Dr. Peter Lock
European Association for Research on Transformation e.V.

Review of the Small Arms Surveys

The institutional context

The Small Arms Survey has established itself as a substantive research institute in Geneva because the major institutes and initiatives focusing on arms control and disarmament, like SIPRI, IISS, WMSE, BICC, IDS, IDDS among others were covering this issue but marginally. As a follow-on initiative of the successful campaign, which led to the Anti-Personnel Mines Convention, an impressive group of like-minded governments agreed to accompany Switzerland in funding research focused on small arms, which were increasingly seen as a serious threat to international stability. Like the landmines the proliferation of illicit small arms was perceived to be instrumental in sustaining persistent armed conflict and violence causing profound human suffering on all continents. It was widely acknowledged that without tackling this issue little progress in human security and development was likely to materialise. The institute was established in 1999 and it is reasonable to assume that the funding governments eventually expect the research to inform their policies.

This renewed interest in small arms was also mirrored within the UN-system. The issue figured prominently on different institutionally competing agendas like crime, public health and arms control in particular. The Small Arms Survey positioned itself in the context of the arms control debate and presented its first Survey (2001) in July 2001 at the special session of the General Assembly, which led to the Programme of Action to Prevent, Combat and Eradicate the Illicit Trade in Small Arms and Light Weapons in All its Aspects in July 2001. Incidentally the third Survey was presented at the review conference of this Programme in July 2003. Accordingly crime and public health issues related to small arms are being dealt with only marginally in the three Surveys published so far. The main focus as also echoed in the advocacy community has been stockpiles, production and the monitoring of the international initiatives to control the illicit proliferation and diffusion of small arms.

Emerging paradigms of the survey

In a review of the first Survey I observed that the existing body of information on small arms has been aptly collected and integrated into a highly readable standard source. I also noted that five of the seven chapters replicated the analytical structure of SIPRI Yearbooks designed to support arms control of major weapon systems. Though the chapters on brokers and the effects of small arms availability broke new ground. They had the potential to orient future data collection away from pointless comprehensiveness of data collection towards thematic priorities. I deemed such an evolution highly desirable, because the compilation of global stockpiles, like offering an estimate of 550 million small arms as a major feature of the first Survey, looked impressive, but was analytically of little explanatory value. The social functions of small arms ownership are of such diversity embedded in different social contexts that their global summation is not supported by a single meaningful quest.

In contrast the inquiry into the shady sphere of brokering was promising and likely eventually to result in recommendations, which actions to take to improve the control of the illicit trade in small arms. Finally I also noted that mainstream narrative reproduced in the press articles should always carefully be tested with respect to its economic plausibility among others and that some seemingly authoritative sources like the Jane's Yearbooks should not be accepted at face value as they tend to reproduce the information supplied by the producers.

The common features of the three Surveys published so far are extensive chapters covering production, stockpiles, transfers and a comprehensive description of political agendas dealing with small arms respectively. Assessment of effects on humanitarian work and human security as well as three country studies in the 2003 edition complement the contents.
The second Survey 2002 pursued the collection of data and "improved" the estimate of global stockpiles leading to a global stockpile at minimum of 639 million small arms, having "discovered" another 89 million. This figure was duly reproduced by the international press and absorbed into UN-reports and from there on into the swathes of advocacy pamphlets and became a standard reference in the academic production. Once released into the arena any methodological restrictions of such a figure are lost and it starts a life of its own, which is not supported by the rather contentious methodology of its creation.

Unfortunately the meagre funding of the UN results in the permanent danger that false, though seemingly appealing data from advocacy rhetoric of the humanitarian industry slips into the Reports of the Secretary General, which in turn "sanctifies" those incorrect representations of reality. A case in point being the contention of the Secretary General that 80 or even 90 percent of the casualties due to small arms are women and children[1]. Based on the Small Arms Survey the Secretary General now informs us that globally a stockpile of 639 million small arms exists. But what is the message? How can this information possibly feed into political decision making? Suppose the researcher responsible for the global stockpile estimate finds reasons to change the parameters of his guess work, the Secretary General is likely to inform his audience next year that the global stockpile of small arms amounts to 750 or 800 million. It would not matter, since the total of the objects counted under the heading small arms bears little, if any relation to crime, violence and armed violence. The obsession with a total number of rather disparate objects, ranging from Stingers to Shotguns, is largely irrelevant to the search of measures to reduce the misuse of small arms in globally hundreds of distinct social settings and of socially unalike embedded small arms.

