Regime Type, Veto Players, and State Participation in Transnational Public-Private Governance Initiatives

Transcription

1 Regime Type, Veto Players, and State Participation in Transnational Public-Private Governance Initiatives Oliver Westerwinter Department of Political Science, University of St. Gallen June 16, 2017 Draft Do not circulate or cite without the author s permission Abstract Why do states use informal institutions to govern global problems? Extant research on the forms of institutionalization in global governance focuses on formal modes of cooperation, such as formal intergovernmental organizations (FIGOs) and treaties. Formal rules, however, are often inadequate, if not entirely misleading, descriptions of the game that actors play in world politics. Recent work on informal global governance largely focuses on informal governance within FIGOs. Little is known about informal governance outside FIGOs. This paper starts to fill this gap. I focus on a particular type of informal global governance; namely, transnational public-private governance initiatives (TGIs) in which states and/or FIGOs cooperate with business and civil society actors. I argue that states choice to participate in TGIs is driven by forces at the domestic level. Specifically, I argue that the extent to which a state participates in TGIs depends on its political regime type and the presence of domestic veto players. The added value of the flexibility and low legal rigidity of cooperation based on TGIs allows democratic leaders to avoid the cumbersome domestic ratification processes of international agreements. It also allows governments irrespective of their regime type to bypass domestic veto players with opposing preferences. The push toward informal global governance that originates in the presence of domestic veto players, however, decreases as the level of democracy in a country increases. I empirically test my argument using a new dataset with information on 468 TGIs between 1885 and Results suggest that state participation in TGIs is strongly driven by domestic politics. Keywords: informal governance, international cooperation, transnational public-private governance initiatives, domestic politics, regime type, veto players, rational design. This research is funded by the Swiss Network for International Studies. I would like to thank the excellent research assistance of Christian Andres, Ruslan Aybazov, Tino Good, Stefano Jud, Rosie Keller, Laura Leibundgut, Giulia Parini, Dominik Schneeberger, Johannes Schultz, and Keto Schumacher in preparing this paper. I also thank Ken Abbott, Liliana Andonova, Michael Barnett, Tom Biersteker, Tom Hale, Virginia Haufler, Dirk Lehmkuhl, Miles Kahler, Erasmus Kersting, Christopher Kilby, Barb Koremenos, Lisa Martin, Katja Michaelowa, Christine Neuhold, Joost Pauwelyn, Jon Pevehouse, Kal Raustiala, Bernhard Reinsberg, Duncan Snidal, Jonas Tallberg, and Felicity Vabulas for helpful suggestions. I especially thank Sharlene Westerwinter for sharing her thoughts and for always supporting me. Müller-Friedbergstr. 8, 9000 St. Gallen, Switzerland, oliver.westerwinter@unisg.ch, web: oliverwesterwinter.com.

2 Introduction International organizations are at the center of contemporary global governance. Today, they are present in all substantive areas of world politics ranging from human rights and environment to the most sensitive financial and security issues. For decades, realist scholars and their institutionalist counterparts have debated whether international organizations matter for world politics. While realists contend that institutions have little influence on state behavior and are epiphenomenal to outcomes (Grieco, 1988; Mearsheimer, 1994), institutionalists argue that they provide essential functions to facilitate cooperation that would not otherwise occur between rational, utility maximizing actors (Keohane, 1984; Axelrod and Keohane, 1985). Part of the question whether institutions matter for international cooperation is about why sovereign states join them in the first place and much research has focused on finding answers to this question (Martin, 1992; Abbott and Snidal, 1998; Mansfield, Milner and Rosendorff, 2002; Pevehouse, 2003; Minnich, 2005; Mansfield and Pevehouse, 2006; Mansfield, Milner and Pevehouse, 2008; Mansfield and Milner, 2012). Research on international institutions has emphasized formal modes of international cooperation, such as international treaties or formal intergovernmental organizations (FIGOs). Accordingly, works that examine whether states join international institutions have focused on formal international cooperation as well. FIGOs are, however, only part of the increasingly complex patchwork of contemporary global governance. They do not exhaust the institutional variety of international cooperation and are often inadequate, if not entirely misleading, descriptions of the game that actors play in world politics (Achen, 2006; Kleine, 2014). More recently scholars have started to highlight the importance of informal governance in world politics (Stone, 2011, 2013; Kleine, 2013) and examined how informal institutions can affect global governance outcomes (Abbott and Snidal, 2009; Vabulas and Snidal, 2013; Avant and Westerwinter, 2016; Avant, 2016). However, little is known about why states become members of informal international institutions in the first place. 1 This paper contributes to filling this gap. I focus on a specific type of informal international institution; namely, transnational public-private governance initiatives (TGIs). TGIs are an important element of contemporary world politics. They are 1 Two notable exceptions are Andonova (2014) and Andonova, Hale and Roger (2016) who examine the determinants of state participation in transnational sustainability and climate change partnerships. 2

