Dangerous Trade: Arms Exports, Human Rights, and International Reputation

2. See chapter 3 for formal definitions of small arms and light weapons (such as machine guns) and major conventional arms (such as tanks and military aircraft). I refer to “small arms and light weapons” interchangeably by either the internationally recognized acronym SALW or the shorthand “small arms.”

3. Dating back to the mid-1990s, there are simply too many newspaper stories, journal articles, and NGO reports using this terminology to begin to cite them or trace its origins.

4. The three opposed were Iran, North Korea, and Syria. There were twenty-three abstentions.

5. Article 223 of the 1957 Treaty of Rome protects members’ right to produce and sell arms in the interest of their national security.

6. On this relationship, see G. Adams 1981; Cooling 1981; Dunne 1995; Eisenhower 1961; Hartung 1996; Keller 1995; Kolodziej 1979; Kurth 1971; Markusen et al. 1991; and Silverstein 2000.

7. SIPRI (2012) estimates that five countries—the United States, Russia, Germany, France, and the United Kingdom—dominate the global market in MCW, accounting for 75 percent of global sales between 2007 and 2011. Seventy-four of the top one hundred arms companies in 2010 were from the United States and western Europe.

8. On these effects, see Boutwell and Klare 1999; Boutwell, Klare, and Reed 1995; Craft 1999; Craft and Smaldone 2002; Eavis 1999; Harkavy and Neuman 2001; Hartung 2001a; Karp 1993, 1994; Klare 1994–1995; Klare and Rotberg 1999; Musah 2002; Renner 1997; SAS 2001, 2002, 2003, 2006, 2007, 2009; Sislin and Pearson 2001; Stohl and Grillot 2009.

9. On this connection between arms and poor human rights, see, for example, Blanton 1999, 2001; Craft 1999; Craft and Smaldone 2002; Neuman 1986.

10. For the full text of the ATT, see http://www.un.org/disarmament/ATT/docs/ATT_text_(As_adopted_by_the_GA)-E.pdf.

11. A multimethod approach “aims to improve the quality of conceptualization and measurement, analysis of rival explanations, and overall confidence in the central findings of the study” (Lieberman 2005:435).

12. Chapter 3 discusses the statistical findings; appendix B provides coding details; and appendix C provides full statistical results tables.

13. Gerardo Munck suggests “matching cases on independent variables,” which “serve the same purpose as statistical control” (2004:104).

14. Audie Klotz and Cecelia Lynch recommend research on both levels of analysis in order to trace more thoroughly “the spread of norms across state boundaries and into societies” (2007:95).

1. Some scholars acknowledge a social dimension to international reputation (e.g., J. Busby 2008; Johnston 2008; Keohane 1984; O’Neill 2006; Paul 2009; Sharman 2006). However, the concept remains theoretically underdeveloped, and its consequences are either discounted or underexplored.

2. See Bromley 1993 for an overview of common usages of the term.

3. Many scholars conclude that reputation can lead to more frequent participation in conflicts but debate the rationality of going to war to “save face.” See, for example, Crescenzi, Kathman, and Long 2007; Hugh-Jones and Zultan 2012; Lebow 1981; Mercer 1996; Press 2005; Schelling 1966; Tang 2005; Walter 2009; Wolford 2007.

4. That is, “while adversaries can get reputations for having resolve, they rarely get reputations for lacking resolve; and while allies can get reputations for lacking resolve, they rarely get reputations for having resolve” (Mercer 1996:10; see also Huth 1988).

5. This use of reputation to inform states’ choices to cooperate with other states corresponds with game theoretical approaches to reputation, which observe that the “information conditions” of repeated games create the very “possibility of reputation” and can motivate compliance to acquire material gain even in the absence of formal institutions (Alt, Calvert, and Human 1988:449; Greif 2006; Milgrom, North, and Weingast 1990).

6. J. C. Sharman (2006) engages this question only in passing. Michael Tomz (2007) does find that in the case of debtor–creditor cooperation, investors’ willingness to make loans—and the terms of those loans—are connected to borrowers’ payment histories. Interesting parallels can be drawn between these approaches to reputation and those of corporate management. Reputation matters in business because of its impact on an undetermined number of “future trading opportunities” and the profits attached (Kreps 1990). It is a strategic approach to spread positive information about a company and its products in order to increase its profits (Davies et al. 2003; Fombrun 1996). In fact, corporate strategies of branding and marketing linked to reputation are spreading to countries by explicit national branding efforts (Aronczyk 2013; Jones and Subotic 2011; van Ham 2008).

7. Other authors share a similar perspective: see Finnemore and Sikkink 1998; Hurd 2007; Johnston 2008; and Schimmelfennig 2001. Nicolas Jabko (2006) refers to a related approach as “strategic constructivism,” in which strategic actors exploit norms and ideas in pursuit of political goals.

8. As Alastair Iain Johnston (2008) points out in the case of China, social pressures can also encourage nondemocracies engaged in international institutions to adopt policies in line with international expectations. Patterns of states ratifying human rights treaties but not complying with them lend support to this argument (Hafner-Burton and Tsutsui 2005; Hathaway 2002).

9. Legislatures are largely uninvolved in arms export decision making. Many European parliaments have called for arms export oversight, but their input is usually post hoc at best. Only in the United States, where oversight on deals over a certain dollar figure is allowed but rarely exercised, and in Sweden, where oversight committees have regular input, do legislatures have any regular role.

10. This concern relates primarily to elites’ perceptions but can extend to that of foreign publics as well (Goldsmith and Horiuchi 2009).

11. Assigning human qualities to states—interests, identities, beliefs, the capacity to act, and so on—is standard practice in IR (Jackson 2004; Ringmar 1996; Wendt 1999, 2004; Wolfers 1962). Alexander Wendt (2004) argues that this view of “states as people” enables states to possess a collective intentionality that cannot be reduced to its individual citizens and that has associated needs and interests (see also Searle 1990; Wendt 1999). I take the common view that states can be ascribed personlike traits accounting for their collective behavior, which allows reputation to be carried by a state as an entity, but also by its leaders and diplomats.

12. A public commitment to humanitarian arms control in general, human rights treaties, nonproliferation policies, climate-change initiatives, and opposition to whaling are all additional examples of issue areas that states, international organizations, NGOs, and other international actors have associated to varying degrees with good international citizenship. The concept itself was introduced by Australian foreign minister Gareth Evans in the 1980s and is often associated with the English School in IR theory (Wheeler and Dunne 1998).

13. International reputation here serves primarily as a motivating factor for states’ commitment to international norms and rules. On the influence of reputation on compliance with international law, see Chayes and Chayes 1995; Downs and Jones 2002; R. Fisher 1981; Franck 1988; Goldsmith and Posner 2005; Guzman 2002; Henkin 1968; Norman and Trachtman 2005. The discussion of arms trade transparency (or lack thereof) later in this chapter addresses why reputation provides less of a “compliance pull,” as Thomas Franck (1990) puts it.

14. Noel Kaplowitz states, “The aspects of national self-imagery which influence foreign policy behavior include how a people sees itself, what it likes and dislikes about itself, the ways in which it may want to change, how it views its history, the ‘lessons’ it has learned, and its conceptions of national purpose and interest” (1984:376).

15. Governments are not the only source of states’ external images. Nonstate actors, such as NGOs, hoping to persuade governments to adopt certain policies or practices, may engage in their own impression-management strategies to make states look good or bad (J. Busby 2007, 2008). Domestic events outside a government’s control can also influence a state’s international reputation. Consider, for example, the Swiss government’s embarrassment over the results of the 2009 referendum to ban minarets, a referendum that it worried would harm Switzerland’s image as a country of tolerance, refuge, and respect for international law.

16. Robert Jervis defines a decision maker’s image of another actor as “those of his beliefs about the other that affect his predictions of how the other will behave under various circumstances” (1970:5). Like the credibility approaches to reputation, image is important in “determining whether and how easily the state can reach its goals” (6). Unlike those approaches, however, Jervis argues that a state’s image is not “completely dependent on the major actions it took” (8), making it “possible for a state to consciously influence others’ images of it without paying the price of altering its basic behavior and sacrificing other goals” (10). In line with my point here, he emphasizes the psychological and subjective nature of states’ images and states’ ability to shape, with varied success, their desired images through “minor and relatively cheap” changes in their policy or behavior (11).

17. Christian Reus-Smit refers to this recognition as an actor’s legitimacy, which describes “not just the capacity to act, but the right or entitlement to act” (2007:158). Wendt notes that “persons,” especially moral and legal persons, are externally constituted by social recognition, which conveys “all the rights and privileges of that status” (2004:293, 294).

18. The American Political Science Association Task Force on U.S. Standing defines standing as “a particular kind of reputation” with “a more relative connotation in that it refers to position within a particular setting—in this case the society of states or ‘international community’” (American Political Science Association 2009:4, 6). Joshua Busby (2008), in turn, defines prestige as standing. Prestige and honor are also often used interchangeably in IR as related to esteem, standing, or glory (e.g., Donelan 2007; Kagan 1998; Lebow 2008; Tsygankov and Tarver-Wahlquist 2009). Barry O’Neill, however, argues that prestige, unlike honor, depends on a group’s recognition of it: “A person might possess personal honor while others do not know about that quality”; prestige, in contrast, stems from a group member’s being “generally admired” and gaining influence in the group because of a good reputation (1999:193; see also O’Neill 2006).

19. Cecilia Ridgeway and Henry Walker summarize theories of social hierarchy formation and call social hierarchies “an enduring feature of human interaction and a fundamental aspect of the organization of social behavior” (1995:281).

20. Robert Gilpin, for example, notes the existence of a “hierarchy of prestige among states” linked to but distinct from material power and resting on the perceptions of other states (1981:30–31). However, soft power can also enhance reputation. For more on hard power, soft power, and standing, see American Political Science Association 2009.