"At least 500 000 people are killed each year by small arms and light weapons" was the opening statement of the Small Arms Survey 2001. This figure became the First Commandment of the small arms advocacy community. The fundraising of a growing humanitarian NGO-industry exploits this horrifying data. However the scientific genealogy of this estimate is nowhere followed up in the three editions of the Survey, instead the foreword of the 2003 edition presents a substantially lower figure. "Easily transportable, cheap, and difficult to monitor in an increasingly globalised world, these weapons (=small arms) are implicated in more than 300 000 deaths each year, primarily in the world's poorest countries.". The attentive reader is left in the dark with the puzzle of this significant discrepancy in the presumably authoritative source on small arms.

This is all the more surprising because other aspects like production are presented in enervating detail in all three volumes. In 2001 the reader is informed that "small arms are legally produced by more than 600 companies in at least 95 countries worldwide" and that "the small arms industry is the most widely distributed sector of the global defence industry." By 2003 the hunter instinct of the researchers brought a total of "1134 companies involved in some aspect of small arms production in 98 countries" into the flashlight of the Survey. Given the fact that small arms are a mature technology, which allows reverse engineering at almost every technological level, from craft production to automated lathes, it is an industrial activity with an exceptionally low entry barrier[2]. As a result licence production has lost its importance and would it not be for the prestige of certain brands no licence would be honoured any longer. Moreover "some aspect of small arms production" is a rather loose definition; it starts with bending and welding scrap metal and ends with sophisticated machine tools in a metallurgic/polymer plant. A similar spread from refilling used cases to automated filling lines applies to the manufacture of ammunition. Impressive as it appears this finding is a mere artefact, void of explanatory value and has become an end in itself. To inform policy making one would need to identify those production capacities, which ultimately fuel the misuse of small arms and illicitly supply ammunition.

There is also some conceptual confusion in the presentation when the small arms production is presented as the most widely distributed sector of the global defence industry. The majority of small arms produced are commercial firearms and large numbers of manufacturers have no relation to the defence industry. In the 2003 Survey a list of thirty countries classified as significant small arms producers is presented (p.14), but even this list does not spell out whether all of these significant producers (no quantitative criteria is given) are active in the defence sector.

Finally before discussing the recent 2003 edition of the Survey in some detail I will address the apparent obsession of the research team with comprehensive global data on production, stockpiles and to a lesser extent transfers. The Small Arms Survey bases its data on the political compromise reached by a United Nations Panel of Government Experts on Small Arms in 1997 (Survey 2001, p.7) in defining Small Arms and Light Weapons (SALWs). This poses considerable conceptual problems for a meaningful data collection, because for all practical purposes the only relation shotguns and shoulder-fired rockets have is the UN-definition. The summation of all items falling into the UN-definition does not contribute to curbing illicit weapons or to improve control strategies[3]. Instead policies confronting the violence caused by small arms must set priorities and target types of weapons posing an imminent danger. Regional variations as to which weapons pose a threat are likely. It is important to recognise that most small arms are more or less safely embedded in cohesive social systems and do not pose a significant threat to human security.

There are certainly alternative, analytically productive options of approaching the analysis of stockpiles. Among others it would make sense to distinguish the lethality as well as the efficiency of different types of weapons, if in the hands of violent actors. Lethality might be defined on the basis of technical criteria. Efficiency, in turn, would capture the attraction of certain types of weapons in the diverse socio-economic contexts of crime, violence and armed conflict. Among the properties to be considered would be price per unit, serviceability, availability of ammunition, "stealthiness" in trade and application among others.

Hence it is indeed doubtful whether the Small Arms Survey was well advised by rigorously sticking to the politico-bureaucratic definition of small arms in its analytical work. The UN-definition is negatively loaded with the resistance of major players to allow for a more transparent regime under the authority of the United Nations. But it is not only the blighting effect of the range of arms somewhat arbitrarily shuffled into the UN-definition, there are more topical means of violence, which are not included in the UN-definition. Currently explosives are the most terrifying weapon with which mostly civilians are being targeted. From the perspective of improving human security, it would be equally important to trace the international networks providing explosives. To all likelihood black markets for small arms and explosives are linked and it would make sense to apply the singular research expertise of the Small Arms Survey to the investigation of explosives as well instead of sticking to the UN-definition.

