In the wake of the global financial and the European debt crisis, the recognition that credit rating agencies (CRAs) play a critical role in guiding the allocation of capital in financial markets has moved beyond academic circles to enter broader public de-bates about the state-of-affairs and the future shape of the global and European financial systems (see Langohr/Langohr 2008). Despite the growing interest in CRAs' power, authority and lack of accountability (see Bolton et al. 2012; Bruner/Abdelal 2005; Gartner et al. 2011; Gras 2003; Kerwer 2005; Levich et al. 2002; Nolke 2004; Rona-Tas/HiB 2011; Sinclair 2001, 2005; Strulik 2007) as well as in the regulation of CRAs (Andrieu 2010; Dittrich 2007; HiB/Nagel 2012; Pagliari 2012), little attention has been paid to how and why states have bolstered CRAs' influence and authority through the public regulatory use of private credit ratings (but see HiB/Nagel 2012: 49-50, 72-73; Partnoy 2006, 2009; White 2010).
Credit ratings have been used by public regulatory bodies, inter alia, to increase the risk sensitivity of investment restrictions for certain financial institutions (e.g. banks, pension funds), to define differential disclosure requirements for issuers of rated bonds with reference to the rating obtained from CRAs, and to adjust capital reserve requirements for banks and institutional broker-dealers to their credit risk exposure (Kerwer 2002: 304; Sinclair 2005: 42-4). Ratings-dependent regulation further boosted CRAs' power in financial market governance by creating additional demand for credit ratings (Partnoy 2006), authorizing CRAs' role as de facto gate keepers to financial markets and, most specifically, publicly enforcing CRAs' private standard of credit-worthiness. In the absence of regulatory recognition, CRAs' standard of credit-worthiness was "merely" based on CRAs' reputation as experts and their ensuing legitimacy in the eyes of other financial market actors (Nolke 2004: 163-4; see Nolke/Perry 2007). The regulatory use of credit ratings for public purposes did not only reinforce the expert authority and the perceived reliability and legitimacy of CRAs; it also made it mandatory for financial market actors to observe CRAs' standard of credit-worthiness (Kerwer 2002: 303-4; Sinclair 2005: 46; see Brunsson/Jacobsson 2002: 134). Thus, by relying on credit ratings, public regulators delegated governance tasks--i.e. risk measurement for the purpose of flexible risk-sensitive regulation--and (quasi-)regulatory authority to private expert agents, namely CRAs. (1) Once their ratings were used in financial regulation, CRAs were turned into publicly authorized agents of the principal "public regulator" and, through their credit-risk assessment activities, performed a governance task on behalf of public regulators (see Hawkins et al. 2006; White 2010)--without facing adequate public oversight and control.
While in the two decades before the global financial crisis the regulatory use of credit ratings spread from the US across OECD countries and finally also reached reluctant Germany, there continued to be remarkable differences between US and Continental European regulatory systems concerning the quantitative and qualitative extent to which financial market regulators and supervisors came to rely on CRAs' credit risk assessments (Richter 2008; Rosenbaum 2009: 22-3). In this article I draw on resource dependence theory (RDT)--supplemented by insights from principal-agent theory (PAT) and the varieties of capitalism (VoC) approach--to explain both common trends and marked differences in the regulatory use of credit ratings in the US and Germany before the global financial crisis. I argue that, through the regulatory use of credit ratings, state actors delegated (quasi-) regulatory authority to private expert agents, i.e. CRAs, because public regulators lacked essential analytical resources to cope with financial market uncertainty. The higher the degree of public regulators' dependence on CRAs' analytical resources, the higher would be the degree of public regulators' use of CRAs' ratings in regulation. Thus regulatory differences between Germany and the US can be explained with different levels of US and German regulators' dependence on CRAs' analytical resources. The different extents of public dependence on private expertise were, in turn, conditioned by different macroinstitutional settings, i.e. varieties of capitalism. An Anglo-Saxon variety of capitalism (liberal market economy) creates pressures for a considerably higher degree of regulatory reliance on CRAs (private expert agents) than a "Rhenish" variety of capitalism (coordinated market economy).