21. According to Aaron Wildavsky, “Reactions [from others] are also indispensable in that the only way the self knows what it is is by feedback from others (1994:139–40). Christine Ingebritsen, for instance, argues that the increasing international scrutiny to which Scandinavia has become subjected “helps redefine its own understanding of self” (2006:107). Kristen Monroe observes that “maintaining self-esteem seems to play a key role in maintaining identity,” which requires a general consistency between perceived self-image and behavior (2001:498). For more on the concept of identity as a social category and its measurement, see Abdelal et al. 2009.

22. D. B. Bromley states, “Although reputations are simplified and distorted representations of persons, they describe people’s social identities. They provide an external standard or criterion against which, through social ‘feedback,’ we try to assess our social identities. Reconciling personal identity and social identity is a continuing problem of adjustment” (1993:11).

23. Scholars note the importance of regular interaction by diplomats in international institutions for international socialization (Adler and Barnett 1998; Checkel 2001; Johnston 2008). See also the International Organization special issue “International institutions and Socialization in Europe” (2005).

24. The need for dense and frequent interactions within a network to assign reputations to actors is also noted by anthropology (Bailey 1971; Colson 1974); sociology (Raub and Weesie 1990); and corporate management (Fombrun 1996). For larger groups, regimes and institutions can help to spread the information necessary for reputation to function as an enforcement mechanism, although this does not take into account the social pressure generated by interactions within a group (Ahn, Esarey, and Scholz 2009; Milgrom, North, and Weingast 1990).

25. Johnston argues that small or consensus-driven groups are more likely to lead to states’ internalization of new norms but concedes that norm adoption without internalization can be a powerful social incentive in large groups where rules are majoritarian and put members’ policy support on record (2008:31–32). Although majoritarian decision rules make it easier for NGOs and other actors not in the room to criticize a state’s policy positions, they are not necessary to establish its reputation among the diplomats debating a policy themselves. This key audience for international reputation knows who dissents from a popular policy initiative. In fact, dissent under consensus rules may be even more controversial and subject to reputational fallout because it can single-handedly prohibit an initiative from passing (or weaken it), unlike dissent under majoritarian rules.

26. The UN Institute for Disarmament Research assembles and publishes information on states’ contributions to goals set out in the 2001 UNPOA.

27. See, for example, Abdelal et al. 2009; Chayes and Chayes 1995; Johnston 2008; Keohane 1984; Simmons 2009; Wheeler 2000; Young 1992; and Zarakol 2011.

28. This dichotomy of interests and expectations can also lead to cognitive dissonance between instrumentally beneficial policies and socially beneficial policies. Leaders will try to resolve this tension within their community’s existing normative framework, to the extent that a norm is defined broadly enough to make this possible (Shannon 2000).

29. Image management is common in corporate branding. Consider, for example, attempts by U.S. banks to use public-relations campaigns, personnel changes, new policies, and repayment of government loans to distance themselves from their “fat cat” reputations following the 2008 financial crisis.

30. And of consistency between policy and action to the extent that others can observe action. Without transparency, criticism of discrepancies between state policy and action can only be limited. As Nicholas Wheeler points out, “Changing norms provide actors with new public legitimating reasons to justify actions, but they do not determine that an action will take place” (2000:9). I discuss this point further in the context of states’ domestic reputations and scandal sensitivity.

31. See Brooks and Wohlforth 2008; Downs and Jones 2002; R. Fisher 1981; Norman and Trachtman 2005.

32. Whether a reputation can carry across issue areas is a matter of some debate. Jonathan Mercer argues that “a general reputation has an enduring character trait that reappears in different types of situations,” whereas “a specific attribution applies not across types of situations but within specific types of situations” (1996:37; see aso R. Fisher 1981). Similarly, James Lebovic and Erik Voeten posit that a good reputation in one area can help states establish good reputations in other areas (2006:868). Michael Tomz and Mark Wright (2010) find evidence of “reputational spillover” from states’ default on sovereign debt to their expropriations of foreign direct investment. However, these issues may still be considered closely related within economic relations and cannot address the broader question of spillover from economic to security affairs, which Stephen Brooks and William Wohlforth contend lacks empirical support (2008:163). “Similar situations” may be defined very narrowly, suggesting that reputation is considerably more complex than other definitions might imply (e.g., Downs and Jones 2002; Norman and Trachtman 2005).

33. This perspective tends to be associated with realist views of international law (Hathaway 2002), although some liberal scholars also argue that without incorporating costly material punishments, treaty commitments may still essentially be cheap talk (Hafner-Burton 2005). In addition, once formalized into binding agreements, cheap talk can lead to expected and unexpected compliance costs (Clark 2013; Goodliffe and Hawkins 2006; Greenhill 2010; Schimmelfennig 2001; Simmons 2009).

34. This would be the general expectation of constructivism.

35. Many scholars observe similar gaps between the expectations and realities of states’ follow-through on their international humanitarian commitments (e.g., Hafner-Burton 2005; Hathaway 2002; Simmons 2009; Smith-Cannoy 2012; Wheeler 2000).

36. I focus on the government as a whole and its top leaders. This focus is justified for two reasons. First, the media and public are less able to differentiate between (and take an interest in) bureaucratic versus executive responsibility. Second, governments decide whether to adopt a particular arms export policy. Top officials are often also responsible for deciding whether to approve the most controversial/political arms export applications.

37. More specifically, John Thompson labels scandals as actions, events, or circumstances in which there is (1) “the transgression of certain values, norms or moral codes”; (2) “an element of secrecy or concealment”; (3) a disapproval of the events or actions by some nonparticipants; (4) the public denunciation of the events or actions by some nonparticipants; and (5) “the disclosure and condemnation of the actions or events [that] may damage the reputation of the individuals responsible for them” (2000:13–14). See also A. King 1984; Mancuso 1998.

38. Scandal cannot simply be defined by equating it with corruption; more corruption does not predict more scandals. Instead, levels of corruption tend to remain relatively constant, whereas scandal levels and type change over time (Lowi 2004:70–71).

39. On this variation in what makes a scandal, see Dobratz and Whitfield 1992; Garment 1992; A. King 1984; Williams 1998.

40. Belgium, with its delicate governing arrangements, is a noteworthy exception. The British Arms to Iraq scandal also contributed to Labour positioning itself as a more transparent and accountable alternative in the 1997 election.

41. See Bowler and Karp 2004; Funk 1996; Meinke and Anderson 2001; Williams 1998.

42. See A. King 1984; Meinke and Anderson 2001; J. Thompson 2000; Williams 1998.

43. See Hirshberg 1993; Monroe 2001; Shannon 2000.

44. Such partnerships are often referred to as “the new diplomacy” or “new multilateralism.” See Axworthy 1998; Malone 2002; McRae 2001; Petrova 2007.

45. For example, Amnesty International has sought to maintain neutrality from governments in order to preserve the credibility of its message and mission (Clark 2001; Dezalay and Garth 2006).

46. This suggestion goes beyond the contention that norm violators can be shamed for not keeping international commitments (J. Busby 2008:83) or because they violate norms of international society (Lebovic and Voeten 2006:869).

47. Archon Fung, Mary Graham, and David Weil (2007) pinpoint the 1980s as the start of U.S. moves toward increasing transparency in corporate governance and financial markets. Others connect the global spread of transparency to trends of democratization, globalization, and financial crisis in the 1990s (Best 2005; Florini 1998; Relly and Sabharwal 2009).

48. See Broz 2002; Chayes and Chayes 1995; Florini 1998; R. Mitchell 1998; Relly and Sabharwal 2009.

49. See Apodaca 2007; Broz 2002; Fearon 1994; Fung, Graham, and Weil 2007.

50. Jeannine Relly and Meghna Sabharwal note that authoritarian regimes have also adopted some fiscal transparency to attract international investment and aid (2009:151). However, whether transparency initiatives generally make good economic sense has been debated: see Alt and Lassen 2006; Broz 2002; Dranove et al. 2003; Florini 1998; Fung et al. 2004; Fung, Graham, and Weil 2007.

51. See J. Thompson 2000 on the rise of the “mediated scandal.”

52. These commitments can be both national and multilateral. Certainly, the public has to care about arms trade practice in order for the transparency–accountability dynamic to function. Given the high volume of information available to the average media consumer and the general disinterest in arms trade policy, it is only where there is a clear disregard of fundamental national values, I argue in chapter 5, that concentrated public attention over conventional arms exports will arise.

53. Domestic defense production is important to national security: “The capacity to restock supplies or modify equipment from domestic sources is often cited as an important wartime asset. Domestic production can reduce vulnerability to boycotts or blockades. Moreover, domestic arms production and arms transfers are widely viewed as important sources of national prestige and diplomatic leverage” (Moravcsik 1991:200).

54. Joseph Grieco (1993) argues that state interests are based on their power position in the international system. States are wary of cooperative arrangements, particularly because relative gains might favor their partners.

55. This failure does not mean that realists predict an unregulated arms trade. According to Joanne Gowa (1994), states are acutely aware of the security externalities connected to improved material power capabilities brought on by foreign trade. As a consequence, they will seek to control the flow of defense goods and technologies to ensure that only their allies receive the strategic benefits. Allies are often given preferred treatment in national export legislation, with faster decision-making processes and less intense scrutiny of applications. Such favorable terms are codified, for example, in Germany for its NATO allies as well as between the United States and Canada.

56. From 1989 to 2005, the United States was the most frequent intervener, with thirty-eight incidents of international military intervention, followed by France (thirty-three incidents) and the United Kingdom (thirteen) (Kisangani and Pickering 2007).

57. Legally binding arms embargoes may be the one exception in which international commitments effectively limit arms export practices (Erickson 2013c).

58. Although the influence of industry may to a degree wax and wane over time, the general observation from this perspective is that it remains a steady fixture of importance in security and defense-related policy making.

59. On the defense industry’s influence, see G. Adams 1981; Eisenhower 1961; Kurth 1971; Markusen et al. 1991.

60. On the constructivist position, see Finnemore 1996a, 1996b; Katzenstein 1996; Klotz 1995.

61. See, for example, Hafner-Burton 2005; Hafner-Burton and Tsutsui 2005; Lebovic and Voeten 2006; Ron, Ramos, and Rodgers 2005; K. Smith 2001; Smith, Pagnucco, and Lopez 1998.

