The Effectiveness of the Present System of Firearms Control

5.1 Strengths of the Present System

Vetting for Personal Fitness

The 1983 Act proceeded from the premise that it should be possible, by careful investigation of applicants’ suitability to possess firearms, to avoid, or at least much reduce, the likelihood that firearms would fall into the wrong hands, and accordingly that systems which sought to link particular weapons to particular individuals could be abandoned.

The reliance which that policy placed on personal vetting required the development of new and stricter vetting procedures. In addition to computer tests conducted to ascertain whether the police records contained anything unfavourable to an applicant, the applicant was required to give detailed personal information and to provide the names of referees with whom the Police could check the correctness of the information given and the possible existence of known unfavourable characteristics.

However, the full extent of these procedures applied only to new applicants, and applicants for 1983-style licences who had licences or certificates of registration or a permit to buy a firearm issued under the previous legislation were subjected only to computer checks. If these checks did not show any matter of concern, no other vetting took place. In the result by the far greater majority of those who received lifetime licences in the 1983 to 1985 Project Foresight licensing were subjected only to relatively casual investigation.

The 1992 Amendment was in large part a response to the Aramoana tragedy. David Gray, of Aramoana, had at the time a current firearms licence, and it was thought by many people (though not by the Police Complaints Authority) that had he undergone a complete vetting the tragedy might have been avoided. The 1992 Amendment was accordingly designed to provide for the regular vetting of all licensees, commencing with a re-vetting of the existing licensees over the period 1993 to 1997. Those seeking special endorsements were to be the most intensively investigated and were to provide additional referees. A new national standard checklist provided "a guide to the bare minimum requirements". Vetters were advised that visits to applicants’ addresses would be "invariably mandatory" for all endorsement seekers, and "usually mandatory" for A licence applicants.

By mid-1994 it was apparent that the Relicensing Project was more expensive and resource-intensive than had been estimated and that a very low level of compliance was being achieved. After a conference held to consider how to "streamline" the re-licensing process, a more discretionary vetting procedure was settled. While it was still to be obligatory to check with the first referee (a spouse, partner or next of kin), this might be conducted by telephone. It would then be within the discretion of the vetter whether a second referee were spoken to or some other check made on the applicant’s suitability. Those procedures applied to a large majority of the applicants.

In 1995 a new system was introduced which enabled a record (or "flag") to be kept on the Wanganui computer of persons who appeared to a police officer to be unsuitable to hold firearms. Its purpose was to ensure that if an application for a licence were made subsequently, the opinion recorded and the reasons stated for the opinion would come to the attention of the officer required to consider the application.

Analysis of Police Files

In order to get some idea of the quality of the vetting process, and the significance both of the discretionary jurisdiction to refuse or revoke licences and of the practice of "flagging" people as unsuitable for the receipt of licences, three sets of files were obtained from the Police. These comprised: 89 revocation files, randomly selected from a variety of districts, from the period 1993 to 1996, which represented 5 percent of the files recorded as revocations in the Wanganui system over that period; 72 refusal files randomly selected from as far back as 1992 and representing 3.5 percent of the refusals recorded over that period; and 67 files relating to persons flagged as "unsuitable", again randomly selected from a variety of districts from 1995 and 1996, which represented 22.6 percent of the total files flagged over that period.

The two matters arising from the examination of those files which are significant at this stage are:

  • that the analysis of the 228 files disclosed a generally conscientious and cautious approach by the Police to their duty if anything the officers appeared to lean towards the "better safe than sorry" approach, both when considering applications and revocations and when creating an "unsuitable flag"; and

  • that while some of those who were refused licences or had their licences revoked may have had a case for the reconsideration of such decisions (and of course, could have sought a reconsideration by the Courts if they had thought fit to do so), a reading of the files left no doubt that the majority of the decisions made were thoroughly justified.

While it is impossible to know how many crimes were averted or tragedies avoided by those decisions, while the vetting system may be amenable to further refinement, and while either in its present form or with such refinement as can be devised the system cannot be wholly effective in eliminating the unsuitable, it would be plainly wrong to discontinue the personal vetting process. It is outstandingly the most useful feature of the present system.

It is appropriate to compare our vetting system with those operating in the United Kingdom and Australia. The English system is discussed at length in the Cullen Report. That report describes a careful but less detailed examination of applicants than ours, but at least as close a re-vetting on renewals, when English licensees are expected to justify the continuance of their licences and to prove continuing use. In Australia the latest reforms propose closer examination of applicants for licences, but as yet the procedures remain unsettled. To this point, save possibly in Western Australia, the New Zealand vetting process appears to have provided a closer and more effective examination than those which have operated across the Tasman.

