Organizational Fields and International Regimes

Multilateral export control, I will show, constitutes such an organizational field, and the process of institutional isomorphism by which standard practices are diffused within fields be

Organizational Fields and International Regimes

Michael LipsonPolitical Science Department

University of Pennsylvania

June 2001

Christopher H Browne Center for International Politics

University of PennsylvaniaWorking Paper Series #01-03

Organizational Fields and International Regimes

Michael Lipson*

Introduction

The study of international regimes has been a central area of research into international relations in the last twenty years.1 This literature has grown exponentially in the quarter century since the phrase “international regimes” was coined.2 Yet fundamental questions remain unanswered or inadequately explained This paper focuses on two such questions First, how do interactions between regimes affect their development? I suggest an

answer, derived from sociological institutionalist theory and based on empirical research into recent nonproliferation export control cooperation This answer also holds

implications for a second gap in regime theory: explaining the development of areas, to which regimes correspond

issue-In recent decades, multilateral export control regimes addressing nuclear,

chemical, biological, and conventional technologies with weapons applications have come to interact amongst themselves with increasing regularity This interaction has produced marked convergence in export control structures and processes across regimes and their member states Rationalist explanations portray this convergence as driven by optimization to the most efficient means of export control This account, however, is

* I would like to thank Michael Barnett, Roland Paris, Rudy Sil, Avery Goldstein, Mark Pollack, Mark Suchman, and Bruce Cronin for comments on drafts of this paper and the dissertation from which it is derived I have also benefited from comments on presentations of this research at the University of

Colorado and the University of Pennsylvania All errors of fact and interpretation are, of course, my own.

1 Ruggie 1975: Krasner 1983; Haggard and Simmons 1987, Rittberger 1993; Hasenclever, et al 1997; Levy, et al 1995.

2 Ruggie 1975, 569.

incomplete and only weakly supported by available evidence A better explanation can befound in the concept of “organizational fields,” drawn from sociological institutionalism

Within organizational fields—communities of related organizations—common practices are adopted across organizations based on conformity with legitimized

standards rather than evidence of effectiveness Multilateral export control, I will show, constitutes such an organizational field, and the process of institutional isomorphism by which standard practices are diffused within fields better accounts for convergence acrossexport control regimes than do rationalist explanations This suggests that cross-regime interactions can have profound effects on regime development and functioning

Furthermore, interaction between the export control organizational field and the various export control regimes has also redefined the boundaries of the nonproliferation issue-area, suggesting a mechanism by which issue-areas, which define regime scope, candevelop: Fields and issue-areas can coevolve In the sections that follow, I address these gaps in extant regime theory by illustrating effects of cross-regime linkages in export control on issue-area development and regime dynamics

In the first section, I describe recent developments in multilateral export control, and show why they are puzzling from a rationalist perspective and how they highlight gaps in theories of international regimes I then outline an “organizational field” approach

to this puzzle, and explain how it addresses both the empirical puzzle and gaps in regime theory regarding cross-regime linkages and issue-area development

In the next section, I describe the emergence of an export control field among international regimes and their members in the nonproliferation issue-area I show how the institutional character the organizational environment of export control regimes was

conducive to the development of a field, and document the development of the field through regime membership growth and cross-regime organizational convergence As a growing number of actors came to interact more routinely and intensively, standard processes and structures diffused through the field based on their legitimacy-conferring properties rather than their efficiency The next section explains the mechanisms by which the export control field developed Finally, I conclude with a discussion of the theoretical implications of this model for our understanding of international regimes, issue-areas, and international cooperation.

Section I: Competing Perspectives on Export Control

a The Empirical Puzzle

The first multilateral export control regime, the Coordinating Committee on Multilateral Export Control (CoCom), was established by Western states in 1949-50 to coordinate controls on trade with the Soviet Union.3 An East-West trade regime rather than a

nonproliferation regime, CoCom remained the sole international export control

organization until the establishment of the Zangger Committee and Nuclear Suppliers Group in the 1970s The 1980s and 1990s, however, saw a burst of activity in multilateralexport control, with the establishment of new regimes to address chemical and biological weapons (the Australia Group), ballistic missile technology (the Missile Technology Control Regime), an EU dual-use technology control system, and a nonproliferation-

3 Both CoCom and COCOM are acceptable spellings I follow Mastanduno 1992, among others, in using the alternate case spelling, which is also used in published Congressional hearings Various dates are given for CoCom’s founding, from 1944 (Buchan 1994) to 1951 (Cleverly 1989) Mastanduno 1992 and

Yasuhara 1984, among others, give the conventional account–that states agreed to establish the regime in late 1949, and that CoCom began to function in January 1950.

oriented replacement for CoCom (the Wassenaar Arrangement) In addition, the Nuclear Suppliers Group (NSG) was revived in 1991, after having been dormant since 1978.