62. The determinacy with which norms influence state behavior is unclear. Constructivists suggest that “norms enable new or different behaviors; they do not ensure such behaviors” (Finnemore 1996a:158). Nevertheless, they firmly assert that “norms create patterns of behavior in accordance with their prescriptions” (Finnemore 1996b:23) and make certain behaviors more likely than others (Wendt 1999). Determining conditions for when and where norms are influential is therefore an important question for IR research (Legro 1997).

63. See Blanton 1999; Bueno de Mesquita et al. 2005; Davenport 1995; Howard and Donnelly 1986; Mitchell and McCormick 1988; Poe and Tate 1994; Rummel 1995.

64. Martha Finnemore and Kathryn Sikkink (1998) argue that instrumentalist motivations may dominate early in norm life cycles but become more deeply embedded over time.

65. On analytical eclecticism, see Abbott and Snidal 2002; Jupille, Caporaso, and Checkel 2003; Katzenstein, Keohane, and Krasner 1998; Katzenstein and Sil 2004; Sil 2000; Wildavsky 1994; Zürn and Checkel 2005.

66. Jay Goodliffe and Darren Hawkins (2006), for example, find a significant role for norms in explaining compliance with the Convention Against Torture. However, they conclude that their quantitative analysis cannot parse out whether their findings are due to rationalist mechanisms (reputational costs) or constructivist mechanisms (state identity and socialization). Because of states’ poor compliance with new arms export policies, the qualitative and quantitative evidence presented here lends itself well both to combining and to separating these mechanisms: states are socialized into policy expectations, to which they respond instrumentally out of concern for their reputation; however, they have not internalized new norms.

1. During the Cold War, only West Germany and the European neutrals sought explicitly to limit arms transfers to areas of instability and conflict. The United States under President Carter also had a policy (followed loosely at best) to prohibit exports to human rights violators.

2. On this optimism, see Anderson 1992; Cornish 1995; Goldring 1994–1995; Jones and Rees 1994; Kemp 1991; Klare 1991; Neuman 1993; Nolan 1991; Pierre 1997; Spear 1994.

3. UN General Assembly Resolution A/61/394 passed by a vote of 153 to 1 (the United States) on December 6, 2006. There were 24 abstentions, including Russia and China. After an unprecedented ninety-eight responses to the secretary-general’s call for states’ views on an ATT in 2007 (instead of the usual ten to fifteen responses for similar exercises), a Group of Governmental Experts met in 2008 to recommend a draft treaty for negotiation. I discuss the ATT process in detail in chapter 4.

4. For most governments, public reporting began only in the 1990s. The United States and a handful of states made reporting improvements during the late 1970s and 1980s. Austria, for example, began issuing reports in 1977, and Sweden followed suit in 1984 (Haug et al. 2002).

5. These problems are magnified substantially with the Eastern Bloc, which had a separate and extremely secretive system of trade (Brzoska and Ohlson 1987; Brzoska and Pearson 1994; Catrina 1988; Keller 1995; Pierre 1982).

6. In particular, China, Israel, Russia, South Africa, and Turkey.

7. SAS anticipates that Comtrade—although an extremely valuable source and the primary source for Cold War export data—generally underestimates the quantity and value of small arms transferred (2009:7).

8. On the properties and use of time-series dyadic data, see Beck and Katz 2001; Green, Kim, and Yoon 2001; G. King 2001.

9. Because the Arms Trade Data Set is intended to assess governments’ arms export decision making, it does not include black-market sales (i.e., sales not authorized by the government). Although illicit arms commonly originate in the legal market, the absence of a measure of illicit sales does not affect the data or analysis here.

10. To accommodate inconsistencies in information reported (price paid/value/amount purchased), SALW transfers are coded as a dichotomous variable, taken from raw Norwegian Initiative on Small Arms Transfers data. See appendix B for a more detailed discussion.

11. MCW data are taken from SIPRI, which uses a dollar-based standardized “trend indicator value” to measure the quality and quantity of weapons transferred. See appendix B for details.

12. The absence of data makes it impossible to assess arms export practice during earlier attempts at multilateral export controls.

13. Knowledge about Soviet arms transfer policy and practice is limited. Research finds that the Soviet Union viewed arms exports as an instrument of foreign policy in the struggle between socialism and imperialism in the Cold War superpower competition (Brzoska and Ohlson 1987:40–44; Catrina 1988:87–88).

14. The 1890 General Act for the Repression of the African Slave Trade (the Brussels Act) also contained arms clauses prohibiting arms transfers within a defined zone in Africa in order to disarm local resistance and restrict competing colonial powers’ activity. Its effects in practice were mixed. See Beachey 1962; Bromley, Cooper, and Holtom 2012. For an overview of pre-twentieth-century regulations (primarily national level), see Krause and MacDonald 1993.

15. David Stone points out that the major powers, although supportive in principle, still had security and economic concerns. The United Kingdom, for example, won exemptions related to supplying armed forces needed to maintain its empire (2000:222).

16. The Nye Committee was formally the U.S. Senate Special Committee on Investigation of the Munitions Industry. For U.S. diplomatic discussions about the proposal, see Foreign Relations of the United States Diplomatic Papers 1934 (“General, British Commonwealth” volume). At home, President Franklin Roosevelt billed it to Congress, when introducing the Nye Committee, as saving both the international community from war and strife “due in no small measure to the uncontrolled activities of the manufacturers and merchants of engines of destruction” and “the peoples of many countries” from “being taxed to the point of poverty and starvation in order to enable governments to engage in a mad race in armaments” (Roosevelt 1934:9095).

17. See Brzoska and Ohlson 1987; Keller and Nolan 1997; Klare 1984; Neuman 1986; Stanley and Pearton 1972

18. U.S. and Soviet arms transfers were also thought to have economic benefits, but their primary motivation was nevertheless political (Brzoska and Ohlson 1987:43; Pierre 1982).

19. COCOM lost its purpose with the end of the Cold War. In 1996, it was replaced by the Wassenaar Arrangement, which primarily promotes arms export transparency.

20. The United States sometimes struggled to make its allies adopt economic containment; West European states preferred to maintain trade links with their Eastern neighbors (Mastanduno 1992).

21. Arms sales also captured international attention with the UN’s 1963 voluntary arms embargo against South Africa, made mandatory in 1977. The embargo was intended to isolate the Pretoria regime, denounce apartheid, and change the government’s policy. However, it was neither strictly implemented nor enforced (Catrina 1988:142–44). In one example among many, the United States was accused of exporting spare parts for military use and turning a blind eye to some cases of technology acquisition and defense trade to enable monitoring of Soviet activity in the region (Klare 1981; Phythian 2000a). The embargo was one of the few mandatory UN sanctions during the Cold War and was lifted in May 1994.

22. Although policy makers introduced human rights into the discussion, public concern was for the destabilizing effect of “too many” arms exports to the developing world, not “to whom” the arms were specifically being sent.

23. On PD-13, see Ball and Leitenberg 1979; Brzoska and Ohlson 1987; Durch 2000; Hartung 1993; Hoffmann 1977–1978; Pierre 1982; Spear 1995.

24. The studies cited examine military aid rather than arms transfers as a whole (which also include the purchase of defense goods and may not include some parts of military aid). Unlike U.S. law on arms transfers more broadly, U.S. law on military aid has mandated since 1974 that a country’s human rights violations should adversely affect its receipt of military aid. Although these results therefore cannot directly substitute for an analysis of arms transfer practices, they nevertheless shed some much needed light onto the role of human rights in one major exporters’ military aid–giving practice during an important period of the Cold War.

25. Although David Cingranelli and Thomas Pasquarello (1985) do find that human rights modestly affect military aid to Latin America in fiscal year 1982, these findings have been contested based on their operationalization of human rights and selection of cases included in the analysis (Carleton and Stohl 1987; Mitchell and McCormick 1988).

26. Full results are reported in appendix C. The statistical analyses use ordinary least squares regressions (for the continuous MCW variable) and logit regressions (for the dichotomous SALW variable). In addition, panel-corrected standard errors must be used with the annual dyadic data to avoid an understatement of the errors due to the high number of error parameters involved in panel data, including panel heteroscedasticity and temporal dependence (Beck 2001; Beck and Katz 1995). I select control variables based on the theoretical relationship between the independent variable of interest (human rights) and the dependent variable (MCW/SALW). Results from models featuring unnecessary variables are often highly contingent on precise specifications and make it more difficult to explore nuances of the relationship between variables of interest and the outcome (Achen 2005; Berk 2004). Appendix B covers variable coding and justification for inclusion in statistical models.

27. See the sources cited in note 2 for this chapter.

28. Although still bound by bloc politics, European and other smaller producers, such as Brazil and Israel, sought to export arms to support their national arms industries, boost their balance of payments and employment, and reduce production costs (Catrina 1988; Pierre 1982; Stanley and Pearton 1972; Taylor 1994; Wentz 1987–1988). This is not to suggest that arms sales did not have economic benefits for the superpowers. In fact, arms were a significant part of Soviet foreign trade, though the Soviet Union supposedly did not intentionally seek out financial advantages from its arms exports (Brzoska and Ohlson 1987:43).

29. On this debate, see Chalmers et al. 2002; Hartley and Martin 2003; Hartung 1996; Ingram and Isbister 2004; Levine et al. 1997; S. Martin 1999; Mayhew 2005; Mintz and Huang 1991; Pierre 1982; Wentz 1987–1988.

30. See Anthony 1997; Brzoksa 2004; Durch 2000; Gold 1999; Harkavy 1994; Kapstein 1997; Keller and Nolan 1997

31. See Cornish 1995; Gold 1999; Kapstein 1997; Neuman 1993; Wulf 1993.

32. See Anthony 1997; Kapstein 1997; Neuman 1993; Wulf 1993.

33. Canada and Japan also proposed a Group of 7 (G7) working group to monitor arms transfers, but France objected to the G7 as the institutional platform. The G7 met in July 1991 and discussed, in part, the United Kingdom’s arms register proposal, which was subsequently pursued under UN auspices. Mikhail Gorbachev was invited to the 1991 G7 summit, and although China was not present, it was briefed in advance on the arms control initiatives to be discussed (Phythian 2000b).

