Rethinking epistemic communities twenty years later

Abstract

The concept of epistemic communities – professional networks with authoritative and policy-relevant expertise – is well-known thanks to a 1992 special issue of International Organization. Over the past twenty years, the idea has gained some traction in International Relations scholarship, but has not evolved much beyond its original conceptualisation. Much of the research on epistemic communities has been limited to single case studies in articles, rather than broader comparative works, and has focused narrowly on groups of scientists. As a result, it is often assumed, erroneously, that epistemic communities are only comprised of scientists, and that the utility of the concept for understanding International Relations is quite narrow. Consequently, an otherwise promising approach to transnational networks has become somewhat marginalised over the years. This article revisits the concept of epistemic communities twenty years later and proposes specific innovations to the framework. In an increasingly globalising world, transnational actors are becoming progressively more numerous and influential. Epistemic communities are certainly at the forefront of these trends, and a better understanding of how they form and operate can give us a clear demonstration of how knowledge translates into power.

Introduction

The concept of epistemic communities – professional networks with authoritative and policy-relevant expertise – is well-known thanks to a 1992 special issue of International Organization, entitled ‘Knowledge, Power, and International Policy Coordination’. Over the past twenty years, the idea has gained some traction in International Relations (IR) scholarship, but has not evolved much beyond this original volume. Most of the research on epistemic communities has actually restricted the empirical scope of the concept, focusing narrowly on groups of scientists, and examines single case studies instead of undertaking the broader comparative work that might reveal something new about the nature of epistemic community influence.¹ There have only been a few recent exceptions to this.² As a result, students of IR often assume, erroneously, that epistemic communities are only comprised of scientists or technicians, like environmentalists or economists, and that the utility of the concept is quite limited. An otherwise promising approach to transnational networks in a globalising world has become somewhat marginalised.

In revisiting the concept of epistemic communities twenty years later, I seek to respond to its critics, to clarify the original intentions of the research programme set forth in 1992, and to put forward specific innovations to the framework. The existing literature is somewhat unclear about what kinds of groups constitute epistemic communities, and too narrow when it comes to the types of empirical cases that have been explored. Through this analysis, I make three overarching points. First, I stress the growing importance of these types of actors in an increasingly globalised world. They not only influence governments, but also other non-state actors with decision-making power. Epistemic communities are at the forefront of recognised trends towards transnational governance, and they are a major means by which knowledge translates into power. Second, I argue that we must pay more attention to the internal dynamics within an epistemic community to understand its strength or weakness. Epistemic communities do not simply exist or not exist, but have varying degrees of influence; and to establish this, more comparative research is necessary. I hypothesise that the more internally cohesive an epistemic community, the more likely it will achieve a high degree of influence on policy outcomes. Third, I reconceptualise the framework of epistemic communities to explain why knowledge and uncertainty – the scope conditions for epistemic community influence – should be understood more broadly; how the relationships between governments and epistemic communities are often highly synergistic; and why the central attribute of epistemic communities is their professionalism, which can be measured according to a number of factors.

The article proceeds as follows. In the first section, I address the added value of the approach by considering the broader literature on global governance and other forms of knowledge-based actors and networks. In the second section, I review the emergence and development of the epistemic communities concept, as well as its chief criticisms. The third section clarifies and reconceptualises the framework, and offers some specific innovations to increase its utility and explanatory power.

Transnational global governance

Epistemic communities are significant players in an environment where transnational processes are continuously growing and evolving. Many scholars have studied recent cases of ‘transnationalisation’ in particular areas of global governance, like in the creation of business safety standards for the air, ship, motor vehicle, food, pharmaceutical, and telecommunications industries,³ or in cases of corporate social responsibility and corporate citizenship.⁴ These areas of global governance can lead both to more regulation – in the cases of ozone-depleting substances, chemical safety, oil spills, acid rain, whaling, and so on – as well as to less – for exchange rate controls, tax competition, corporate law, tariffs, and so on.⁵ Djelic and Quack argue that transnational communities of different types impact the global governance of migration, social and political activism, and expertise.⁶ In terms of expertise in global governance, they argue that professional communities, including epistemic communities, are growing in importance through increasing transnationalisation in the context of globalisation.

In this robust literature, it is clear that epistemic communities not only seek to persuade states, but also a wide variety of non-state actors. They are not only underpinning specific government policies, but also shaping governance more broadly.⁷ In effect, they are continuously strengthening the very transnationalism that they represent. Much of the literature on epistemic communities assumes that they must only direct their efforts at states. To a growing extent, however, their target audiences are much broader. Indeed, both states and non-state actors increasingly engage with and listen to transnational non-state actors that have a claim to expertise.⁸

Situating epistemic communities in the broader literature of transnational global governance also highlights the fact that there are many other actors and networks that construct the rules and norms of the international system in addition to epistemic communities. Slaughter's transgovernmental networks of regulators, judges, and legislators,⁹ Keck and Sikkink's transnational advocacy networks,¹⁰ and Adler and Pouliot's communities of practice¹¹ are often alluded to in the field of IR. There are many other examples put forward in other disciplines that illustrate just how important actors like epistemic communities are for shaping an order in which both states and non-state actors are powerful. Such non-state rule-makers include non-governmental and international organisations, multinational corporations, and advocacy coalitions, among others.