This broader range of cooperation was matched by striking levels of

organizational convergence across both regimes and export control agencies of member states, especially in the 1990s.4 Regimes have converged in both membership and in theirstructures and processes National control systems have increasingly adopted

standardized practices

Figure 1 demonstrates that the membership of the various regimes became

increasingly congruent throughout the 1990s

Figure 1 Regime Membership Growth

0 5 10 15 20 25 30 35 40

1949195319571961196519691973197719811985198919931997

Year

Membership

CoCom Zangger NSG AG MTCR WA EU

AG: Australia Group

CoCom: Coordinating Committee for Multilateral Export Controls

EU: European Union Community Regime for Dual-Use Export Controls

MTCR: Missile Technology Control Regime

NSG: Nuclear Suppliers Group

WA: Wassenaar Arrangement

Zangger: Zangger Committee

Sources: Sipri Yearbook and www.sipri.se (various years)

4 Convergence in organizational form and practice has not been matched by equivalent convergence in the content of national control lists and licensing policies However, the focus of this article is the form of export control agencies and regimes See Deltac/Saferworld 1995 on variation across national control lists.

Of the 41 states that are members of at least one of the five main control regimes, 26 are members of all.5 Of the remaining 15 states, 8 are members of at least 3 regimes.6

Moreover, while data is spotty, evidence suggests that the structures and procedures of member states’ national export control systems have increasingly converged on a

common standard The following data, compiled by the Center for International Trade and Security, demonstrates increased congruence of national export control systems of four former Soviet states with an ideal score representing an international standard. 7

Table 1 Longitudinal Evaluation of Export Control Development in Belarus, Kazakstan,

Russia, and Ukraine (percent of ideal score)

Source: Craft and Grillot 1997, 14.

5 The EU dual-use control system is not included here, as only the 15 EU members are eligible.

6 Sipri, “Table of membership of multilateral military related export control regimes,”

<http://projects.sipri.se/expcon/natexpcon/country_matrix.html>, Accessed April 13, 2001 (Information current as of October 5, 2000); Mutimer notes that regime membership overlaps closely with the 28 states

of the OECD Mutimer 2000, 87.

7 Craft and Grillot 1997, 14 In Craft and Grillot 1997, the data in this table is labeled “preliminary results.” Date from 1999 show regression since 1996 in the scores for Belarus, (73.22), Kazakhstan (59.26), and Russia (76.29), with Ukraine showing further improvement (85.79) All except Belarus, however, still score substantially higher than their 1994 scores Belarus’s 1999 score approximates its 1994 score Scores for Kazakhstan, Russia, and Ukraine can be found in the table at

http://www.uga.edu/cits/ttxc/nat_eval_nec_table.htm I thank Michael Beck for providing me with a more complete table of 1999 results Richard Cupitt has recently argued that export control cooperation reached

a plateau around 1996 Richard T Cupitt, “Multilateral Nonproliferation Export Control Arrangements in 2000: Acheivements, Challenges, and Reforms,” Working Paper No 1, Study Group on Enhancing Multilateral Export Controls for US National Security, April 2001 Available from http://www.stimson.org/ tech/sgemec/paper1.pdf, Accessed May 22, 2001 This raises the question, not addressed in this working paper, of the limits to convergence in organizational fields.

Structural and procedural convergence is exemplified by the diffusion of so-calledcatch-all clauses These clauses, which originated as a rule in the 1991 U.S Enhanced Proliferation Control Initiative (EPCI), were rapidly adopted by the Missile Technology Control Regime (MTCR), EU system, and in national regulations of states such as

Russia.8 Regimes such as the MTCR and NSG, which had been limited to controlling technologies related to the production and delivery of nuclear weapons, extended their controls to cover non-nuclear payloads (in the MTCR’s case) and dual-use technologies (in the NSG’s), thereby conforming more closely to the other regimes In the 1990s, the NSG, AG, MTCR, and Wassenaar Arrangement all adopted variants of “no undercut” rules preventing members from allowing export licenses refused by another member

Mainstream theories of international relations—realism and neoliberalism—explain development in export control cooperation as the product of rational, interest-driven behavior.9 The creation of new regimes and membership growth of existing regimes is explained as a response to a perceived increase in proliferation threats, or to provision of side-payments And the standardization of organizational form and

procedures is explained as driven by optimization: agencies and regimes converge on the most efficient methods of export control Empirical evidence discussed below provides only mixed support for these claims Though more incomplete than wrong, these

explanations are ultimately unsatisfactory

First, converging export control practices do not appear to be driven by a commonperception of an increased proliferation threat Common threat perception should produce

8 Anthony and Zanders 1998, 396; Cornish 1995, 41; Anthony and Zanders 1998, 398.

9 I use the terms neoliberalism or rational institutionalism to refer to the rational-choice approach to the study of international institutions championed by Robert O Keohane (1984, 1993), who now calls the approach “institutionalist.” In this paper, “institutionalist” refers to sociological institutionalism.

agreement on items to be controlled and targets to be proscribed National control lists, however, vary significantly, and states’ decisions to join regimes often seem driven by considerations other than security threats.10 The efficiency-driven convergence thesis is further weakened by the near absence of evidence that standard practices are, in fact, more efficient than alternatives (More on this below.) Finally, the claim is

counterintuitive It’s not clear why the same methods would be most efficient for the different tasks undertaken by the different regimes It is a very different thing to control the export of a chemical with commercial application than to limit transfer of complete weapons systems or production facilities One would expect different methods to be appropriate for each task Thus, rationalist explanations are, at best, incomplete

An alternative explanation, based on constructivist theory, holds that states develop national export control systems meeting international standards, and join

multilateral regimes, due to their identification with the liberal international community Empirical research has found mixed support for this “liberal identity” explanation, and itsexplanatory power seems to vary over time and across countries.11 However, national identity does not, in any case, explain the content of the standard practices adopted by states Even if the adoption of certain export control practices is norm-driven, how do those particular practices acquire this normative sanction?

