Enriching the RIT Framework

Regulation as a Normative Enterprise

Regulation is an inherently normative enterprise; introducing intermediaries makes the normative issues both more vivid and more complicated. The standard R → T model does not usually problematize its relatively simple normative structure: R represents the public interest, and its task is to shape T’s behavior to better serve that interest. To be sure, the “public interest” (like any collective interest) is a messy concept that is hard to define and operationalize, but the RT model largely brackets those complexities, focusing on the practical problems involved in R controlling T to attain R’s public interest objectives.

While R → I → T regulation can also be viewed as a technical matter (with the objectives simply being R’s goals), the inclusion of an intermediary opens up more normative space. Some intermediaries are technical practitioners (e.g., lawyers, accountants) with no intrinsic interest in the substantive regulatory problem. Yet they promote forms of regulation that require and facilitate their services, and these “technical” details may have important policy and legitimacy implications. Other intermediaries (such as human rights and environmental nongovernmental organizations [NGOs]) seek involvement precisely because they want to affect the character of public regulation itself. But even intermediaries with “noble” objectives may not reflect all social values. Consider the extreme environmentalist who seeks to impose strict environmental regulations. Such regulations might in fact be socially inefficient if the costs of compliance are seen as excessive, not only by the targets, but also by citizen beneficiaries who would prefer less stringent regulations that allow for the production of cheaper goods.

While these interests may make intermediaries helpful allies for R, they also expose R to pressure from intermediaries (and perhaps to indirect pressure from targets transmitted via the intermediaries) over which policies, forms, and levels of regulation to pursue. As we discuss below, intermediaries (not merely targets) might even capture the regulator to pursue their own private interests. These conflicts between intermediaries and regulators, even more than the unsurprising conflicts between regulators and targets, show that the very concept of “public interest”—and therefore of appropriate regulation—is contested.

In short, the inclusion of intermediaries in the regulatory framework raises new normative issues and makes the importance of familiar ones all the more salient. How we evaluate regulatory outcomes depends both on our own values and on the values and goals of the actors that we are studying. RIT does not provide answers to these normative questions, but it can provide useful guidance for evaluating them.

The Diversity of Regulatory Actors

The empirical articles in this volume apply the RIT model to regulation across an extraordinarily broad set of fields, ranging from sustainable fisheries to medical devices, from transnational production to international crime and beyond. Viewing them collectively, the cases demonstrate the value-added of explicitly incorporating intermediaries into the study of regulation.

Given this diversity of fields, it is hardly surprising that we also find a great diversity among the Rs, Is, and Ts involved in the cases. The regulators include international governmental organizations (IGOs), nation-states, national regulatory agencies, and private schemes. The intermediaries may be NGOs, such as Amnesty International or Human Rights Watch; courts, such as the International Criminal Court; government agencies, such as the Food and Drug Administration; private agencies, such as Standard and Poor’s; or private organizations, such as the U.S. Green Building Council (USGBC). The targets include pharmaceutical firms, nation-states, carpet exporters, and even beneficiaries such as the carpet weavers themselves. This diversity is enhanced because for any given field there may be multiple regulators, intermediaries, and/or targets, with varying interactions both within and across these actor categories.

Intermediaries play an equally diverse set of roles in regulation. The RIT framework identifies four key theoretical categories of capacities that regulators seek to acquire through intermediaries (operational capacity, expertise, independence, and legitimacy), and the cases put empirical meat on them. These capacities translate into a wide range of intermediary roles, from gathering information and assisting implementation to clarifying norms, from monitoring to verifying compliance, from enforcing to providing feedback and advice to regulators.

One thing that comes through clearly in the cases is how significant intermediaries are for the ongoing interpretation and further development of regulations. Graeme Auld and Stefan Renckens showed in their contribution that intermediaries often acquire superior information that provides the basis for feedback about how rules work in practice; that information is valuable for facilitating and monitoring target compliance and, thereby, valuable for rule reformulation. Jean-Pierre Galland discussed how, as auditors gain expertise and influence, they advise regulators and accreditation bodies and provide best practice guides for third-party certifiers. These functions can be very important, and indeed Auld and Renckens argue that intermediaries are often coregulators in the early stages of regulatory schemes.

Thus, intermediaries may play very different roles, and these may change and develop over time. In addition to sometimes acting like regulators, at other times intermediaries can be targets, as when their activities as monitors are themselves monitored by other intermediaries. Moreover, the boundaries among R, I, and T are not always clear-cut; the roles overlap and interact. One of the values of the RIT framework is that it is a guide to examining the range of actors that participate in any regulatory process and their various interactions.

