"REVIEW OF FIREARMS REGISTRATION--TR1994-9e" This paper covers the "Executive Summary" of Technical Report TR1994-9e, REVIEW OF FIREARMS REGISTRATION, which became available in October, 1994. The Summary is a "systematic assessment of the Firearms Registration and Administration System (FRAS) maintained by the RCMP." (Page vii paragraph 2) HIGHLIGHTS: [KEYNOTING ADDED throughout] FRAS has a staff of 29 in 5 work units for: "...processing restricted weapon registration applications..., issuing [registration] certificates [83,913 in 1993], issuing inter-provincial carry permits [statistics hidden], reviewing and processing revocations of certificates [1 in 1993] and permits [1 in 1993], maintaining a national list of firearms dealers [7,839 in 1993], operating a firearms tracing program, preparing the annual Firearms Report [for] Parliament and publishing the National Firearms Manual [revised 226 times from 01 Jan 78 to 26 Jan 94, or once every 26 days for 16 years]." (vii 4) "USER REQUESTS are received and routed within FRAS on an ad hoc basis, with no central screening or clearing point. A paper-file query log is kept, but cannot be analyzed for trends." (vii 4) For $6.9 million per year, that seems poorly organized. THE RESTRICTED WEAPON REGISTRATION SYSTEM (RWRS)... contains records of approximately 1.2 million firearms and 565,000 owners." (viii 5) The purpose of one RWRS record is to uniquely identify one firearm, and indicate its immediate previous possessor, its present possessor, and the location where it is normally kept. The RWRS cannot do that well enough to be cost-effective. Each RWRS record uses one data field for the name and address of the current possessor and one data field for the name and address of the immediate past possessor of one firearm. Each RWRS record uses seven data fields to identify the firearm as one of a group of firearms which share those seven identifying characteristics, and one more data field to uniquely identify one particular firearm with that group--by serial number. "RCMP... STAFF HAVE ESTIMATED THAT A SIGNIFICANT NUMBER OF RECORDS IN THE RWRS DATABASE CONTAIN INFORMATION WHICH IS ENTERED IN NON-STANDARD FORM, OR IS STALE OR INCORRECT." (ix 2) Limited testing of the system indicates that at least 30 per cent of the 1.2 million registration records are useless trash, and that a high percentage of the rest contain inaccurate data. The method used for "uniquely" identifying firearms does not work, and cannot be made to work well enough for an FRAS record to be admissible as evidence in a court of law. Attempts at finding a firearm through FRAS often fail due to the poor quality of the records. The RWRS cannot and does not fulfill its purposes. "There are many variations in the 'make', 'model', 'calibre', and 'barrel' fields, WHICH WERE NOT EDITED AT THE TIME OF DATA-ENTRY. The 'type' and 'action' fields, which were edited, do not appear to contain intrinsic errors." (ix 2) The information in the latter sentence is very unlikely to be true. It has long been a habit of FRAS clerks to automatically type in "HG" (for HandGun) even when the firearm is not one, simply because well over 90 per cent of all entries are "HG." A surprising degree of firearms expertise is required to correct or even notice such an incorrect entry. "An August, 1993 evaluation of RWRS indicated that a significant number of user complaints are legitimate and recommended rewriting the system and cleaning up the existing database... However, as a result of FRAS staff reductions, there will only be a partial cleanup of the existing database..." (ix 3) The database cannot be "cleaned up." The problems are inherent in the fact that diverse firearms cannot be uniquely identified by any known system, and the system would have to be operated exclusively by world-class firearms experts in order to prevent the current shambles from recurring quite quickly. The only way to "clean it up" would be to have every firearm physically examined and compared to its TRWS record by an expert in firearms identification. It cannot be made cost-effective. This is a serious and complex problem, caused by the wording of the law itself and by the very nature of firearms. The definition of "firearm" in CC s. 84(1) is, briefly: "'firearm' means any barrelled weapon..., and includes any frame or receiver... and anything that can be adapted for use as a firearm." Therefore, if parts are stripped off a firearm, one eventually ends up with a pile of uncontrolled spare parts and a naked "frame or receiver" which is still, in law, a "firearm." Removing some spare parts from a firearm and substituting others is lawful, and part of the ordinary behavior of firearms owners. The resulting firearm may be very different from the firearm that existed before the changes--but it is the same firearm, because the