Safety Training

There can be little doubt that the work of the Mountain Safety Council team has played a significant part in reducing the incidence of firearms accidents in New Zealand. Part 6.2.2 of the report considers whether the training courses presently run could be extended to incorporate practical training and refers to recent Australian experience of a similar development. Whether or not that development is approved, the size and value of the commitment of the Mountain Safety Council team, and the value of requiring training as a pre-condition of the grant of licences, should be seen as a strength of the present system.

Use of Club Disciplines

The only formal incorporation of club structures into arms control is in relation to pistol shooting (as described in part 2.6). Only those who are members of pistol clubs recognised by the Police may be issued with a pistol endorsement, and pistol shooters are required to take part in club activities at least 12 times a year.

Possible reforms of the pistol provisions are considered in part 6.1.1 of the report. At this point it is appropriate to record that the Police’s control of handguns in New Zealand is very much supported by the disciplines imposed on their members by pistol clubs, acting under the supervision of the New Zealand Pistol Association. Those disciplines have achieved a truly commendable record of safety in the use of handguns in club situations. Since this is a real strength in the present system, consideration should be given to the practicability of enlisting other clubs’ support. This would depend upon their acceptance of the appropriateness of rules imposing similar standards of discipline upon their members, and of reporting to the Police any fall from those standards.

5.2 Weaknesses of the Present System

The principal systemic weaknesses are described in this section; weaknesses in the administration of the system are considered in part 6.4.

5.2.1 Absence of control over firearms, as distinct from shooters

The present system registers particulars only of restricted weapons, handguns and MSSAs, which together comprise no more than 4 percent of the total armoury. The consequences of that situation are:

  • that the "arms controllers" have no idea of the numbers of firearms within the country or within the control of any particular individual, or where those firearms are;

  • that they have no effective control over dispositions of firearms, either by sales, transfer or other dispositions between living persons, or over the manner in which firearms are dealt with upon the death of a licensee;

  • that the fact that particular firearms cannot be traced to their owners gives no encouragement to owners of firearms to secure their firearms properly, or to report the theft of guns, which might cause an inquiry as to the security under which they had been held;

  • that the system cannot generate enough information to inform and develop firearms policy; and

  • that some evidence which would be useful for crime prevention and detection is lostū less frequently than many would expect, but in cases where it could be of major importance to the solution of a serious offence.

Details of the arguments for and against the registration of firearms are set out in part 6.2.1 of this report, and will not be reviewed here. For present purposes it is sufficient to note the adverse consequences which flow from its absence.

5.2.2 Ten-year licences

The stated police preference in 1980, before their belated turn away from a combined licensing/registration system to one which depended wholly on licensing and the vetting of licensees, was for a combined system with licences to be renewed every three years. Instead the 1983 Act moved to lifetime licences.

In 1992 the police preference was for three- or five-year licences. They were advised that political considerations required that licences be for a ten-year period, notwithstanding that it was understood by Government that the rationale for abandoning lifetime licences was to ensure sufficiently frequent reviews to avoid the risks involved in a long period without re-vetting licensees. The same record also shows that the particular problems likely to arise from changes of address over any lengthy licence term were canvassed at that time.

Subsequent experience has shown that the ten-year period has caused major problems in maintaining accuracy of the licence register. Various calculations have been made of the size and significance of the address problem. In the report Policing Gun Laws prepared for the New Zealand Police Association, Mr Philip Alpers reported police estimates of errors in addresses ranging from 30 to 50 percent. A December 1995 draft review prepared by PNHQ calculated that 19.66 percent of the call-in letters sent to licensees with the initials C to J were being returned. In a review of this topic prepared on 9 December 1996 for this Review by the Firearms Licensing and Vetting Co-ordinator, he estimated that for every call-in an average of 20 percent of letters were returned to the Police, but said that the Police considered 25 percent to 30 percent "the most likely" overall error rate.

Because one of the key questions which must be considered before proposing the introduction of new methods of administration must be whether or not existing police records are accurate, the Review commissioned Mr Reece Walters of the Institute of Criminology to make an independent examination of police records and estimate the error rates in respect of:

  • address information contained on PNHQ licensing records; and

  • details in the MSSA and pistol registers.

His research was based on an analysis of 159 licence files in respect of both first licensing and renewal applications, randomly selected from the years 1993 to 1996.