3 institutions in which states and/or FIGOs cooperate with business actors and non-governmental organizations (NGOs) to govern global problems. In contrast to traditional state-based governance, in TGIs private actors are not only the objects of governing, but are involved at the center stages of the governance process including decision-making, monitoring, and enforcement (Reinicke and Deng, 2000; Börzel and Risse, 2005; Abbott and Snidal, 2009). By examining why states participate in TGIs, I contribute to the analysis of the factors that affect states membership in the institutions of informal global governance. I also contribute to the literature on informal governance in world politics more generally by highlighting the importance of informal governance outside FIGOs, something that scholars have only recently begun to explore systematically (Vabulas and Snidal, 2013). I argue that states choice to participate in TGIs is driven by domestic political processes. Specifically, I argue that the extent to which a government chooses to participate in the creation of a TGI depends on its political regime type and the strength of the domestic veto players it is confronted with. The added value of the flexibility and low legal rigidity of informal international institutions allows democratic leaders to avoid the often cumbersome and costly domestic ratification processes of international agreements. Informal international institutions also permit governments to bypass domestic veto players with opposing preferences due to their often lower level of political salience and lack of transparency. In other words, informal international institutions are instruments that allow governments to reduce the domestic costs of international cooperation and to work toward global governance outcomes that are as close as possible to their own preferences. As a result, governments using informal international institutions to pursue their foreign policy preferences can achieve their goals easier and at lower costs because of the relative absence of institutional and political constraints. To test this theoretical argument empirically, I conduct a large-n statistical analysis of states decisions to join the creation of new TGIs. I use a new dataset that contains detailed information on states participations in 468 TGIs in a broad range of issue areas (Westerwinter, 2017a). I find that countries with democratic political regimes and strong domestic veto players have an increased probability to participate in the founding of new TGIs. I also find that the positive effect of domestic veto players on state participation in TGIs is more pronounced for democracies than for non-democracies and decreases in a country s level of democracy. 3

4 Figure 1: TGI and FIGO growth, Net cumulative growth, FIGOs & TGIs, Number of institutions created Count 200 Type FIGO TGI Count Type FIGO TGI The remainder of this paper proceeds as follows. First, I briefly map the recent increase in the number of TGIs in world politics over time and and across issue areas and compare it to the growth of FIGOs. Second, I introduce my theoretical argument about the domestic politics of state participation in TGIs. The third section presents my empirical strategy including the data and the operationalization of the variables used in the empirical analysis. Section four reports the results of the empirical tests of my argument. The last section concludes and links the findings and their implications to the emerging literature on informal global governance and international cooperation more broadly. Mapping TGIs in world politics The institutional architecture of global governance has undergone dramatic changes in the past decades. The number of informal international institutions in the form of TGIs has been growing rapidly since the 1990s, both in absolute terms and even more so relative to FIGOs. I base my exploration of the growing importance of TGIs in world politics on a new dataset that allows for the comparative analysis of FIGOs and TGIs over time and across issue areas (Westerwinter, 2017a,b). The data on FIGOs comes from the recently updated Correlates of War dataset on intergovernmental organizations (Pevehouse et al., 2015). 2 2 I am grateful to Jon Pevehouse for sharing the data prior to its public release. 4

5 TGIs experienced a rapid growth in the past decades. From 63 in 1990, within 24 years the number of TGIs increased by about 590 percent to a total of 435 in This evidence is consistent with other studies of transnational governance initiatives (Abbott and Snidal, 2009; Abbott, Green and Keohane, 2016). According to Abbott and Snidal (2009), transnational governance networks are a recent phenomenon. While few such arrangements existed before 1994, since then their number has increased steadily (Abbott and Snidal, 2009, pp ). Likewise, Abbott, Green and Keohane (2016, p. 248) find that private transnational regulatory organizations formed by different mixtures of business and civil society actors have proliferated in the past decades. Although the absolute number of FIGOs was still higher at 316 in 1990, TGIs became the most frequent form of cooperation in the dataset as of In addition, between 1990 and 2014, the growth of FIGOs was much slower at about 6 percent (see figure 1). Importantly, the decrease and flattening out of the growth of FIGOs co-occurred with the beginning of the sharp increase in the importance of TGIs as a major form of cooperation in the late 1990s. This observation is further supported if we look at the number of new formations of international institutions over time. As the right-hand panel of figure 1 shows, the growth rates of FIGOs were consistently higher than those of TGIs between 1950 and Only in the second half of the 1990s, FIGO growth rates began to plummet, while at the same time the number of newly created TGIs increased dramatically. While the recent growth of TGIs is striking (see figure 1), it is not universal. In some issue areas we observe more of a turn towards TGIs as institutional arrangements for governing global policy issues than others. In 2014, about 51 percent of all organizations in our data are concerned with environmental issues, including climate change and sustainable energy-related problems; 40 percent deal with development. Sixty-four percent address social problems, and 25 percent deal with health problems. Transnational public-private governance initiatives have also begun to enter the realms of high politics, with 25 percent of the TGIs in our data dealing with trade and commerce, 5 percent with finance, and 4 percent with security issues. FIGOs, by contrast, are most prominent in the areas of social affairs, development, trade and commerce, and technical issues (see figure??). These figures should be interpreted cautiously. Nevertheless, we can see that TGIs are not equally distributed among issue areas, and that they are not limited to low politics issues. They are also important for states that seek to address economic and 5

