¹ See International Organization for Standardization (ISO), Friendship among Equals: Recollections from ISO's First Fifty Years (Geneva: ISO Central Secretariat, 1997)Google Scholar; and Loya, Thomas and Boli, John, “Standardization in the World Polity,” in Boli, John and Thomas, George, eds., Constructing World Culture: International Nongovernmental Organizations since 1875 (Stanford, Calif.: Stanford University Press, 1999)Google Scholar.

² Compatibility standards, for example, have risen in prominence with the explosive growth of information and telecommunications technologies. People want to participate in networks that allow them to share databases, have access to large selections of compatible software, exchange documents, combine products made by different vendors, or simply communicate directly. See Besen, Stanley and Farrell, Joseph, “Choosing How to Compete: Strategies and Tactics in Standardization,” Journal of Economic Perspectives 8 (Spring 1994)CrossRefGoogle Scholar.

³ Organization for Economic Cooperation and Development, Regulatory Reform and International Standardization (Paris: OECD Working Party of the Trade Committee, 1999), 4Google Scholar.

⁴ GATT TBT Agreement, article 2.4. The TBT Agreement is reprinted inter alia in Sykes, Alan, Product Standards for Internationally Integrated Goods Markets (Washington, D.C.: Brookings Institution, 1995), 169–214Google Scholar.

⁵ More specifically, product standards cover properties such as interoperability, interconnectability, levels of safety, conformity, materials, systems of classification, methods of testing, the operation of systems, and quality assurance. For an excellent overview of different types of international standards, see Abbott, Kenneth and Snidal, Duncan, “International ‘Standards’ and International Governance,” Journal of European Public Policy 8, no. 3 (2001)CrossRefGoogle Scholar.

⁶ For the sake of brevity we henceforth refer to international product standardization simply as international standardization.

⁷ Martin, and Simmons, , “Theories and Empirical Studies of International Institutions,” International Organization 52 (Autumn 1998)CrossRefGoogle Scholar.

⁸ Keohane, , After Hegemony (Princeton: Princeton University Press, 1984), 93Google Scholar.

⁹ Mattli, Walter, “Public and Private Governance in Setting International Standards,” in Kahler, Miles and Lake, David, eds., Governance in a Global Economy: Political Authority in Transition (Princeton: Princeton University Press, 2003)Google Scholar; see also idem, “Private Justice in a Global Economy: From Litigation to Arbitration,” International Organization 55 (August 2001)Google Scholar.

¹⁰ For a good review, see David, Paul and Greenstein, Shane, “The Economics of Compatibility Standards,” Economics of Innovation and New Technology 1, no. 1 (1990)Google Scholar.

¹¹ ISO is the product of a merger between the International Federation of National Standardizing Associations (ISA), established in 1926, and the United Nations Standards Coordinating Committee (UNSCC) of 1944. The UNSCC had been set up by the United States, the United Kingdom, and Canada to bring the benefits of standardization to bear on the war effort and the work of reconstruction. Many of the statutes and rules of procedure of ISO were adopted from ISA. See ISO (fn. 1), 15.

¹² For historical information about IEC, see www.iec.ch/about/history/hentry-e.htm, accessed March 1, 2003.

¹³ See www.iso.ch/iso/en/aboutiso/isoinfigures/January2003-p2.html and www.iec.ch/news__centre/iec_figures/figures-e.htm, accessed March 1, 2003.

¹⁴ For further details on these organization, see www.iso.ch and www.iec.ch.

¹⁵ See, e.g., Brunsson, Nils and Jacobsson, Bengt, eds., A World of Standards (New York: Oxford University Press, 2000)Google Scholar.

¹⁶ Article 3.1.1 of the Statutes of ISO.

¹⁷ In addition, both organizations offer other forms of membership for very small or developing countries.

¹⁸ This is an official estimate. Some interviewees contend that the number is as high as one hundred thousand; authors' interviews with anonymous sources, Washington, D.C., August 31,2000, and Paris, October 5,2000.