Braithwaite and Drahos argue that in the area of global business, states have increasingly become ‘rule-takers rather than rule-makers’.¹² Indeed, they find that transnational policy communities, like the International Conference on Harmonization and the International Telecommunication Union, come up with many rules that legislators vote on after they are already presented as fully-formed laws. Certain powerful corporations like Microsoft, Motorola, and Boeing also work as transnational policy communities, shaping laws for international bodies that then impose these standards on individual states. Braithwaite and Drahos describe this as a process of ‘global privatization of public law’¹³ by these powerful ‘webs of influence’.¹⁴ Tsingou also finds that as financial markets become more transnational, traditional state-based regulation is declining while private, transnational policy communities are gaining more control in decision-making by virtue of their expertise and economic strength.¹⁵

Similarly, Graz and Nölke examine the phenomenon of ‘transnational private governance’, defined as non-state actors engaged in global governance. While they note the limitations of this form of global governance – firms must voluntarily choose to abide by the norms these transnational actors advance – they provide strong evidence of ‘the ability of non-state actors to cooperate across borders in order to establish rules and standards of behaviour accepted as legitimate by agents not involved in their definition’.¹⁶ These transnational policy elites, such as rating agencies, investment banks, consultancy firms, professional associations, trade unions, elite clubs, and trade exchanges, can be either formal or informal, knowledge-based or not, and they are often the enforcers of the rules once they are established. Thus, private transnational policy elites are distinctive from epistemic communities, but demonstrate the importance of actors like them.

Epistemic communities fit well into broader research on the phenomenon of transnational global governance. Non-state actors, whose influence often rests on shared knowledge, are involved in many aspects of transnational governance. They may not always comprise epistemic communities, indeed they may be competitors to them, but it is clear that actors with recognised expertise, shared policy goals, and a willingness to act are becoming increasingly influential.

The study of epistemic communities deserves a significant rethinking. Numerous IR research programmes could potentially make use of this concept and its propositions, such as theories of civil war settlement, alliance formation, identity change, and so on. In an increasingly globalised world with considerable advances in transnational interaction, the value of expertise and knowledge and the networks of professionals who develop and sustain it are ever more apparent. Over the past two decades, research carried out using an epistemic community framework has demonstrated its promise. Scholars have used it to shed light on everything from EU integration to mitigating climate change to dealing with AIDS in Africa. Epistemic communities are likely far more numerous and influential than has been recognised, but research into the weight and scope of their influence has lagged. A reinvigorated epistemic community research programme would do much to enhance our understanding of the successes and failures of international policy formation.

Epistemic communities then and now

Emergence of the concept

Ludwik Fleck's idea of the thought collective in his book, Genesis and Development of a Scientific Fact, provided the seed for the idea of epistemic community, as did Michel Foucault's adaptation of the Greek concept of episteme.¹⁷ Thomas Kuhn, Burkart Holzner, and Ernst Haas played important roles in advancing the idea further.¹⁸ In his 1962 book, Kuhn explored the notion of scientific community, defining it as a group of individuals from a particular discipline whose work revolves around a shared paradigm – a set of beliefs and methodological standards for the pursuit of scientific research. In the field of sociology, Holzner was the first to use the term epistemic community in 1968, while Ernst Haas introduced the concept to IR in order to understand groups of scientists. The latter influenced today's major epistemic community scholars, John Ruggie and Emmanuel Adler, who were Haas's students, as well as Peter Haas, his son.¹⁹

In 1975, Ruggie drew upon Foucault's episteme,²⁰ and broadened the scope of Kuhn's scientific community idea, arguing that epistemic communities can arise from ‘bureaucratic position, technocratic training, similarities in scientific outlook and shared disciplinary paradigms’.²¹ According to Ruggie, epistemic communities share intentions, expectations, symbols, behavioural rules, and points of reference. The episteme, which holds an epistemic community together, ‘delimits … the proper construction of social reality’ for its members, and if successful, for international society.²²