34. Human rights did return to U.S. national arms export policy in 1995 with Clinton’s Presidential Decision Directive 34, which includes human rights (and other political and economic considerations) in its arms export criteria. Janne Nolan (1997) notes, however, that U.S. practices at this time did not change much from those of the Reagan and Bush administrations.

35. Countries were invited to include SALW starting in 2003.

36. Consensus rules for the 2001 UN small arms meeting made it possible for the United States to block a legally binding document that included multilateral transfer controls. I explore its position in detail in chapter 4.

37. Chapter 4 explores these major multilateral policy initiatives in detail.

38. The first sixty-seven signatory states to the ATT include Albania, Antigua and Barbuda, Argentina, Australia, Austria, Bahamas, Belgium, Belize, Benin, Brazil, Burkina Faso, Burundi, Chile, Costa Rica, Côte d’Ivoire, Croatia, Cyprus, Czech Republic, Denmark, Djibouti, Dominican Republic, Estonia, Finland, France, Germany, Greece, Grenada, Guyana, Hungary, Iceland, Ireland, Italy, Jamaica, Japan, Latvia, Liechtenstein, Lithuania, Luxembourg, Mali, Malta, Mauritania, Mexico, Montenegro, Mozambique, Netherlands, New Zealand, Norway, Palau, Panama, Portugal, South Korea, Romania, Saint Lucia, Saint Vincent and the Grenadines, Senegal, Seychelles, Slovenia, Spain, Suriname, Sweden, Switzerland, Tanzania, Togo, Trinidad and Tobago, Tuvalu, United Kingdom, and Uruguay. The United States signed on September 25, 2013. The ATT goes into effect after fifty states deposit their instruments of ratification on December 24, 2014.

39. It can also be argued that foreign policy has always been an expression and tool of a state’s values and ethics and that all that is new here is the explicit articulation and labeling by some states of their foreign policy as “ethical.” This debate was most prominent under the British Labour government between 1997 and 2010.

40. The inclusion of small arms on the international agenda in connection with states’ broader acceptance of international humanitarian norms was not an intentional “grafting” effort at the start, as in the landmines case (Price 1998), although the Control Arms Campaign would later do so in making its case for the ATT.

41. On these conflicts and arms control, see Eavis 1999; Goose and Smyth 1994; Hartung 2001a; HRW 2003a; Human Security Centre 2005; International Committee of the Red Cross 1999; Karp 1994; Klare 1994–1995; Misol 2004; Sislin and Pearson 2001; Wezeman 2003. See also chapter 1.

42. For a selection of contributions to this debate, see Brzoska 2001; Durch 2000; Erickson 2013c; Fruchart et al. 2007; Haass 1997; Hufbauer, Schott, and Elliot 1990; Kaempfer and Lowenberg 1995; Naylor 1999; Pape 1997; Phythian 2000a; Sislin and Pearson 2001.

43. See, for example, Blanton 1999, 2001; Boutwell and Klare 1999; Boutwell, Klare, and Reed 1995; Control Arms 2003; Craft and Smaldone 2002; Dhanapala et al. 1999; Goose and Smyth 1994; Harkavy and Neuman 2001; HRW 1999, 2003a; Human Security Centre 2005; International Committee of the Red Cross 1999; Karp 1993, 1994; Keppler 2001; Klare 1984, 1994–1995; Klare and Rotberg 1999; Maniruzzaman 1992; Misol 2004; Muggah and Berman 2001; SAS 2001, 2002, 2003, 2006, 2007, 2009; Sislin and Pearson 2001; Smith and Tasiran 2005.

44. Afghanistan and Mali were also key “affected” states. Denise Garcia (2006) provides a historic overview.

45. APL are a discrete category of weapon with limited economic or military importance. They are also inherently indiscriminate and can therefore be banned on the basis of international humanitarian law. In contrast, small and major conventional arms are broad categories of weapons with a legally recognized place in the provision of state security. Moreover, they have a significant worldwide economic base. Finally, unlike for landmines, there are both groups that advocate for and groups that oppose more restrictive controls of conventional weapons, which are not inherently indiscriminate weapons. See Brem and Rutherford 2001; D. Garcia 2006; O’Dwyer 2006.

46. See chapter 2. For materialist perspectives on compliance, see Downs, Rocke, and Barsoom 1996; Goldsmith and Posner 2005; Keohane 1984; Morrow 2007. From the normative or constructivist perspective, see Chayes and Chayes 1993; Checkel 2001; Finnemore and Toope 2001.

47. The following institutes in particular: SIPRI, the U.S. Arms Control and Disarmament Agency, and the International Institute for Strategic Studies in the United Kingdom.

48. Small arms transfers are tracked by the Norwegian Initiative on Small Arms Transfers, established in 1997.

49. The press and many NGOs do call states on questionable export deals, but these publicity incidents are typically isolated and more attention grabbing than a systematic investigation of practices over time (see chapter 5).

50. Blanton’s findings are potentially important—especially given the lack of research otherwise—but, as a superpower, the United States may be an exceptional case. Its policies have often gone against the grain of world politics, whether in support of or in opposition to shared arms export restrictions. The problematic design of Blanton’s Heckman models also calls her findings into question. Heckman models are subject to specific restrictions (see Achen 1986; Sartori 2003), which Blanton does not meet. Blanton 2000 uses the full model for both equations, instead of dropping at least one variable in stage 2. Blanton 2005 reverses model specification requirements by including all independent variables in the stage 2 and excluding a pair of variables (Saudi Arabia and GDP) from stage 1. She provides no explanation for her decision to reverse the model. Moreover, the dummy variable for Saudi Arabia, as Blanton’s model itself acknowledges, is of little value for the selection stage and should therefore be excluded from the model entirely (all variables in stage 2 must be included in stage 1). Finally, GDP is a highly influential variable that can significantly change the results of an equation depending on whether it is included or not. Excluding GDP leads to a poorly specified model and misleading results.

51. Reestimated models of five-year blocks or “windows” are used for each dependent variable. The models start with 1982, the first year for the lagged independent variables, and add an additional year to each regression until a five-year window has been reached. To illustrate, the first five-year window is 1982–1986, the next is 1983–1987, and so on. There are twenty-nine consecutive five-year windows in total, ending in 2010. For recent applications of models of windows in political science, see Adolph 2004; Kayser 2007; Kwon and Pontusson 2005.

52. This result is consistent with different iterations of exporters included in the model, such as democratic exporters only and EU exporters only (subject to the EU Code since 1998).

53. Although there was a drop with “very bad” human rights in 2010, the coefficient is insignificant, and a single point is insufficient to establish whether this drop is the start of (or return to) a trend.

54. See, for example, Capie 2004; Cloud 2007; V. Garcia 2003; HRW 2002; Myerscough 2006; Waldermeirin 2007; Wright 2007.

55. Control Arms (2003), for example, argues that the United Kingdom expanded arms availability to Indonesia and relaxed arms export controls more generally in response to U.S. arms export policy in the war on terror. Although the results for MCW transfers do not show a similar uptick for bad human rights performers, MCW export practice has generally shown less variation over time and may be slower to adapt.

56. However, recent findings suggest that legally binding export restrictions in the form of arms embargoes do significantly reduce SALW and MCW transfers (Erickson 2013c).

1. The United States opposed a broader mandate for the meeting, along with China, Israel, Russia, Cuba, India, and Pakistan. The latter three countries are not major exporters. By the ATT vote in 2006, only the United States was still outright opposed to shared arms transfer controls.

2. For this view, see Anders 2007; Efrat 2010; Karp 2002; Krause 2002; Miller et al. 2003; O’Dwyer 2006; Stohl and Hogendoorn 2010; Wyatt 2002.

3. All of the cases reviewed are full democracies and among the top-five small arms and/or MCW exporters. See chapter 1 for more on case selection.

4. Studies connecting arms export policy to the MIC and defense industry preferences include G. Adams 1981; Cooling 1981; Dunne 1995; Efrat 2010; Eisenhower 1961; Hartung 1996; Keller 1995; Kolodziej 1979; Kurth 1971; Markusen et al. 1991; Moravcsik 1991, 1992, 1993; Silverstein 2000.

5. Although born in the United States (Eisenhower 1961), this concept travels in Europe and elsewhere in the world (Cooling 1981; Kiss 1997; Stanley and Pearton 1972).

6. For example, the defense industry in the former Eastern Bloc enjoyed an equal—if not greater—status in domestic politics. Better conditions for defense industry employees during the Cold War “created a deep attachment to defence industrial activity,” which has carried over into the difficult post–Cold War years (Kiss 1997:137).

7. See Beier and Crosby 1998; International Committee of the Red Cross 1996; Petrova 2007; Price 1998; Rutherford 2000.

8. On the French MIC, see Boyer 1996; Chatillon 1983; Graves 2000; interview 63308220; Labbé 1994; Moravcsik 1992; Stanley and Pearton 1972.

9. On the French view of the arms trade, see Chatillon 1983; J. Clarke 1981; Défense et sécurité nationale 2008:279, 283; Graves 2000; Guay 1998:108; interview 63308220; Kolodziej 1987; Sarkozy 2008; Scaringella 1998.

10. After Labour governments nationalized “large swaths of the British defense industry,” Thatcher reprivatized companies during the 1980s (Graves 2000:60). Labour renounced its call for common ownership in 1995.

11. On the British MIC, see Bishop and Megicks 1996; Higham 1981; Jones and Rees 1994; Phythian 2000b.

12. Ian Davis (2002) provides a detailed history and description of both. See also Cooper 1997; Guay 1998; Jones and Rees 1994; Mayhew 2005; Phythian 2000b; Spear 1990.

13. On party support of the defense industry, see Cook, Foss, and Scott 2004; Jones and Rees 1994; Norton-Taylor, Lloyd, and Cook 1996.