¹⁹ The ISO central secretariat, for example, has a full-time staff of 163 from 25 countries.

²⁰ See www.iso.ch/iso/en/aboutiso/isoinfigures/archives/January2003.pdf, accessed March 1,2003.

²¹ The IEC process is practically identical.

²² Individual national standards bodies within the ISO membership take on the administrative responsibilities for different technical committees through holding “secretariats” for technical committees.

²³ If these conditions are not met, the draft standard is sent back to the relevant committee for reconsideration in the light of the technical reasons submitted in support of negative votes.

²⁴ We here use 2 × 2 games with ordinal payoffs, where 1 indicates the least, 4 indicates the most desirable of the possible outcomes. These models are surely too simple as a description of the coordination problems of actual international standardization, which involves multiple actors and often multiple possible strategies, but they are very useful for heuristic purposes. As Duncan Snidal has pointed out, increasing the number of actors in coordination games (even with divergent interests as in Figure 3) does not impede cooperation “to nearly the same extent” as in a Prisoner's Dilemma game, and it may even “facilitate cooperation in some cases”; Snidal, , “Coordination versus Prisoners' Dilemma: Implications for International Cooperation and Regimes,” American Political Science Review 79 (December 1985), 936CrossRefGoogle Scholar; see also Stein, Arthur, “Coordination and Collaboration: Regimes in an Anarchic World,” in Krasner, Stephen, ed., International Regimes (Ithaca, N.Y.: Cornell University Press, 1983)Google Scholar.

²⁵ The game in Figure 2a has two Pareto-optimal Nash equilibria, and the game in Figure 2b has two Nash equilibria, one of which is Pareto optimal and one of which is not.

²⁶ Note that even in the version depicted in Figure 2b, simple coordination is not a “harmony” game.

²⁷ Scientists tend to consider metric units clearly technically superior for computational reasons. Coordination, however, may be even more important since the lack of coordination raises the likelihood of mistakes, such as when Lockheed Martin Astronautics, using English units internally, overlooked converting one set of figures into metric units while working as a subcontractor for NASA's Jet Propulsion Laboratory, which uses metric units throughout. The loss of NASA's $125 million Mars Climate Orbiter in September 1999 was widely attributed to this mistake; “Engineers’ Lapse Led to Loss of Mars Spacecraft,” Washington Post, October 1,1999, AlGoogle Scholar.

²⁸ For early applications to international relations, see Schelling, Thomas, Strategy of Conflict (Cambridge: Harvard University Press, 1960), 58–67Google Scholar,135–36, 291–301.

²⁹ Gibbons, Robert, Game Theory for Applied Economists (Princeton: Princeton University Press, 1992), 11–12; and David Kreps, Game Theory and Economic Modeling (New York: Oxford University Press, 1990), 100–102Google Scholar.

³⁰ Loya and Boli (fn. 1).

³¹ Meyer, John and Rowan, Brian, “Institutionalized Organizations: Formal Structure as Myth and Ceremony,” American Journal of Sociology 83 (September 1977)CrossRefGoogle Scholar; Meyer, John, “The World Polity and the Authority of Nation State,” in Bergesen, Albert, ed., Studies in the Modern World-System (New York: Academic Press, 1980)Google Scholar; and Meyer, John, Boli, John, Thomas, George, and Ramirez, Francisco, “World Society and the Nation State,” American Journal of Sociology 103 (July 1997)CrossRefGoogle Scholar. For overviews of sociological institutionalism, see Powell, Walter and DiMaggio, Paul, eds., The Nevi Institutionalism in Organizational Analysis (Chicago: University of Chicago Press, 1991)Google Scholar; Hall, Peter and Taylor, Rosemary, “Political Science and the Three New Institutionalisms,” Political Studies 44 (December 1996)CrossRefGoogle Scholar; and Finnemore, Martha, “Norms, Culture, and World Politics: Insights from Sociology's Institutionalism,” International Organization 50 (Spring 1996)CrossRefGoogle Scholar.