serial number identifies only the "frame or receiver" and it is unchanged. Interchanging of spare parts is uncontrolled and almost certainly uncontrollable even if the law is changed. @NEWPAGE As an example of the problem, the following table is a possible printout when the RWRS is requested to print out a list of all registered handguns ("Type: HG") with the serial number 123456: SERIAL | MAKE | MODEL | ACTION | CAL. | BBL. ----------------------------------------------------------------- 123456 | REMINGTON RAND | 1911 | SA | .45 | 127MM ----------------------------------------------------------------- 123456 | COLT REMINGTON | US ARMY | SA | .45 | 128MM ----------------------------------------------------------------- 123456 | SPRINGFIELD | 1911A2 | SS | .308 W | 379MM ----------------------------------------------------------------- 123456 | PACHMAYR | DOMINATOR | SS | .223 R | 254MM ----------------------------------------------------------------- 123456 | COLT | GOVERNMENT | SA | 9X19MM | 152MM ----------------------------------------------------------------- 123456 | COLT | SERVICE ACE | SA | .22 | 130MM ----------------------------------------------------------------- 123456 | AUTO ORDNANCE | 1911A1 | SA | .40 S&W | 115MM ----------------------------------------------------------------- 123456 | COLT | GOVERNMENT | SA | .32 | 100MM ----------------------------------------------------------------- 123456 | COLT | GOVERNMENT | SA | .380 | 89MM ----------------------------------------------------------------- 123456 | COLT | 1903 | SA | .380 | 100MM ----------------------------------------------------------------- 123456 | BROWNING | 1910 | SA | .32 CAP | 85MM ----------------------------------------------------------------- 123456 | FABRIQUE N'TNL | 115 | SA | 7.65MM | 88MM ----------------------------------------------------------------- In the above cases, the serial number appears only on the frame of the first 7 pistols, and identification of the maker appears only on the upper parts--which are uncontrolled interchangeable spare parts without serial numbers. In the above table, the first 6 pistols are probably all the same firearm, differing because spare parts have been interchanged. The 7th is possibly the same firearm. It would take a physical examination to be certain, because other makers produced exact or almost exact copies of the frame of this pistol, and the serial number ranges sometimes duplicated the serial numbers of the original versions. The fact that the RWRS contains those 7 records does not mean that there are 7 pistols; there may be 1, 2, or 3--and the only way to tell is to attempt to physically examine each of the 7 that are supposed, by the RWRS, to exist. There are certainly not 7 pistols. The RWRS is riddled with this type of problem. It is extremely difficult to detect and correct. It certainly cannot be done by examining the records and not the firearms. The 8th and 10th pistols are definitely not the same firearm as the 1st through 6th, but may or may not be the same firearm. The 9th pistol is definitely not the same as the 8th or 10th. The 11th and 12th pistols are the same firearm, described first in US fashion and then in European fashion. When FRAS is presented with an application to register, and the terms used as identifying entries on the application do not match with the terms used on the existing record, the possibility of a masked error arises. The clerk may create a new record for the new owner, without deleting the old record from the old owner. Then on paper, the firearm is registered to two people--and the RWRS cannot detect that the error has occurred. "UNLIKE MOST REGISTRATION SYSTEMS, SUCH AS DRIVER AND MOTOR VEHICLE LICENSING WHICH HAVE EXPIRY DATES, THERE IS NO EFFECTIVE MECHANISM FOR ENSURING THAT RWRS DATA IS UPDATED." (ix 4) It is not possible to use expiry dates to update firearms possession licensing documents, whether called firearms registration certificates or by some other name. That was established by the Supreme Court of Canada in 1985. "THE RWRS REWRITE IS BASED SOLELY ON AN INTERNAL EVALUATION AND WAS NOT PRECEDED BY A REQUIREMENTS DEFINITION IDENTIFYING SYSTEM USERS AND NEEDS..." (ix 5) That is hardly surprising. FRAS services very few outside requests from law enforcement agencies or personnel, and has always been very reluctant to work with firearms dealers and firearms owners, its main customers. In part, this has been necessary to mask the high error rates, impossibility of identifying firearms within the data base, and inability of the system to detect it when a record becomes trash. For example, if a firearms possessor moves or dies, and he or his heirs fail to notify FRAS, all relevant records become trash because they no longer describe the situation. The RWRS cannot detect it when that happens. The