Admittedly some separate estimates of world totals for military firearms and shoulder-fired rockets are given in the 2002 Survey (p.73). However, the apparent obsession of the researchers with world totals provides us with a worldwide stockpile of 22 m, ridiculously spelled out to the last digit, while two pages before we were informed that this very figure is the estimate of all shoulder-fired rockets produced through 1999 as compiled by a commercial consulting company serving the American military-industrial sector. The way this data is applied in the Survey requires the unreasonable assumption that not a single rocket has been fired to date. More generally the chapter on stockpiles in 2002 contains a considerable number of political assessments, which go far beyond the evidence offered.

Discussing mortars (p.72/2002) the Survey contends that "several European countries keep production lines, apparently in hope of export orders to developing countries", while "the United States closed the last of its production lines in the early 1980s." In fact mortars are a low-tech product amenable to import substitution by any standard metalworking factory making it unlikely that many private companies in Western Europe ( and not countries as the Survey insinuates) incur costs to keep production lines open. Technically it does not take much of a "production line" to manufacture mortars.

In the 2001 Survey the modelling of military firearms holdings relied on a multiplier based on the ratio of firearms per soldier of 2,25 in Canada and the bold assumption that this ratio would provide a conservative (lower end) estimate at the global level. The rules of empirical research would require to broaden the extremely weak base for the parameter (multiplier) used to put the estimate on methodologically safer ground. However, neither the 2002 nor the 2003 Survey proceed to improve the methodological base of the calculus presented. Sound empirical research would endeavour to significantly increase the number of hopefully representative countries for which a ratio can be established and then identify either the median or the arithmetic mean value. Instead the reader is informed that "it is significant (sic !) that the Small Arms Survey has been informed by the Canadian government that the number of its uniformed military personnel no longer is 107,500, as it was only a few years ago. As of 2001 its authorized strength was 93,500 for all services... This troop level leads to a current multiplier of 2,5 firearms for every man and woman in uniform for Canada."(p.79/2002).

The modelling of firearms holdings in China in the 2002 Survey is breathtaking and recalls the Cold War guestimates of Soviet military holdings. The comments on China in the 2003 Survey (p.86) reveal at best the arbitrary picking of single sources and the spinning of theories around them without even checking the plausibility of the conclusions drawn. A single CNN text reporting that 1,34 million illegal firearms were confiscated in a campaign labelled "Strike Hard" in June 2001 is taken at face value. One wonders that a tightly controlled authoritarian regime were all arms production has been controlled by the state should have allowed the illegal diffusion of millions of firearms out of state control and even it were the case, admit the facts. Another hypothesis to be tested before taking the CNN figure into account would be whether the systemic trend of institutions in a communist system to inflate their achievements was at play in this case. Finally could it be that the translator behind the CNN text simply got the figures wrong?

The specific flaws of the 2003 survey: chapter two

The chapter on stockpiles in the 2003 Survey confirms the bias of this research, which brings the entire research of the Survey into disrepute. I will discuss only the cases where I checked the figures presented and was embarrassed to find that the apparent zeal of the author was to maximise his figures, not heeding the rule that more than one independent source is needed before an information can be taken for granted and used to construe a statistic.

The author of the stockpile chapter takes his level of information as the standard knowledge and contends that "after Erfurt and a series of similar (sic?) events in Europe, new information became available on firearms ownership." "Contrary to the common assumption that Europeans are virtually unarmed" the author continues to assert that within the EU there are 84 m firearms, of which 67 m are in civilian hands. He is also surprised that September 11 had little effect on the global stockpile of known small arms, he apparently nurtures a causal link, which he does not bother to disclose.

A subtitle of the chapter reads "The year of Europe" and starts with a table of seven shootings in Europe. They are readily taken to reflect a global agenda. Attention is less on terrorism and the criminal effects of the growing size and technical sophistication of global small arms proliferation. Europe's gun problem is a matter of rising violent crime and mass murder, the author contends. No attempt is made to substantiate this contention by consulting crime statistics among others. My reading of crime statistics does not discern such a general trend. The statement is worthy of electoral platforms of rightwing populist parties and should not figure prominently in the Survey presumably reflecting social research of high quality.