³² The potential of “negative” or conflictual role identities impeding international cooperation is brought out more explicitly by some constructivist scholars, for example, Wendt, Alexander, “Anarchy Is What States Make of It.” International Organization 46 (Spring 1992), esp. 418CrossRefGoogle Scholar.

³³ See, e.g., Meyer et al. (fri. 31), 148. Because these models become operational only when they are internationally and transnationally “intersubjective,” or shared, they are at any given point in time largely “beyond the discretion of any individual participant”; Meyer and Rowan (fn. 31), 343.

³⁴ Wendt (fn. 32), 419.

³⁵ See March, James and Olsen, Johan, Rediscovering Institutions: The Organizational Basis of Politics (New York: Free Press, 1989), esp. 23–24,160–62Google Scholar; and idem, “The Institutional Dynamics of International Political Orders,” International Organization 52 (Autumn 1998)Google Scholar.

³⁶ This sets world society theory apart from constructivism, which as such makes no substantive assumptions about the beliefs that constitute the international social structure and leaves open the possibility that actors' material interests will counteract the normative/ideational demands created by social structure. Otherwise, many elements of world society theory exhibit close affinities to constructivist theories in IR. See, e.g., Checkel, Jeffrey T., “The Constructivist Turn in International Relations Theory,” World Politics 50 (January 1998)CrossRefGoogle Scholar; Lapid, Yosef and Kratochwil, Friedrich, The Return of Culture and Identity in IR Theory (Boulder, Colo.: Lynne Rienner, 1996)Google Scholar; Ruggie, John Gerard, Constructing the World Polity (London: Routledge, 1998)CrossRefGoogle Scholar; and Wendt, Alexander, A Social Theory of International Politics (Cambridge: Cambridge University Press 1999), esp. 177–78,181–84,224–33, and ensuing debateCrossRefGoogle Scholar.

³⁷ Meyer (fri. 31); Meyer et al. (fii. 31,1997).

³⁸ Meyer et al. (fn. 31), 144. It is the assumed universality of the models of world society that allows this approach to explain similar behavior by dissimilar actors or actors with different interests; see Finnemore (fn. 31), esp. 334.

³⁹ Meyer (fn. 31), 145.

⁴⁰ Ibid, 165.

⁴¹ Loya and Boli (fn. 1), 192.

⁴² Ibid., 196.

⁴³ Ibid., 194.

⁴⁴ Meyer and Rowan (fn. 31), 341,355–59; Meyer et al. (fn. 31), 149,154–56.

⁴⁵ Loya and Boli (fn. 1), 193.

⁴⁶ Ibid., 191.

⁴⁷ Ibid.

⁴⁸ Ibid., 188.

⁴⁹ See Finnemore (fn. 31), 339–42. Note that in IR constructivism, there are conscious attempts to incorporate at least ideational contestation, based on Habermas's notion of discourse; see Risse, Thomas, “Let's Argue: Communicative Action in World Politics” International Organization 54 (Winter 2000)CrossRefGoogle Scholar.

⁵⁰ For recent overviews of works in the realist tradition, see Walt, Steven, “The Enduring Relevance of the Realist Tradition,” in Katznelson, Ira and Milner, Helen, eds., Political Science: State of the Discipline (NewYork: W. W. Norton, 2002)Google Scholar; and Elman, Colin and Elman, Miriam Fendius, eds., Progress in International Relations: Appraising the Field (Cambridge: MIT Press, 2003)Google Scholar.

⁵¹ Gowa, , “Bipolarity, Multipolarity, and Free Trade,” American Political Science Review 83 (December 1989)CrossRefGoogle Scholar.

⁵² On relative gains and the fungibility of power, see the debate in Baldwin, David, ed., Neorealism and Neoliberalism: The Contemporary Debate (NewYork: Columbia University Press, 1993)Google Scholar.