In the early 1990s, Peter Haas published a book on the Mediterranean and epistemic communities.²³ Then together with Emmanuel Adler (who had earlier theorised about epistemic-like communities in The Power of Ideology),²⁴ he operationalised the concept of epistemic community further in the 1992 International Organization special issue, defining it as ‘a network of professionals with recognized expertise and competence in a particular domain and an authoritative claim to policy-relevant knowledge within that domain or issue area’.²⁵ In other words, epistemic communities are networks of experts who persuade others of their shared causal beliefs and policy goals by virtue of their professional knowledge. Their policy goals must derive from their expert knowledge, not some other motivation, otherwise they lose authority with their target audience, usually elite governmental decision-makers. Their reliance on expert knowledge, which they validate within their group, is what differentiates them from other actors that seek to influence policy. For example, other transnational actors may rely on shared values deriving from idealism (advocacy networks),²⁶ self-interest (multinational corporations), a fixed agenda (lobbying groups), methods of interpretation (interpretive community),²⁷ argumentation (argumentative or rhetorical community),²⁸ or shared practices (communities of practice).²⁹ Ultimately, epistemic communities seek to ‘benefit human welfare’ by persuading decision-makers of the goals they share in their particular domain of expertise.³⁰

Other contributors to the 1992 special issue – including Raymond Hopkins, Ethan Kapstein, M. J. Peterson, and others – put forth a number of case studies that show how epistemic communities are indeed at work in a variety of contexts, and how they influence policy outcomes. For example, William Drake and Kalypso Nicolaïdis argued that an epistemic community of services experts convinced governments to include trade in services as part of the General Agreement on Tariffs and Trade's (GATT) overall trade liberalisation efforts. Despite states’ entrenched preference for protectionism, the services experts informed governments about how liberalising trade in services would benefit the public good across the board. At the outset of this process, governments did not even speak of services as part of the discourse on trade, and saw services liberalisation as a risky proposition. Once they were convinced of the idea, the influence of the epistemic community waned.³¹ John Ikenberry argued that an influential epistemic community of British and American economists and policy officials, all experts in and proponents of Keynesian economics, were able to persuade governments of their policy goals during the negotiations leading to the Bretton Woods agreement.³² Emanuel Adler suggested that a domestic, US-based epistemic community of arms control experts was able to persuade the international community to form a regime under the 1972 antiballistic missile arms control treaty. Although this was a domestic rather than transnational epistemic community, it was still able to translate technical and scientific knowledge into actual international policy.³³ Other cases included epistemic communities of whalers, environmentalists, and central bankers.³⁴ All together the special issue still makes for a rich collection of detailed empirical evidence, convincing elaborations on the concept of epistemic communities, and a solid research programme for future scholars to follow.

Development of the concept

This early interest in epistemic communities seemed to promise a significant research agenda that would advance the literature on transnationalism and networks. Subsequently, scholars have made advances in establishing under what circumstances epistemic communities matter. At the same time, however, the epistemic communities research programme has fallen short of its promise in that subsequent works on the topic have been few and far between, and have applied the concept somewhat narrowly to single cases of environmentalists, economists, and other scientists.³⁵

To name some examples of the literature that followed the 1992 special issue, Amy Verdun argues that a group of monetary experts, known as the Delors Committee, constituted an epistemic community important to the creation of the European Monetary Union.³⁶ She finds that the process of European Union (EU) integration has been for the most part more technocratic and expertise-driven than political. Anthony Zito explores the role of an environmentalist epistemic community that influenced the EU's acid rain policy.³⁷ As a sceptic, he admits to purposefully choosing an easy test for the impact of epistemic communities, and his goal is to find out the extent to which the structure of EU institutions restricts their ability to exercise influence. He argues that we should expect the EU to be a favourable environment for epistemic communities: there are many access points for influence and innovation is valued. However, EU decision-making is also rather fragmented and there are many points at which policy ideas can be struck down. He concludes that European epistemic communities are more likely to be influential at early stages of policymaking when the costs of change are still not known, existing policies are unclear, and a crisis has caused new levels of uncertainty among decision-makers.³⁸

One question that has been well considered in this literature is: under what conditions do epistemic communities matter? Zito provides a valuable summary from several different studies that I have condensed and categoried into the following table (see table 1).³⁹

Table 1. When epistemic communities are persuasive: summary of the literature⁴¹

What can we make of this second wave of research? First of all, it is clear that uncertainty on the part of governments continues to stand out as a significant scope condition for epistemic community influence. Haas, Radaelli, and Hall argue that uncertainty arises from crises or a particular issue that is surrounded by continuous uncertainty. Second, an epistemic community's relationship to decision-makers is key, according to a number of scholars. One aspect of this is whether the community has ready access to decision-makers. Another aspect is whether the epistemic community's policy goals are more or less in line with existing norms. The argument is that decision-makers are looking for solutions that solve problems without being too disruptive. Third, there is also an emphasis in this literature on the nature of the issue itself. Several scholars argue that if the issue is more technocratic, quantitative, and scientific in nature, then epistemic communities have a straighter path to influence. Finally, if there is little competition from other actors epistemic community influence is more likely.

