ENCUESTA DIRIGIDA A REPRESENTANTES LEGALES TABLA

Theories of international politics have often conceived of power as arising from the structure of relationships in the global system, but measures of power have been limited to monadic or dyadic characteristics of states. This paper argues that the structure can

be operationalized in terms of statistics describing prominence in complex networks, and provides an application to the global banking system. Besides offering precision, such an operationalization offers novel conclusions regarding the performance of the system in the wake of the most severe financial crisis since the 1930s: despite being the epicenter of the crisis, the United States has increased in prestige according to some measures and remained at the core in others. European financial centers have declined by most measures, while the rest of the countries have remained peripheral. Notably, this includes emerging markets like the BRICS – Brazil, Russia, India, China, South Africa – which have not increased in prominence in global banking according to these metrics.

The complex network conceptualization of power can be applied to substantive areas other than banking. Strange (1987) conceived of structural power as existing in four areas: finance, production and exchange, security, and knowledge generation. Future analyses may be able to demonstrate whether American preeminence in banking has extended to other areas of the global financial system as well as these other areas. Such an approach might provide a different picture than metrics now in common usage such as the Composite Index of National Capabilities (CINC) scores, which counter-intuitively suggests that the Soviet Union was more powerful than the United States during the 1980s and that Chinese power eclipsed American power in the late-1990s.

Moreover, a complex network approach provides expectations regarding future perfor- mance of the system. While such dynamics have barely been mentioned here, inferential models of networks allow researchers to formally test whether mechanisms such as pref- erential attachment, triadic closure, or other structural processes exist in networks. Such models are increasingly common in political science, but have yet to be utilized to an- alyze the distribution of power in the system. These models allow for the inclusion of country-level and dyad-level covariates which also impact development and change within

processes in network formation. For this reason, applications of complex network science – especially in its dynamic form – may complement and enrich other conceptualizations of power, and may unify disparate strands of the literature.

This analysis has barely scratched the surface. Many other measures of network promi- nence exist and can be applied to studies of the global political economy and security sys- tem. Among others, such possibilities include analyzing when crises are likely to arise and spread, the conditions under which states move from the periphery to the core, the mani- festations of structural power in bargaining over global governance structures and security situations, and many other topics. By providing an empirically precise and theoretically rich environment in which to study the global political system as a system, a complex network approach can improve our understanding of many phenomena of interest.

2 GLOBAL BANKING AS A COMPLEX POLITICAL ECONOMY

2.1 Introduction

To this point international political economy (IPE) discussion of “the most virulent global financial crisis ever” (Greenspan 2010, 3) has contained almost no analysis of global finance, instead focusing on the pre-crisis innovations of micro-level units and the post- crisis response behaviors of governments. What IPE conversation has occurred has mostly been speculative (Helleiner 2010; Drezner and McNamara 2013; Oatley et al. 2013) or sociological (Cohen 2009b; Keohane 2009; Helleiner 2011; Oatley 2011). The only well- established theoretical framework in the empirical literature which seeks to explain dy- namics at the global level is a “race to the bottom” driven by global competition, which is often presumed rather than empirically demonstrated. This despite the fact that “race to the bottom” theories have been called into serious question or in other ways complicated in many other issue-areas of the global political economy.1 _{Nevertheless, such theories}

remain predominant in the literature (Meseguer and Gilardi 2009).

Such a post-crisis lack of interest in the politics of global finance is at odds with prior theory, which maintains that predominance in world financial markets is advantageous be- cause it confers a broad base of power (Strange 1998; Cohen 2000, 2006) which may be employed to provide advantage for domestic firms through the coercion or neglect of their foreign rivals (Oatley and Nabors 1998; Simmons 2001). Conceptualizations of power

1_{Also in finance. More than a decade ago (Simmons 2001, 590) noted that “[t]heories of ‘races to the}

bottom’ are of little help” in explaining patterns in global capital markets. The recent crisis, which was cen- tered around AAA-rated financial instruments and supposedly low-risk debt of OECD sovereigns, provides support for the Simmons thesis as it indicates that many financial actors believed they were acting safely

in financial systems have been imported either from neo-Marxist structuralism (Strange 1982) or from bargaining models of international relations (Drezner 2007). Both agree that a state’s ability to attract interest from foreign financial firms is a indicator of power, but the former places primary importance on the qualitative nature of the global structure of financial relationships while the latter emphasizes quantitative attributes such as the im- portance of the size of internal markets in bargaining conducted between major powers. Thus, in existing literature there is a tension between the complex and the parsimonious, the accurate and the precise.2 _{Perhaps it is this tension which has led to the relative lack of}

post-crisis attention from IPE scholars.

This paper provides a bridge which may connect the two while providing extensions to both. I argue that power in global banking can be conceptualized as emerging from a complex adaptive system comprised of national units and the linkages between them, theorized in terms of the linkages as well as the units, and quantitatively analyzed using recent inferential network methodologies. This approach creates opportunities for theoret- ical arguments which situate national political economies within a global context in unique ways, and for empirical analyses which uncover patterns in the global banking system that would be obscured using common methodologies. As such, this paper seeks to unify vari- ous strands of the IPE of finance literature while providing extensions to them. Rather than treating the global financial system as an “abstract force ‘out there”’ (Helleiner 2011, 78), it problematizes the system by situating national political economies within a global mar- ket (Cohen 2009b; Palan 2009) which is modeled statistically. The central argument is that global banking is a complex system which emerges through the interaction of processes which occur at multiple strata.3

2_{These divisions roughly correspond with the divisions between the “American” and “British” schools of}

international political economy (Cohen 2008, 2009b; Helleiner 2011).

3_{In this context, “strata” does not correspond with the traditional levels of analysis proposed by Waltz}

I focus on global banking. Following Oatley et al. (2013), I conceptualize the global banking system as a weighted and directed dynamic network. I do this because observed bank relationships comprise an interdependent market which more closely resembles a network than a series of independent and identically distributed monads or dyads, so a network specification is most appropriate for analysis of this kind of system. I employ in- ferential network methodologies to uncover the determinants of the global banking system, which is the outcome variable.4 _{Determinants of this system include country-level po-}

litical institutions, macroeconomic fundamentals, national policies, dyadic relationships, and endogenous structural processes. The relative effect of these variables is analyzed us- ing recently-developed extensions of exponential random graph models (ERGM), which I compare to linear regression models in order to highlight the utility of my approach. The findings suggest that the ability to attract cross-national financial relationships, which con- fers power in the world economy, is not a pure function of monadic or dyadic attributes of states. I show that in some ways the importance of internal market size, key elements of political systems, and shared characteristics (such as regional proximity, similar levels of development, or common political systems) have been overstated by previous studies. The importance of dynamic structural processes such as preferential attachment and triadic closure has been understated or only vaguely identified.