⁵³ Grieco, , Cooperation among Nations: Europe, America, and Non-Tariff Barriers to Trade (Ithaca, N.Y.: Cornell University Press 1990), 10Google Scholar.

⁵⁴ Krasner, , “Global Communications and National Power: Life of the Pareto Frontier,” World Politics 43 (April 1991)CrossRefGoogle Scholar.

⁵⁵ Ibid.; Krasner (fn. 24); Baldwin (fn. 52); Mearsheimer, John, “The False Promise of International Institutions,” International Security 19 (Winter 1994–95)CrossRefGoogle Scholar.

⁵⁶ Dan Drezner, “The Global Governance of the Internet: Bringing the State Back,” Political Science Quarterly (forthcoming).

⁵⁷ As a further alternative, realism may predict that if existing institutions fail to accommodate the interests of the most powerful actors, those actors will create and privilege rival institutions, leading to the demise of the institutions that fail to adapt to the distribution of power. We address this issue in the conclusion.

⁵⁸ Martin and Simmons (fn. 7); Krasner (fn. 54), 365.

⁵⁹ See, e.g., Thelen, Kathleen, “Historical Institutionalism in Comparative Politics,” Annual Review of Political Science! (1999)CrossRefGoogle Scholar; and Pierson, Paul, “Increasing Returns, Path Dependence, and the State of Politics,” American Political Science Review 94 (June 2000)CrossRefGoogle Scholar.

⁶⁰ Büthe, Tim, “Taking Temporality Seriously: Modeling History and the Use of Narratives as Evidence,” American Political Science Review 96 (September 2002)CrossRefGoogle Scholar.

⁶¹ The idea also plays an important role in much of the literature on varieties of capitalism and in many second image reversed arguments in international political economy, such as Katzenstein, Peter, ed., Between Power and Plenty: Foreign Economic Policies of Advanced Industrial States (Madison: University of Wisconsin Press, 1978), esp. 323—32Google Scholar. Recent works in the second image reversed tradition that are related to standardization and/or regulatory issues include Kollman, Kelly and Prakash, Aseem, “Green by Choice? Cross-National Variations in Firms' Responses to EMS-Based Environmental Regimes,” World Politics 53 (April 2001)CrossRefGoogle Scholar; and Newman, Abraham and Bach, David, “Self-Regulatory Trajectories in the Shadow of Public Power: Resolving Digital Dilemmas in Europe and the United States,” Governance 17, no. 3 (2004)CrossRefGoogle Scholar.

⁶² See Hopner, Martin, “What Connects Industrial Relations with Corporate Governance? A Review on Complementarity” (Manuscript, Center for European Studies, Harvard University, January 2003)Google Scholar; and Hall, Peter and Soskice, David, eds., Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford: Oxford University Press, 2001), 17CrossRefGoogle Scholar. Note that we conceptualize complementarity as being on a continuum.

⁶³ Notions of institutional complementarity were introduced in economics to analyze the fit between organizational structure and work practices within firms, e.g., Chandler, Alfred Dupont, Strategy and Structure (Cambridge: MIT Press, 1962)Google Scholar; Milgrom, Paul and Roberts, John, “Complementarities and Fit: Strategy, Structure, and Organizational Change in Manufacturing,” Journal of Accounting and Economics 19 (March-May 1995), 179–208CrossRefGoogle Scholar. In comparative political economy, notions of complementarity play an important role of the varieties of capitalism literature, especially Aoki, Masahiko, “The Contingent Governance of Teams: Analysis of Institutional Complementarity,” International Economic Review 33 (August 1994), 657–76CrossRefGoogle Scholar; Amable, Bruno, Ernst, Ekkehard, and Palombarini, Stefano, “How Do Financial Markets Affect Industrial Relations? An Institutional Complementarities Approach” (Manuscript, University of Paris, August 2001Google Scholar, http://pythie.cepremap.ens.fr/~amable/instcomp.pdf, accessed June 30, 2003); and Peter Hall and Daniel Gingerich, “Varieties of Capitalism and Institutional Complementarities in the Macroeconomy” (Paper presented at the International Seminar on Institutional Complementarities and Dynamics of Economic Systems, Universite de Paris, April 5–6, 2002). None of these, however, considers the implications of institutional complementarities for the ability of actors at a lower level to affect coordination efficiently at a higher level of aggregation.