This article makes several contributions to the literature. First of all, it provides the first systematic and theory-guided comparative analysis of the regulatory use of credit ratings in Germany and the US, thus explaining an important, but still under-researched and, in particular, under-theorized facet of CRAs' power. Secondly, my analysis of CRAs points to, and explains, both (limited) convergence and continued differences in US and German financial market governance. Thirdly, the proposed theoretical framework, which centrally builds on resource dependence theory (RDT), enhances the analytical rigor of extant accounts of expertise-based delegation and states' capacity gaps in financial market governance (see section 2). The empirical analysis demonstrates the usefulness of RDT for explaining varying levels of delegation to private experts. In fact, the macroinstitutionally embedded resource-dependence perspective of this article should be helpful for capturing transatlantic commonalities and differences in the delegation of political authority to private actors far beyond the case of CRAs. Finally, the analysis underlines in one more issue-area the usefulness of the varieties of capitalism approach when it comes to capturing transatlantic differences in financial market governance (see Nolke/Perry 2007).
The remainder of this article is structured as follows: Section 2 introduces the theoretical framework for analysis in more detail and derives a specific testable hypothesis. Section 3 provides a comparative overview of the use of credit ratings by US and German regulatory bodies. In section 4 I show that both intertemporal and cross-country (US and German) patterns in the regulatory use of credit ratings correspond to the expectations of the proposed theoretical framework. Section 5 draws on this framework to account for US and German positions in the negotiations that led to the inclusion of provisions for ratings-dependent banking regulation into the 2004 Basel II Accord. The conclusion (section 6) sums up the main findings, discusses the broader relevance and limitations of these findings and briefly reflects on some regulatory changes after the global financial crisis.
2 Explaining the Transfer of (Quasi-)Regulatory Authority: Resource Dependence and Macroinstitutional Context
2.1 A Resource-Dependence View on Delegation to Expert Agents
In general terms, a lack of policy-relevant information and expertise on the part of state actors has commonly been considered a key rationale for the delegation of governance tasks and political authority to specialized--national, international or transnational--expert agents (Hawkins et al. 2006: 13-15; Pollack 2003: 23, 28-9; Thatcher/Stone Sweet 2002). However, scholars in the tradition of principal-agent theory (PAT) have so far failed to explicate and operationalize the resource dependence argument which implicitly underlies the notion of delegation to a specialized agent due to states' limitations in expertise and informational resources. In the following, I seek to show that resource dependence theory (RDT) advances our understanding of delegation to private actors in that it provides a clear-cut, empirically testable hypothesis specifying why, under what conditions and to what extent public regulators will delegate (quasi-)regulatory authority to private expert agents such as CRAs. Taking recourse to RDT ameliorates (self-declared) shortcomings of PA approaches that are deficient in terms of causal-theoretical saturation (Hawkins et al. 2006: 9-10; Thatcher/Stone Sweet 2002: 3). (2) Moreover, it puts wide-spread, but often under-theorized ad hoc arguments about missing state capacities for effective financial market governance (Eichengreen 1997; Major 2009; Speyer 2006; Strange 1988; Tsingou 2008; see Mayntz 2010 for a recent more nuanced view) on a theoretical footing that allows for systematic comparative analysis over time and across countries. My resource dependence argument thus enhances existing underspecified explanations for delegation and theorizes claims about (varyingly large) capacity gaps of states.
RDT, which emerged as a major approach of interorganizational analysis in the late 1970s but was (re-)discovered by political scientists only quite recently (see Bruhl 2003; Kruck 2011; Liese 2009; Nolke 2004; Rittberger et al. 2010: 332-9; Rittberger/Kruck 2010; Steffek 2013), is based on a view of organizations as rational self-interested actors that are oriented toward the effective and efficient attainment of specific organizational goals (Pfeffer/Salancik 1978: 23; Scott 1981: 57). The underlying rationale for establishing relationships with other organizations is an organization's lack of self-sufficiency, i.e. its need for access to specific external resources which are crucial to achieve its particular objectives. This may include material resources, e.g. funds, technical material, and personnel, as well as immaterial resources, e.g. information, expertise, and legitimacy. Thus, organizations which, for the accomplishment of their organizational goals, depend on (access to) scarce resources controlled by another organization in their task environment will be prone to establish relationships with this...