Mathias Koenig-Archibugi and Kate Macdonald made the valuable move of incorporating the beneficiaries of regulation into the analysis and examining their shifting actor roles. Beneficiaries may be targets (as when regulation aims to solve their collective action problem); at other times beneficiaries may carry out intermediary functions (as when they serve as fire alarms monitoring target behavior); and at still other times beneficiaries play the role of regulators (as producer organizations have direct representation on the Fairtrade Standards Committee). Tetty Havinga and Paul Verbruggen made the broader point that many intermediaries have a “chameleonic” character, enabling them to change “color” (that is, regulatory role) in different circumstances.

As discussed above, the goals of intermediaries are a key determinant of how they fulfill the various roles that they are given (or take) in the regulatory process. While the regulator’s ideal intermediary might be a neutral party that simply implemented the regulator’s policies, intermediaries are also agents with substantive and organizational objectives. Just as the character and roles of intermediaries vary widely, so too do their goals. Some are professionals who intermediate as a job; others are activists who do so as a mission; others may be obstructionists who do so to frustrate regulatory intent. This variation can be appreciated by considering the many ways in which intermediaries enter the regulatory process.

Who Selects the Intermediary?

Once we bring an intermediary into the analysis, the question of who selects the intermediary, and for what purposes, becomes important. In the traditional R → T model, the initial presumption is that the regulator enforces the regulation. Principal-agent (P-A) variations recognize that the regulator needs to select an intermediary (agent) to operate on its behalf in implementing, monitoring, enforcing, and evaluating the regulation. The emphasis in P-A then shifts to the regulator’s difficulties in monitoring its intermediary when there is informational asymmetry, that is, where the regulator cannot fully observe whether the intermediary is doing its job properly. Such asymmetry creates incentives for the agent to shirk its duties or pursue its own private objectives. Regardless of how well the regulator can manage these problems, the presumption is that it chooses the intermediary.

In fact, however, the regulator often leaves it to the target to select the intermediary. This might be because the target is better positioned to choose an intermediary with the right expertise; it might instead be a way for the regulator to placate the target (and perhaps fool third-party audiences) by ensuring that the intermediary will not be overly zealous. This is very common when states are the targets and do not want international regulation to be overly intrusive on their national sovereignty. Tom Pegram pointed out in his article, for example, that allowing states to designate their own national intermediaries creates a significant risk that they will not properly implement the Optional Protocol to the Convention against Torture.

Allowing targets to select the intermediary is equally common at the domestic level and in private transnational regulation: target firms often “intermediary shop,” picking their own auditors, credit rating agencies, and other intermediaries. This creates an implicit competition among intermediaries to please the targets to get and retain their business; it likely undermines regulatory intent. Moreover, even if the target does not choose the intermediary, the same effect may occur where intermediaries depend on targets for their remuneration, as is common with private intermediaries. Intermediaries also often must have extensive ongoing relations with targets and need their cooperation to perform their tasks. For all these reasons, the intermediary may become increasingly close to and reliant on the target over time, which raises thorny questions about intermediary independence and therefore the effectiveness of regulation.

In other cases, intermediaries “self-select,” inserting themselves into the regulatory process. They may do so to influence the policy goals being sought, either to assist the regulator in successfully regulating the target, or because they want to shape or shift the content of the regulation itself. They may also do so because it serves their organizational goals (although they will typically present themselves as enhancing regulatory quality). In their article, Axel Marx and Jan Wouters showed how some actors gain legitimacy by working as intermediaries for well-regarded regulators such as the International Labour Organization. In many cases, however, intermediaries’ primary goal is remuneration, as with lawyers, accountants, or credit rating agencies. The temptation to promote lax regulation to keep and attract business may be very high in these cases, although even here intermediary behavior may be tempered by professional standards and codes.

In sum, the processes by which intermediaries are selected constitute an important source of regulatory variation. The selection process can produce intermediaries that pursue the regulator’s interests, are more closely allied with targets, or act in ways that mainly serve their own interests. To understand the impact of intermediation, then, we need to understand from where intermediaries come.

RIT Pathologies

The different types, goals, and origins of intermediaries open up a wide range of regulatory “pathologies.” From the regulator’s perspective, an R-I-T relationship may appear pathological whenever regulatory outcomes deviate from that of its idealized regulatory setting, in which the intermediary simply mediates between the regulator’s intentions and the target’s behavior. That is roughly the story in Nicole De Silva’s analysis in this volume of how the International Criminal Court (ICC) uses NGOs as intermediaries. Early in the life of the ICC, NGOs largely shared the court’s objectives; there was no major goal divergence. Later, however, NGOs developed somewhat different goals, and became increasingly critical of many facets of ICC operations. The ICC came to see its “intermediaries as liabilities,” forcing it to rethink its relations with NGOs.