Organizational field theory was developed to explain precisely the sort of puzzle presented here: the diffusion across organizations of practices and structures only loosely

10 On national list variation, see Deltac/Saferworld 1995 On non-security reasons for joining regimes, see the studies in Bertsch and Grillot 1998, which find that “military security concerns play little role in state export control decision-making in the FSU.” Bertsch and Grillot 1998, 219 These findings are supported

by responses in interviews I conducted with current and former U.S government officials in January of 1999.

11 Bertsch and Grillot 1998 See also Cupitt and Grillott 1997; Chafetz 1995; Mutimer 2000

coupled to efficiency criteria Field theory, a branch of sociological institutionalism, explains the process by which these organizational features acquire legitimacy and become widely adopted Field theory is particularly well suited to the study of the export control developments discussed above for two reasons First, it is a research program focused on just the kinds of interorganizational processes encountered in cross-regime interactions such as those among export control organizations Also, field theory focuses

on cognitive and intersubjective aspects of organizational environments And it is in intersubjectively constituted aspects of these environments—issue areas and

organizational fields—that conventional regime theory exhibits gaps

b Gaps in the Regimes Literature

Regime theory is seriously underdeveloped in its understanding of the effects of regime linkages and the origins of issue-areas As one recent review asserts,

cross-For the most part, analyses of international regimes have focused on individual arrangements on the assumption that they are self-contained or stand-alone

institutions to be studied in isolation from one another…Yet it is apparent that

institutional linkages are widespread—and becoming more so—in international society.12

Regimes do not exist in isolation In an increasingly institutionalized international

society, regimes routinely interact with other regimes, and are affected by each other’s activities Such cross-regime interactions have received little scholarly attention, and theirconsequences are poorly understood Recognizing this, Oran Young and others have recently called for greater attention to such institutional linkages across regimes.13

12 Levy, et al 1995, 317-8.

13 Young 1996.

Regime theory’s lack of an account of issue-area development is particularly serious given the fact that the concept of issue-area is a central element in the

conventional definition of regimes: “principles, norms, rules, and decision-making procedures around which actor expectations converge in a given issue-area.”14 While the existence of an issue-area is a precondition to regime creation, we know little about the processes by which issue-areas emerge, or what implications this may hold for regime formation Noting this gap, a leading scholar of international relations has observed that

“[T]he way in which issues become aggregated into an issue-area has to become an explicit part of regime analysis.”15

Thinking of cross-regime interactions in terms of organizational fields links the two problems, showing how cross-regime linkages and issue-area evolution are related phenomena.16 Through the development of an organizational field the one can, in fact, produce the other

c Rationalist Explanations of Export Control Cooperation

14 Krasner 1983, 1, 2.

15 Kratochwil 1993, 83, paraphrasing Haas 1975, 1980.

16 An organizational field roughly corresponds to Young’s concept of a nested institution Young 1996.

Rationalist theories provide less complete explanations of the extension and convergence

of export control regime activity.17 Such explanations—both realist and rational

institutionalist—account for increased cooperation primarily as a response to perceived increased common proliferation threats, and explain convergence as optimization to the most efficient means of control Several recent export control efforts have, indeed, been responses to the development or use of WMD in the Iran-Iraq War of the 1980s, and the

1991 Persian Gulf War And export control agencies do try to determine more effective means of controls

However, this account is difficult to reconcile with several other observations If increased export control cooperation in the 1980s and 1990s were a result of a common perception of increased threat, one would expect greater uniformity of national control lists Yet significant variation persists in items controlled and targets proscribed

Differences in the 1990s between the U.S and its allies over using sanctions against

“rogue” states have reflected disagreement not only over the efficacy of sanctions but also over the degree of threat posed by potential proliferators It is in the form, more than the content, of national and multilateral control systems that convergence has occurred

Much recent export control cooperation seems driven by factors other than

security threats.18 The two main alternatives to security threats proposed as motives for adhering to international export control standards are side-payments and identity States may develop national control systems and seek to join regimes in order to obtain rewards both tangible—access to markets and technology—and intangible The latter motivation

17 See Lipson 1999 for a more extensive discussion of hypotheses derived from institutionalist and

alternative explanations.

18 Beck, others in Bertsch and Grillot 1998

was described in several interviews with U.S officials as the desire (particularly on the part of former Soviet and Warsaw Pact states) to “join the club” and be accepted as legitimate members of the Western international community.19 These explanations for increased cooperation better account for recent trends States are certainly motivated by the economic benefits of greater access to Western technologies and markets that can come from being a participant in, rather than target of, export control cooperation This side-payment explanation, derived from rational institutionalism, is certainly part of the explanation U.S officials involved with efforts to improve control systems in the former Soviet Union and Central and Eastern Europe, however, generally judged norms and identity—the desire to “join the club”—to be equally or more important than economic motivations in the decisions of these states to develop international standard control systems While these factors, highlighted by neoliberal and constructivist theories, largely explain the increased export control cooperation of the 1990s, it is unclear that they would have translated into such extensive growth in regime membership and

participation without the prior development of the export control field States have been brought into the regimes largely through field-constructing outreach activities such as seminars and exchanges conducted by regimes and member states In any case, side-payments and identity factors do not explain the convergence of national and multilateral export control practices