only way to detect it is to do periodic checks. No such check has been done since 1934. "Application forms have no unique identifier and are not sequentially numbered when received by FRAS. COMPLETENESS AND ACCURACY OF DATA CONTAINED IN THE APPLICATION INCLUDING INFORMATION CONCERNING THE WEAPON AND OWNER IS MANUALLY CHECKED. Identity of the Local Registrar who approved the transaction is not verified. Applications which fail verification are returned to the Local Registrar by mail... there is no record of the rejection rate... There is no mechanism for verifying that stale certificates are destroyed." (x 2) "ACCURACY OF DATA... CONCERNING THE FIREARM" regarding a particular firearm is "MANUALLY CHECKED" by an FRAS person who has no access to the firearm itself. That is rather futile. Losses in the mail sometimes lead to loss or duplication within the RWRS--either the firearm is registered to two (or more) people simultaneously, or the registration simply vanishes. The concept of "stale certificates" has no legal basis. A registration certificate is valid until it is revoked, and FRAS does not use the revocation procedure to eliminate "stale certificates." "INFORMATION FROM DEALER INVENTORY REGISTRATIONS IS ENTERED AS PROVIDED BY THE DEALER, UNLESS THERE ARE OBVIOUS DESCRIPTIVE ERRORS." (x 4) The same US service pistol might, then, be entered as "Make: COLT (REMINGTON)", "Make: REMINGTON", "Make: REMINGTON (COLT)" or "Make: COLT"; "Model: 1911", "Model: Government", "Model: US ARMY"--or in dozens of other patterns, none of which are "OBVIOUS DESCRIPTIVE ERRORS." That is not "uniquely" identifying a firearm. "Since they are given a lower priority, it is not unusual for a registration application from an individual purchasing from a dealer to be in process before the dealer registration is processed. Firearms officers responsible for dealer inspection described FRAS dealer inventory records as LACKING CURRENCY AND SUBJECT TO A HIGH RATE OF ERROR." (x 4) In one carefully-documented case, a dealer who had meticulously obeyed all of the rules for transfers found that, after four years, FRAS was showing 106 restricted weapons as part of his inventory after they had been properly transferred. Some dealers now refuse to answer queries from FRAS because FRAS does not correct its records. "FRAS USES 26 DIFFERENT FORMS (not including those used in the field by other agencies.) User comment was generally critical and a comprehensive review was often suggested." (xi 2) The system is currently under attack in the courts, and FRAS may lose a great deal of its power over the next 3 years. This excessively bureaucratic mass of overly-complicated red tape is not a sign of cost-effectiveness. "FRAS CONDUCTS APPROXIMATELY 100 WEAPONS TRACES PER YEAR... [and] most traces do not produce a positive trace." (xi 5) Compute the cost-effectiveness of registering 1.2 million firearms for over half a million owners (over $99 million) in order to be able to do traces on "APPROXIMATELY 100" of them--and then fail to trace more than 50 of the 100. That is over $2 million per successful trace, or enough money to hire at least another 50 police officers at $40,000 per year. Which is more valuable--one successful trace, or 50 more police officers? Registering 1.2 million firearms at $82.69 each (TR1994-9e, Table 2, page 27) cost taxpayers over $99 million--not counting lost wages, travel expenses, and other costs not mentioned in TR1994-9e. Was that cost-effective? One wonders what percentage of the traces--successful and unsuccessful--were run in an attempt to trace an error in processing within the system as opposed to what percentage were run as part of a criminal investigation. If the former percentage is higher, it would appear that the system exists mainly to create work for itself--tracing its own errors--and that is clearly not a cost-effective methodology. "There is no mechanism for formal feedback from requesting police departments as to whether information produced by a trace was useful. No records are kept of correlation between traces and offence type... The tracing program cannot currently provide national information about weapons tracing or smuggling activities. The lack of coordination which has reduced the effectiveness of the program appears to be encouraged by rivalry and distrust between Canadian Police agencies and within BATF [US Bureau of Alcohol, Tobacco and Firearms]..." (xi 6-xii 1) For this we spend millions of tax dollars every year? 81,913 firearms were registered in 1993. According to TR1994-9e, the cost of that (at $82.69 each, Table 2, page 27) was over $6.9 million. The missing records are vital in determining cost- efectiveness of the RWRS and FRAS itself. "Practices with regard to restricted weapons transactions vary so widely across Canada, and even within a single province, that