14. The Financial Times called the second Al Yamamah deal “staggering both by its sheer size and by its strategic importance, not only for defence relations but also for investment and trade links between the two countries” (White and Mauthner 1988; see also “Arms and the Arabs” 1988). An investigation into the deals was suppressed in 1992, but the Serious Fraud Office opened another in 2004 to look into British Aerospace (BAE) and UK Ministry of Defense side payments. Although the deals were made during Thatcher’s tenure in office, Blair stopped the inquiry in 2006, arguing that it would lead to the loss of British jobs and “a vital strategic partnership” (“Saudi Prince” 2007). In a distinct change of tone, the press responded that the defense industry is worth preserving, but “not at any price” (“The Bigger Bang” 2007). BAE paid £286 million in criminal fines in 2010 to settle investigations in the United Kingdom and the United States without admitting to bribery or corruption (“BAE Handed” 2010).

15. Interviews 32107200, 35207200, 34207200, 36307200, 39307200.

16. Interviews 35207200, 37207200, 39307200.

17. Both the Allies and the German public were largely convinced of the MIC’s guilt, and the largest companies were broken apart in the early postwar yeas as a result (Homze 1981:76–77). The dismantling stopped by the late 1940s in response to the Soviet threat (Homze 1981:77), and a resurgence began by the 1950s (Homze 1981:79; see also Davis 2002, Pearson 1986).

18. Interview 44307255. See also Davis 2002; Freedman and Navias 1997; Moravcsik 1992; Stanley and Pearton 1972.

19. It is worth noting that much discretion surrounds the interpretation of German legal export restraints as well as a preference for coproduction arrangements with countries such as France that are subject to less-restrictive export guidelines and for selling production licenses instead of finished weapons (Brzoska 1986, 1989; Cowen 1986; Davis 2002; Graves 2000; Guay 1998; Pearson 1986; Wulf 1996).

20. The three geographic regions in Belgium—Wallonia, Flanders, and Brussels—now hold arms export–licensing competence. The Brussels share of Belgian business is extremely tiny and therefore not discussed here. Regionalization has created difficulties at the regional level—licensing and reporting mechanisms had to be developed from scratch—and problems for representation in multilateral fora at the international level. See Flemish Peace Institute 2007b.

21. Hassink 2000; interviews 20307211, 21407211, 22407211, 23407211, 31307211.

22. Over time, dual-use technology, such as visualization screens and military imaging and electronic equipment, has been the most valuable Flemish defense good (Duquet 2008; Duquet, Castryck, and Depauw 2007).

23. Interviews 21407211, 22407211, 23407211, 31307211.

24. See Guay 1998; Hartung 1996; Silverstein 2000; Wolpin 1991.

25. Arms exports were briefly emphasized as a political issue in the United States in the mid-1990s to draw out production lines and protect jobs (Hartung 1996; Nolan 1997). Yet, according to Terrence Guay, “many of the [U.S.] companies that have chosen to remain in the defense business now find themselves in a monopoly or near-monopoly position. Most contractors are generating strong cash flows, and many cost-cutting defense firms have seen their stocks rise. While defense companies’ revenues have fallen sharply in the 1990s, profitability has risen” (1998:90).

26. For more on the jobs–defense industry connection, see G. Adams 1981; Hartung 1996; Markusen et al. 1991; Stanley and Pearton 1972; Wolpin 1991.

27. One NGO representative commented that some defense firms had even privately expressed an interest in establishing an international code of conduct, but the American political climate was unfavorable to make investing the resources into it worthwhile (interview 48207002).

28. Johnston attributes states’ responses to new expectations to “social influence,” which he claims can include “backpatting, opprobrium or shaming, social liking, status maximization, and so on” (2008:20). Social influence allows for “public conformity without private acceptance,” unlike persuasion, which requires both public conformity and private acceptance (25), a dynamic strongly suggested by the gap between states’ arms transfer policies and practices.

29. These institutions are the most relevant for the major arms exporters. Others include ECOWAS, the Organization of American States, and the South African Development Community.

30. I do not mean to imply here that a reputation for good international citizenship—and the enhanced social status that might accompany it—cannot also indirectly contribute to material benefits.

31. Operating under the same consensus rules, the EU agreed to the legally binding Common Position to succeed its politically binding Code of Conduct in 2008. The UN, however, had to use majoritarian voting in the General Assembly between 2006 and 2013 to pass the ATT.

32. It is worth noting that in such cases less-powerful dissenting states can often hide behind the opposition of a superpower such as the United States. Supporting states and NGOs focus their opprobrium on the superpower’s opposition when it is present rather than on that of “less-significant” players.

33. Although states’ obligations are essentially the same under both documents, the Common Position does make some formal additions. See M. Bromley 2012.

34. Unanimity is not always required. Some issues (although not foreign and security policy) fall under Qualified Majority Voting (majority voting weighed by member population) as designated by EU treaty law. Nevertheless, between 75 and 80 percent of decisions in the 1990s were taken unanimously, even where majoritarian rules could have been applied (Mattila and Lane 2001:40).

35. Appeasing France so that any document could pass was a central concern of the UK presidency, which worked with France in order to minimize its opposition. As a result, the 1998 Code of Conduct was not legally binding and mandated only that members report exports annually to the EU, but not that reports be made public. In fact, France strongly opposed any references to public accountability proposed by the British Labour government, which was fresh off a national election in which transparency and accountability were key issues (SIPRI 1999:439–41).

36. On the EU debate over lifting the arms embargo to China, see Erickson 2013b.

37. The EU has facilitated further policy convergence by establishing a common list of military equipment; increasing the amount of information exchanged between member states; and creating a user’s guide to clarify criteria application (M. Bromley 2008:9).

38. European Parliament 2008; French Ministry of Foreign Affairs 2008; interviews 27207211, 37207200, 43107200, 61208220, 64208220.

39. Consensus rules are widely cited as the reason for the small arms conferences’ lack of success (SIPRI 2007:432). As one government official said, there were just “so many damn countries” (interview 16107255), making consensus perhaps an impossibly high bar to clear.

40. News sources and informal conversations with participants suggest that China, Cuba, Egypt, India, Iran, Israel, Pakistan, and Russia also opposed a more rigorous UNPOA (Anders 2007; SIPRI 2007:432).

41. See chapter 3 for a historical overview of the emergence of the international small arms agenda as well as SAS 2001, 2002 and UN Department of Disarmament Affairs 2002 for discussions of the origins of the 2001 conference.

42. See UN Department of Disarmament Affairs 2002:79 n. 17 for a full list of invited NGOs.

43. The United States has had a reputation in the UN General Assembly both for commonly flouting the majority (Smouts 2000:48) and for being an obstructionist force on the small arms processes in particular.

44. Covering illicit small arms transfers in “all their aspects,” the UNPOA is a comprehensive document acknowledging the deep complexity of the problem but also making consensus a next-to-impossible feat.

45. This group primarily included China, Cuba, Egypt, India, Iran, North Korea, Pakistan, Russia, Syria, Venezuela, and Zimbabwe. In the final ATT vote, however, only Iran, North Korea, and Syria voted in opposition.

46. Defense cooperation agreements are exempt from the treaty, for example, because of Indian lobbying. However, a ban on arms transfers to nonstate actors is not mentioned because of unresolvable disagreements between the United States (opposed) and India, Nicaragua, Russia, and others (in favor) that would have otherwise prevented treaty progress.

47. The ATT was adopted by majoritarian vote in the UN General Assembly after being blocked from consensus adoption in the final negotiation by Iran, North Korea, and Syria, which also voted against the treaty in the General Assembly. Some importing states abstained because they felt the ATT favored exporters over importers (e.g., Indonesia), whereas others abstained because they felt their substantive concerns had not been addressed (e.g., India on nonstate actors). Argentina, an ATT lead state, was the first to sign on June 3, 2013. Iceland was the first state to ratify the ATT, depositing its instrument of ratification on July 2, 2013.

48. To preserve the anonymity of the large majority of interviewees who requested it, I do not name any participants.

49. A similar point might be made about “rogue” states, whether by choice or not.

50. In particular, the United Kingdom’s announcement to renew its push for an ATT in September 2006 following the 2006 UN small arms conference was met with immediate and widespread support. According to Foreign Office minister Kim Howells, “There were ambassadors literally queuing up to express their support for the proposals we announced this morning and they ranged from very small countries like Mauritius to very large countries like Canada” (Associated Press 2006).

51. Interviews 39307200, 34207200, 37207200, 58107255.

52. For more about British leadership on this issue, see Erickson 2013a.

53. For example, see Blair 1997; R. Cook 1997a, 1997b, 2002; Cooper 2000; Dunne and Wheeler 2001; Wheeler and Dunne 1998.

54. Interviews 37207200, 35207200, 40107200, 41107200 43107200.

55. Interviews 37207200, 32107200, 35207200, 39307200, 43107200.

56. One NGO representative noted that although the idea of an ATT came from NGOs, he acknowledged that the scope and form of the current ATT was “invented” by Straw and continued by his successors (interview 33207200).

57. NGOs report that in the United Kingdom they engage in “behind-the-scenes advocacy, policy recommendations, and policy-based research” and “identify points of leverage” on the issue of arms exports (interview 37207200). It is not clear whether they remain behind the scenes by choice or if doing so has been forced by lead governments seeking credit.

58. Interviews 16107255, 19407255, 45207255, 58107255.

59. One German arms trade expert reports that Germany was “disappointed with the ammunition thing and was hopeful that this would be the German thing” (interview 19407255). Promoting the ATT during Germany’s 2007 EU presidency seemed like a promising opportunity for leadership, though the United Kingdom had already claimed leadership a few years earlier. Germany has been, according to many, a more passive supporter in practice and has “missed some good opportunities to seize the field to play a prominent role” (interview 58107255).