⁶⁴ Since national-level SDOs are part and parcel of ISO and IEC, they constitute quite unambiguously a set of institutions “more or less ‘naturally’ involved with the set of activities directly concerned”; Amable, Bruno and Petit, Pascal, “Identifying the Structure of Institutions to Promote Innovation and Growth,” CEPREMAP Working Papers no. 9919 (Universite de Paris, October 1999), 3Google Scholar. We thus use the international institutional structure itself to provide an analytical focus of most pertinent institutional complementarities in the dense web of political-economic social institutions of advanced capitalist democracies.

⁶⁵ As noted before, the international standardization organizations are nongovernmental organizations, whose members are national-level standards organizations, not states/governments.

⁶⁶ Regarding the ability of some types of domestic institutions to provide resources that help actors adopt to change while others inhibit change, see Hall, Peter, “The Movement from Keynesianism to Monetarism: Institutional Analysis and British Economic Policy in the 1970s,” in Steimo, Sven, Thelen, Kathleen, and Longstreth, Frank, eds., Structuring Politics: Historical Institutionalism in Comparative Analysis (New York: Cambridge University Press, 1992)Google Scholar. Among the imperatives for international standardization are trade agreements that privilege international standards and the increase in international trade, which has been disproportionately concentrated in manufactured goods; McKeown, Timothy, “The Global Economy, Post-Fordism, and Trade Policy in Advanced Industrialized States,” in Kitschelt, Herbert, Lange, Peter, Marks, Gary, and Stephens, John, eds., Continuity and Change in Contemporary Capitalism (New York: Cambridge University Press, 1999)Google Scholar.

⁶⁷ This is not to say that the domestic institutions for standardization are identical across Europe; for a rich discussion, see Jay Tare, “National Varieties of Standardization,” in Hall and Soskice (fn. 62), but the difference between the U.S. and the Europeans is much larger than any difference among the Europeans.

⁶⁸ We are not arguing that any one system is superior in general. As Hendrik Spruyt shows historically, welfare-enhancing standardization does not require a single mode of governance for the standardization process; Spruyt, , “The Supply and Demand of Governance in Standard-Setting: Insights from the Past,” Journal of European Public Policy 8, no. 3 (2001), 382–83CrossRefGoogle Scholar. Similarly, we have found few systematic differences in firms' assessments of the technical quality of U.S. and European countries' domestic standards per se.

⁶⁹ On this bias, see, for example, Vogel, David, Kindred Strangers: The Uneasy Relationship between Politics and Business in America (Princeton: Princeton University Press, 1996)Google Scholar, esp. 29–72.

⁷⁰ See NIST, Public Hearing Proceedings: Improving U.S. Participation in International Standards Activities, cited in U.S. Congress, Office of Technology Assessment, Global Standards: Building Blocks for the Future (hereafter cited as Global Standards Report), TCT-512 (Washington, D.C.: U.S. Government Printing Office, March 1992)Google Scholar, 15. (NIST is an agency of the U.S. Commerce Department's Technology Administration). See also the presentation by James Thomas, president of ASTM, at a 1998 conference entitled, “Toward a National Standards Strategy to Meet Global Needs,” in Leuteritz, Krista Johnsen and Leight, Walter, eds., Conference Report (Washington, D.C.: U.S. Department of Commerce, February 1999), 53 (http://nist.gov/ts/htdocs/210/ir6290.pdf, accessed June 30,2003)Google Scholar.

⁷¹ SRI, Industrial Standards (Menlo Park, Calif.: SRI, Long Range Planning Service, 1971), 3Google Scholar.