The author finds the time sequence of control measures in Germany quite typical for Europe. Again he does not corroborate that the introduction of tougher registration in Germany in 1972 has its correspondence in other European countries. The German move was a direct reaction to the RAF, a terrorist movement operating in the FRG only. For historical reasons and distinct cultural traditions( i.e. hunting) the weapon cultures in Europe display a wide diversity. Changes in laws were often instituted in reaction the specific national events.

The core of the exposure of European stockpiles is a table giving estimates of "Known civilian firearms in the European Union" broken down into the minimum of registered and unregistered civilian firearms. The German figure of unregistered firearms is taken from a publication of the World Federation of Sport Shooting Associations, which in turn quotes a dissertation from 1994. The figure itself was constructed on the basis of questionnaires sent to 14 associations representing interest in firearms by a committee of the parliament in 1972. The estimates provided by these associations ranged between 8 million and 25 million unregistered firearms of all kind. The government concluded on the basis of the estimates provided that there might be 17-20m unregistered firearms in the FRG. Among others it was generally assumed that the disarmament by the Occupation forces after 1945 failed to collect personal weapons while former soldiers were said to have been reluctant to comply with Allied orders. However the large majority of unregistered firearms were hunting rifles, shotguns, collectibles and sport rifles and pistols. Contrary to the contention of the Survey Dobler, the author of the indirectly quoted dissertation did not present an estimate of his own. He merely reported the proceedings of the parliamentary committee in 1972. However, it is reasonable to assume that all parties providing estimates of unregistered firearms at the time were interested to impress the parliamentary debate with largely inflated figures. All parties involved were interested to discourage laws making registration obligatory as long as the presumed high density of unregistered firearms did not translate into violent crime. To the best of my knowledge no other source for the number of unregistered firearms has turned up so far. Thus, no serious data is available.

The same applies to registered firearms. The Survey presents a figure quoted by the Spiegel in 2002 and asserts at the same time that this is the minimum. However, according to a leading firearms lobbyist the figures quoted by governmental entities in recent years vacillated between 3 and 10 million depending on the political context. Due to the federal structure of the German political system no compilation and no central registry of licences (Waffenbesitzkarten) is available. There are strict regulations concerning the storage and transport of these sport-, hunting and collection weapons (Absolutely no automatic military firearms are allowed.). What is crystal clear and known with relative precision is that only around 1000 private citizens are entitled (by means of a 'Waffenschein') to own and carry an automatic pistol. All other entitlements ('Waffenschein') pertain to persons acting on behalf of the government.

But the most breathtaking information on Germany to be found in the Survey insists that counted per capita almost as many weapons are being sold as in the United States, namely one million firearms are added annually to the stock in Germany (p.70). While again no exact figure is available, my inquiry with the producer association and leading firearm lobbyists confirmed my hunch that this figure is totally out of proportion. The single source for this adventurous contention is Virginia Ezel, who administers the heritage of Edward Ezel an internationally renowned small arms enthusiast. If the author had bothered to check his chapter for cohesiveness he would have augmented his estimate of the German stockpile given an annual addition of no less than one million.

The reckless manipulation of sources is not restricted to the data on Germany however. On page 71 the reader is baffled by the asserted volume of small arms smuggled into the EU. "One of the largest transfers of illegal small arms and ammunition was intercepted by the Italian police in early 2001. This reportedly weighed13,000 tons."(p.71). Checking the source given reveals no mention of small arms: "Inter-agency and inter-police force co-operation this year also achieved success, resulting in the seizure of over 13,000 tonnes of arms and the break up of a "colossal" international arms racket" (p.47, Davis, Ian, Chrissie Hirst, and Bernardo Mariani, 2001. Organized Crime, Corruption and Illicit Arms Trafficking in an Enlarged EU, London Saferworld.). And even this statement is based on single newspaper article: Il Sole-24 Ore, 20/4/01. In a diligent research process this seemingly extraordinary data would have been checked against other sources.

No wonder that the author observes a trickling of easily concealed weapons at an unrelenting pace, which "has led to a progressively more alarming build-up throughout Europe."(p.71). One only wished that the disclosed rearmament in Europe, apart from giving consolation to American readers affiliated to the NRA, were corroborated by a credible source in Europe.