This second wave of literature has put forward case studies that demonstrate the value of the concept in explaining why decision-makers choose some policies over others, why new ideas gain traction, and what makes some epistemic communities more persuasive than others. However, it has fallen short in that there has not been much consideration of the concept itself and its utility for understanding International Relations, such as when epistemic communities and the norms they espouse are more or less likely to impact state behaviour or regime formation. Rather than providing innovations to the approach, this second wave of research has served to refine it, and in the process, limit it. Claire Dunlop observed in 2000 that ‘the approach as it stood in the 1992 Special Edition remains largely the same today’.⁴² The same could be said more than ten years later still.

Criticism of the concept

There are several categories of criticism of the epistemic community concept that have emerged in the literature alongside its further development. David Toke and Ronald Krebs argue that the influence of epistemic communities has been overstated. They contend that it is too often assumed that these groups have unproblematic access to those in the position to make policy decisions.⁴³ What if there are other actors who compete for policy influence who are more convincing than epistemic communities? A second criticism along these lines is that epistemic communities are not necessarily better at solving complex policy problems than governments themselves. Toke argues that governments are arguably more knowledgeable about navigating crises of uncertainty.⁴⁴ A third area of criticism has to do with the motivations of epistemic community members. Krebs argues that such members may just be after their own personal or professional self-interest. He asks, what if they are simply reflecting their own domestic cultures and strategic interest rather than professional expertise?⁴⁵ A fourth criticism stems from the belief that epistemic communities can only be influential if they are part of a political coalition. Sebenius and Dunlop argue that epistemic communities must inevitably engage in political activism. Thus, rather than persuading by virtue of their authoritative claim to knowledge, they end up relying on bargaining, just like all other political actors.⁴⁶ Fifth, Dunlop points out that Haas's original framework is too ambiguous about which of the criteria for epistemic community influence is most central to explaining their preferences: shared principled beliefs, shared causal beliefs, shared notions of validity, or a common policy enterprise.⁴⁷

Despite their scepticism, these critics actually provide several constructive suggestions for improving the viability of the epistemic community framework. They argue that future research in this area should account for a number of things, including: domestic politics (why some epistemic communities’ ideas gain traction over others);⁴⁸ competition among epistemic communities or with other actors;⁴⁹ the context within which epistemic communities operate, especially the major political interests of a given time period; the varying degrees of power that epistemic communities might have; and the relationship between scientific knowledge and political preferences.⁵⁰

Along these lines, a recent collection of articles by Emanuel Adler and Vincent Pouliot attempt to subsume epistemic communities into a broader idea: communities of practice.⁵¹ They define practices as patterned, meaningful action that knowledgeable actors engage in within a particular organisational context. Depending on what one wants to explain, practices can be nearly anything from war to diplomacy to tactics to everyday interactions among professionals.⁵²

While the idea of communities of practice is certainly an intriguing new line of research, I would argue that subsuming epistemic communities within it risks throwing out the distinctive utility of the concept. Members of communities of practice may not even be aware that they are engaged in a set of practices, and they may not even have a shared motivation, let alone specific policy goals. Unlike epistemic communities, they do not even necessarily deliberate, coordinate, or self-identify. The practices approach tends to prioritise identifying the practices over understanding the community of actors itself, the internal dynamics that characterise the community, and the great variety of actions – beyond practices – its members take individually or collectively to impact policy outcomes. In so doing, it tends to imbue a path-dependent quality to practices, and underplays the role of human agency. While this approach does resolve the problem of defining epistemic communities too narrowly, the emphasis on practices is at the same time very broad (that is, encompassing nearly every type of observable phenomena in IR) and limiting (downplaying the actors themselves and their agency). This approach also does not account for the degree of internal cohesion or degree of success, which I suggest are a central part of understanding the role of epistemic communities.

With the exception of Adler and Pouliot's work, there is a surprising tendency amongst scholars to hold closely to Haas's original definition, and even interpret it very narrowly. Perhaps this explains why scepticism surrounding the epistemic community framework has grown, and has itself met with little response.

Unless a response to these criticisms is forthcoming, this still promising area of research may continue to be marginalised. Indeed, an implicit, overarching response to the epistemic community paradigm has been that the concept is simply not a necessary addition to the literature, and that other theories are sufficient to explain policy outcomes or norms emergence. This marginalisation would be unfortunate, as I argue below, because epistemic communities are making an increasing contribution to the development of preferences and worldviews of states and international regimes. Indeed, in an age of increasing transnational activity, epistemic communities have growing potential to contribute to our understanding of International Relations. The criticisms that have been made are in fact not that difficult to overcome.