⁷² Hamilton, Robert, “The Role of Nongovernmental Standards in the Development of Mandatory Federal Standards Affecting Safety or Health,” Texas Law Review 56 (November 1978), 1343Google Scholar.

⁷³ The main national SDO often has a sister organization responsible for electrical and electronic standards.

⁷⁴ For details, see Falke, Josef and Schepel, Harm, eds., Legal Aspects of Standardisation in the Member States of the EC and of EFTA (Luxembourg: Office for Official Publications of the European Communities, 2000)Google Scholar.

⁷⁵ Consensus does not mean unanimous consent but implies that objections have been reconciled, or are not sustained, or are considered to be of such minor significance as not to warrant further delay.

⁷⁶ For example, there are about 36,000 for DIN and 20,000 for BSI.

⁷⁷ The acronyms stand for Comité Européen de Normalisation and Comité Européen de Normalisation Eléctronique.

⁷⁸ CEN, for example, has 274 technical committees handling 8,842 standardization work items.

⁷⁹ In this section we draw primarily on previous work on standardization and on information from sixty-six interviews that we conducted with standards experts mostly between March 1999 and September 2001. We draw to a lesser extent on information relayed to us in response to two open-ended questions, which we posed at the end of our survey to gather more detailed information of the type usually reserved for qualitative interviews.

⁸⁰ These are the Vienna Agreement of 1991 between the ISO and CEN, and the Dresden Agreement of 1996 between the IEC and CENELEC (a revision of the Lugano Agreement of 1991). A working agreement also exists between the IEC and the European Telecommunications Standards Institute (ETSl). Under the Vienna Agreement, CEN and ISO have agreed upon detailed procedures for the exchange of information and their cooperation in the drafting and adoption of standards. For example, when a new international standards project is proposed, members in the relevant ISO technical committee decide by majority vote whether CEN or ISO should take the lead in developing the standard. In the vast majority of cases ISO takes the lead, but voting on the final draft standard is parallel with CEN. However, the lead is likely to go to CEN if specific requirements of European directives or regulations must be reflected in a standard, or if the commission mandates that a standard be written by a certain target date, or if the affected businesses are primarily European. Voting on the draft standard takes place in both CEN and ISO. If adopted by both, the European-made standard will become an international standard without further technical discussion at the ISO. The agreement between CENELEC and IEC contains similar provisions; see Inside the IEC (Geneva: IEC Central Office, 1990), 22–24Google Scholar. For a detailed recent analysis of EU-level standardization, see Egan, Michelle, Constructing a European Market: Standards, Regulation, and Governance (New York Oxford University Press, 2001)CrossRefGoogle Scholar.

⁸¹ Global Standards Report (fh. 70), 13.

⁸² Raymond Kammer, statement before the House Committee on Science Subcommittee on Technology, Hearing on International Standards: Technical Barriers to Free Trade, April 28, 1998 (www.house.gov/science/kammer_04–28.htm, accessed June 30,2003).

⁸³ The examples are drawn from ANSI, American Access to the European Standardization Process (New York: American National Standards Institute, 1996), chap. 3Google Scholar.

⁸⁴ The survey was internet based, though paper copies by mail or fax were available for those (few) participants who preferred hard copy. Participants, typically managers or standards experts in firms, were contacted by e-mail or phone and were directed to our website via gateway URLs (www.standards-survey.com or www.standardssurvey.com), equipped with a username and password. Each questionnaire listed thirty-six queries with multiple-answer options and took between fifteen and twenty-five minutes to complete, depending on the extent to which respondents used the write-in options. Translated questionnaires were used in Germany and Spain. By the click of the mouse, the results were electronically transmitted for coding and evaluation. Between September 2001 and May 2002, nearly fourteen hundred questionnaires were collected, about two hundred to three hundred per sector. The response rate was about 32 percent, a rate almost twice as high as most business surveys that do not work with panels of regular participants. No compensation was paid to participants. The convenience of online participation and our use of reminder messages to nonresponding participants appear to have boosted participation, though the promise of a free copy of the final report with executive summary surely also helped.