Rationalist accounts see this convergence as the result of national agencies and regimes recognizing and adopting the most efficient means of implementing controls This explanation, while initially plausible and intuitively satisfying, is surprisingly weak

19 Author’s interviews with U.S government officials, Washington, D.C., January 4-14, 1999.

An efficiency-driven convergence explanation leads us to expect to find evidence that thepractices adopted are, in fact, more efficient than alternatives Not only is such evidence lacking, but the difficulty of determining the relative efficiency of alternative practices is sufficient to insulate export control organizations from such technical criteria In such an institutional environment, organizational convergence is better explained as in large part

a result of the copying of legitimized models under conditions of uncertainty This mechanism, derived from organizational field theory, is generally consistent with

rationalist expectations Thus, institutionalist and rationalist explanations in this case are not mutually exclusive but complementary However, the rationalist explanation alone is incomplete, failing to explain how certain practices rather than others become legitimizedand diffused Institutionalism, by placing the question in the context of institutional isomorphism within an organizational field, offers a more complete account of recent export control convergence, and identifies mechanisms to explain it

d Organizational Fields

The standard definition of an organizational field is “Those organizations that, in the aggregate, constitute a recognized area of institutional life: key suppliers, resource and product consumers, regulatory agencies, and other organizations that produce similar services or products.”20 This definition reflects the concept’s origins in the study of industrial sectors However, the field concept is broader than that More generally, “The notion of field connotes the existence of a community of organizations that partakes of a common meaning system and whose participants interact more frequently and fatefully

20 DiMaggio and Powell 1991, 65.

with one another than with actors outside of the field.”21 The term is meant to “signify both common purpose and an arena of strategy and conflict."22 The concept is flexible as

"A field is always an analytical construct and how one defines it depends upon the phenomena in which one is interested."23 Organizational field theory developed within the institutionalist branch of sociological organization theory, which focuses on the socialand intersubjective aspects of organizational environments.24 As a prominent

institutionalist explains,

Institutional theory emphasizes that organizations are open systems—strongly

influenced by their environments—but that it is not only rational or efficiency-based forces that are at work Socially constructed belief systems and normative rules

exercise enormous control over organizations—both how they are structured and how they carry out their work.25

Organizational field theory, then, centers on explaining convergence of

organizational form and practice as the consequence of institutional, legitimacy-driven isomorphism within an organizational field Applied to international cooperation, this explanation suggests improvements to theories of international regimes by specifying ways in which institutional linkages can produce changes in regimes, and how this process can redefine issue-areas

Specifically, interactions between organizational elements of different regimes within an organizational field cause isomorphism across these regimes Formal

21 Scott 1995, 56 Thus, Slaughter’s (1997) concept of transgovernmentalism describes organizational fields comprised of counterpart agencies of different states.

organizations within the different regimes will adopt common structures and practices, which, though collectively legitimized, will likely not be optimally effective Further, as

an organizational field develops through interaction between organizations within

separate regimes, these organizations will come to see themselves as involved in a

common project This perception of a common project can be institutionalized as an issue-area broader than those defining the specific regimes Thus, institutional linkages within organizational fields not only produce isomorphism across regimes within the field, but can also institutionalize a new, broader issue-area This processes makes

possible the development of a single regime corresponding to the expanded issue-area In the section that follows, I describe the organizational field of nonproliferation export control

Section II: The Emergence of Institutionalized Export Control as a Field

a The Institutional Environment of Export Control

Organizational fields develop within what are known as institutional environments The distinction between institutional and technical environments is a key concept of

institutionalist analysis.26 In technical environments, organizations are selected for their effectiveness, according to technical efficiency criteria such as profit-loss margin or market share In institutional environments, organizations are relatively insulated from such efficiency criteria, and are instead selected for legitimacy criteria As Steven Weber has put it, institutional environments “are characterized by relationships built on

noneconomic and nontechnical standards of legitimacy that have less to do with

26 Neo-institutionalist organizational analysts, exemplified by DiMaggio and Powell 1991, typically refer to their research as “institutionalist.” I use the term to refer to this branch of organization theory, not

neoliberal institutionalism, or any other claimant to the label of institutionalism.

efficiency than with the structure of extant organizations in the environment.”27

According to W Richard Scott, “In institutional environments, organizations are

rewarded for establishing correct structures and processes, not for the quantity and quality of their outputs.”28 Isomorphism in these environments is due primarily to

adoption of those features accorded collective legitimacy.29 This tends to be particularly the case among nonprofit organizations such as hospitals, public schools, and governmentagencies, upon which institutionalist research has focused.30 Institutionalist researchers have come to see the technical-institutional environment as more of a continuum than a dichotomy, moving away from efficiency versus legitimacy debates to analysis of the effects of both rational-technical and institutional aspects of organizations’ environments.Thus, more recent research has argued that “institutional rules provide the context and frame within which effectiveness criteria are constructed.”31 However, the technical-institutional environment distinction has proven to be a fruitful one for institutionalist researchers

Export control organizations face an institutional environment Like schools and nonprofit organizations, export control regimes are insulated from clear criteria of

successful and unsuccessful outcomes The lack of publicly available evaluations of the effectiveness of particular export control methodologies—such as import

certificate/delivery verification (IC/DV) systems or “catch-all” clauses—is striking

27

Weber 1994, 5.