6 Figure 2: TGIs and FIGOs across issue areas, 2014 FIGOs & TGIs across issue areas, Percent Type FIGO TGI 20 0 Development Environment Finance Health Human rights Security Social affairs Technical Trade & commerce security problems. Another source of variation in international institutions is the pattern of state participation across them. Starting with TGIs, we can see that in 2014 a small number of states, such as the United States, the United Kingdom, Canada, and Germany participate in a large number of TGIs. Others, including Russia, China, Brazil, and India, as well as many African and Latin American countries, are much less involved (see figure 3). The top twenty TGI participant states in 2014 included the US, UK, Germany, Netherlands, and Switzerland, but also Kenya, Nigeria, South Africa, and the Philippines. Compared to TGIs, the pattern of state participation in FIGOs is more balanced, with a larger number of states being members of a larger number of organizations. How can we explain this striking variation in states participation in transnational public-private governance initiatives? Explaining state participation in TGIs: Democracy and veto players Informal governance is a strategy. States select informal governance because they perceive it as the best way to structure particular interactions and to govern global problems, allowing them to achieve their objectives (whether substantive, political or organizational) to the greatest extent possible. As with any strategic choice, the selection of informal modes of global governance takes place in situations character- 6

7 Figure 3: State participation in TGIs and FIGOs, 2014 Transnational public private governance initiatives, 2014 Formal intergovernmental organizations, 2014 Count Count ized by particular sets of constraints and opportunities, which shape the costs and benefits derived from available institutional designs (Koremenos, Lipson and Snidal, 2001; Koremenos, 2016). The decision to govern global problems using informal modes of cooperation, rather than more formalized structures, is thus shaped by a potentially large number of variables. I argue that an important driving force of state participation in informal international institutions resides in countries domestic political arenas. Of particular importance is the political regime type of a country. The governments of democratic states have to adhere to a range of legal procedural requirements when it comes to policy making (Tsebelis, 1995). This does not only apply to policy making at the domestic, but also at the international level, and to the participation in FIGOs and international treaties (Milner, 1997; Mansfield and Pevehouse, 2006; Mansfield, Milner and Pevehouse, 2008; Mansfield and Milner, 2012). At several stages of the policy making process, a democratic government has to achieve agreement with international as well as a range of domestic veto players (Putnam, 1988). This can cause considerable costs and make it more difficult for the government to realize policies that are close to its preferences (Milner, 1997; Martin, 2000). One way for democratic leaders to reduce these costs and to realize policies that are as close as possible to their own ideal point is to opt for modes of international cooperation that are less demanding 7

8 in terms of domestic politics than FIGOs and international treaties. Informal international institutions in the form of TGIs are one such mode. Because of their informal character, TGIs require less, or even no involvement of domestic players in foreign policy making. The rules and standards created by TGIs do not typically take the form of international treaties and therefore do not require domestic ratification to become effective (Abbott and Snidal, 2000). Furthermore, informality at the international level is associated with reduced visibility and transparency of policy making which reduces the likelihood that the domestic opposition takes note of government efforts which in turn reduces the likelihood and therefore expected costs of domestic opposition to foreign policy efforts(stone, 2011). This creates an informational advantage for the executive vis-à-vis its domestic opposition and makes achieving foreign policy goals more likely and less costly. This leads to my first hypothesis: Hypothesis 1: Democratic states are more likely to participate in TGIs. Rather than democracy in general, it may be that more specific institutional characteristics of a country s political regime shape its choice to participate in TGIs. Specifically, it may be the configuration of domestic veto players and their relationship to the government that pushes an executive toward informal forms of international cooperation. If the domestic veto players in a country are strong, they have the ability to effectively influence policy making (Tsebelis, 1995). However, the existence of effective veto players does not automatically increase the difficulties of the government to create policies that are close to its own preferences. As long as the veto players have policy preferences that are close to those of the government, the additional costs for the government that are created by their involvement are marginal. Yet, once the preferences of the veto players and the government differ, the complexity of the policy making process increases. Now the government has to reach agreement with actors with diverging preferences which it cannot easily bypass because they occupy institutional veto positions. In such a situation, it becomes difficult for the government to realize policies that are close to its preferences. At the very least, achieving policy goals becomes more costly for the government (Tsebelis, 1995; Milner, 1997). When a government leader anticipates that the domestic opposition would block a particular foreign policy initiative, the adoption of a new international treaty, or the creation of a new intergovernmental organization, she can try to bypass this resistance by choosing a mode of international cooperation that 8