Other pathologies result in even greater deviations from successful regulation in the public interest. The framework article at the start of the volume discusses some of these in largely theoretical terms; here we highlight some empirical instances that emerge in the cases. The basic problem of regulatory capture, in which the target turns the regulator to its own ends, is well known. The RIT model incorporates this possibility, but adds nuance by showing that additional avenues for capture exist when intermediaries are present.

One important possibility is that an intermediary may provide another point of access for the target to gain control of the regulatory process. Jeroen van der Heijden illustrated this through the case of the carpet industry in his article; after failing to persuade government regulators to accept its (weak) voluntary Green Label Standard, targets worked through USGBC, an intermediary. USGBC incorporated the standard into the requirements for LEED (Leadership in Energy and Environmental Design) certification, which is accepted by many local, state, and national regulators. Tellingly, many USGBC board members are involved in the construction trades and, at the time of adoption, the chairman of its board was employed by a multinational carpet company. Other sectors whose products have not been included in LEED requirements have lobbied against government recognition of LEED certification, so as not to give competitors an advantage. Thus, the choice of intermediary has become a battlefield for regulatory contestation.

Another important form of capture highlighted by the RIT model is capture of the regulator by the intermediary. The consequences may sometimes seem secondary, as when auditors press for management-based standards that they can audit uniformly and cheaply. But in some cases, such regulatory decisions favor targets, lessening regulatory impact. At other times, intermediaries capture the regulator primarily to expand their business. In his article, Jean-Pierre Galland shows how the private auditor Bureau Veritas has used its expertise on complicated regulatory issues, and its multinational connections, to work with, through, and sometimes against the French government to secure a position in the broader European and international regulatory regimes.

Such activities may go beyond competing for existing regulatory business to embedding private actors in the regulatory system. Andreas Kruck showed how the regulatory use of private credit rating agencies, such as Moody’s or S&P, has entrenched those agencies as key players in the financial system, generating high profits and lucrative related business. The oligopolistic structure of intermediaries in this sector has resulted in flawed credit ratings and a lack of transparency, which do not serve either the regulator’s goals or the public interest. Even when regulators recognize the problem (as in the 2010 Dodd-Frank Act), they face a “selection dilemma” that leaves them too reliant on a few expert intermediaries to fundamentally change the system.

This reminds us that intermediation is not always an unmitigated improvement to regulation. Jeroen van der Heijden provided a further illustration of the “darker side” of intermediation, where intermediaries pursue their private interests and largely ignore the public interest. Intermediaries such as the German (DIN) or Netherlands Institute for Standardization (NEN) push not only for privatization of regulation, but also for the creation of new, more complex, and frequently revised regulations, primarily to expand the market for their services, domestically and internationally. As with credit rating agencies, when regulators come to rely heavily on intermediaries, they may not be able to properly control them. Canadian local councils have discovered that they cannot discipline misbehavior among private building inspectors without stalling construction in their localities. Here, ironically, private behavior has become the bane of public regulation.

Limiting Capture through Regulatory Design

Fortunately, the empirical articles in this volume provide important guidance as to how capture and other pathologies can be mitigated through better regulatory design. One important technique is for regulators to require the use of accreditation bodies to monitor and ensure the competence of third-party certifiers, so as to prevent a race to the bottom as certifiers seek to please the firms that pay them (see Galland). Allison Loconto showed how private transnational regulators of sustainable agriculture carefully select intermediaries that are committed to following its rules and then rely on diverse government and private accreditors to maintain the credibility of auditing, attestation, and determination of compliance. In his piece, Timothy Lytton described how accreditors rely on the peer review procedures of a meta-accreditor—the International Accreditation Forum—to ensure their individual and collective reliability.

Pegram suggested a different approach: regulatory stewardship, in which intermediaries monitor one another (and the governments and other actors that create or empower them) to prevent capture. He outlined a set of regulatory rules and intermediary practices that would make this approach effective. Alternatively, beneficiaries may be brought into the regulatory scheme as fire alarm monitors of target performance (see also Koenig-Archibugi and Macdonald).

Of course, effective regulatory design depends on the precise circumstances of the RIT actors involved. Martino Maggetti, Christian Ewert, and Philipp Trein discussed how institutional designs must take into account the potential ability of the target to capture the intermediary. When an industry has a high potential for capture, the regulator will need to structure intermediary relationships and monitor intermediaries carefully, to avoid regulatory slippage. De Silva similarly argued that when an intermediary (or secondary intermediary) has divergent goals, then the regulator (or primary intermediary) should use delegation rather than orchestration to manage their relationship more strongly.

