|
The Central Firearms Registry:
Assessing data accuracy
According to a recent survey the Central Firearms Registry (CFR) is less than 30% accurate. This means police have a three in ten chance of tracking the holder of a licenced firearm. The CFRs computer system needs urgent upgrading, and training will improve the accuracy of data capture. Mechanisms should be developed to verify and update information supplied by licence applicants.
The Central Firearms Registry (CFR) is a department of the South African Police Service (SAPS) that approves and regulates all licences, permits and authorisations pertaining to firearms in South Africa. As such it is vital for controlling and managing firearms. The Registry records the personal details of all firearm licence holders, as well as the contact information about the firearm (its make, tracking serial number and history). The police rely on these records to trace firearms and link individuals to the crimes committed with firearms.
The CFR is currently staffed by 245 members with 105 positions vacant. The Registry comprises 15 sections that handle applications from the different provinces, security companies, dealers, imports and exports of firearms. In 1998, each employee received an average of 70 license applications per month. Of these, 59 licences were approved per employee.
Not all applicants receive a licence (Figure 1). In 1998, 173744 of the 205 897 firearm licence applications were approved. In other words, 26% were turned down. Applicants can appeal against this decision. In 1998, 32% of those who were refused licences made successful appeals.
There are currently 3527035 firearm owners registered at the CFR. Of these, 655425 have more than one firearm. Based on these figures, a conservative estimate puts the number of licensed firearms in South Africa at 4,2 million. Table 1 outlines the number of firearms registered in each province.
Accuracy of the CFR
The accuracy of information recorded by the Registry has been questioned by the police and by other organisations. Data has become unreliable over time as a result of several factors:
Figure 1: Firearm applications received and approved by the CFR

Source: CFR
- Licence holders fail to notify the CFR about changes in their personal details and those of their firearms. This is the main reason for poor data quality. Licence holders may be unaware of their legal obligation to inform the CFR of any such changes.
- It could also be an intentional action by licence holders.
- The CFRs computer system is outdated and requires urgent capital investment to function optimally. The system was developed in a piece-meal manner and does not serve the current needs of the user. It is also outdated, inefficient and slow. The menu driven software is largely written in Afrikaans, which limits the CFRs scope for employment and frustrates users.
- Staff at the CFR are not appropriately trained in the operation of the computer system.
- With the reintegration of the former homelands into South Africa, information on firearm holders in the homelands was merged with the CFRs database. In many cases records were lost during this process, or incorrectly captured.
Table 1 Number of firearm licenses approved by the CFR
|
1994
|
1995
|
1996
|
1997
|
1998
|
|
KwaZulu-Natal
|
30 746
|
18 018
|
24 016
|
24 982
|
24 351
|
|
Gauteng
|
83 363
|
53 502
|
62 291
|
70 228
|
61 263
|
|
Northern Cape
|
6 040
|
4 058
|
4 407
|
4 749
|
4 455
|
|
North West
|
14 646
|
9 931
|
13 010
|
13 398
|
12 871
|
|
Western Cape
|
30 854
|
21 406
|
22 743
|
19 717
|
18 254
|
|
Eastern Cape
|
12 588
|
8 361
|
26 017
|
15 580
|
13 874
|
|
Northern Prov
|
8 890
|
7 033
|
8 682
|
9 643
|
10 152
|
|
Mpumalanga
|
15 442
|
10 635
|
12 676
|
13 699
|
14 663
|
|
Free State
|
15 887
|
9 747
|
11 533
|
12 194
|
13 861
|
|
Total
|
220 450
|
144 686
|
187 371
|
186 187
|
175 742
|
Source: CFR
Survey of licence holders
At the request of the department for safety and security, the Institute for Security Studies (ISS) undertook an audit of 2458 licence holders in all nine provinces. The survey, conducted in June and July 1999, aimed to measure the accuracy of the CFR.
The survey was undertaken at 61 points across the country. For each sample point and each year between 19821999, the CFR randomly selected 10 to 15 licence holders names. From this larger sample, the ISS selected between 210320 respondents in each province.
Members of the SAPS Illegal Firearm Units administered the survey. Officials visited the addresses given by licence holders, where they verified the information contained in the CFR. Details such as the name, address, and identity number and whether the licensed firearm was still in possession of the licence holder were checked.
Of the 2458 names given to the SAPS, only 1138 (46%) were verified. A few of the outstanding 1320 records were not verified because respondents declined to be interviewed. Seventeen percent of licence holders were not at home when the police arrived and the remaining 35% of questionnaires were not returned by the SAPS. In anticipation of these problems, the sample was increased at the outset to ensure valid numbers on return.
Of those license holders whose whereabouts police could determine, only 28% were living at the address reflected on the CFRs database (Figure 2). This means that police have a 28% chance of successfully tracing the owner of a legal firearm.
Figure 2: Result of police attempts to locate licence holders at addresses registered with the CFR (n = 1138)

Source: ISS Integrity of CFR survey
In several cases, the name, address and identity number of those licence holders that were successfully traced by police were incorrectly recorded on the CFRs database (Figure 3). Furthermore, only 84% had all the correct firearm(s) in their possession. The 313 licence holders had licenses for 796 firearms an average of 2.5 firearms per person. Of these firearms, 91% were with their lawful licence holders. Of the 72 firearms that were not in the possession of the licence holder, 19% had been stolen, 10% had been sold and 5% were lost (Figure 4). In all these cases, owners failed to notify the CFR of the events.
Figure 3: Accuracy of licence holders' personal details
(n = 323 )

Source: ISS Integrity of CFR survey
Figure 4: Reason why firearms were not in licence holders possesion
( n = 74 )

Source: ISS Integrity of CFR survey
Main problems
The main source of concern is the failure of licence holders to notify the CFR of a change in residential address. The department for safety and security should ensure that gun owners are aware of this responsibility when they register for, and obtain, a licence.
The fact that 15% of the licence holders addresses did not exist suggests that applicants details are not being verified by the CFR as part of the registration process. The survey also found that 20 licence holders were deceased. An executor of an estate is responsible for notifying the CFR of a licence holders death, but this information is not always transferred.
Clare Hansmann and
Ettienne Hennop,
Institute for Security Studies
This article is sponsored by the following funders:
Royal Norwegian Government
Royal Swedish Government
Government of the Netherlands
Government of Switzerland
European Union

|
|
|