60. Interviews 14107255, 16107255, 15107255, 58107255.

61. Interviews 10407255, 14107255, 16107255, 19407255, 58107255.

62. This comment suggests the use of “rhetorical entrapment” as an NGO strategy and perhaps an unintended consequence of states’ support for “responsible” arms transfer policies and related multilateral initiatives (Schimmelfennig 2001).

63. Whether Germany’s civilian power will or can persist is a matter of debate (Harnisch and Maull 2001; Hockenos 2007).

64. Or, according to some critics, French foreign policy is “long on symbolism but woefully short on substance” (Gordon 1993:134).

65. The idea that France is “destined to play leading global and European roles” has been a foundation of French foreign policy since De Gaulle (Gordon 1993:17).

66. Among French elites, consensus over France’s support of and role in the EU is accepted and explained as “Mitterrand’s European legacy” (Wood 1997:131). A similar argument can be made that France promotes the UN’s “prestige and capabilities,” strengthening its “own status and influence” in the process (Gregory 2000:170; see also Utley 2000).

67. The head of state prevails on major diplomatic questions and keeps a “direct relationship” with the minister of foreign affairs, “which enhances coordinated action” (Enjalran and Husson 1999:66).

68. France’s arms trade NGO culture—or lack thereof—is detailed in chapter 5.

69. Belgian federalism provides a separation of competences but no hierarchy between the federal and regional governments, “designed to save rather than destroy national unity” (Fitzmaurice 1996:145).

70. The problems for Belgian foreign-policy making were initially more practical. Federal and regional officials were unclear how Belgium would be represented in international institutions and whether regional officials, who now implement arms export policies, would acquire equal representation alongside federal officials, who retain foreign-policy-making competences.

71. The Ministry of Foreign Affairs values its role in the EU and UN. Marc Houben attributes this attitude to Belgium’s “self-image of vulnerability,” which is “deeply rooted in society and widely shared by the different political parties” (2005:31). Arms export decision making aside, foreign policy and European integration are considered matters of consensus in Belgian politics (Fitzmaurice 1996).

72. The Belgian federal government signs in the name of all governments and is “very pro [arms control]” but has not put the issue to the regions for input in practice. The regions, however, reportedly “don’t see a reason not to follow” (interview 21407211). External relations, including with the EU, are typically delegated to the competence of the federal government (Fitzmaurice 1996:148, 245–50).

73. Interviews 20307211, 52207211, 53207211, 56107211, 57207211.

74. This satisfaction with Belgian policy does not necessarily extend to the regions. In particular, NGOs hope that Wallonia will improve transparency measures. See also Amnesty International 2007 for broad policy objectives.

75. On this reputation, see Adam 1989; Inter Press Service 1987; Lowther 1991; Vranckx 2005.

76. Please note that conversations with U.S. government officials inform the analysis but are not for attribution.

77. This theme was certainly clear in interviews outside of the United States.

78. See the discussion of Israeli policy in chapter 6.

79. Stohl, it should be noted, was nominated by Senator Diane Feinstein (D–CA) for a spot on the 2006 U.S. delegation as the choice of the pro-control NGO community. She was turned down in favor of three NRA members, as also happened with the 2001 delegation.

80. Despite attempts over several years, I was unable to interview any representatives from the NRA or the World Forum on the Future of Sport Shooting Activity, which represents the NRA at the UN.

81. Some NRA advocates suggest that the right to bear arms is a universal human right (Schmidt 2007). Other experts, however, have contested this claim in a comparative context, finding that some states have explicitly rejected Second Amendment–like rights in their own national laws, while international law protects states’ right to self-defense, not individuals’ right to self-defense or to bear arms (Frey 2006; Miller et al. 2003).

82. Of course, the NRA membership—or the “pro-gun” constituency more broadly—is not homogeneous. Some members believe it should stay out of issues such as the arms trade and focus on firearm safety and hunting. Moreover, as Kristin Goss (2006) points out, its membership base is small compared to the majority of Americans who consistently support gun control.

83. A Gallup poll from early October 2009 found that although the Obama administration has not pursued a domestic gun control agenda, 55 percent of gun owners and 60 percent of Republican-leaning respondents indicated that they believed Obama would attempt to ban gun sales in his time as president (Newport 2009).

84. Following the first small arms meeting in 2001, World Forum on the Future of Sport Shooting Activity representative Thomas Mason, however, warned that some states would “seek to resurrect the issue of civilian possession of firearms” (2002:205). Since then, the UN has consistently reaffirmed civilian possession as a domestic issue outside of the scope of discussions. Yet for the NRA, there may be more to the story. As Aaron Karp observes, “[The NRA’s] exclusive concern was protecting the rights of U.S. civilian gun owners to buy and sell firearms as they pleased.… As it learned that domestic disarmament was not part of the [UN] agenda, it grew more comfortable” (2002:190). He suggests instead that NRA leaders have used “the UN process as a whipping-boy” to feed into the distaste of “extreme supporters” to “anything connected to the UN” (190). Wayne LaPierre also notes that the NRA used the issue in the 2000 election to rally support to its cause and preferred candidates as well as to solicit additional funds from its members (2006:53; see also Feldman 2007).

85. ATT preamble, available at http://www.un.org/disarmament/ATT/docs/ATT_text_(As_adopted_by_the_GA)-E.pdf.

86. For recent discussions about the origins and consequences of U.S. exceptionalism, see Chayes 2008; Fuchs and Klingemann 2008; Ignatieff 2005; Koh 2003; Kohut and Stokes 2006; Lipset 1996; Madsen 1998; Sperling 2007.

87. Both Carter and Clinton relied heavily on case-by-case exemptions despite policies that may have appeared more in line with current calls for “responsible” arms export controls.

88. Polling suggests that Obama’s handling of foreign policy within the first year of his presidency quadrupled his predecessor’s approval ratings among key European allies (“Transatlantic Trends” 2009).

89. A domestic backlash against the decline of U.S. world standing was palpable throughout the 2008 election process, leading the candidates to declare the need to improve U.S. reputation (Koh 2003; Kohut and Stokes 2006; “Only in America” 2008; Walt 2005). However, the ATT was unlikely to win points at home except perhaps indirectly by gaining points abroad (see chapter 5).

90. Michael Ignatieff notes that the stability of U.S. politics and institutions has led in part to its “legal isolationism” and reluctance to look to international law and foreign precedents in U.S. law (2005:8–9) as well as to a sense of self-sufficiency and lower incentive “to stabilize its own institutions with foreign treaties,” in contrast to the European powers (17).

1. In 1995, Ecuador was involved in a border war with Peru. Argentina was charged as a neutral guarantor of a 1942 peace treaty between the two parties. A UN embargo prohibited arms sales to Croatia from 1991 to 1996.

2. The Argentine Senate must lift Menem’s immunity before he can serve his seven-year sentence.

3. Yet 42 percent of the Argentine public were opposed to giving the UN the power to regulate the international arms trade, and only 36 percent were in favor (Chicago Council on Global Affairs and WorldPublicOpinion.org 2007).

4. With a permissive public attitude toward the arms trade and the high level of secrecy surrounding it, scandals were rare during the Cold War. The U.S. Iran– Contra Affair and the Swedish Bofors Affair are two exceptions.

5. Only Sweden and the United States allow for advance legislative scrutiny of arms export deals (those worth more than $50 million in the United States).

6. On this decline, see Baum 2002; Melanson 2000; Small 1996.

7. On the public and foreign policy, see Almond [1950] 1965; Baum 2002; Holsti 2004; Risse-Kappen 1991; Rosenau 1961; Sobel 2001; Wittkopf 1990.

8. See O’Dwyer 2006; Petrova 2007; Price 1998; Rutherford 2000.

9. Funding needs may also contribute to this difficulty. NGOs work on many issues related to SALW to keep donors interested, but this multitasking makes creating a clear, coherent, and simple message for campaigning difficult (interview 2206225).

10. For further information about the interviews, see chapter 4 and appendix B.

11. Interviews 19407255, 11407255, 17207255, 46307255.

12. As I show later in this chapter, it is typically through NGO press releases and prompting that the media pick up news stories about arms export decisions.

13. Interviews 20307211, 21407211–23407211, 26207211, 29207211, 31307211.

14. Interviews 37207200, 33207200, 36307200, 39307200, 43107200.

15. The 1970s were an unusual interlude to this trend. In his presidential campaign, Jimmy Carter responded to the U.S. public’s apparent concern regarding an overabundance of arms transfers (see chapter 3). The 1970s also saw among an attentive European public a generally negative attitude toward arms transfers, but no corresponding policy changes.

16. Chicago Council on Foreign Relations 2004; Chicago Council on Global Affairs 2008; Chicago Council on Global Affairs and WorldPublicOpinion.org 2007.

17. Interviews 59108220, 60108220, 63308220, 64208220. In 2006, 77 percent of French people supported giving the UN the power to regulate the international arms trade (Chicago Council on Global Affairs and WorldPublicOpinion.org 2007). Interview participants, however, seemed unaware of such figures. One NGO representative observed that no one was really sure where the public stood on the issue and that a study would be needed to find out (interview 61208220).

18. Domestic structure comprises “the nature of the political institutions (the ‘state’), basic features of the society, and the institutional and organizational arrangements linking state and society and channeling societal demands into the political system” (Risse-Kappen 1991:484).

19. I identify arms trade scandals by widespread media, NGO, and government use of the label scandal. See chapter 2 for a theoretical overview of the concept, literature, and the effects of scandal on democratic politics.

20. Low-level transparency exists where governments do not make publicly available their arms export reports. Although news of “irresponsible” exports can break due to investigative journalism or NGO fieldwork, it is extremely rare. See chapter 2 for a conceptual discussion of transparency.

21. Because the size, resources, and structure of national NGO communities vary by case, I define an “active pro-control NGO community” as the presence of groups engaged in sustained lobbying and public mobilization with the goal of regulating the national arms trade. They may be home-grown groups or national affiliates of international NGOs such as Amnesty International.

22. On uneven execution of information sharing, see Abramson 2008; M. Bromley 2012; Holtom 2008; Laurance, Wagenmakers, and Wulf 2005; Lebovic 2006.

