I, Eric Martin Larson, make the following declaration pursuant to 28 U.S.C. § 1746.

1. I am currently a Senior Evaluator at and have been employed full time by the U.S. General Accounting Office (GAO) in Washington, D.C. since 1987. My professional duties at GAO are primarily the statistical analysis and evaluation of computerized data, including the validity and reliability of administrative data records. Many of these data (such as those pertaining to various classes of legal immigration, and the reporting of diseases) are complex must be interpreted according to various decision rules regarding their definition(s), data entry procedures, editing and quality control checks for errors. During three of the 11 years I have been employed at GAO, I have received the top outstanding achievement award in my division or group based upon my abilities to analyze and interpret complex data. It should be emphasized, however, that all statements by me in this Declaration and my analyses of data gathered and reported by the Bureau of Alcohol, Tobacco and Firearms under the National Firearms Act of 1934, as amended, 26 U.S.C. § 5801, et seq. ("NFA") is the result of work by me as a private individual and does not represent the policy or position of the U.S. General Accounting Office.

2. In addition to the above profesional experience, I have earned the following degrees from the University of Texas at Austin: Bachelor of Journalism with Honors, 1974; Master of Arts (journalism), 1976; Master of Science in Community and Regional Planning, 1979; Master of Arts (sociology), 1984; and Doctor of Philosophy (demography), 1987.

3. Since 1986, as a firearms collector and in furtherance of my personal collecting and research interests, I have been studying certain firearms originally manufactured in the United States in or before 1934 (but not replicas thereof) which are currently defined as "any other weapon" ("AOW") under the NFA. These AOWs are primarily specialized firearms designed for trappers, and odd firearms such as knife-pistols which fall under the NFA largely for technical reasons, rather than any history of use as weapons by criminals. Today, virtually all of these firearms are valuable collector's items because they: (1) were created at a time when there were virtually no laws regarding firearm design; (2) are very specialized firearms that had a limited commercial market even at the time they were manufactured; (3) unlike any other NFA firearm, the Congress repeatedly lessened controls on them, although the NFA virtually destroyed the retail market for these types of firearms; (4) represent a unique niche in U.S. firearms evolution, design and genealogy, because there is nothing else like them; and (5) are extremely rare-- I estimate that fewer than 17,000 exist from an original group of about 170,000 that were manufactured in the United States in or before 1934.

4. After I began publishing my research in 1993, privately as well as in various major firearm magazines and reference books, collectors and others contacted me with disturbing information regarding the pre-1934 AOWs I have been studying. Specifically, in at least six instances, persons told me that they had inherited a registered AOW that was in their family for many years, yet BATF claimed the firearm was not registered. Upon attempting to transfer ownership to the lawful heir the BATF stated that the firearm was not registered in the National Firearms Registration and Transfer Record (NFRTR) and thus was contraband that must be involuntarily forfeited. When the heirs located original registration paperwork for the firearms, however, the BATF stated that a mistake had been made and then processed the ownership transfer. In other instances, lawful heirs have told me that they specifically recalled the firearm was registered (i.e., IRS agents came to the house to drop off registration documents), but could not locate their copy of the registration form (often some 30 or more years later), and were forced to abandon their firearm to BATF because BATF claimed the firearm was not registered.

5. Based upon a careful analysis of statistical data regarding NFA registration activity released publicly each year by BATF since approximately 1990, and written and statements to me by BATF personnel, it appears that BATF has been adding firearms to the NFRTR after being confronted with lawful registration documents by owners (or heirs) of NFA firearms. I testified about obvious errors in the NFRTR, and the serious legal problems created by missing or destroyed NFA firearm registration documents, before the House Subcommittee on Treasury, Postal Service, and General Government Appropriations in 1996, 1997 and 1998 (references hereinafter will be to "1996 testimony" and so forth, indicating page numbers in the published hearing records). While my primary research focus and interest is upon pre-1934 AOWs, I recognize that the problems identified with the NFRTR data potentially affect any firearm that is registered under the NFA.

6. The NFRTR data indicate potentially serious problems with original NFA firearm registrations. Specifically, in analyses of data publicly released by BATF, I found that during 1992 to 1996, BATF may have added 119 or more firearms to the NFRTR which were originally registered on Form 1 or Form 4467 during 1934 to 1971, for which BATF lost or deliberately destroyed the original records. In theory, the number of original registrations in past years should not change over time; however, these registrations--some 27 years old--have increased. From 1992 to 1996 (the last year for which BATF has publicly released data), inspection of these annual data discloses that the number of NFA firearms registered on Form 4467 has increased by 36 (1997 testimony, Table 1, page 63; further interpretation of errors and additions of firearms is on pages 57 to 62, and pages 84 to 88). There are other discrepancies and impossibilities reflected by these data. For one, in 1996, there were 164 Form 4467 registrations before 1968; this is impossible, because Form 4467 did not exist before 1968. For another, the 1996 data show 146 Form 4467 registrations for years after 1971; however, the U.S. Supreme Court prohibited such registrations on April 5, 1971. In a telephone interview in early April 1996, BATF employee Gary Schaible told me that he could not explain post-1971 registrations, and in response to my question (repeated three times to make sure of its accuracy) whether BATF had ever added firearms to the NFRTR because BATF lost the original registration and a lawful owner produced his or her copy of the registration, he stated: "Yes. I assume that's happened." In 1997, NFA Branch Chief Nereida W. Levine stated in a letter to me that if a registered NFA firearm was not recorded in the NFRTR, the NFRTR would be amended to reflect the registration (1996 testimony, pages 88 to 96; 1997 testimony, page 97.