Reconceptualising epistemic communities

In this section, I endeavour to push the boundaries of Haas's original definition of epistemic community to enhance the concept's utility and explanatory power. I argue that the actors that comprise epistemic communities can be governmental or non-governmental, scientific or non-scientific, and that their persuasiveness rests in large part on their degree of internal cohesion and professionalism. I hypothesise that if an epistemic community is not internally cohesive, then it is less likely to be as persuasive as one that is. This internal cohesion ultimately provides the group with an episteme, a shared worldview that derives from their mutual socialisation and shared knowledge.⁵³ Moreover, when a group of professionals with recognised expertise is able to speak with one voice, that voice is often seen as more legitimate because it is based on a well-reasoned consensus among those in the best position to know.⁵⁴

In response to the earlier approaches and subsequent criticisms, I have identified four innovations, which address: (1) the variation in internal cohesion within epistemic communities and the central importance of professionalism; (2) the role of uncertainty in understanding epistemic community influence; (3) the relationship between epistemic communities and governments; and (4) the nature of knowledge. This last point also includes a central clarification regarding epistemic communities: non-scientific knowledge is just as – if not more – influential as scientific knowledge in influencing policy goals. Overall, I argue that if these innovations are taken into account, the universe of possible epistemic community cases grows significantly, and a more nuanced understanding of their activities and impact is possible.

Variable internal cohesion and the importance of professionalism

It is important to recognise that epistemic communities do not simply exist or not exist, as has been the assumption in much of the literature. They can be strong or weak in relation to each other as well as in relation to other actors. In order to better understand the impact of epistemic community influence, it is also necessary to examine their internal dynamics and to engage in broader, comparative research that will tease out the relative strength or weakness of the identified communities.⁵⁵

Critics have pointed out that research on epistemic communities must do a better job of taking into account the international or domestic political context within which they exist. Many factors are part of this context: coexisting epistemic communities, leaders’ personalities, security threats, economic costs, and so on. To be sure, external context does matter. It indicates how high the barriers are for an epistemic community to be successful, that is, to achieve its policy objectives. A number of theoretical propositions can be tested in future research. For example, if multiple epistemic communities put forward conflicting evidence then the barriers for success are likely to be higher for any one epistemic community. If national security threats have suddenly increased, epistemic communities specialising in this issue area are more likely to be consulted. If the economic cost to change a particular policy is high, epistemic community persuasion is likely to be more difficult. If a coalition government is in power, epistemic community influence may be more likely.

Taking external context into consideration, I would suggest that when policy change is highly controversial, even some level of epistemic community influence is significant. Comparative and qualitative research enables scholars to take this into account, and to assess relative influence, given a particular context. Once the context is established for a particular geographic region or issue, it is possible to isolate the independent impact of epistemic communities through case studies, process-tracing, and interviewing.

Beyond the overarching comparative political context, the next step is to look at an epistemic community's internal dynamics. In this respect, a major hypothesis I put forward is that a strong epistemic community that has a greater potential for influence is one that not only possesses a high degree of recognised expertise, but is also internally cohesive. The original conceptualisation of epistemic communities focuses more on establishing the presence or absence of expertise and the validity of the shared knowledge across a group. By contrast, I argue that socialisation, relationships, and persuasive processes within the epistemic community are even more important in ultimately determining its strength or weakness. Substantial research on policy networks in the comparative politics literature has long demonstrated the importance of personal relationships, shared values, and a sense of community within these networks in impacting policymaking more generally.⁵⁶ These dynamics reveal how well epistemic communities are able to frame social reality collectively. Even if there is initial disagreement on substance, robust social cohesion can enable an epistemic community to overcome internal differences and thus be more externally persuasive.

On a broad level, constructivism provides valuable insight into both the external role and internal cohesion of epistemic communities.⁵⁷ The primary basis of constructivist theory is that social interaction explains actors’ behaviour and defines their preferences and interests.⁵⁸ One cannot assume a priori that facts have a fixed interpretation, objects have a given value, and actor preferences are inherent. Epistemic communities are one important kind of actor that helps to assign meaning to things. As Haas puts it:

Epistemic communities are important actors responsible for developing and circulating causal ideas and some associated normative beliefs, and thus helping to create state interests and preferences, as well as helping to identify legitimate participants in the policy process and influencing the form of negotiated outcomes by shaping how conflicts of interest will be resolved.⁵⁹

Beyond this, I argue it is not the nature of their knowledge that determines how these processes play out (discussed in more depth below), but their professionalism.

The process of professionalisation itself is one way in which internal cohesion is established. There is a robust and long-standing literature on professionalism in the field of sociology.⁶⁰ Magali Larson argues that members of a profession continuously establish, refine, and re-establish the role and status of their profession.⁶¹ In so doing, they endlessly define and redefine the common vision, shared standards, and professional identity that mark them as a recognisable, coherent whole. This can be achieved to varying degrees of success. For example, historical research has shown that the professional fields of legal work, engineering, and medicine have been internally divided.⁶² Following from this definition, Andrew Abbott suggests that the social underpinning of a profession is its jurisdiction over a particular kind of expertise and the type of work that goes with it.⁶³ This leads to variation in the legitimacy of expert groups as they compete with each other for their own niche and status as professional groups.