23. Transparency has also increased within the arms industry in response to pressure from company stakeholders, shareholders, NGOs, and governments. Small arms companies that are not listed on the stock market tend to be less transparent (Weidacher 2005).

24. See Phythian 2000b; Ponting 1990; Tomkins 1998; Vincent 1998.

25. The Gesetz über die Kontrolle von Kriegswaffen (known widely as the KWKG) regulates the export of “weapons of war” as defined by German law. The Aussenwirtschaftsgesetz (known as the AWG) regulates military-related technology and armaments, including dual-use technology. See Davis 2002 for a summary.

26. Richard Moose and Daniel Spiegel attribute this concern to “the end of the Vietnam war, the personal style of a Secretary of State [Kissinger], the oil price hike of 1973, and the recession” (1979:228).

27. For more on sleaze and the 1997 election, see Lee 1999; Vincent 1998; Worcester and Mortimore 1999.

28. See Aaronovitch 1996; “The Scott Report” 1996; and Stephens 1996, among many (indeed most) others.

29. On Labour’s promises, see Blair 1997; Cook 1997a, 1997b; Labour Party 1997.

30. In the United Kingdom, the Campaign Against the Arms Trade had long advocated an arms trade ban. However, it has taken a back seat in the current debate as a result of other groups’ willingness to compromise (regulate, not ban). Anna Stavrianakis argues that the main NGOs in the current debate are “reformist” rather than “transformist,” upholding “the close relationship between arms capital and the state” and seeking to “[regulate] away [the arms trade’s] worst excesses within the existing system” rather than to overhaul it radically (2010:9).

31. Interviews 32107200, 37207200, 40107200, 41107200, 42107200, 43107200.

32. Broek 1997, Control Arms 2004, Oxfam 1998, 2002, and Saferworld 2007 are examples of NGO reports seeking to improve arms export practices, which have also picked up a following in the press.

33. See also UK House of Commons 1999, debate on November 3, and 2000, debate on May 4, among others.

34. See the All-Party Parliamentary Group on International Corporate Responsibility website at http://appg-icr.org/.

35. Interviews 35207200, 37207200, 38207200, 39307200.

36. The exception has been support for more liberal East–West trade as a part of Ostpolitik (Davis 2002:159).

37. The 1982 guidelines (Cowen 1986:269) held until new guidelines reflecting the 1998 EU Code of Conduct were adopted in Germany in January 2000. Current criteria include provisions to disqualify applicants who are located in areas of tension, are engaged in an armed conflict, or show “reasonable grounds to suspect serious human rights violations.” For the full text of the new guidelines, see http://www.bmwi.de/DE/Themen/Aussenwirtschaft/aussenwirtschaftsrecht.html.

38. The domestic economy is typically the point of greatest concern for Germans (Klein and Kuhlmann 2000). In periods of economic trouble and high unemployment, this concern can translate into tolerance for more arms transfers (Brzoska 1989; Davis 2002). Even then, however, governments have sought to avoid criticism by maintaining “the public image of a restrictive arms transfer policy” (Brzoska 1989:171).

39. Of course, this aspect of German identity can easily clash with its identity as an export-oriented economic powerhouse with respect to the arms trade (Davis 2002).

40. Alternative sources of information included reports in the foreign press from other intelligence sources, data collection and publication from open sources by organizations such as SIPRI, and individual groups’ efforts in Germany to monitor port activity. The BUKO Kampagne: Stoppt den Rüstungsexport (Campaign to Stop Arms Exports), which sought to ban the arms trade completely, especially engaged in data collection and dissemination strategies.

41. Of course, common standards of export control today are meant in part to keep suppliers from following this coproduction strategy. Germany nevertheless supports these policies, recognizing their importance for its international reputation (see chapter 4).

42. Chemical weapons and dual-use technologies present a useful comparison. In the early 1990s, scandal emerged with knowledge that Germany had supplied Iraq with chemical weapons and infrastructure in the 1980s (Phythian 1997). The government immediately responded with legal reform to place stricter controls on dual-use technology (Davis 2002; Müller et al. 1994). Scandal accelerated the reform process by authorities “[a]nxious to repair the tainted reputation of Germany” (Müller et al. 1994:4).

43. Activist Andrea Kolling notes that the German government got around the “areas of tension” criterion by declaring in 1984 that there was no Indonesian war against East Timor (1997:58).

44. “Support for NATO” means support both for Turkey as a NATO ally and for the United States, whose policy supported Ankara and labeled the Kurdish Workers’ Party a terrorist organization.

45. In the 1992 Schleswig-Holstein election, this candidate—CDU leader Ottfried Hennig—did not take the majority from the SPD. Yet both parties lost ground to right-wing extremist parties (May 1992).

46. On Stoltenberg, see Casdorff 1992; M. Fisher 1992; Kinzer 1992; “Unter dem Druck” 1992. Stoltenberg had also been implicated in covert transfers of supplies of former East German arms to Israel—an area of tension in the Middle East—in 1991, a factor that certainly increased public distaste for his continued role as head of the Defense Ministry (M. Fisher 1992; “Kritik an Stoltenberg” 1992).

47. Interviews 14107255, 17207255, 19407255, 44307255, 45207255.

48. For a sample of overview pieces on German civilian power, see Harnisch and Maull 2001 as well as Hockenos 2007.

49. See Gaetner 1987 and Pontaut 1987 for the articles that (re)broke the would-be scandal.

50. All interviewees pointed out the lack of an active arms trade NGO community in France. Its absence is considered symptomatic of an absence of an established French NGO lobbying culture more broadly. Although foreign development and aid groups are prominent, the proliferation of NGO activity in domestic politics on foreign-policy issues is more recent and a point of contention in French politics (S. Cohen 2004).

51. On French arms supplies to Rwanda, see Alusala 2004; Austin 1995; Braeckman 1994; Callamard 1999; Dorn, Matloff, and Matthews 1999; HRW 1994; Melvern 2000, 2004.

52. See Patrick De Saint-Exupéry’s (1998a, 1998b, 1998c, 1998d) four-part series in Le Figaro in January 1998 and Rousselin 1998. The reports accused France of complicity in the genocide and openly criticized the content and conduct of France’s Africa policy. See also Le Figaro in general from March 30 to April 3, 1998.

53. For the Quilès Commission report, see http://www.assemblee-nationale.fr/dos-siers/rwanda/r1271.asp.

54. The Rwandan government has been dissatisfied with French responses to the scandal and has since made allegations that France played an active role in the 1994 genocide (“Report Done” 2007; “Rwanda Accuses France” 2008).

55. Note later in the chapter similar public sentiments in the United States during the Iran–Contra Affair.

56. The Rwanda Affair did, however, expose major flaws in France’s Africa policy and helped to instigate reform (Utley 2002). And although the arms trade angle remained untouched, it later spurred French activism to control illicit small arms flows in Africa (interview 60108220; Utley 2002).

57. Primarily Amnesty, Agir-Ici–Oxfam, Comité Catholique contre la faim et pour le développement, l’Observatoire des transfers d’armements, and Secours Catholique—Caritas France.

58. Boland I prohibited the use of funds “for the purpose of overthrowing the Government of Nicaragua or provoking military exchange between Nicaragua and Honduras” (qtd. in Draper 1991:18). Boland II “prohibited any military or paramilitary support for the Nicaraguan contras for the period of October 3, 1984 to December 19, 1985” (Draper 1991:23–24) and remained in effect into 1986.

59. The plan was to focus questions on “North’s diversion scheme and [shy] away from the larger issues of administration policy toward Iran and the Contras” (Kornbluh and Byrne 1993:310). Attorney General Edwin Meese testified in 1989, “I was concerned that the two major policy issues within the Administration at the time would be merged together and that this would—could complicate the ability of the President in both of the issues” and might constitute an impeachable offense if revealed by a source outside the administration (qtd. in Draper 1991:521).

60. On these possible repercussions of Iran–Contra, see, for example, Barnes 1987; R. Busby 1999; Cohen and Mitchell 1988; Draper 1991; Kornbluh and Byrne 1993; Thelen 1996; Trager 1988; Walsh 1997; Wroe 1991.

61. For the full reports, see http://www.gwu.edu/~nsarchiv/NSAEBB/NSAEBB365/index.htm.

62. Note the ongoing covert arms transfers (primarily dual-use technologies for biological and chemical warfare) to Iraq during this time, also against U.S. law and policy. However, it was a matter of debate “just how much of a scandal Iraqgate was,” and the lack of any newfound accountability following Iran–Contra was clear (Jentleson 1994:9; see also Kornbluh and Byrne 1993).

63. To the extent that some NRA leaders are interested, they argue that governments should supply more arms to areas of conflict and genocide (LaPierre 2006).

64. The U.S. NGO community focused on domestic gun control has not worked on export policy. NGOs advocating arms export controls (among other issues) include Amnesty U.S., the Federation of American Scientists, HRW, the Quakers, and a few think tanks, such as the former Center for Defense Information.

65. On the “supergun,” see Lowther 1991; Naylor 1999; Toolis 1990.

66. Bull sent U.S.- and Canadian-made shell casings to PRB because Canadian law did not require end-user certificates for “inert” shells to going European countries. The manufacturer “in turn could arm the shells and sell them to anywhere that [the much less restrictive] Belgian law would allow” (Lowther 1991:107).

67. Cools’s death initially appeared connected to the supergun. However, investigations concluded in 1995 that “the personal secretary to the senior minister of the Wallonian regional government had been using his office as a cover for a gang stealing securities, credit cards and paintings.” Cools had pushed for an investigation, and in response the official hired an assassin (Naylor 1999:308–9, 311).

68. Past restrictions in arms trade to recipients engaged in conflict or internal repression were policy, not law (Inter Press Service 1987).

69. See also Adam 1989; Inter Press Service 1987; Vranckx 2005.

70. Reports indicated that the Flemish Volksunie Party refused to authorize Belgian arms sales to Saudi Arabia (Agence France Presse 1991a), which caused a split between Flemish- and French-speaking ministers in the coalition and dissolved Parliament (Agence France Press 1991b).