The reviewer is unable to replicate this probing exercise for the entire chapter, but these samples impart little credibility to the chapter as a whole. The detailed discussion of small arms in Africa contains many conjectures, which would require detailed sociological scrutiny. At the same time comparing the total of small arms holdings with other regions misses to take the extreme poverty of this continent into account, which explains to a large part the lower totals. Deducting the hunting- and sport firearms in other regions, which are associated to high income levels and not affordable to the African population at large would provide a more realistic parameter to compare holdings in Africa with. Thus, the finding of an Africa only lightly armed is flawed, if the endeavour of the Survey aspires to address public health, crime, violence and armed conflict as well as on methodological grounds.

The discussion of Latin America is imprecise and does not make use of available empirical data (OAS, CEPALC, VivaRio among others).

Observing Israel the author reports: "Requests for Israeli gun licences went up from 4,170 requests in 2000 to 7,790 in 2001. Of the latter, 4,500 or 58 per cent were granted. Despite ongoing fighting this remains the lowest known approval rate on any western country." (p.77). This observation highlights the flawed conceptual approach of the author again. While gun licences in Israel pertain to military style, including automatic weapons, the approval rates[4] he apparently bases his comparison on apply mostly to hunting- and sport weapons, but never to military style automatic weapons. In most countries the weapons licensed in Israel are out of reach for private citizens.

Finally the compilation of stockpiles on the basis of a definition, which was deliberately flawed for political reasons in the negotiating process at the UN is a waste of research time if not broken down into categories related to public health, crime, violence and armed conflict. And not least, the spectacular figures derived at dramatize the problem beyond proportion, which may even discourage concrete measures as the problem appears beyond any solution. Summing up, the funding governments clearly deserve better research on this important aspect.

The Other Chapters: production

It has already been highlighted that the identification of 98 countries involved in some aspect of small arms production is of little relevance and most likely arbitrary because this soft criteria applies to more countries though those might lie beyond the reach of the "avaricious" British scholars (OMEGA Foundation), who register this data.

By estimating the United States "to have one of the world's biggest domestic markets" (p.16) the Survey leaves the reader guessing, which the other similarly big domestic markets are (an information, which the next survey will hopefully provide). Another suggestion for the next Survey would be to take a closer look at India given that the Indian small arms industry has begun to place advertisements in international military journals.

In a subchapter on small arms technology the Survey reproduces the technology forecast provided by the editor of Jane's Infantry Weapons. While some of this might be useful, it seems that the author was carried away by the military-technological debate when he insinuates that little innovation is in the pipeline. This may or may not apply to the purely military dimension of small arms. Instead in the third year of this publication the reviewer would have expected a detailed evaluation of emerging technologies making firearms safer, since only public pressure can bring about any innovation in this important field. What are the prospects of personalised locks protecting any weapon from being used by an unauthorised person. Consider how important such a device would be for police officers, who from time to time are injured or killed with their own weapons. There is clearly a wide area of innovation and standardisation (separation of official and private weapons on the basis of distinct calibres) to be explored with the aim to reduce the misuse of firearms. This is a debate where the Survey should lead the argument.

The next subchapter deals with illicit craft production. While this dimension is certainly worth a one-off discussion in the context of the survey, its importance appears slightly overstated. The reviewer particularly doubts the contention, not corroborated by empirical data, that "... illicit craft production ... is characterized by the desperation of its buyers and increased likelihood that its products will be used to kill."(p.26). Obviously the comparative of the increased likelihood are manufactured weapons like the famed AK-47. The reviewer submits the hypothesis that craft production is also prevalent where arms are an expression of social status in the first place or are embedded in a stable set of social rules as demonstrated in the chapter on Yemen.

The subchapter on CEE countries supplies timely data on a region where restructuring is imminent. Not all the data reported appear to be consistent, the turnover per worker oscillates between 7,000 US $ and 50,000 US $. The lower figure is hardly viable even taking the low wages into consideration. The data on Bosnia and Herzegovina begs improvement. Given the semi-colonial status of this territory more direct data should be accessible and the Survey would be the institution to retrieve such important data instead of piecing together disparate quotes from other secondary sources.

In the conclusion of this chapter the Survey asserts "Sales are down, but the number of firms is up" (p.49). One wonders whether this counterintuitive contention is not owed to improved registration of firms rather than a real increase of the numbers of companies entering the fray. Again the assertion that craft production weapons are the dominant source of weaponry for marginalized groups begs further empirical study. Is it really corroborated by crime statistics or is it not merely a deductive argument?