It is important to remember that epistemic communities cannot be comprised of whole professions. There may be competing epistemic communities from the same profession, advancing different goals. In addition, many groups within a profession may not be motivated by their knowledge-based expertise and desire to support the public interest, thus precluding them from being epistemic communities. But professionalism is at the heart of the internal dynamics of any epistemic community.

Professionalism can be broken down and operationalised in a number of ways for the purposes of measuring internal cohesion within an epistemic community. Here, I emphasise four variables that are particularly universal in application: (1) selection and training; (2) meeting frequency and quality; (3) shared professional norms; and (4) common culture. All four of these variables lie on a strong-weak continuum rather than having a specific sufficiency minimum. I suggest that the variables altogether are best thought of as a so-called family-resemblance model.⁶⁴ That is, not all of the variables need to be strong for an epistemic community to exist. This would create an artificial dichotomy. Instead, the overall strength of these variables together says more about the strength or weakness of a given epistemic community.

Selection and training provide insight into the origins of epistemic community membership, and largely determine the status of a profession. When selection and promotion are highly competitive, this ensures that those who eventually constitute an epistemic community have a high level of expertise. Similarly, if these individuals undergo rigorous and extensive training their expertise is not only more likely to be recognised by others, but they are also more likely to have developed a sense of cohesion. Training can come in a variety of forms, but when standards are consistent across national borders, transnational epistemic communities are more likely to be cohesive.

The quality and frequency of meetings point toward the nature of interaction among members of an epistemic community. The more time they have together face-to-face, the more likely they are to build strong ties, strengthen shared professional norms, and cultivate a common culture. In particular, informal meetings in smaller groups enable a richer environment for socialisation and the development of a common culture.⁶⁵ Frequent meetings solidify a body of shared professional norms that concern the protocol, procedure, and standards of consensus-building within an epistemic community. These professional norms arise early on through training, and evolve as individuals encounter various circumstances. Even when members of a group disagree about certain substantive issues, their professional norms give them a common basis of understanding that they can count on, and this makes it easier for them to eventually compromise or reach consensus on substance.

Common culture comprises the sense of purpose, identity, symbolism, and heritage within the community. It is more than simply esprit de corps, but a sense of identifying with one another. An epistemic community with a strong common culture is far more likely to remain cohesive regardless of the circumstances they face. Member-state ambassadors to the EU, for example, have real incentives to engage in hard bargaining, rather than deliberation and persuasion. They work on the basis of instructions from their capitals, which have significant powers to sanction wayward diplomats. However, the ambassadors’ strong common culture prevents them from engaging in bargaining.⁶⁶ Rather, they tend to sell their national capitals on the benefits of common action and stress that arguments that smack of national self-interest would not be persuasive. Similarly, senior EU military officials will often not even show up at a meeting if their capitals have given them firm negotiating positions which they know will run afoul of an existing consensus. In these circumstances, they send their deputies to read out instructions, rather than opposing the sense of common culture that typically prevails.⁶⁷ All together, the importance of professionalism and the variables discussed above show why it is necessary to look beyond the external role of epistemic communities to the dynamics within them. Future research can examine whether strong internal qualities, as defined above, are key to understanding what makes epistemic communities more influential, and hence persuasive.

Uncertainty

The central avenue for epistemic community influence is generally understood to be post-crisis conditions of uncertainty for decision-makers.⁶⁸ Haas writes, ‘New ideas will be solicited and selected only after crises, for crises will alert politicians to the need for action and will seek to gather information about their interests and options.’⁶⁹ A useful innovation to the framework is to take a broader perspective on uncertainty.

Uncertainty is actually a built-in and nearly constant feature of the international system, whether it is objective or perceived. Most issue areas in IR are works-in-progress, whether they involve dealing with global pandemics, capturing pirates off the coast of Africa, sharing intelligence on terrorist threats, mitigating climate change, or increasing accountability in international institutions. Uncertainty is everywhere, not just in circumstances that might be labelled after the fact as having constituted major crises.⁷⁰ Even issues that are objectively certain, like climate change, for example, may be perceived to be uncertain. States are bound together through overlapping regional and international organisations, and are increasingly interdependent in multifaceted ways. The world is getting smaller and interactions more complex. Uncertainty is a normal state of affairs.

Rather than looking for epistemic community influence at specific critical junctures and post-crisis periods of uncertainty, it is more likely that when epistemic communities exist they are always at work, shaping and reshaping the nature of regimes, policy choices, and norms. Moreover, widespread global uncertainty in the twenty-first century – with the global financial crisis, rising powers, nuclear proliferation, and climate change among other things – means that there are many rapidly changing policy issues that would lend themselves well to the continuous presence of epistemic communities. Empirically, some issues, like how to deal with credit default swaps, are considered more uncertain than others, like routine decisions on how much to appropriate for the Justice Department each year. Theoretically, it may be more likely that epistemic communities will be called upon in the wake of a crisis, but this does not mean that they are more likely to successfully influence decision-making. I argue that uncertainty can create space for epistemic community activity, but the degree of uncertainty, however one tries to measure this, does not necessarily correlate with the likelihood of epistemic community impact.