28 Scott 1991, 167.

29 DiMaagio and Powell 1991.

30 Meyer and Hannan 1979; Meyer, Ramirez, and Soysal 1992.

31 Scott 1998, 117.

While this may just be an indication that such reports are secret, there is reason to believeotherwise Though current and former policy-makers I interviewed indicated that the U.S.intelligence community actively monitors the extent of weapons-related exports to states and regions of concern, these analyses apparently focus more on the technology

acquisition of particular states rather than the effectiveness of particular control

systems.32 I am aware of only one published effort to assess the effectiveness of a

particular export control method While the study is seriously flawed, its authors deserve credit for trying to measure something so elusive that no one else has apparently even tried.33 However, this study is the exception that proves the rule Despite the impossibility

of proving a negative, there is little or no evidence of the relative effectiveness of

alternative procedures for export control This absence is a hallmark of institutional environments

32 Author’s interviews with current and former U.S government officials, Washington, D.C., January 4-14,

1999 Interview subjects were, without exception, extremely careful to avoid revealing sensitive

information (States other than the U.S have more limited intelligence capabilities, and were said to rely, to varying degrees, on information provided by U.S intelligence.)

33 Craft and Grillot 1999 To assess the effectiveness of transparency provisions in national export control systems (required of CoCom and Wassenaar members), this study treats the absolute number of regime- proscribed exports from the UK, France, Germany, and the United States reported in a congressional report

as a proxy for the relative effectiveness of each national export control system From the fact that the report found more companies conducting illegal business from countries with more transparent systems (the U.S and Germany), Craft and Grillot conclude that Wassenaar’s reliance on transparency measures is likely to

be ineffective in controlling proliferation However, this proxy measure fails to control for the size of each national economy, or more relevant here, each country’s technology export sector Thus, the 69 American violations could well be evidence of greater, not lesser, export control effectiveness than France’s 27 violations Under the measure used, a country with 10 violations would be judged less effective at export control than one with 5, even if the first country’s technology export sector was 100 times larger than that

of the second A more useful measure would be violations as a percentage of total exports covered by licensing regulations, but such a measure would be exceedingly difficult to construct.

Interestingly, if Craft and Grillot are correct, their findings would undermine the claim that export control is an institutional environment—as such measurement tends to be unfeasible in institutional environments, accounting for organizations’ insulation from efficiency measures However, their results would support the claim that the diffusion of features such as transparency provisions across the export control field is driven by factors other than their effectiveness at controlling exports, since they conclude that transparency is ineffective As transparency measures do carry legitimacy, institutional isomorphism would explain their diffusion better than optimization On the development of transparency norms, see Florini 1996.

While measuring organizational effectiveness tends to be exceedingly difficult in institutional environments, this does not mean that organizations in these environments are not subject to any evaluative criteria In institutional environments and organizational fields, field “auditors” set and monitor compliance with widely accepted standards that emphasize correct process.34 As James Ron has noted, “Over time, organizations learn what types of responses to audits are legitimate and what types of external structure they should adopt to maintain the proper aura of respectability The emphasis on proper procedure, or organizational structure, is especially pronounced in the case of

bureaucracies, which produce no clear quantifiable product.”35

Accordingly, publicly stated criteria for the evaluation of export control systems

do exist but are almost exclusively focused on process rather than outcome This is not particularly surprising, given the inherent difficulty of judging export control

effectiveness The exercise is inescapably one of counterfactual analysis, requiring answers to questions such as “Would this control not now in effect have prevented this export?” and “Would this export have occurred in the absence of an existing regulation?” Such questions cannot be conclusively answered As one former policy-maker said,Probably maybe the very first question I was asked when the new [Clinton]

administration came to town—after they took office—was how can we judge the effectiveness of our export controls And you know what? Hard to measure Very hard

to measure because you don’t know what didn’t go, you don’t know what slipped through that you haven’t picked up in your enforcement mechanism, you don’t know what items you denied that a foreign company supplied instead, so really it’s a very simple and a profound question, and it’s very hard, and it’s a very important question but it’s very hard to answer And I remember I must have been sent back to review that one-pager five, six, seven times and it is just very hard to answer So I think the short answer is there is no systematically reliable way to measure.36

34 Ron 1997.

35 Ron 1997, 277 Ron’s reference is to organizational fields, rather than institutional environments.

36 Author’s interview with former U.S government official, Washington, D.C., January 14, 1999.

Given the difficulty of evaluating the effectiveness of export controls, existing evaluative criteria deal with procedural and administrative features of control systems: process not outcome.37 Most such criteria derive from an export control “common

standard” developed in CoCom in the early 1980s This standard required:

1 Acceptable import certificate and delivery verification (IC/DV) documents

2 Control over reexports of CoCom-origin, controlled goods

3 Control over indigenous exports of CoCom controlled goods

4 Cooperation in prelicensing and postshipment checks

5 Enforcement cooperation (includes cooperation in policing transshipments and free-trade zones.)38

In the early 1990s, the U.S Department of Commerce developed a system for export control evaluation derived from CoCom’s common standard effort.39 The Export Control System Appraisal (ECSA) program considers countries’ export control law, licensing, enforcement, administration, information processing, and technical training.40

The exclusive focus on procedural and legal-administrative matters, to the exclusion of outcomes, is striking

37 According to Frank Dobbin, “Scott (1992:16-17) has more clearly specified the distinction [between

institutional and technical environments], arguing that most organizations depend simultaneously on output controls linked to technical considerations and process controls linked to institutional considerations….”