it is impossible to describe a typical registration process... inconsistent and inappropriate practices continue to exist..." For this we spend millions of tax dollars every year? Local Registrars receive little or no training and, with the exception of Quebec, do not generally occupy the position long enough to acquire any particular expertise... it would appear that the objectives of the firearms provisions of the Criminal Code are not necessarily understood or supported by the Local Registrars." (xii 4) For this we spend millions of tax dollars every year? "Local procedures appear to have been influenced by a policy decision... It was understood to be RCMP policy... Local Registrars hold problematic applications for as long as possible, in the hope that the application will eventually be abandoned... Although RCMP policy may have changed..." (xii 5 xiii 1) In law, each issuer of a firearms control document is a statutory decision-maker, or independent quasi-judicial licensing tribunal, at any time when he is working on the issuance of refusal of a firearms licensing document. He is not a police officer; police officers do not have such quasi-judicial powers. As a statutory decision-maker, he may refuse to issue "in the interests of the safety of the applicant or any other person." That is a clause which occurs repeatedly throughout the firearms control sections of the Criminal Code. It defines the limits of the decision-making powers vested in each such person to refuse on the basis of opinion judgement. The statutory decision-maker may be precluded from issuing by a statutory provision, but that is neither a refusal nor an opinion judgement matter. Where a person is vested with the power to make decisions on the basis of opinion judgement, the law carefully circumscribes his powers. The firearms control system is out of control, and its issuers and refusers have been behaving unlawfully for years. Instead of acting quasi-judicially as required by law, Local Registrars, Chief Provincial/Territorial Firearms Officers, and Firearms Officers act on the basis of "policy" set by someone else. That is forbidden by law. Where a statute vests a person with judicial or quasi-judicial authority, that person, and that person alone, must exercise that authority--or the appeal against his opinion-judgement decision will be directed against the wrong person, and the law will be brought into disrepute. "Quebec has the only provincially integrated process." (xiii 4) The process as described is in violation of the principles of law, in that opinion judgement matters are farmed out to several offices, the authority vested in particular officers by the law is ignored, "policy" reigns, and the appeal process is fouled beyond recognition by scattering the decision-making power. In fact, the issuance or refusal of every firearms registration certificate, insofar as issuance may be refused, is placed, by law, in the hands of one person: the Commissioner of the RCMP [CC 109(6) and (7) plus 112 (3) and (8)]. Quebec ignores that. Additionally, the Quebec system is the highest-cost system. While the national average cost of issuing 1 registration certificate is $38.56 [RCMP] + $44.13 [Average] = $82.69, the Quebec cost is $38.56 [RCMP] + $65.46 [Quebec] = $102.02. TR1994-9e recommends that all provinces adopt the Quebec system. If "cost recovery" is added, the additional costs incurred by handling, processing and accounting for cash, cheques and credit cards must be added to the current cost of issuing registration certificates. That will probably boost the cost of issuing each registration certificate by $20.00 or more. Many police officers interviewed characterized the firearms provisions as a nearly impenetrable maze." (xiv 4) The conclusions of Mr. Justice Gibbs, Supreme Court of BC and Provincial Court Judge Gordon [quoted] in Hurley v. Dawson (Dawson had refused Hurley a permit which would have cost Hurley his job, for a non-violent marijuana offence), 1986, demonstrate that it isn't only the police and citizens who are confused: "This is a difficult case. Not the least of the difficulties is due to the tortuous language of the gun control provisions of the Criminal Code. In Regina versus Neil, Judge Gordon was moved, with some justification, to refer to those provisions as 'one of the most horrifying examples of bad draftsmanship that I have had the misfortune to consider', as 'so convoluted that even those responsible for enforcing the provisions are apparently unable to understand them', and as 'a challenge to one's sense of logic.'" As an illustration of the problem of galloping obfuscation, this is one recent addition to the Criminal Code: Bill C-17, new CC s. 91 (4.1): "Subsection (2) does not apply to a person to whom a permit to possess a particular restricted weapon has been