Transfers

The chapter on transfers, also belonging to the standing features of the Surveys is plagued by the bias to demonstrate the outstanding role of Europe and conjectures with little if any plausibility. It is a useful exercise to exploit the detailed UN trade statistics for a better understanding of small arms transfers. However these statistics do not allow to segregate military-style weapons. Thus the export as well as the import values mainly reflect the consumption of private arms by upper and middle class people in OECD countries (with the exception of Japan). But even this marginally relevant data is not adequately presented in the text[5]. "Although the United States is the largest market for imported civilian small arms (...), North America is second in terms of the value of both exports and imports. On both counts, first place goes to the 15 members of the EU."(p.101) To reach this conclusion the author had to employ double counting in Europe, namely to cumulate intra-EU-trade and EU-exports. Since both North America and the EU supply data to the UN it would have been easy to provide a regional comparison after deducting the internal trade. The outcome would have been a huge difference the other way around[6].

"The scale of EU small arms imports supports the conclusion that European civilians are more heavily armed than commonly assumed." (p.103) The figure under discussion is 241 million US-$. Had the author bothered to collect some rough information on shooting sport in Europe and the average annual consumption of ammunition by these sportsmen, he might have regretted the conclusion drawn.

On the other hand the unequivocal statistical identification of Cyprus as a turntable of grey small arms trade confirms what has been contended ever since the times of embargoes against parts of former Yugoslavia.

The tendency of making unfounded maximising conjectures surfaces again when the Czech small arms exports to Yemen are being discussed. 6 million US $ is apparently the value of exports for 2002. Then the author implies that this figure may be low, because the Czech press reports the export of between 20,000 and 30,000 pistols to Yemen. A rough price calculation reveals that the 6 million US $ and the higher figure for pistols exported are absolutely compatible and nothing suggests that the monetary data is "low".

Moldova[7] is accused of exporting small arms to Iran, but nothing indicates that this sale was not government-to-government and hence a perfectly legal transaction. It is correct that the US-State Department objects to this deal and puts pressure on the government of this impoverished country. Thus, should the Survey really take a position on such issues?

The Survey documents a dramatic increase of imports of pistols and revolvers in Guatemala. This data is taken to "challenge the orthodox assumptions that Central America is saturated with post-conflict guns." (p.114). While this may be the case, this hypothesis should fist be tested against the plausible assumption that "post-conflict guns", mostly automatic rifles are not the type of weapons in demand by the upper- and middle classes, who seek personal protection in a climate of perceived high crime rates and armed violence. Pistols and revolvers are indeed more likely to be the weapons of choice in this social setting.

A final personal remark concerning a photograph on page 112 aiming to illustrate the arms transfer chapter. It shows a kneeling Afghan woman duly covered by a burka apparently trying to sell a simple, most likely rather old rifle. My reading of this photograph, though I don't have first hand experience in Kabul, is that of a woman living in abject poverty desperately trying to earn some income by selling an old gun in a social context saturated with automatic weapons. I suggest that the illustrations should reflect more humility and respect for the social context and rather highlight representative and important aspects of the respective subject.

The remaining chapters

The other chapters three of which are country studies contribute to a better understanding of problems caused by illicit small arms availability. The exploration of the relationship between small arms and development presents interesting attempts to approach this relationship more methodologically. Two small observations on ambivalent ratios might be helpful. The downfall of the share of agriculture in Angola's exports (p.143) is not automatically explained by an implosion of agricultural production. One would have to check price data and the absolute change of oil-based export income. The author highlights the social bias in domestic violence against women in poor households, but leaves the reader guessing whether this violence has anything to do with small arms. (p.147).

The case study on Yemen puts the quantity of weapons into perspective and will help to usefully inform future case studies in other countries. To a lesser degree this applies to the case study on Georgia. Anthropologists, who studied Georgia would have to make important additions[8] to this case study, which mainly discusses the turmoil following the dissolution of the Soviet Union on the territory of Georgia. The rigorous quantitative methodology attempted will contribute to the debate how data can best be generated in configurations lacking transparency. A satisfying solution obviously requires continued efforts in the coming year to mature.