⁸⁵ Stratified random sampling uses one or more characteristics of the units to subdivide the sampling frame into groups, then draws from each group a sample proportional to its share of the total population. Within each group, random selection is used, which justifies the standard assumptions underpinning statistical inference. We implemented stratified random sampling through making the probability of selection of each firm proportional to its contribution to the total employment in the industry (a variation of what is also known as “dollar sampling”). For a more detailed discussion, see Weisberg, Herbert, Krosnick, Jon, and Bowen, Bruce, An Introduction to Survey Research, Polling, and Data Analysis, 3rd ed. (Thousand Oaks, Calif.: Sage, 1996), 38–76Google Scholar; and Lohr, Sharon, Sampling: Design and Analysis (Pacific Grove, Calif.: Duxbury Press, 1999)Google Scholar.

⁸⁶ It is based on the response to the survey question: “When a new international standard has been proposed and is being discussed in the appropriate technical committees/working groups of the international standards developing organization for such standards (ISO, IEC, and so on), [how frequently] does your company try to get involved in the specification of that standard?” The response options are never/rarely (1), sometimes (2), about half of the time (3), often (4), and very often (5).

⁸⁷ Respondents selected among seven size categories, ranging from “less than 50” to “more than 1500” employees.

⁸⁸ In addition to anecdotal evidence, there also is a serious theoretical argument developed by Jack Knight that implies that large firms should play a dominant role; Knight, , Institutions and Social Conflict (Cambridge: Cambridge University Press, 1992)CrossRefGoogle Scholar.

⁸⁹ Five response categories, ranging from “rarely” to “very often.”

⁹⁰ Four response categories, ranging from “very inexpensive” to “very expensive.”

⁹¹ Respondents had to choose among eight response categories, ranging from “0%” to “more than 50%.”

⁹² The regressions were estimated in Stata, version 7.

⁹³ Four response categories, ranging from “very dissatisfied” to “very satisfied.”

⁹⁴ The insignificance of the Size variable is consistent with ISO/lEC standardization being an institutional process of harmonization rather than a market process, but it is nonetheless surprising. Conventional wisdom and anecdotal evidence suggest that large firms in particular try to use the standardization process to get their own, often proprietary technical specifications adopted as the industry standard; see, for example, Austin, Marc and Milner, Helen, “Strategies of European Standardization,” European Journal of Public Policy 8, no. 3 (2001)CrossRefGoogle Scholar. Similarly, one might expect that the current practice of market leaders (which tend to be larger firms) provides a focal point in the standardization process. Interestingly, an ordered logit model of involvement with only firm size as the independent variable produces a positive coefficient that is substantively and statistically highly significant, indicating that larger firms are more likely to be involved in international standardization than smaller ones. However, this positive effect of Size evaporates with the inclusion of other independent (control) variables. The insignificance of Size persists even when the original seven-category ordinal variable is replaced by dummy variables for each category or by an ordinal variable based on recoding the seven categories into three, four, or five groups. This finding is consistent with views expressed by Hamilton (fn. 72), who questions the conventional wisdom about the dominance of big firms (p. 1381).

⁹⁵ We used the CLARIFY software, which (among other quantities of interest) provides estimates for the change in predicted probabilities (first differences) and confidence intervals around the point estimates, based on Monte Carlo simulations. See Tomz, Michael, King, Gary, and Wittenberg, Jason, CLARIFY: Software for Interpreting and Presenting Statistical Results, version 2.1 (May 1, 2003)Google Scholar (http://gking.harvard.edu, accessed June 30,2003). See also King, , Tomz, , and Wittenberg, , “Making the Most of Statistical Analyses,” American Journal of Political Science 44 (April 2000)CrossRefGoogle Scholar.

⁹⁶ See Abbott, Kenneth and Snidal, Duncan, “Hard and Soft Law in International Governance,” International Organization 54 (Summer 2000)CrossRefGoogle Scholar.