9 makes the involvement of the domestic opposition in foreign policy making more difficult or avoids it altogether. Specifically, she can decide to use informal international institutions, such as TGIs, to govern a particular problem. This makes the new foreign policy initiative less visible and more difficult to influence for her domestic opponents. Formal international agreements typically require national legislatures to become involved in the process to ratify and implement an agreement, which increases the level of public attention and scrutiny (Lipson, 1991; Yarbrough and Yarbrough, 1992). Informal institutions, by contrast, carry no need for legislative approval and are therefore less likely to become captive to domestic conflicts. Moreover, informal cooperation reduces the incentives among domestic veto players to mobilize against agreements. Since formalized agreements are often characterized by high levels of rule precision, more formalized arrangements provide better information about the distributional implications of cooperation (Goldstein and Martin, 2000). This, in turn, increases the incentives for negatively affected domestic groups to mobilize in opposition to such agreements. Due to their lower level of rule precision, the distributional implications of TGIs often remain obscure, which reduces the incentives for the domestic opposition to mobilize against them (Eilstrup-Sangiovanni, 2009). These features of informal cooperation are attractive when domestic veto players with preferences opposed to the government exist in a country. I hypothesize: Hypothesis 2: States with legislative veto players whose preferences differ from the executive s preferences are more likely to participate in TGIs. Finally, the effect of domestic veto players on a country s participation in transnational public-private governance initiatives depends on a country s regime type in complex ways because of the interplay between the strategic considerations of governments and the domestic institutional context. On the one hand, the positive association between domestic veto players and a country s participation in TGIs is more pronounced for democratic states than for non-democratic states. For a government, the strategic advantages and benefits of pursuing foreign policy goals by using informal institutional structures at the international level are most pronounced when this government is confronted with domestic veto players in an institutional context that provides these veto players with ample opportunities to intervene in the foreign policy making process. Whether this is because of the requirement to inform the parliament of 9

10 foreign policy strategies and cooperation efforts, the involvement of opposition parties, or the requirement of ratification of international agreements, in democratic countries, veto players are, relative to nondemocracies, able to use a range of institutional procedures and structures to participate in and shape the outcomes of foreign policy making. In non-democratic countries, such institutional provisions for the inclusion of veto players in foreign policy making are less developed or entirely absent. Thus, even if veto players with foreign policy preferences different from the preferences of the government exist, they will have less opportunities to become involved and affect the outcomes of foreign policy making. On the other hand, for democratic countries the extent to which domestic veto players exert a positive effect on a country s participation in TGIs decreases with an increase in democracy. Democracies are characterized by a thick institutional structure on which the making, implementation, monitoring, and enforcement of regulations is based (Tsebelis, 1995). Part of this domestic institutional architechture are rules and procedures as well as more general norms that govern the interaction of different institutions and particularly the interaction and use of different forms of institutional power both in the form of veto player positions and government office. Rather than on material capability, expertise, or other individual attributes of players, veto positions in democracies are based on institutional procedures and structures and these institutions do not only create veto positions, they also provide rules, procedures, and norms that govern and limit veto power and its use (Tsebelis, 1999). These meta rules and limitations on the exercise of institutional veto power reduce the domestic costs of international cooperation that stem from domestic veto players and thus, everything else being equal, reduce the incentives of the executive to choose informal modes of international cooperation to bypass domestic veto players. The same logic applies to governments. Even though with an increase in democracy governments have an increased incentive to bypass domestic veto players to reduce the domestic costs of international cooperation, high levels of democracy also make it more difficult for governments to bypass domestic veto players by using informal international cooperation. As the institutional and legal provisions that allow veto players to participate in foreign policy making become more and stronger, the more difficult and costly it is for a government to circumvent these provisions in order to achieve foreign policy outcomes that are closer to its own preferences. In addition, social norms of participation and transparency are likely to be stronger and more widely appreciated among political and societal actors in democracies 10