issued under subsection 110 (1) where the person is not the person mentioned in the registration certificate issued in respect of the restricted weapon, when the person to whom the permit has been issued possesses the restricted weapon at the place authorized by the permit." "Earlier registration" (xiv 5) For 100 traces per year, with over half of them unsuccessful, is the major expenditure to do this cost-effective? How many real cases have demonstrated that the current system is failing, and what were the real consequences in each such case? Do we know? "Updating records" (xiv 6) The only way to check is to send a police officer to physically check that the registered firearm is at the registered address in the hands of the registered possessor. This is enormously expensive, and probably futile; criminals don't register their firearms, so only the law-abiding would be checked. In New Zealand, the police found that such a check could never be completed, because registrations already checked were "going bad" before the check could be completed. Is that cost-effective? Cost recovery (xiv 7) This matter is currently before the courts. There are good arguments that this "cost recovery" is actually a thinly- disguised form of taxation, and that this form of taxation is not within the lawful powers of Parliament. No assumptions should be made that this would be an easy or practical--or even lawful-- thing to do. It will take years for the courts to decide it, and, if the courts so rule, all monies collected may have to be returned to those who paid the "cost recovery" fees. Database management (xv 1) The problem here is one of "clear standards." There are no "clear standards" in this field. For example, the Colt .45 carried by all cowboys in the movies can be, and is, legitimately registered as "Model:" P, Peacemaker, 1873, M1873, Army, SA, Single Action, SAA, Single Action Army, Frontier, etc., etc. "Implementation of the Quebec administrative structure in all jurisdictions could be considered." (xv 2) As noted above, this structure is unlawful. It is also very costly, and it is very doubtful that elaborate record-keeping of the firearms in the hands of the law-abiding--which misses virtually every firearm in the hands of a criminal--is cost-effective or even worth doing at all. Training of Local registrars... particularly in identifying firearms... (xv 3) The author of this paper is a serious collector who has actively studied firearms for over 50 years. He can identify, by sight, probably 80 per cent of firearms placed before him--but he is still surprised on a regular basis. To be useful, such training would have to be comprehensive--and it could not be comprehensive and cost-effective simultaneously. Screening procedures (xv 4) Effective screening can only be done by a firearms instructor who observes the applicant actually using firearms, under supervision, for a reasonable period of time. Expecting a person who has never seen the applicant handling a firearm to render a useful judgement is ludicrous. LAN technology (xv 5) Technology cannot correct the fact that much of the database is garbage, and cannot be significantly improved. Processing garbage results only in better-organized garbage. It would still be inadmissible in any court of law, and so useless that only about 100 requests for traces would come in per year. It cannot be made cost-effective. Query processing (xv 6) Same problems. Forms (xv 7) The earlier analysis makes this the best recommendation in TR1994-9e. Only one well-designed form is actually needed. Data capture (xv 8) Once again, technology is suggested without looking at the basic problem of unique identification. Trying to get a group or firearms experts to agree on one "correct" set of identification entries for a particular firearm is a real problem. It is rather like trying to nail jelly to a tree. Suggested way to test the possibility of "clear standards": 1. Obtain 20O photocopies of complete registration certificates (RWRS records), selected randomly from the RWRS files. Make enough photocopies of the 200 RWRS records to give one set to each available firearms expert, and give one set to each of them. 2. Tell the experts that in future, there will only be one "correct" entry per data field for each "Make", "Model", "Type", "Action", and "Calibre" entry. 3. Explain that where the data field for "Calibre" shows ".380 ACP" in one registration certificate and ".38", ".380", ".380 CAP", "9MM", "9MM CORTO", "9MM KURZ" "9MM BROWNING", or "9MM BROWNING SHORT" on others of the same "Calibre," 7 of the 8 entries are wrong--because all are describing the same cartridge, and there can only be one "correct" entry--without specifying which entries are "incorrect" or "not enough data to tell." 4. Give each expert a highlighter pen, and have them highlight each "Make", "Model", "Type", "Action", and "Calibre" entry in the 200 copies which is "incorrect." 