71. Prominent and active NGOs include Amnesty Flanders and Wallonia, International Peace Information Service, Network Vlaanderen, Oxfam–Solidarity Belgium, and research groups such as the Groupe de recherche et d’information sur la paix et la sécurité (Research and Information Group on Peace and Security) and the Flemish Peace Institute.

72. The Code states, “Before any Member State grants a license which has been denied by another Member State or States for an essentially identical transaction within the last three years, it will first consult the Member State or States which issued the denial(s). If following consultations, the Member State nevertheless decides to grant a license, it will notify the Member State or States issuing the denial(s), giving a detailed explanation of its reasoning” (Operative Provision no. 3). For the full text of the Code, see http://www.consilium.europa.eu/uedocs/cmsUpload/08675r2en8.pdf.

73. Flemish Green Party (Groen) health minister Magda Aelvoet resigned.

74. The section heading refers to the 2003 regionalization of export controls (Otte and Verschelden 2006).

75. See also interviews 21407211, 24107211, 27207211, 52207211, 55107211.

76. Interviews 29207211, 21407211, 27207211; Mampaey 2002:13.

77. Export denials noted in the Flemish press following regionalization include exports to Chile, Ethiopia, Saudi Arabia, Suriname, Tanzania, Turkey, and Venezuela (Brinckman 2005; Verschelden 2006a, 2006b).

78. In the case of Liberia, the Belgian government noted that Belgium had been granted an exemption by the UN in order to supply local police under UN auspices (Brinckman 2005).

79. This is not to say that NGOs have no access to the Walloon government. One official notes that the government sometimes has contact with NGOs in addition to meetings with companies (interview 55107211).

1. See, for example, Amnesty International 2010; Blanton 1999, 2001; Boutwell, Klare, and Reed 1995; Craft and Smaldone 2002; Eavis 1999; Goose and Smyth 1994; Harkavy and Neuman 2001; Hartung 2001a; Karp 1993, 1994; Klare and Rotberg 1999; Musah 2002; Neuman 1986; Sislin and Pearson 2001; Verwimp 2006.

2. Attention to arms exporters in the post–Cold War scholarly literature is limited, perhaps because of sparse media attention or a widely shared mindset that not much has changed over time or both.

3. The United States was the sole vote against the ATT process until late 2009, a move that contributed to its obstructionist international reputation. I attribute U.S. opposition primarily to its long-term reluctance to concede to external export constraints and exceptionalist attitude toward international institutions. See chapter 4.

4. Israel was ranked seventh among recipients and tenth among MCW exporters from 2002 to 2006 (SIPRI 2007:418, 422). SAS lists Israel in the lower end of the top-thirty small arms exporters (2006:71).

5. On postapartheid South Africa’s search for international legitimacy, see Becker 2010.

6. South Africa developed its defense industrial base during the apartheid era. It is now a midsize MCW exporter, ranked twentieth from 2002 to 2006 (SIPRI 2007). It is also in the lower end of the top-thirty SALW exporters (SAS 2006).

7. South Africa released annual arms trade reports from 1995 to 2002, which were subject to civil society criticism on specific cases (e.g., sales to Rwanda and Pakistan). Legislative reform in 2002 required annual export reports, but none was presented from 2003 to 2006. In response, NGOs “expressed concerns that South Africa may be returning to apartheid era secrecy.” Reports for 2003–2006 were released in 2007 but were viewed as too late to allow any real critique or action (Lamb 2007).

8. Known South African arms exports to human rights violators from 2000 to 2004 include those to China, Colombia, Indonesia, Nigeria, Pakistan, Saudi Arabia, Uganda, and Zimbabwe (Nathan 2005:371; SAS 2007:98–107).

9. International Action Network on Small Arms (IANSA) affiliates include Amnesty International, Centre for Conflict Resolution, Coalition for Peace in Africa– Action Support Centre, Congolese Association for Peace and Development, Gun Free South Africa, Institute for Security Studies, the Peace and Security Program at the University of Witwatersrand, and SaferAfrica.

10. Brazil ranked thirty-third in MCW exports from 2002 to 2006 (SIPRI 2007:418).

11. Pro-gun groups do not dominate Brazilian politics on this issue as the NRA does in the United States. IANSA lists a number of active, established pro-control affiliates: Movimento Paz Espirito Santo, Desarme.org, Viva Rio, Children and Youth in Organised Armed Violence, and Instituto Sou da Paz. Many pro-control groups are more focused on domestic than international issues. Viva Rio and Sou da Paz are involved in both.

12. Matthew Baum sees a similar trend of broadening public attention to political issues through soft news, which can frame them as “compelling human dramas” and capture the imagination of a public that would ordinarily not be tuned into or interested in such events (2002:91).

13. See Schimmelfennig 2001; Wheeler 2000.

14. Edward Kwakwa (2003) observes that separating the interests of the United States and the interests of the international community requires the exclusion of U.S. social interests, which affect its reputation and social standing.

15. This is a primary reason ICBL leaders chose to pursue the landmine treaty outside of the UN. Similarly, the EU consented to the politically binding Code of Conduct for Arms Exports in 1998 because of French opposition to a legally binding document (France withdrew its opposition to legalization in 2008).

16. Research generally finds that norms will diffuse more easily when there is a cultural match between international and domestic norms (Checkel 1999). Leaders’ choice of framing, the use of naming and shaming, and the wielding of social and material incentives to adopt norms can also affect norm diffusion (Finnemore and Sikkink 1998; Nadelmann 1990; Price 1998), as can norm-taking states’ domestic structures (Checkel 1999). Norm adoption by “critical states” can also speed and spread the process (Finnemore and Sikkink 1998; Florini 1996; Nadelmann 1990), as happened with British support for “responsible” arms export controls.

17. See http://www.un.org/disarmament/ATT/docs/ATT_text_(As_adopted_by_the_GA)-E.pdf.

1. Although thirteen states were absent for the UN General Assembly vote on Document A/67/L.58, all exporter states in the data set were present. Of the twenty-three formal abstentions, only China and Russia abstained among the exporters included in the data set. Only three states formally voted against the ATT (Iran, North Korea, and Syria), none of which are major exporters and are therefore included in the data set only as importer states.

2. No database provides a comprehensive combined figure for annual SALW and MCW transfers disaggregated by recipient outside of U.S. commercial arms sales.

3. SIPRI is the only public source for annual conventional arms trade disaggregated export–import data for non-U.S. exporters. Although there is little quantitative work on the conventional arms trade, researchers often use SIPRI data to illustrate trends in the trade. Post–Cold War examples include Durch 2000; Erickson 2013b, 2013c; Golde and Tishler 2004; Harkavy 1994; Khanna 1992; Kinsella 1994; Sanjian 1998; Wulf 1993.

4. I thank Mark Bromley at SIPRI for providing me with the initial data from the SIPRI Arms Transfer Database in 2006 (see SIPRI 2006). The current version of this database is now available at http://www.sipri.org.

5. Re-exports authorized by the original recipient state should appear with its transfer data.

6. For a complete description of SIPRI’s TIV calculation methodology, see SIPRI 2007:429–30.

7. SIPRI argues that identifying an accurate and reliable price figure for weapons is impossible for three reasons: “First, in many cases no reliable data on the value of a transfer are available. Second, even if the value of a transfer is known, in almost every case it is the total value of a deal, which may include not only the weapons themselves but also other items related to these weapons … as well as support systems … and items related to the integration of the weapon in the arms forces. Third, even if the value of a transfer is known, important details about the financial arrangements of the transfer (e.g., credit or loan conditions and discounts) are often unavailable” (2007:429).

8. I accessed NISAT in spring 2006, fall 2011, and summer 2013 from http://nisat.prio.org. See NISAT n.d.

9. SAS calls UN Comtrade “the most comprehensive source of comparable data on the international trade in small arms” (2003:99). However, Comtrade relies on voluntary submissions by governments and tends to be incomplete in the area of SALW, a particularly sensitive export. SAS concludes that Comtrade data “[appear] to be the strongest on the exports and imports of western countries, principally North America and western Europe” (99).

10. See, for example, Forsythe 2006; Krain 1997; Lopez 1986:90; Poe and Tate 1994.

11. The presence of oil may cause a government to repress its citizens to protect its access to the resource (HRW 2003b; Ross 2004); fund repressive regimes, thus allowing them to stay in power (Chen 2007; HRW 2003b); attract companies that may violate human rights (Forsythe 2006; UN Economic and Social Council 2006); and cause other states to turn a blind eye to a government’s repressive practices in order to retain access to an important resource (Chen 2007; Forsythe 2006).

12. See, for example, Blanton 1999; Bueno de Mesquita et al. 2005; Davenport 1995; Howard and Donnelly 1986; Mitchell and McCormick 1988; Poe and Tate 1994; Rummel 1995.

13. See, for example, Davenport 1995; Davenport and Armstrong 2004; Mitchell and McCormick 1988; Poe and Tate 1994; Wolpin 1986.

14. See, for example, Banks 1986; Carleton and Stohl 1987; Cingranelli and Richards 1999a, 1999b; Goldstein 1986; Hafner-Burton and Ron 2009; Human Security Centre 2005; Poe, Carey, and Vazquez 2001; and Stohl et al. 1986.

15. Sabine Carey (2007) also uses this approach.

16. The primary advantage of the Cingranelli–Richards data is that they allow researchers to disaggregate types of physical integrity violations. However, this level of detail is too fine-grained for a study of arms export practice. See Hafner-Burton and Ron 2009 for comparisons between the Cingranelli–Richards data and PTS data.

17. The correlation between variables from the two sources is nevertheless high (0.808).

18. To illustrate, 54 percent of Amnesty-coded dyad-years fall from 3 to 5 on the PTS in contrast to 45 percent of DOS-coded dyad-years.

19. Nevertheless, Emilie Hafner-Burton and James Ron note that although democracies “are better protectors of human rights,” measures of human rights are not highly correlated with measures of democracy, suggesting that countries can be both democratic and abusive (2009:379, 365).