Dobbin 1994, 126.

38 National Academy of Sciences 1991, 66.

39 Author’s interview with U.S government official, Washington, D.C., January 8, 1999.

40 U.S Department of Commerce, “Export Control System Appraisal.” Mimeo, n.d This document was the Department of Commerce contribution to a National Research Council handbook prepared for foreign export control officials My thanks to Suzette Grillot for providing me with a copy (Elsewhere the ECSA acronym is used to refer to Export Control System Assessment, rather than appraisal U.S Department of Commerce 1996, II-94.)

These evaluative systems do not just gather dust sitting on shelves Significant resources are expended to promote adherence to these standards The Bureau of Export Administration (BXA) of the U.S Department of Commerce administers a

Nonproliferation and Export Control (NEC) Program to assist other countries in

developing and upgrading their national export control systems to meet internationally promulgated standards similar to those outlined in the ECSA document described

above.41 Through this program, BXA works “with representatives of approximately 23 or

24 countries to help them strengthen or establish an international standard export control system.”42 Material support for this and related programs has been substantial Under the Nunn-Lugar Cooperative Threat Reduction program, approximately 39 million dollars were provided for NEC cooperation Since 1994, funding for these activities has varied between 10-20 million dollars annually through the State Department’s Nonproliferation and Disarmament Fund (NDF).43 Other countries conduct similar programs.44

Another widely known system for evaluating export controls has academic origins Experts at the Center for International Trade and Security (CITS) at the

University of Georgia have developed a system for evaluating the effectiveness of

national export control systems.45 The CITS national export control evaluation system,

41 The NEC program’s evaluation categories are: legal foundation and regulatory development, licensing procedures and practices, preventive enforcement mechanisms, industry-government relations, and

program administration and system automation U.S Department of Commerce, “Nonproliferation and Export Control (NEC) Cooperation Program,” Brochure.

42 Author’s interview with U.S government official, Washington, D.C., January 8, 1999.

43

Author’s interview with U.S government official, Washington, D.C., January 7, 1999; U.S State

Department, Nonproliferation and Disarmament Fund, http://www.ndf.org/html/fundcontrol.html.

44 Mutimer 2000, 88.

45 Grillot and Craft 1996; Craft and Grillot, et al 1997; Grillot 1998 While CITS is organizationally part

of a state university, its personnel view it as an NGO

based on the CoCom common standard, consists of a 72-item questionnaire addressing ten elements of an export control system: licensing system, regime adherence, training, custom authority, penalties, control lists, catch-all clauses, bureaucratic process,

import/export verification, and information gathering/sharing.46 Each of these is

considered in terms of the development of policy and legal underpinnings,

institutionalization, and implementation.47 Export control officials responding to the questionnaire evaluate their systems, and responses to questions concerning different elements are assigned weights to reflect their relative importance The result is a

numerical index used to rank the relative congruence of different national systems with international standards.48 While the CITS system yields a quantitative score, it is a whollyqualitative evaluation, thoroughly process-oriented like the methods on which it is based And CITS is explicit that GES scores measure export control system development, not effectiveness As one explanation of the system states,

A perfect score simply suggests that a country’s export control system consists of Western style export control policies, institutions, and procedures that are in place and in practice The score does not reflect, in other words, an individual level of

export control effectiveness—only an individual level of development.49

Thus, export control organizations are evaluated for their conformity with widely legitimized standards rather than for their effectiveness, and therefore exist in an

organizational environment significantly more institutional than technical

46 This system, and the project devoted to it, has gone by different names Previously referred to as the Global Evaluation System (GES), it has recently been renamed the National Export Control Evaluation (NECE) project Craft and Grillot, et al 1997, 9: http://www.uga.edu/cits/ttxc/nat_eval.htm.

47 Craft and Grillot, et al 1997.

48 The resulting scores are available on the CITS web page (http://www.uga.edu/~cits), and in publications

by CITS-affiliated authors, such as Craft and Grillot et al 1997 and Bertsch and Grillot 1998

49 Grillot 1998, 29, note 57

b The Export Control Field

Organizational fields develop within institutional environments This development occursthrough a process of institutional definition, or structuration, the extent of which is conventionally assessed in terms of four indicators: increased organizational interaction, development of intra-field structure, increased information load, and sense of a common enterprise.50 In this section, I first describe the record of multilateral export control cooperation since 1949 I then demonstrate the development of an export control field along each of the above four dimensions, and illustrate mechanisms by which

isomorphism is produced within the field

Efforts to control transfers of arms and related technologies date to at least the sixteenth century.51 Modern export control is of more recent provenance, and some of its practices seem to have originated in the British “navicert” system of World War II.52

However, multilateral export controls originated with the formation of the Coordinating Committee on Multilateral Export Control (CoCom) in 1949 CoCom was established by the U.S and its allies to coordinate restrictions on the export of strategic goods to the

50

DiMaggio and Powell 1991, 65 This is not necessarily an exhaustive list of dimensions of field structure According to W Richard Scott, “To these indicators others can be added, including extent of agreement on the institutional logics guiding activities within the field, increased isomorphism of structural forms within populations in the field, increased structural equivalence of organizational sets within the field, and increased clarity of field boundaries.” Scott 1995, 106 According to Scott, DiMaggio and Powell take the concept of structuration from Giddens, but “define it more narrowly as referring to the degree of interaction and the nature of the interorganizational structure that arises at the level of the organizational field.” Scott

1995, 106.