The name and address details were first checked against TESSA (Telecom Emergency Services Support Application) data and the 1996 Electoral Roll. Follow-up letters and telephone calls were made to those licensees whose details were not established by the TESSA and Electoral Roll search. This disclosed 85 percent correct addresses, a somewhat better result than police personnel within firearm licensing had predicted, and markedly better than that which had been suggested in Policing Gun Laws. Some of the difference arises from the more recent material he was considering. At the same time, the nature of the errors in the address details for the 15 percent not cleared was considered to indicate that two-thirds were "effectively untraceable".

The inquiry then proceeded to consider the accuracy rate for name, address and date of birth, all data important for tracing and endeavouring to correct irregularities or non-compliance. The additional factor reduced the number of correct recordings to 80 percent.

Mr Walters then obtained from the Land Transport Safety Authority, the Ministry of Fisheries, the Fish and Game Council of New Zealand and the Maritime Safety Authority advice about the accuracy of their records and the effect
of longer or shorter licence periods. That advice showed rates of accuracy in roughly inverse ratio to the length of the licence period, ranging from 50 percent accuracy for lifetime licences to over 90 percent accuracy with annual licences. Details provided in the report appear to justify the conclusion that:

Information systems which are regularly accessed and updated are more likely to increase accuracy and hence efficiency than data which remains "inactive" for long periods of time.

A further problem arising from the ten-year licence period is necessarily the accumulating effect of deaths of licensees. Part 6.2.1 of the report recommends protocols to deal with that problem, but clearly a shorter licence period would prevent the accumulation of such defects, which would accrue at the rate of 1.5 percent per annum if licensees are affected by the national mortality rate for persons over 18.

5.2.3 The competition of arms business with all other police business

Time and time again I was advised by police officers or former police officers with experience of the Arms Office business that pressures on the Police to respond to other more dramatic and urgent needs had resulted over recent decades in the arms business being given a progressively lower priority, and becoming under-resourced. Other common concerns were that there was a reluctance at district or regional level to devote resources to monitoring compliance with the arms code or to devote resources to the detection and prosecution of other than the most blatant breaches of it.

It is impossible for me to make any satisfactory measurement of the extent of the support allocated to different aspects of police work, let alone trace the trends in such allocations.

It is, however, appropriate to note:

  • that this view was expressed not only by those presently working in arms offices, but also by others who had moved out of such positions and are now back in the main stream of police work, but are aware of the strains there;

  • that it would not be difficult, if faced with the responsibility of apportioning resources across the whole gamut of police responsibilities, to see arms control business as less urgent than some other police duties; and

  • that the history of recurring underestimation of the needs of arms control is hard to understand unless that work has indeed come to be seen as a residual obligation, rather than an area with its own particular needs which have to be provided if it is to function properly.

It is accordingly a matter of some importance that, whatever variation on the current arrangements is selected, it incorporates some means of ensuring that competition with other work does not in the future result in a similar suppression of the development and maintenance of an efficient arms control system.

5.2.4 Complexity and awkwardness of the arms code

The arms code is contained in the Arms Act 1983 (a 44-page statute with 78 sections), the 1992 Amendment (a 21-page statute with 39 sections) and the Arms Regulations 1992.

The 1983 Act is in itself difficult to construe, and its construction is certainly not helped by the circumstance that the 1992 Amendment was clearly compiled in haste and relates uneasily to the principal Act.

Harding’s 1981 review of the Australian gun legislation criticised its structure:

As in so many other areas of law, most jurisdictions have allowed a situation to develop where the sources are so numerous or so obscure that the ordinary citizen cannot reasonably be expected to be able to ascertain what the operative law actually is.

Since that was written several States have amended their gun laws. It is a sad commentary on the state of the New Zealand code that, by comparison, most Australian legislation is markedly more user-friendly and comprehensible. If there are to be major reforms it is important that time be taken to rewrite the legislation in modern form, and that another layer of complexity is not simply added to what is already complex and obscure legislation. This is recommended in part 6.2.5.

5.3 Viability of the Present System

It will be apparent from the content of parts 5.1 and 5.2 that in my view the weaknesses of the present system far exceed its strengths.

The recent acceptance by the Police that their earlier support of the present system is no longer appropriate and that the system is in need of radical reform, reduces the need to justify my assessment. However, it is sufficiently important to call for some additional justification. That can be found in the facts:

  • That the present 1992 Relicensing Project is not meeting its stated principal objectives. This was recognised in a series of reports by the Firearms Co-ordinator, most plainly in a report dated 21 March 1996 which advised that the intended compliance rate was unachievable, that instead of being self-funding costs were exceeding revenue by 80 percent, and that the project was generating a large number of surplus firearms and there was no strategy to deal with that problem.