11 compared to non-democracies. On top of this, channels for the transmission and spread of political information are more difficult to control in democracies. Together, this creates an environment in which it is more difficult for governments to hide foreign policy projects that are contentious at the domestic level in informal institutional arrangements and, if they are caught, governments will pay a higher cost in form of damaged reputation, withdrawl of resources, or loss of votes. Thus, despite the strategic incentives governments have for using informal international cooperation, the costs of pursuing this strategy increase with higher levels of democracy. As a result, the added value of cooperating informally with other governments diminishes. From this discussion, I derive two additional hypotheses: Hypothesis 3: The positive effect of domestic veto players on a country s probability to participate in TGIs is more pronounced for democracies than for non-democracies. Hypothesis 4: The positive effect of domestic veto players on a country s probability to participate in TGIs decreases in its level of democracy. Data and methods To test my hypotheses, I estimate logistic regression models on a time-series, cross-section sample of all states from 1885 to The unit of analysis is the country-year. The dependent variable in my analysis, T GI f ormed it, is an indicator variable that is coded 1 if country i participates in the creation of at least one new TGI in year t and 0 otherwise. Data on state participation in TGIs between 1885 and 2014 is drawn from a new dataset on transnational public-private governance initiatives (Westerwinter, 2017a). This data contains information on 468 TGIs in a broad range of issue areas and covers the time period between 1885 and No undisputed definition of the universe of TGIs in world politics exists and it is unclear how many initiatives this universe of cases contains. While there is consensus that the number of transnational public-private governance arrangements has risen dramatically over the past decades, numbers range across sources. Kaul (2006, p. 219), for example, shows that the number of transnational public-private partnerships has increased from 50 in the mid-1980s to at least 400 in the 2000s. Similarly, Abbott and Hale (2014) document the existence of 223 global solution networks of varying mixtures of states, business, NGOs, and other actors. Focusing on environmental governance, Andonova (2010, p. 25) reports 11

12 the creation of over 400 so-called type II partnerships in the aftermath of the 2002 UN World Summit on Sustainable Development and the initiation of more than 150 collaborations between UN agencies, states, and non-state actors between 1998 and 2008 under the auspices of the UN Fund for International Partnerships. Finally, Widerberg and Stripple (2016) show that even within a single issue area, such as climate change, the number of existing governance arrangements reported varies considerably across databases. This ambiguity is at least partially due to the lack of a widely shared definition of transnational public-private governance arrangements (Börzel and Risse, 2005; Abbott and Snidal, 2009). Therefore, building on existing research, I develop my own definition of TGIs and use it to specify the criteria for selecting initiatives into the sample. I define transnational public-private governance initiatives as institutions that 1) involve at least one state and/or FIGO, one business actor, and one civil society organization; 2) perform tasks that are related to governing global problems; and 3) are institutionalized to the extent that they create a basis for shared expectations about behavior and are observable. First, I focus on institutions that are of a multistakeholder nature and bring together actors from the public sector, the private for-profit sector, as well as the private non-profit sector. In other words, my definition of a TGI zooms in on the center of the governance triangle proposed by Abbott and Snidal (2009). State actors are governments, government agencies, or representatives of government agencies. Institutions in which the only public participant is a local government actor, such as a municipality or a city, are not included in the sample. FIGOs may participate in a TGI either through their main secretariat or their organizational branches. Business actors encompass firms, business associations, and business foundations. Civil society actors may be NGOs, NGO coalitions, or universities and research institutes. Second, TGIs are built to fulfill a task or set of tasks that is related to providing governance at the global level in a broad sense. Governance tasks that TGIs may be concerned with include the creation of rules and standards that govern the behavior of states, corporations, and other actors, the implementation of rules and standards, the financing of projects, as well as the facilitation of information exhange and networking related to a global problem. Other governance tasks that may be part of the activities of TGIs are agenda setting as well as the monitoring, enforcement, and adjudication of rules and standards (Abbott and Snidal, 2009; Avant, Finnemore and Sell, 2010). In other words, TGIs are explicitly focused 12

13 on contributing to governing global problems or the provision of global public or collective goods. As a consequence, the sample of TGIs used in this paper does not include, for example, information sharing platforms without a focus on a particular governance problem. Third, TGIs are characterized by a minimum level of institutionalization that generates stable shared expectations about the behavior of the actors involved. This minimum level of institutional structure is critical for distinguishing TGIs from a range of instances of cooperation at the global level that are of a less regular nature. For example, once off, ad-hoc meetings of public and private actors are not captured by my definition of a TGI. The definition also does not include project-based collaboration between states and/or FIGOs, business, and NGOs. The minimum level of institutionalization of TGIs is also important from a practical research perspective. TGIs are often characterized by a lower level of institutional formalization than other forms of global governance, such as FIGOs. There is a limit to the extent of the informality of global governance institutions that researchers are able to observe. One of the hallmarks of informal institutions is that they are difficult to trace empirically because they leave little or no publicly accessible paper trail (Christiansen and Neuhold, 2012; Koremenos, 2013). At the lower end of the continuum of institutional formalization, governance efforts become difficult to observe and many global governance institutions of a highly informal character may never be observed by any researcher. To address this problem and to avoid bias toward more institutionalized forms of governance in the dataset, I theoretically exclude from the definition of a TGI, governance efforts that feature very low or no institutionalization. All institutions that meet these three requirements are part of the population of TGIs in world politics as understood here and therefore included in the data collection. An example of a TGI that governs issues related to global health is the Global Fund to Fight AIDS, Tuberculosis and Malaria (GF) (Liese and Beisheim, 2011). Based on the participation of states, the private sector, and NGOs, the GF garners, manages, and disburses resources to fight HIV/AIDS, Tuberculosis, and Malaria. Founded in 2002, the GF has a governing board and its work is supported by the Global Fund Secretariat, the Global Fund s Office of the Inspector General, as well as other institutional structures. 3 In the security area, the International Code of Conduct for Private Security Service Providers Association (ICoCA) was founded in 2013 (Avant, 2016). Building on the International Code of Conduct for Private Security Service Providers 3 See accessed:

14 process that started in 2009, the ICoCA brings together states, private security companies, and a range of civil society organizations that collaborate to create and monitor standards for private security service providers and how states make use of their services. The ICoCA meets on an annual basis in the form of its general assembly, has its own secretariat, and working bodies focused on different substantive aspects of the governance of private security providers. 4 The data used in this paper is a sample of the population of TGIs in world politics. The creation of the sample took place between 2015 and 2016 and involved a team of researchers. To identify TGIs, the definition of a TGI was applied to seven source databases that were identified as containing information about transnational governance arrangements. These databases include the Global Solution Network (GSN) database, 5 the United Nations Environment Programme (UNEP) Climate Initiatives Platform, 6 and the UN World Summit on Sustainable Development (WSSD) partnerships database. 7 The team of researchers also searched all organizations that are recorded as networks in the Yearbook of International Organizations. 8 Further, the research team applied the TGI criteria described above to the organizations documented in the databases of Andonova, Hale and Roger (2016) as well as ongoing research projects at the University of Zurich and Duke University. 9 Finally, the research team searched websites of TGIs that were included in the sample for references and links to identify additional organizations that meet the TGI definition. Thus, in total eight sources were used to identify cases for the dataset. Across these sources, 468 TGIs were identified for inclusion in the dataset (for a full list of the TGIs in the sample, see table A-I in the appendix). To test whether democracies are more likely to participate in the creation of a new TGI, I create a variable, Regime type it, that captures the extent to which a country is a democracy. I use the PolityIV data (Gurr, Jaggers and Moore, 1989) and its 21-point scale to measure a country s political regime type and its level of democracy. To examine differences in the effects of domestic veto players for democracies and non-democracies, I create two additional variables both based on PolityIV s 21-point scale regime 4 See accessed: See accessed: See accessed: See accessed: See accessed: I thank Katja Michaelowa and her team at the University of Zurich as well as Ben Collins and Suzanne Katzenstein (Collins and Katzenstein, 2016) and their team at Duke University for sharing their databases. 14

15 type measure. The first variable, Democracy it, is coded 1 for states that have a regime type score of 7 or higher, and 0 otherwise (Mansfield and Milner, 2012). This variable identifies whether a country i is a democracy in a given year t or not. The second variable, Autocracy it, measures whether a country i is an autocracy in t or not. It is coded 1 for countries that have a PolityIV regime type score of 7 or lower in year t, and 0 otherwise. In robustness analyses, I use unified democracy scores (Pemstein, Meserve and Melton, 2010) to capture a country s regime typ and different cutpoints (6 and 6) (Pevehouse, 2003) for the democracy and autocracy indicators. These alternative operationalizations do not substantively alter the results of my main analysis. For measuring domestic veto players, I follow existing studies of the relationship between veto players and international cooperation (Mansfield, Milner and Pevehouse, 2007; Mansfield and Milner, 2012; Lupu, 2015) and use the PolCon III measure developed by Henisz (2002). The measure is especially useful for teh purpose of testing my theoretical argument because it is designed to quantify the difficulties executives face when making policy changes. Based on a spatial model of interaction between political actors, the measure takes into account three factors: (1) the extent to which there are effective legislative veto points; (2) the extent to which these veto points are controlled by different parties from the executive s; and (3) the extent to which the majority controlling each veto point is cohesive. The measure therefore contains information not only about institutional veto points, but also about the extent to which those are controlled by opposition groups, which is crucial to testing my hypotheses. The measure is continuous, with possible values ranging from 0 to 1. The largest values are given to country-years that feature effective, cohesive legislatures with divergent preferences from those of the executive. I use this measure to test whether countries with domestic veto players are more likely to participate in the creation of new TGIs and expect a positive sign on the coefficient of the variable Domestic veto players it. To test hypothesis 3 and 4, I include interactions between Regime type it, Democracy it, and Autocracy it respectively and Domestic veto players it in my models. I expect the interactions between Regime type it and Democracy it and Domestic veto players it to have a negative sign and the interaction between Autocracy it and Domestic veto players it to be statistically insignificant. In addition to the main independent variables, I control for a range of other possible driving forces of states participation in TGIs. I include a variable that captures the number of TGIs in wich a country 15

