities—in which vigilance and knowledge of the business of others is fairly great and thus the possession of a firearm is difficult to conceal. When this is combined with the role of traditional authorities, for whom the level of respect is likely to be high, the pressure to abide by the law will be great. However, three major categories of people would be by-passed by this formal registration system. These include the refugees in Kigoma, Kagera and Morogoro, Somali traders and the nomadic Masaai. In this section we will attempt to provide a profile of firearm ownership in Tanzania. Due to the fact that police conducted the survey and respondents usually under-report illegal firearm possession in quantitative surveys, it is assumed that the following results indicate a minimum level of firearm ownership in Tanzania.

Knowledge of a friend with a firearm

With regards to the question, 'do you know a close friend or family member who has a gun?', it is best to first look at those respondents that indicated 'no' as it sometimes the case that respondents choose the 'do not know' response when they know the firearm in question is acquired and/or possessed illegally.

Legal firearm ownership is lowest in Zanzibar and Kigoma. The indication that 95% of the respondents in Zanzibar do not know someone that owns a firearm is clearly problematic data. It is felt that due to the secessionist sentiments in Zanzibar and the resultant perceptions among the local population of the police headquarters on the mainland, it may be best to disregard the Zanzibar responses to this question. The low firearm ownership in Kigoma, when considered in light of respondents from Kigoma indicating the highest increase in firearms across all sample regions, would suggest that the increase in firearms is largely in illegal weapons. Another possible explanation to bear in mind, however, is that firearm ownership was previously very low.

Legal firearm ownership would appear to be highest in Pwani, Arusha and Ruvuma. It is interesting to note that two of these regions, Pwani and Arusha, surround major urban centres—respectively, the cities of Dar es Salaam and Arusha. This suggests that the location (urban or rural) may be a factor influencing the level of firearm ownership. The high level of legal firearm ownership in Ruvuma may, in part, also be explained by its location. Ruvuma borders Mozambique and Malawi via Lake Nyasa along which there is a great deal of trading—both legal and illegal—and as with the other border lakes of Tanganyika and Victoria, is insufficiently policed.

Figure 12: Knowledge of a close friend or family member with a firerarm (percentage)

Firearm licensing

It is not surprising that the vast majority of the respondents across all regions indicated that they suspected that the firearm in question was licensed. The reason for this is that respondents in this type of survey are reluctant to admit to, or implicate themselves or close friends or relatives, in illegal activity, i.e. possessing an unlicensed weapon. The tendency will have been compounded by the fact that the survey was conducted by police officers. The highest level of firearm licensing was reported in Pwani, Ruvuma and Morogoro, while the lowest level of licensing was reported in the regions of Kilimanjaro, Manza and Kagera.

The regions with low levels of licensing could be identified as priority areas for public awareness and education campaigns when the government starts to address issues related to firearm licensing.

Table 30: Level of firearm licensing (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Definitely licensed	53	58	41	73	32	68	78	49	46	67	81	76
Most probably licensed	32	25	31	10	31	15	14	17	47	24	9	3
Sub-total licensed	85	83	73	83	63	83	92	67	94	91	90	79
Definitely	4	6	11	3	15	13	1	15	4	4	6	19
Most probably unlicensed	11	11	17	13	22	5	6	18	3	4	10	2
Sub-total unlicensed	15	17	28	17	37	17	8	33	6	9	10	21

Access to a firearm

The regions with the highest level of firearm access were Kigoma (where 24% of the respondents had access to a firearm), Mwanza (19% of the respective respondents), Tanga (18% of the respondents), Kagera (15% of respondents) and Arusha (17% of respondents). Both Kigoma and Kagera are regions of high insecurity. Similarly the residents of Tanga and Mwanza (located on the shores of Lake Victoria) are increasingly obtaining firearms for protection. The northern part of Tanzania, where these regions are located, are also some of the richest areas for hunting in Tanzania, which may further explain the relatively higher levels of access to firearms. Explaining Arusha's relatively high apparent incidence of firearm possession is slightly more difficult, as it borders neither a country openly riven by conflict nor a major lake, as Kagera, Kigoma and Mwanza do. In addition, Arusha does not appear to suffer from above average levels of insecurity. However, possible influences on the level of firearm possession in Arusha might include the presence of the major urban centre of the city of Arusha, that the region is the centre of tanzanite mining and that organised criminal groups are based in the region.

The lowest level of firearm access was reported in Mbeya, Pwani and Kilimanjaro, where less than 5% of the sample indicated they had access to a firearm if they needed one. The low level of firearm ownership in Kilimanjaro is expected, as this is one of the safest provinces in Tanzania. The main road transport route to Moshi, the town located closest to the tourist attraction of Kilimanjaro, obtains disproportionately more resources and attention than any other route within Tanzania as a result of the high priority the government places upon its tourism industry. The low level of access to firearms in Pwani, however, is slightly surprising given the responses to other questions. Respondents in Pwani indicated a general decline in safety (rising crime and a falling sense of peacefulness) and a relatively high incidence of legal firearm ownership (as indicated by knowledge of others who possess a firearm). Analysis of other indicators should help to illuminate the true picture of firearm diffusion in Pwani.