51 On sixteenth century English efforts to prevent the Spanish from obtaining cast-iron guns see Roodbeen

1992, 3 Roodbeen cites Cipolla 1985, 35.

52 Navicert is a contraction of “naval certificates.” Matson 1994, 3 See Mendlicott 1952, 94-101; Ritchie 1938; Matson 1994; Gordon and Dangerfield 1947, 36-39.

Soviet bloc.53 CoCom established a template for later export control regimes Like later control regimes, CoCom was an informal, non-treaty based, multilateral organization in which national representatives agreed on coordinated lists of items and targets subject to varying levels of export controls to be implemented by national authorities CoCom differed from later regimes in being focused on containment rather than proliferation, and

in its veto rule under which any member could veto another’s export of a controlled item.This was possible due to the shared sense of threat among CoCom’s relatively small membership during the Cold War Later regimes, lacking the veto rule, nonetheless adopted many features developed in CoCom, including methods of control list

construction and procedures for licensing and consultation

CoCom was created in an environment more technical than institutional

Multilateral export control was a new phenomenon, and no collectively legitimated set of practices existed to guide it Nor was there a community of organizations to confer legitimacy on such practices Furthermore, CoCom was less insulated than later export control regimes from measures of effectiveness, as success or failure could be relatively easily judged by the capacity of the Soviets to obtain controlled items This was easier tomonitor than the diffusion of more widely available technologies to multiple destinations around the world While CoCom predated the export control field, it was critical to the field’s establishment, as it became a model for later regimes

1 Nuclear proliferation

53 On CoCom’s origins, see: Mastanduno 1992; Roodbeen 1992; Osgood 1957; Sørensen 1991; Yasuhara

1984

The first steps in the development of a technology control field beyond CoCom were the establishment of the IAEA in 1956 and the 1968 nuclear nonproliferation treaty (NPT), which together comprised an international nuclear nonproliferation regime.54 The

inability of this regime to deal with proliferation by non-members led to the first modern multilateral export control arrangements to deal with proliferation rather than East-West trade—the Zangger Committee and the Nuclear Suppliers Group

In 1971, NPT members formed the Nuclear Exports Committee, which came to beknown after its chair, Claude Zangger In 1974 the committee agreed to rules governing nuclear technology exports Zangger Committee decisions are adopted at the discretion ofmember states, and have “no status in international law.”55 The Zangger Committee established a so-called “trigger list” of items the export of which triggers IAEA

safeguards and requires assurances that they will not be used in explosive devices or exported to recipients not accepting these restrictions.56 Because France was a major nuclear supplier that had not (until 1992) signed the NPT, the Zangger Committee had a serious gap Largely to address this problem, the Nuclear Suppliers Group was formed in

re-1975 Since then, the NSG and Zangger Committee have functioned alongside each otherwith overlapping memberships In 1997, China joined the Zangger Committee but not theNSG The Zangger Committee’s primary responsibility has been updating the trigger list

Because the NSG originally held its meetings in London, it came to be known as the London Suppliers Group or London club Its original members were France, Canada, West Germany, Japan, Britain, the U.S and the Soviet Union Within two years, the

54 Smith 1987.

55 International Atomic Energy Agency 1990.

56 Van Ham 1994, 14; Anthony and Zanders 1998, 380.

regime more than doubled its size as eight more countries joined.57 The NSG contrasted with CoCom, which targeted the Soviet Union, Poland, Czechoslovakia, and East

Germany, while these states were members of the nuclear regime Nonetheless, the NSG drew on export control methods developed in CoCom

Developing countries to which nuclear exports were proscribed were sharply critical of what they saw as the NSG’s “technological apartheid,” which asserted was a violation of the NPT’s requirement that its signatories not be denied access to nuclear technology for peaceful uses.58 While formally unaffiliated with the NPT, the NSG acquired greater legitimacy in 1978 when the IAEA published its guidelines as INFCIRC/

254 However, the protests of developing countries were a major factor in the NSG’s failure to meet between 1978 and 1991.59 The Zangger Committee continued to update itstrigger list during this period, and NSG members did meet informally Some non-

members agreed to accept NSG guidelines in the 1980s.60 Responding to alarm over Iraq’s nuclear program and concerns over the security of nuclear materials in the former Soviet Union, the NSG began to meet on a regular basis again in 1992.61 At the 1992 plenary in Warsaw, NSG members developed a second set of guidelines covering dual-

57 These were Belgium, East Germany, Italy, the Netherlands, Poland, Sweden, and Switzerland Van Ham

1994, 15.

58 Van Ham 1994, 16 Article IV of the NPT calls for “the fullest possible exchange of equipment,

materials, and scientific and technological information for the peaceful uses of nuclear energy.”