  • That the reassessment of costings made by Coopers & Lybrand for the purposes of this Review found that actual costs of the licensing project considerably exceed police estimates.

  • That in its present form the system cannot provide sufficient information to allow its administrators, or the government, to recognise and take account of changes in gun use and ownership.


The question is not whether the present scheme should be sustained, but the extent and timing of necessary change.

5.4 Effect and Costs of the Enhancements Proposed in the May 1996 Review

The cost estimates in the May 1996 Review were not easy to construe. Estimates of the cost of operating the existing system appeared to be net both of income and of some allowance for "ordinary" arms work, but the amounts of the offset in each case were not stated. Similarly the cost of the proposed enhancements was not clearly spelt out, the largest cost figure given being a sum of $400,000. When further particulars were sought I was advised that the Police had decided to ask Coopers & Lybrand to cost both the present scheme and the suggested enhancements, and that that firm had been asked to confer with me about any particular costing issues arising in the course of the Review.

That finally resulted in my receiving from Coopers & Lybrand two reports. The first, dated March 1997, estimated the cost of:

1. the present scheme; and

2. the enhancements to that scheme proposed by the May 1996 Review.

The second, dated May 1997, estimated the cost of:

3. a range of possible "buy-back" options; and

4. a combined licensing and registration system, with three-year licences.

I requested that the estimate of the cost of the present scheme include the cost of issuing the final batch of call-in notices and of taking steps to complete the call-in programme, so far as that can be achieved, by 30 June 1998, as I agreed with the police view that it would be wrong to stop the programme at this time. That work apart, the first estimate looked at an existing situation. Accordingly, although it must have been affected to a degree by the state of the arms records, there is no unusual factor which would call for particular caution in receiving and applying that estimate.

In summary Coopers & Lybrand estimated the current cost of firearms control to be $11.1M per annum, made up of $7.5M for licensing activities and $3.6M for monitoring and enforcement activities. The main reason for the increase from the May 1996 police estimate of $7.2M was said to be "the different costing methods used, ie. the somewhat disparate nature of previous costing exercises".

The other three costing exercises were all affected, in one degree or another, by the circumstances that on a number of central factual issues there was insufficient reliable information to permit precise assessment, and that I found it impossible to signify in advance of the completion of the inquiries being made for the Review what my recommendations were going to be. A perusal of the two reports shows a commendable professionalism in their endeavour to overcome those difficulties and provide helpful guidance on the principal cost issues. In my view the reports will do that provided their readers note and keep in mind the advice that:

Given the high number of assumptions and the broad level of process definition, costs indicated in the report should be regarded as broadly indicative of the level of expected expenditure.

I return then to look directly at the subject of this part of the report, the effect and cost of the enhancements proposed in the May 1996 Review. While that Review recommended a large number of changes, as befitted a report which declared the current system to be effective and to have a high level of integrity, the amendments or "enhancements" it proposed were preceded by advice that "none of these matters are major". I would agree with that assessment of the enhance-ments so far as it relates to their effect on the system. For the greater part they are sensible proposals which could be expected to have some beneficial effect on the existing system, but their overall effect on its inadequacies would be slight.

Some dealt with technical or drafting questions and do not need consideration here. Those which have more general significance can be grouped and considered under the following headings.

Military Style Semi-Automatics

Recommendations 3, 4, 5, 6 and 7 concern the importation and purchase of MSSAs and parts and a requirement that those obtaining replacement parts surrender the worn parts to the police without compensation. They proceed from the stated assumption that the decisions made in 1992 to allow MSSAs to be licensed as such, or to be "sporterised" by removing military characteristics and thus attaining a "sporting configuration", mean that "the arguments about MSSAs having any use in New Zealand are largely irrelevant".

As the discussion of public and shooters’ attitudes in part 2.7 indicates, that is not the view of a significant part of the general public, nor even the view of a majority of shooters. In part 6.1.1 it is concluded that the balance of public interest is in favour of a ban, which supersedes the need for full examination of the May 1996 recommendations about MSSAs.


Recommendations 8 and 17 propose that it be an offence to fail to comply with security conditions. The only sanction currently available for non-compliance is revocation. As is pointed out in part 5.2, that penalty is inappropriate for minor breaches, and is very seldom invoked. Both recommendations are clearly appropriate and are dealt with in part 6.2.3. It may be an area in which instant fines would be an appropriate procedure.