Adding Complexity to RIT

The RIT model is deliberately very simple, although the framework article suggested several possible extensions, many of which were informed by earlier drafts of the articles in this volume. Here we consider the complexities that time, chains of actors, and networks introduce before concluding on the virtues and limits of a simple RIT model.

Regulation is an inherently dynamic process, unfolding as actors learn about their circumstances, relationships, and outcomes, and adapt accordingly. Auld and Renckens demonstrated this clearly with their analysis of the Marine Stewardship Council. Its initial rules and procedures were not highly developed, leaving auditors wide discretion to “translate” broad rules into ones more applicable to specific targets. Indeed, in that case auditor actions went beyond merely “translating” to developing regulatory substance and procedures; the intermediaries effectively served as coregulators. Kruck examined the dynamic process of financial regulation, highlighting the path dependence inherent in financial regulators’ continued and deepening reliance on the expertise of a small number of credit rating agencies over time. This dependence became progressively normalized, while regulators’ capacity for direct regulation atrophied. Thus, even after the major failures of credit rating agencies in the run-up to the financial crisis, regulators found it difficult to alter or fully control their relation with their intermediaries.

A number of articles illustrated chains of regulators and intermediaries; some used the language of primary and secondary regulators and intermediaries to indicate their sequential relationship. Marx and Wouters showed the prevalence of regulatory chains in global labor governance, even where the primary regulator (the ILO) has not delegated authority to secondary regulators. De Silva showed how the ICC, as primary intermediary, engages NGOs as secondary intermediaries to increase its capacity and influence, vis-à-vis both its criminal targets and its state principals. In each case, actor chains enhance regulation, as different links in the chain provide different capabilities needed for effective regulation.

Havinga and Verbruggen drew an important distinction between serial and parallel configurations of intermediaries, illustrating their different implications for food safety regulation. They argued that serial configurations entail greater dependence among intermediaries, whereas parallel arrangements allow intermediaries to act independently of one another. Havinga and Verbruggen found a mixture of the two configurations in various areas of EU food regulation, creating significant regulatory complexity.

Several of the articles also introduced actor networks. Jacint Jordana looked at how regional transgovernmental networks in the banking industry intermediate between global and local regulators. Because they are horizontal and weakly institutionalized, networks allow for “collaborative intermediation” that does not threaten state sovereignty. Lytton used networks to analyze food regulation, arguing that a network approach has an important advantage in not privileging government over private regulation. Specific features of networks also may be important; for example, higher density networks may increase transparency and oversight.

Methodologically Dealing with Regulatory Complexity

Dynamics, chains, and networks all entail increased complexity of regulatory regimes. Two articles in this volume proposed more complicated models to capture this institutional multiplicity. Havinga and Verbruggen highlighted the multiple “configurations” of RIT actors that occur in complex, hybrid governance and proposed a polycentric framework that focuses on the “enrollment” of regulatory actors. Lytton argued that networks provide the best approach to understanding complex governance systems.

We agree that these two approaches provide valuable complementary perspectives on regulatory regimes, yet we view the RIT model as a desirable starting point, for methodological and normative reasons. Methodologically, RIT corresponds to the initial perspective of any aspiring regulator—from a state to an NGO—seeking to regulate particular targets. This perspective is essential to understand even if the regulator must also take into account interactions with other regulators or chains. It also provides a basis on which to incorporate the complexity called for by Havinga and Verbruggen and by Lytton. By decomposing regulatory regimes into their constituent actors and relationships, the RIT approach helps us to analyze their composition and problematize their operations. This does not deny the virtues of alternative systemic approaches: every model provides only a partial view of the world, and triangulation of the sort proposed provides a valuable complement to the RIT approach.

Normatively, relying on polycentric enrollment and networks supposes that systems will self-organize in desirable ways. This may be true for evolutionary or other reasons, although those justifications need to be explored. The irony, at least in market settings, is that the very purpose of regulation is to address circumstances where systems (frequently markets) fail to move naturally toward desirably self-organized outcomes. While RIT regulation can be abused (as through capture), introducing intermediaries also opens up the possibility of improving outcomes (in terms of particular regulatory goals). And while the RIT model may fail to capture all aspects of systemic complexity, it provides guidance on improving system performance, by helping us better to understand and manage the goals, roles, and impacts of intermediaries.

Unavoidably, writing a conclusion gives one the last word in a volume. But that is not our intention. We believe that the volume has shown how explicit attention to intermediaries provides an important way to improve and extend the analysis of regulation. Rather than settling questions, however, our hope is that we and the contributors have opened up significant new debates.