16 participates in a given year t 1. I include this variable because a state that is already involved in many TGIs might have less opportunities, resources, and incentives to join new TGIs. Alternatively, it could also be that there are positive spill over effects across institutions that result in an increased likelihood of states joining new TGIs if they are already part of other TGIs. Table 1: Summary statistics: Dependent and independent variables Mean Variance Std. dev. Min. Max. N TGI formed it ,905 Domestic veto players it ,756 Regime type it ,875 Democracy it ,875 Autocracy it ,875 TGI participations it ,905 Log GDP it ,921 Log GDP per capita it ,662 Civil law it ,389 Common law it ,389 Islamic law it ,389 Parliamentary system it ,049 Presidential system it ,049 Western hemisphere it ,117 Europe it ,117 Africa it ,117 Middle east it ,117 Asia it ,117 Furthermore, I include variables that capture a country s GDP and GDP per capita to control for economic power as well as economic development as factors that shape state participation in TGIs. Scholars have argued that the economic power and development of states shapes their incentives to use informal modes of global governance (Stone, 2011; Vabulas and Snidal, 2013). Data on GDP and GDP per capita come from Gleditsch (2002) and the World Bank. I use the natural logarithm of GDP and GDP per capita. I also account for a country s domestic legal tradition. Countries with different legal traditions at the domestic level may differ in their international cooperation behavior as well as their preferences regarding the design of international cooperation (Simmons, 2009; Mitchell and Powell, 2011). The three indicator variables Civil law it, Common law it, and Islamic law it capture the domestic legal tradition of a state and are based on data from Mitchell, Ring and Spellman (2013). I also control for a country s political 16

17 system being a parliamentary or presidential system. The variable Parliamentary system it is coded 1 for countries with a parliamentary system, and 0 otherwise. Similarly, the variable Presidential system it is code 1 for countries with a presidential system, and 0 otherwise. Data on countries political systems comes from the World Bank (Keefer, 2012). Finally, I include regional fixed effects to account for unobserved heterogeneity among states of different world regions as well as a lagged dependent variable to account for serial correlation. All independent variables are lagged by one year. Table 1 provides summary statistics for all variables used in the main empirical analysis. Table A-II in the appendix reports the correlations between all independent variables used in the main analysis. Results The models in table 2 are estimated using logistic regression. As we can see across model specifications, countries with more democratic political regimes are more likely to participate in the creation of a new TGI in a given year. This relationship is strongly statistically significant across models. We observe the same relationship between the domestic veto player configuration in a country and its probability to participate in the creation of a new TGI. The more a government is confronted with institutionally strong veto players with policy preferences different from its own, the more likely it is to engage in transnational public-private governance initiatives. Also this positive relationship is consistently statistically different from 0 across model specifications and estimation techniques. Furthermore, there is strong support for the negative interaction between the level of democracy of a country and the effect of the strength of domestic veto players on state participation in the creation of a new TGI. For more democratic countries, the positive effect of domestic veto players on their likelihood to participate in the creation of a new TGI is less pronounced than for less democratic countries. The coefficients of non-linear models are hard to interpret directly. I therefore compute predicted probabilities for state participation in the creation of new TGIs for different levels of regime type, domestic veto player constraints, as well as for the interaction between the level of democracy of a country and domestic veto player strength. As we can see in the upper left-hand panel of figure 4, as the extent to which a country is a democracy increases, its likelihood of participating in the creation of a new TGI increases. This effect is not only statistically significant, but also of a considerable substantive size. Ev- 17

18 Table 2: The effects of regime type and domestic veto players on state participation in TGIs Model 1 Model 2 Model 3 Model 4 Domestic veto players it (0.341) (0.379) (0.301) (0.333) Regime type it (0.0141) (0.0108) (0.0112) (0.0165) Domestic veto players it political regime type it (0.0440) (0.0380) (0.0386) (0.0474) TGI formed it (0.0862) (0.0955) (0.0920) (0.0693) TGI participations it (0.0109) ( ) (0.0105) Log GDP it (0.0362) (0.0237) (0.0290) (0.273) Log GDP per capita it (0.0864) (0.0652) (0.0582) (0.195) Civil law it (0.208) (0.175) (0.160) Common law it (0.202) (0.192) (0.173) Islamic law it (0.287) (0.190) (0.207) Parliamentary system it (0.244) (0.177) (0.175) (0.276) Presidential system it (0.195) (0.149) (0.143) (0.235) Western hemisphere it (0.205) (0.312) (0.271) Europe it (0.214) (0.315) (0.261) Africa it (0.236) (0.345) (0.270) Middle east it (0.316) (0.359) (0.312) Asia it (0.234) (0.357) (0.265) Constant (1.026) (0.887) (0.749) Wald χ N 4, 073 4, 073 4, 064 5, 036 Robust standard errors in parentheses. All significance tests two-tailed. Model 3 is a rare events logit model (King and Zeng, 2001). Model 4 includes country fixed effects. Time-invariant effects are dropped. The standard errors in model 4 are based on 5,000 bootstrap iterations. + p < 0.10, p < 0.05, p < 0.01, p <