The case study on the Republic of Congo should be read as another building block of an on-going discourse on empirical methodology. Only after a large number of similar case studies will be available can the debate qualitatively and methodologically advance, which for example would make the arbitrary multiplier used in the stockpile chapter eventually obsolete.

A rather disciplined argued chapter looks at the evolution of norms pertaining to small arms. Given the noise produced by the advocacy industry and the pamphlets relentlessly churned out by NGOs and the humanitarian industry this thorough description of the multilevel political process is particularly welcome.

Finally weapon collection is being discussed and various examples compared. The conclusion provides for improved realism in planning measures. One may differ in one or the other assessment, but the sobering evaluation separating facts form the rhetoric of funding organisations is an important contribution to the international debate and may help to contribute to less plunder often associated with such programs at all levels.

The way forward

The Survey urgently requires a clear focus based on the most problematic aspects of illicit small arms availability and misuse of firearms. It should shed all conceptual heritage from the arms control debate and instead identify configurations in which small arms display their harmful side. Public health, crime, violence and armed conflict are conceivable priorities. Admittedly a research institution created by governments eager to better inform their policies on small arms has to strike a balance between basic research and applied policy-oriented research. But the price should not be lack of rigour in the methodology in favour of political comments and themes favoured by the advocacy community. All analyses should be substantiated by empirical findings. The Survey is not well served by presenting bits and pieces of seemingly interesting information in dozens of boxes as is the current custom.

The currently practice of maximising conjectures on the basis of a single source in the stockpile chapter is counterproductive, not only from a social science perspective but also at the political level. Some of the general discussions of military doctrine and the like under this heading would have been only indicated if it had actually contributed to a better understanding of the small arms issue. Thus, the most radical, though probably also the most productive step would be to declare the research on stockpiles accomplished and apply the freed resources to a more focused endeavours of pertinent data collection. Given the exhaustive treatment so far the same applies to the chapter on production. It is time to acknowledge that IANSA's member organisations are spreading around the globe and begin to monitor small arms on the ground. Thus, the Survey will be less likely to miss emerging trends than was the case only three years ago. If needed the Survey can always again take up "stockpiles" and "production". As far as transfers are concerned more rapid shifts are likely, thus this subject should continue to be monitored on an annual basis.

Once the focus or a sequence of foci for the next Surveys has been defined, the current dominance of political scientists among the research staff must give room for a broader participation of criminologists, public health specialists, anthropologists as required by the respective foci. Otherwise the funding governments will not receive analyses and recommendations commensurate with the funds invested in the Small Arms Survey.

Hamburg, 27 August 2003 Peter Lock

Footnotes

[1] In a letter to the author the Director of Communication of the Secretary General has acknowledged that regrettably this "bogus data" had indeed crept into the UN documents.

[2] Glock, originally a small manufacturer of safes, produced its first weapon in the early eighties and is now the market leader for pistols.

[3] It is important to distinguish between arms control and curbing illicit small arms. For arms control it is an obvious priority to create transparency and obtain information on the total of available arms. The objects arms control is dealing with are mostly unambiguously embedded, namely they are identified as instruments of warfare. This allows construing scales aiming at making the objects countable and hence comparable. Such information is expected to inform strategies to avoid arms races. But small arms do not lend themselves to such procedures. The diffuse character of small arms depending on the respective social and institutional context and the disparate social embeddedness makes any (global) total a void figure impossible to interpret meaningfully. Most small arms are safely embedded and controlled by social rules and are not cause of concern.

[4] The author reveals no information on approval rates elsewhere. It is doubtful whether such data is available for many countries.

[5] The tables 3.2 and 3.3 contain a footnote drawing attention to the fact that intra-European trade is included in the European total. However, the author himself ignores this methodological caution.

[6] For some data on trade distribution in Europe see: Peter Lock, Small arms production in Germany, Switzerland and Austria www.Peter-Lock.de/txt/dasui.html.

[7] The record of surplus sales by Moldova is indeed a murky one. Early on after independence the CIA bought half a dozen of the most advance version of the Soviet fighter aircraft SU-27. This sale was not marked by accountability either.

[8] For an overview see: Koehler,Jan, Die Zeit der Jungs Zur Organisation von Gewalt und Austragung von Konflikten in Georgien, Münster und Hambudrg (Lit-Verlag) 2000.