Recommendations 9, 10, 11, 12 and 13 proposed higher levels of security for dealers’ premises and that the current guide to "minimum security requirements" for dealers be rewritten to provide a more informative and user-friendly publication. Recommendations 14 and 15 propose more rigorous inspection of dealers’ premises, and recommendation 16 proposes that districts liaise with medical occupational and health inspectors regarding inspection of dealers’ premises.

Apart from the last of those recommendations, which seems likely to raise as many difficulties as it would solve and to move away from the objective of having arms control officers adequately trained in assessing security requirements, those recommendations would all assist in improving security. However, in my view the need for better security goes beyond the area of dealers’ security and calls for a restructuring of the security provisions of the code to ensure that levels of security relate to the degrees of hazard involved in the numbers and types of firearms being secured. Proposals to this end are discussed and developed in part 6.1.3.

Continuation and Implementation of the 1992 Relicensing Project

Recommendations 19, 20, 21 and 22 follow a brief description of the state of the Relicensing Project in May 1996. Recommendation 19 suggests a further review to establish the adequacy of current fees and make recommendations on any fee increases required. Recommendation 20 suggests a public campaign aimed at increasing licensing compliance which it estimates would cost in the vicinity of $400,000. Recommendation 22 proposes that Firearms Licensing develop a strategy to assist districts in following up on sporterised MSSAs, and Recommendation 23 proposes that districts develop and improve follow up procedures "subject to additional finance being available".

Leaving aside the question of underestimation of the cost of those proposals, it is my view that they underestimated the difficulty of achieving appropriate response levels at this time. Examination of the results obtained by the few districts which were able to apply significant resources to following up non-responders soon after call-in dates showed differing rates of success, and the belief that even with adequate resources being made available to carry out all further lines of inquiry which can now be followed, 10 to 15 percent of former licensees will be untraceable. The figure may be larger, for inevitably the proportion must increase with the passage of time by reason of deaths, people moving overseas and changing addresses within New Zealand, and others making dispositions of their firearms in the informal manner which has become a common occurrence in this country. The study already reported which assessed the rate of error in the current licensing records found at least 10 percent of those licensees to be effectively untraceable. The basic record for the purposes of the re-licensing programme is the record made during Project Foresight, at least ten years earlier.

As to the likely cost of the enhancements the Coopers & Lybrand estimate is that the initial programmes would cost $7.4M, thereby adding 70 percent to 80 percent to the overall cost of the present system, over the next two to three years. In addition the ongoing cost of maintaining the enhancements is estimated to be $1.8M annually, adding 15 to 16 percent to the annual cost of the present system. Again those estimates far exceed those indicated in the May 1996 Review.

While it must be appropriate to try to wind up the Relicensing Project by seeking responses from the 13 percent of licensees who have not yet been sent call-in notices, when that work had been done and after all the proposed enhancements had been added the end result would remain a system with the basic flaws noted in part 5.2. Unless more radical reform than was proposed in May 1996 is made, the system must continue to fall short of achieving reasonable control of firearms and continue to provide inadequate information to develop better controls. That conclusion, which was provisionally reached in November last, has been progressively strengthened by the further information obtained since then and is, of course, supported by the recent Police Response.

The question then becomes what forms of improvement can be sufficiently devised on the basis of the information now available—which is the topic of the next part of this Review.

Return to the Project Table of Contents


  1. Gubb, to Region and District Commanders re Operational Implications of Arms Amendment Act 1992 "Administrative Instruction No. 1".

  2. Coote, circular re Review, 13 October 1994; and Police Executive Conference Notes, 28 October 1994.

  3. Lord Cullen, The Public Inquiry into the Shootings at Dunblane Primary School on 13 March 1996, 16 October 1996. See chapter 8, and particularly paras. 8.8- 8.13, 8.56- 8.65, and 8.74- 8.90 of the report.

  4. See the Proceedings of the 1980 International Shooting Sports Symposium October 1980, and address by Inspector Cook.

  5. Alpers, Policing Gun Laws. A report commissioned by the NZ Police Association, October 1996.

  6. PNHQ, "Draft Management Review", December 1995, at 11.

  7. Walters, "Accuracy of Police Licensing and Firearm Records", Institute of Criminology, Victoria University of Wellington, April 1997, at 16.

  8. Harding, Firearms and Violence in Australian Life, 1981, at 3.

  9. 9 Coopers & Lybrand, Report on Cost Information Required for Review of Firearms Control in New Zealand, 1997, Part 2, at 10.

  10. Operations Support Group (PNHQ), A Review of Firearms Control in New Zealand ("the May 1996 Review"), at 1.