19 Figure 4: Predicted probabilities of TGI participation Pr(TGI participation) Pr(TGI participation) Regime type Domestic veto players Domestic veto players effect on Pr(TGI participation) Regime type Note: Calculations of predicted probabilities based on model 1 in table 2. All other variables held constant at their means. 95% confidence intervals. 19

20 erything else being equal, countries with low levels of democracy are estimated to have a probability of joining in the founding of a new TGI of about 0.2 in a given year. By contrast, governments of more democratic countries are estimated to have about 0.4 probability to participate in the creation of a new TGI. The graph in the upper right-hand panel of figure 4 shows a similar effect of domestic veto players on countries likelihood of joining in the creation of a new TGI. Finally, the plot at the bottom of figure 4 indicates that the positive effect of domestic veto players on a country s TGI participation differs across regime type. While we have a strong positive relationship between domestic veto players and the likelihood of TGI participation for non-democracies and low levels of democracy, the relationship weakens and becomes eventually statistically insignificant for more democratic countries. Together. these findings lend support to hypotheses 1, 2, and 4. We also observe consistently positive and mostly statistically significant relationships between GDP and civil law tradition, on the one hand, and the likelihood of a country to participate in a TGI, on the other hand. The positive association between GDP as a measure of economic power and TGI participation is in line with arguments about power and informal global governance that suggest that informal governance is a strategically valuable instrument of powerful countries since it increases the return on power resources and provides powerful countries with increased opportunities to shape institutional outcomes (Stone, 2011; Westerwinter, 2017b). Moreover, countries whose legal systems are based on a civil law tradition tend also to be more involved in TGIs. Other domestic legal traditions as well as different types of political systems are not statistically significantly related to states participation in TGIs. Finally, we observe that countries that have already participated in TGIs in the past are also more likely to participate in the creation of new TGIs in a given year. This suggests that state involvement in several transnational governance initiatives facilitates rather than constrains involvement in new governance initiatives. The models in table 3 empirically explore hypothesis 3 by including interaction effects of the domestic veto player variable with indicator variables for democratic and autocratic countries. As can be seen in table 3, while the effect of domestic veto players on a country s probability to participate in the creation of a new TGI is as expected negative and statistically significant for democracies, this effect is positive and, more important, statistically insignificant for autocracies. Together with the positive and statistically significant effect of domestic veto players, this suggests that the effect of domestic veto players on state 20

21 TGI participation is overall positive. However, for autocracies the effect of domestic veto players on state TGI participation is not distinguishable from 0, while it is positive and statistically significant for democracies. This supports hypothesis 3 and the argument that if confronted with strong domestic veto players with policy preferences different from the preferences of the government, democracies have more incentives to engage in global governance by informal means than autocracies. While for democracies informal governance provides strategic advantages in dealing with domestic veto players, autocracies are not constrained by their domestic political system to the same extent and therefore have fewer or no reasons to consider cooperationg informally with other states since they can bypass domestic opposition using alternative means. Finally, as model 6 in table 3 shows, being an autorcacy is strongly negatively related to a country s likelihood of participating in the creation of a new TGI. This may reflect that autocracies have not only fewer reasons to cooperate via TGIs, but also try to avoid interacting with private actors, including NGOs and NGO coalitions, at the international level. As autocratic governments often constrain the involvement and participation rights of civil society actors at home, they also seem to avoid institutional arrangements at the global level in which representatives of civil society organizations participate in the governance process. I also estimated a series of alternative models to check the robustness of my results. First, I estimated model 1 in table 2 using alternative operationalizations of domestic veto players (Mansfield, Milner and Pevehouse, 2007; Mansfield and Milner, 2012). As table A-III in the appendix shows, independent of whether I use different operationalizations of the Polcon iii measure or alternative measures of domestic veto players, the results do not differ from those presented in my main analysis. Second, in order to address the potential role of time dependencies in states decisions to participate in the creation of new TGIs, I estimate models that include years since the last participation in the creation of a new TGI including splines as well as time polynomials of years since the last participation in the creation of a new TGI (Beck, Katz and Tucker, 1998; Carter and Signorino, 2010). Table A-IV in the appendix reports the results from these estimations. The results obtained from these models are consistent with those obtained in my main analysis. Third, I estimate models with the number of TGIs in the creation of which a country participated in in a given year as the dependent variable. Given that the distribution of this count 21