Because of this, epistemic communities may not always decline in influence once their ideas have been considered by policymakers, as Peterson, Drake and Nicolaïdis, and others have suggested.⁷¹ Since conditions of uncertainty are the norm rather than the exception, it is more likely that epistemic communities that have emerged and that remain internally cohesive will prove to be enduring, even as their attention shifts with the rise of new issues. Indeed, in order to even have a legitimate voice at the table when a crisis strikes, an actor may have already had to establish itself beforehand as one to be listened to.

Very influential epistemic communities may impact policy goals even when the issue seems quite certain. New evidence or information may come to light that compels epistemic communities to develop new policy goals in previously resolved policy areas. For example, international financiers have begun to suggest that certain types of start-up businesses in the developing world are unable to comply with the payback terms of micro-loans, originally thought to be a highly successful incentive programme.⁷² This could cause an epistemic community of finance experts to push for changes to international micro-financing rules. Or, through ongoing internal deliberation and re-evaluation, epistemic communities may reach consensus on previously contested knowledge. This happened in the late 1890s, when scientists finally came to a consensus on the causes of the cholera epidemic whereas before they could not agree on whether it was a result of a microbial agent or foul vapours. This enabled the international policy community to enact international sanitary guidelines for the first time.⁷³

Or, contextual factors might change, causing decision-makers to be more open to ideas that they had previously not considered viable.⁷⁴ For example, when energy prices went up, alternative energy became viable. In this way, epistemic community influence can be found in gradual policy shifts through everyday efforts to persuade. Gradual change should not be discounted as less impactful, especially in the longer term.⁷⁵ Gradual change has enabled the EU to evolve from a simple coal and steel community to a quasi-federal entity encompassing everything from common passports to a European arrest warrant.

Thus, by exploring even established policy issues that continually evolve – trade agreements between two countries, human rights regimes, defence pacts, and so on – one is likely to find epistemic communities operating behind the scenes. As transnational activities increase, the likelihood of identifying such areas of active epistemic community involvement should also increase. When it comes to uncertainty, a broader interpretation enables a fuller consideration of the range of epistemic community activity under conditions of perceived or objective uncertainty, as well as certainty.

The relationship between epistemic communities and governments

A frequently heard argument is that epistemic communities, as a band of outside experts, have little substantive influence over politics, particularly in the most important or basic areas of international relations decision-making. Krebs writes,

Persuasion may certainly take place in such a political environment, but it is more a rare and contingent outcome than a regular and predictable one … While epistemic communities might dominate governmental decision making with regard to seemingly technical questions about, for example, the consequences of pollution or the details of military force employment, the mechanism seems less accurately to describe how state leaders acquire their information about basic issues in international relations.⁷⁶

He essentially argues that there is a distinction between how leaders gain an understanding of technical details – the domain of epistemic communities – versus more important, political issues for which leaders rarely seek guidance from epistemic communities. Thus, it is important to establish further the extent to which epistemic communities and governments working in synergy are able to craft policy solutions that pertain to both technical and basic issues. There are many examples to support this idea. International development experts under the umbrella of the United Nations Development Program work in tandem with governments around the world to implement national and regional strategies on governance, information technology, poverty, business, and so on.⁷⁷ The same is often true for networks of environmental, business, and labour experts. Parts of governments may be better at crafting policy than epistemic communities – as they understand the protocol, language, and technicalities necessary for policy documents – but this does not preclude them from taking advantage of synergies with intersecting communities of experts.

As to the availability of such advice for more insular agencies, it is important to appreciate that epistemic communities are also often located within government structures, although they continue to exercise independent agency. Rather than making such communities less influential, this may afford them greater access to decision processes. Indeed, in some circumstances, governments are the catalyst, bringing groups of experts together expressly to provide advice. The European Commission often brings together groups of ‘personalities’ or ‘wise men’ to brainstorm and offer advice on the basis of their expertise, such as the development of a European security research programme or the resolution of Russia-EU visa issues.⁷⁸ These groups may or may not evolve into epistemic communities but if they do, they are no different from epistemic communities that come together for professional reasons.⁷⁹ As Drake and Nicolaïdis write, ‘the issue is not where community members sit but instead what they say’.⁸⁰ If epistemic actors are located inside government, then the argument that governments are rarely willing to listen to epistemic communities is less convincing.