5. Gather each RWRS record for a given firearm, and compare which entries have been highlighted as "incorrect" or "not enough to tell." Tracing (xv 9) The only way to trace a firearm is to enter the system with the serial number and no other data, then have a genuine firearms expert scan the resulting list. That method will not always succeed (e.g., where the firearm has 2 or more legitimate serial numbers), but it is the best chance available. Adding any other information before doing the search will markedly lower the probability of success. "the [current] restricted weapons registration scheme... tries to qualify persons by registering weapons, and does neither particularly well." (xvi 4) That is a profound and correct assessment. "The present restricted weapons system demonstrates the operational and cost inefficiencies inherent in a regulatory scheme where implementation mechanisms... are not properly designed to pursue the defined policy objectives." (xvi 5) That is a profound and correct assessment. Now to the question: What are the policy objectives, and how are they defined? As the New Zealand Police learned when they tried to find out of their registration system was meeting its "defined policy objectives," no answer to the question could be found--because no one knew what those policy objectives were, or how they were defined. No one had ever recorded what effects the system would have if it were successful. That raises a point that shows up repeatedly in this paper: No one knows why registration of firearms is desirable. It has never been studied for cost-effectiveness, and, as this paper clearly shows, the use of it for tracing firearms--the most common reason put forward for doing it--is minimal and usually unsuccessful. If, on the other hand, the intent is to deny legal unsupervised access to firearms to those who should not have it, there are better, cheaper and more cost-effective systems than firearms registration. Operating environment... An effort should be made to predict compliance levels... (xvii 1) That is very true. When the government recently enacted Orders in Council and laws which made many more firearms "restricted" or "prohibited," compliance was at about the 10 per cent level. Since then, widespread confiscation of firearms--including utilization of the registration system to locate them--has vastly increased distrust of the registration system. Is it cost-effective to criminalize large numbers of Canadians by expanding firearms registration at a time when they trust neither their government nor the police not to confiscate their property? In sum, this report concentrates far too much on the possibility that advanced technology will allow the data to move farther, faster, and more accurately. It ignores the fact that erroneous data will always be a severe problem, and cannot be kept out of the system. As an example, the report does not even mention the problem of duplicated serial numbers. For example, huge numbers of surplus German military Luger pistols have been imported into Canada, manufactured between 1908 and 1945. On 01 Jan of each year of production, the Germans re-started the serial number range on these pistols from "1"--duplicating serial numbers every year. They made 3.5 million Luger pistols. Duplication of serial numbers is one common way of rendering unique identification impossible. There are others. There are not enough well-trained people in Canada to operate the system, because the Local Registrar is the only official who actually sees the firearm. If he makes an error, it is probably not correctable later. It understates the fact that the system is rarely used to trace crime firearms, indicating that the police find it of very little value in the real world. It completely ignores the fact that the data in the system is largely trash, and that the data cannot be significantly improved. Unique identifications of a firearm is far more difficult than the average person imagines, and keeping the records free of error is practically impossible. There are simply too many ill-trained people involved in the process, making errors, not recognizing errors, and copying errors on into the next generation of records. There is a pressing need for a check on the registration system. It has now been in operation since 1934, and no check of this type has ever been done. If the system is working, the public should know that; if it is a huge waste of scarce and costly police resources, the public should know that. The following test is designed to be as cheap as possible and as informative as possible, with 100 guns per group to ease the conversion to percentages. The numbers are inadequate to give good statistical data, but such a check can be done