Figure 13: Access to a firearm (percentage)

Actual firearm ownership

Respondents were asked to indicate the number of household assets their household possessed. Respondents were asked to indicate the number of each item that they possessed from a list including things such as radios, bicycles and cell phones amongst others, but also firearms.

Figure 14: Possession of a firearm as a household asset (percentage)

Amongst those respondents that had firearms in the household, less than 1% indicated that they had more than a single firearm.

The highest level of firearm ownership was found in Morogoro (19%), Arusha (18%) and Mwanza (15%). The lowest indications of firearm ownership were in Kagera (4%) and Zanzibar (1%).

Willingness to own a firearm

Over a quarter of the sample was willing to own a firearm. The greatest willingness to own a firearm was in the region of Morogoro. The disproportionately large quantities of firearms in this region, as a result of the previously mentioned insurgent training camps, could influence the attitudes of these respondents. It was found that 55% of the Morogoro respondents would be willing to own a firearm.

The Kilimanjaro respondents were the least willing to own a firearm with only 21% of them stating that they were willing to possess a firearm. Reasons for this low willingness can be found in the relatively small amount of land available for hunting (as the many large national parks here prohibit hunting) and the large number of police in service, who bolster levels of safety. Willingness to own a firearm is often linked to the need to own a firearm, although broader social, cultural and historical factors, such as the extent of hunting, previous exposure to weapons and past association with firearms (in fighting independence struggles for example) will also be important.

Table 31: Willingness to own a firearm (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Yes	30	42	28	37	21	31	55	30	29	41	41	5
No	51	40	64	59	66	57	42	53	62	43	58	87
Do not know	19	18	8	4	13	12	3	17	9	16	1	8

Need for improved controls over firearms

Local residents are often in the best position to identify the need for better controls over firearms. They often have a better grasp of the reality of the situation and their perceptions provide reliable indicators to guide government policy.

Besides the relatively low response by the region of Arusha, all the other regions identified the need for improved controls over firearms. However, the respondents from Arusha had the largest 'do not know' category of all the regions. The undecided proportion of the sample implied they still had to make up their minds on the issue.

Clearly this gives the government of Tanzania a mandate to address the issue of firearm proliferation. The lack of control of firearms could extend to include such aspects as the need for tighter firearm licensing, seizure of unlicensed firearms and audits of existing state-owned firearms.

Table 32: Need to improved firearm controls in area (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Yes	56	82	85	89	98	94	96	79	91	90	98	72
No	9	1	9	4	1	3	0	10	4	3	1	6
Do not know	35	17	6	7	1	3	4	11	5	7	1	22

Table 33: Frequency of worrying about injury with a firearm (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Always	4	8	38	32	36	2	42	5	55	17	14	25
Often	4	9	18	6	2	2	13	4	4	4	6	4
Sometimes	17	10	20	21	2	5	5	11	4	4	8	1
Seldom	64	30	19	38	21	40	23	55	29	22	56	25
Never	11	43	5	5	39	51	17	25	8	53	16	45

In a previous section of the report, respondents were asked to indicate shifts in the prevalence of a number of health-related concerns such as HIV/AIDS, cholera and malnutrition. When respondents were questioned about these health-related concerns, firearm injury was one of the issues to be considered. Respondents from Pwani (84%), Zanzibar (77%) and Kigoma (71%) indicated an increase in the number of people affected by firearm injuries. In Kagera (83%), Dar es Salaam (63%) and Mwanza (51%) firearm injuries were considered to have decreased.

Figure 15: Change in the number of firearm injuries (percentage)

The link between fear and risk of injury was not as direct as was expected. The respondents from Pwani indicated the highest level of fear of injury, as well as the highest level of actual firearm injuries taking place. However, this was not a consistent pattern for the other regions.

Concern about firearms scattered throughout the country

The respondents from Kigoma (85%), Kagera (76%) and Pwani (79%) indicated the greatest concern over the fact that there are so many firearms scattered throughout Tanzania. The lowest levels of concern were measured in Arusha (35%) and Mbeya (40%).

Affected by firearms

Respondents were asked to indicate the extent to which they were affected by firearms. This is, in part, an indication of the impact of firearms. The respondents from Kagera, Kigoma and Zanzibar have been the most affected by firearms, with a significantly greater percentage of respondents than from other regions indicating that they had been affected by firearms. It was found that 49% of respondents from Kagera, 33% from Kigoma and 33% from Zanzibar, have been affected by firearms. The impact of firearms in other regions was only negligible, with less than 5% indicating that they had been affected by firearms.

Table 35: Extent of being affected by firearms (percentage)

	Arusha	Dar es Salaam	Kagera	Kigoma	Kilimanjaro	Mbeya	Morogoro	Mwanza	Pwani	Ruvuma	Tanga	Zanzibar
Large	7	2	49	33	2	2	2	2	3	2	2	33
Same as before	16	4	8	4	14	21	2	43	1	20	2	33
Small	77	94	43	63	84	77	96	55	96	78	96	34