Regarding Form 1 registrations, from 1992 to 1996, the BATF has apparently added at least 83 firearms that it reports were originally registered during 1934 to 1971 and during unknown years (1997 testimony, Table 2, page 65; further interpretation on pages 64, 66 and 67).

To try and get a better idea of the magnitude of the possible errors in the NFRTR for all data from 1934 to 1971, including years labeled "unknown" by BATF, I inspected total "REGISTRATION ACTIVITY" for these years using data publicly released by BATF during 1992 to 1996. These data show that BATF added 666 NFA firearms during 1992 to 1996 for the years 1934 to 1971, including unknown years (1997 testimony, Table 8, page 94; further interpretation, and documentation of errors BATF identified in the NFRTR, is on pages 95 to 121).

The NFA data exhibit other logical impossibilities, such as firearm registration activity occuring before 1934, the year the NFA was enacted (1996 testimony, Table 3, page 91).

7. The NFRTR data also indicate potentially serious problems with NFA firearm transfers. Specifically, during 1992 to 1996, BATF added 283 Form 4 (tax paid) transfers to existing data (see 1997 testimony, Table 4, page 71). Note that 192 of these additions of Forms 4 occurred in 1996, for years prior to 1968--that is, 30 or more years ago. This apparently means that during the course of an attempted transfer of the firearm the BATF discovered there was no record of the firearm in the NFRTR, and was forced to add the data to the NFRTR when confronted with valid Form 4 transfer documents retained by the owners of the firearms.

8. In response to my Freedom of Information Act request for a copy of the data entry manuals for the NFRTR, the BATF initially refused to release anything, stating that these manuals are "related solely to the internal personnel rules and practices of an agency." I pointed out that I was asking only for instructions on how NFRTR data were entered and retrieved, including decision rules on classifications, etc., and not information about any specific firearm. BATF ultimately responded by sending me 23 pages (some heavily censored) of materials. Inspection of these 23 pages indicates that the NFRTR data base instruction manual is organized haphazardly at best. There is no index, materials appear to have been randomly inserted, and the documents themselves are not of a uniform format. No information about data fields or variable names is included (1997 testimony, page 121).

9. I complained about the preceding and other errors in the NFRTR to the Office of Inspector General, Department of the Treasury, which responded by referring my complaint to BATF. BATF conducted an internal investigation of my complaint and in a final report dated September 1997 dismissed all of my charges (1998 testimony, pages 67 to 119). I specifically refuted each BATF denial with valid and reliable evidence, and identified evidence that BATF ignored, did not consider, and pointed out that BATF employee Gary Schaible had perjured himself in federal court in 1996 or in the 1997 internal report, because his statements regarding the deliberate destruction of NFA firearm registration documents by BATF employees were blatant contradictions (1998 testimony, pages 120 to 129).

10. In October 1997, the Honorable Dan Burton, Chairman, House Committee on Government Reform and Oversight, requested the Office of Inspector General, Department of the Treasury, to (1) review the BATF internal report, and (2) independently audit the BATF's firearm registration practices. The Treasury Department report was scheduled to be delivered to the Committee in June 1998; however, it has been held up by reviews within the BATF and possibly by other factors, and is unavailable at this time.

11. The issues of errors in the NFRTR, and particularly those of the lost or destruction by BATF of original firearm registration documents and transfer records, potentially affect anyone who owns an NFA firearm. BATF's continued refusals to address these problems is alarming, and must be remedied.

12. Finally, issues of the legitimacy of NFA firearms registrations and ownership are at stake. For example, at least four firearms that I lawfully purchased, as evidenced by approved Forms 4 for each of them, were apparently illegally registerd by the BATF years before I acquired them and without my knowledge. When I requested a statement in writing from BATF as to whether it intended to seize these firearms as contraband because BATF itself had illegally registered them, the BATF advised me that the NFRTR showed me to be the lawful owner of these firearms. Nothing in the NFA says it is illegal to possess a firearm transferred in violation of the NFA; however, any such firearm can be seized as contraband. My repeated requests for a statement from BATF as to whether it regards my four firearms as contraband have been ignored. The reason isn't too hard to figure out. If BATF states the firearms are contraband, then BATF is admitting at least some of what I have alleged: namely, that the accuracy and integrity of the NFRTR has been compromised by BATF itself. If BATF states that the firearms are not contraband, that's an equally difficult position, since information provided by BATF as to the dates of original registration, as well as copies of documentation proving that these firearms were illegally registered by BATF, would clearly refute such a position (1997 testimony, pages 88 to 92; 1998 testimony, pages 130 to 136).

13. It is my considered professional opinion that the NFRTR registration activity data publicly released by BATF during each of the years 1992 to 1996, for the foregoing reasons, demonstrate reasonable legal doubt as to the accuracy and integrity of the NFRTR.

I declare under penalty of perjury under the laws of the United States of America that the foregoing is true and correct. Executed on September 9, 1998.