quickly and cheaply. The test will strongly indicate whether or not a more far-reaching check should be done. The RWRS badly needs to be checked, before it is enlarged to take in rifles and shotguns--a very costly extension. The possibility of a record "going sour" obviously increases with the length of time it has been in the system, so this test is designed to check for the speed at which RWRS records "go bad": 1. Obtain from the RWRS computer data base a printout of the first 100 registration certificates found dated 2 to 3 years ago, the first 100 dated 4 to 5 years ago, then 100 from 6 to 7, then 100 from 8 to 9, and so on to a final 100 with 20 from the batch dating from 20 to 24 years ago, 20 from 25 to 29 years ago, 20 from 30 to 34 years ago, and 20 from 35 to 60 years ago. (The system was first instituted 60 years ago, in 1934.) The search should be controlled to prevent the name of each chosen possessor from being chosen again during the search. 2. Send a police officer or other messenger to the address given on each registration certificate, and have him learn if the person still lives at that address, if the firearm is still in that person's possession, and if the firearm identifying information (found on each registration certificate in the 8 data fields purporting to identify the firearm) is all correct. 3. Present a report to Parliament on the percentage of certificate holders who no longer reside at the given address and the percentage of registered firearms that are no longer in the possession of the registered possessors who do still live at the registered address. Also, each of the 8 data fields on each registration certificate should be checked against the actual firearm, and the percentage of errors found in each data field should form part of the report. Of course, the data from the data fields will be rather messy, because there are many "right" ways to record data in those data fields (see the RCMP's National Firearms Manual, Appendix 4-5, for 118 pages of "common" identification examples). The report should, however, show up gross errors such as recording the "Model" number as the serial number; the "Model" as the "Make" and vice versa; the year of production as the "Model:"; incorrect barrel lengths and incorrect calibres; rifles, submachine guns and machine guns registered as "HG" (handguns), etc., etc. The most severe problems in doing such a check will be preventing the firearms control bureaucracy from burying the results in secrecy or "fudging" the data--for example, by tracing a person who has moved without notifying the system, and then recording his registration as correct, although it wasn't. It would be worthwhile (if possible) to delegate full control of this task to a group with no vested interest in the registration system, such as the Auditor-General's office or the Library of Parliament. That aspect should be researched. REGARDING THIS REPORT: The report suffers from insufficient technical expertise regarding firearms identification methods, and an excessive respect for high technology. The earliest principle of computer data processing has been ignored. That principle is: Garbage in, garbage out. In other words, one cannot get good data from the computer if the data put into the computer is flawed. The authors of TR 1994-9e have made overly optimistic judgements about the improvements that can be made by processing data more quickly, more accurately, and more conveniently. They have missed or ignored the underlying problem that the data itself is so seriously flawed as to be unusable in a court of law. It is true that FRAS experts and RWRS data have often been used in the courts, but that is coming to an end. The National Firearms Association's research into the RWRS has already resulted in sufficient grounds for having evidence from that system barred from use as evidence, due to its very high error rates, inability to identify firearms uniquely, and inablility to solidly connect a RWRS record to a particular firearm. TR1994-9e suggests that the RWRS system will soon be expanded to include all currently unrestricted rifles and shotguns--an expansion to at least 20 times the present size and cost. It is desirable to answer some of the many cost-effectiveness, accuracy, and desirability questions raised by TR1994-9e--before we commit at least $1 billion dollars of taxpayer money to such an expansion. While TR1994-9e speaks of "cost recovery," it is still unclear whether or not such "cost recovery" is lawful. Even if the "cost recovery" is lawful, TR1994-9e is still considering taking at least $1 billion out of taxpayers' pockets in the first year of the system. It just leaves by a different route--by way of fees instead of taxation--but without making the taxpayer any happier. It is suggested that public resistance will be formidable.