59 According to van Ham, France and West Germany in particular wished to avoid further alienating developing states Van Ham 1994, 16 According to Richard Cupitt, for commercial reasons member states also wanted to avoid negotiating further controls on dual-use items Cupitt 2000, 123.

60 Cupitt 2000, 123.

61 See International Atomic Energy Agency 1997; Van Ham 1994 The NSG has since held plenaries in Warsaw, Lucerne, Madrid, Helsinki, Buenos Aires, Ottawa, and Edinborough The “London Club” label is therefore, no longer appropriate.

use items The regime has grown rapidly in recent years, reaching a membership of 39 with the recent additions of Latvia, Belarus, Cyprus, Turkey, and Slovenia.62

2 Chemical, biological, and missile proliferation

The next major step in multilateral export control was the establishment of control

regimes to address chemical and biological weapons, and ballistic missile technology The creation of the Australia Group and the Missile Technology Control Regime

broadened the boundaries of the export control field and extended the concept of

proliferation beyond its original restriction to nuclear technology

In the 1980-1988 Iran-Iraq War, first Iraq and then Iran used chemical weapons incombat In response, Australia took the initiative in the creation of a regime to control

“precursor chemicals” useable in chemical weapons production Founded by 15 states in 1984-1985, the Australia Group has expanded to include 32 states and extended its list of controlled chemicals from 8 to 54 The AG later addressed biological weapons

proliferation as well An informal regime, the AG “has no charter and, apart from the support provided by the Australian Department of Foreign Affairs and Trade as a point ofcontact and the Australian embassy in Paris as a meeting place, it has no institutional foundations.”63 The AG has come under criticism from developing states similar to that directed at the NSG This criticism intensified after the coming into force of the Chemical

62 The NSG still displays concern in its official documents with its perception as a discriminatory cartel One such document expresses a desire to “address concerns that the NSG acts as a culb of developed countries to restrict access top advanced technology This is emphatically not the objective of the NSG and would be contrary to NPT obligations as well as a range of undertakings on peaceful nuclear cooperation made most recently at the 1995 NPTREC [NPT Review and Extension Conference].” Nuclear Suppliers Group, n.d

63 Anthony and Zanders 1998, 387.

Weapons Convention, as “many developing countries view the maintenance of an export control regime outside the CWC as undermining the commitment made in Article XI” to avoid interference with economic and technological development.64 However, calls for the AG’s abolition have been rejected by its members

The early 1980s saw concern over both chemical warfare and the diffusion of missile technology as rocket and missile technology became increasingly available to developing countries Missile technology fell under the rubric of WMD proliferation as several states with missile programs also had nuclear programs.65 Further, “the technical characteristics of ballistic missiles, including the speed with which they can reach targets,their invulnerability to defenses, their adaptability for delivering weapons of mass

destruction, and their particular utility for preemptive military operations, make them inherently destabilizing in regions where combat is likely to occur.”66 In 1982, the U.S initiated discussions with allies to develop common measures to control missile

proliferation By 1985, a few states were coordinating their controls, and in 1987 the G-7 states established the MTCR At first, the MTCR focused only on missiles capable of carrying nuclear payloads Members agreed to control exports of missiles with ranges of

300 km or more, and payloads of 500 kg or more These criteria were believed to cover missiles capable of delivering a nuclear payload to targets in regions of likely conflict.67

MTCR controls cover complete missiles and major subsytems (Category I items, subject

to a presumption of denial) and dual-use items (Category II items, evaluated

case-by-64 Anthony and Zanders 1998, 391.

65 Nolan 1991, 115.

66 Nolan 1991, 8.

67 Ozga 1994

case) In 1993, members added extended MTCR controls to cover missiles capable of delivering chemical or biological warheads, including missiles with ranges and payloads below the original limits At the same time, the MTCR also added a “catch-all” clause.68

The MTCR is even more informal than the other export control regimes It “is not based upon a treaty or an official agreement; it lacks an international executive body that monitors the exports of missile-related technologies on a permanent basis It has no verification provisions and no mechanisms to enforce its guidelines.”69 However, while seen “just a time-winning device,”70 the MTCR has been credited with significant

successes These include bringing about the cancellation of the joint

Argentinian-Egyptian-Iraqi Condor II missile project.71 The regime holds an annual plenary meeting and conducts monthly meetings in Paris between plenaries In lieu of a secretariat, the French Ministry of Foreign Affairs serves as a point of contact MTCR members also hold ad-hoc technical experts meetings and “special” meetings for “recruitment

purposes.”72 Decisions are made by consensus

From the original G-7 states, membership has grown steadily and currently stands

at 32 In addition, several non-members adhere to the guidelines Most major suppliers have either joined the MCTR or adhere to its guidelines.73 China, though not a member,

68 Anthony and Zanders 1998, 396.

69 Van Ham 1994, 25.

70 Sha Zukang 1999, 11

71 The MTCR is also credited with delays in India’s missile program, inhibiting sales of Chinese missiles to Pakistan, promoting stronger controls by Germany and Israel, and the cancellation of Brazilian missile programs Ozga 1994 However, Iran, North Korea, and India have all tested missiles with impressive range in recent years, prompting American efforts to develop a National Missile Defense.

72 Ozga 1994.

73 Mistry 1997, 62.