In light of these possibilities, if our analysis is to be comprehensive, we must be willing to look broadly when identifying epistemic communities. It does not matter whether members of an epistemic community come together organically, are spurred to action by an NGO, or are brought together by governments to form an advisory committee. This has little inherent impact on how the group behaves or what it does once its members come together – that is, whether or not it grows into an epistemic community with the intense interaction and emergent values that this entails. Epistemic communities may sometimes be difficult to differentiate from regular bureaucratic groups, but a close look at internal dynamics will distinguish them.

From a researcher's perspective, it is likely that careful and numerous interviews of group members and those who interact with them will be necessary to determine whether the group is an epistemic community. An epistemic community is likely to have professional relationships that go beyond their formal, bureaucratic role. Key questions to address are: Is an institutional group or committee producing outcomes that go beyond the expectations of its formal functions? Do the committee's members bring a high level of expertise to their institutional positions? Did they know each other or work with each other in previous settings? Do they share training or educational experiences? Do its members meet often outside of work and informally? Do they share a particular culture and professional norms that are independent of their formal function? Not all of these qualities are necessary, but they provide some initial clues as to whether further research might be fruitful. Many committees or groups of professionals that are formally part of government are actually given autonomy to act as part of their jobs. The question is whether the committee takes on a life of its own and evolves into an epistemic community. A group of experts may come together for many reasons, but it becomes an epistemic community at the initiative of its members. Of course, many epistemic communities are not housed within formal structures. But for those that are, direct access to decision-makers is indeed unproblematic.

Moving beyond scientific knowledge

There is sometimes the tendency in academia to resist reconceptualising existing ideas, and to simplify what those ideas really meant in the first place. Over the past two decades, there has developed a widespread assumption that epistemic communities, as Haas originally defined them, are exclusively limited to scientific or technical groups. This is certainly not true, as the originators made clear twenty years ago. Thus, an important clarification is in order. Haas was explicit about this very point:

In this volume, we stress that epistemic communities need not be made up of natural scientists or of professionals applying the same methodology that natural scientists do … By our definition, what bonds members of an epistemic community is their shared belief or faith in the verity and the applicability of particular forms of knowledge or specific truths.⁸¹

Drake and Nicolaïdis take this to heart in their contribution to the special issue, describing what they call a two-tiered epistemic community, comprised of individuals with diverse expertise like journalists, government officials, lawyers, academics, industrial specialists, and others.⁸² In spite of this strand of argumentation, the empirical focus of epistemic communities has continued to emphasise scientific knowledge, and scholars have repeatedly argued that this is where these actors are likely to have more influence.

It is possible that this tendency to limit the scope of the concept arose from its intellectual origins – scientific communities – or perhaps from a subtle orientation in the definition itself. Haas argues that truth-tests and peer-reviewed publications are necessary for epistemic communities to validate their knowledge.⁸³ Even Drake and Nicolaïdis write, ‘Regardless of affiliation, the members’ authority derives from their articulation of causal beliefs that appear to external policymakers to be “scientifically objective”.’⁸⁴ Clair Gough and Simon Shackley argue that ‘scientific knowledge is the “glue” that helps to keep policy actors committed and can be used as a trump card against opponents to the epistemic coalition’.⁸⁵ These statements and others have lent a certain bias in the determination of who can comprise epistemic communities.

While it is certainly true that groups of scientists can and often do form epistemic communities, there is no reason to assume that actors with non-scientific expert knowledge cannot be just as persuasive, and operate according to the same or similar criteria. Diplomats, judges, defence experts, high-ranking military officials, bankers, and international lawyers, among others, all have just as much of a claim to authoritative knowledge as scientists. Indeed, the dynamic within non-scientific epistemic communities is similar. They (1) share professional judgment on a policy issue; (2) weigh the validity of their policy goals in their area of expertise; (3) engage in a common set of practices with respect to the problem area with the goal of improving human welfare; and (4) share principled beliefs.⁸⁶ There is no reason to assume that their shared expertise is less reliable or influential. Professionalism, rather than science, is the glue that holds epistemic communities together, facilitates consensus, and enables persuasion.

High-ranking members of the military profession, for example, clearly have a wealth of tactical expertise – specialised knowledge of how best to devise military strategy on the ground and during an operation. They also share expertise in military doctrine, which is the broader body of knowledge covering force orientation, war strategies, and logistics. Over the centuries, especially in Europe, shared military expertise, culture, and tradition have evolved, facilitating the development of transnational cohesion in this area. There is a long, shared history of military socialisation and professionalisation in Europe. Since at least as early as the seventeenth century, there has been a trans-geographic dissemination of training, culture, and military know-how. Alliance formations and professional learning enabled convergence over time as military officials came into contact with each other, studied each other's strategies, and read the same classic texts. More recently, the emergence of a number of high-profile and internationally-oriented military academies, such as Westpoint, the School of the Americas, Saint-Cyr, and the General Staff Academy in Moscow, have contributed to the dissemination of shared professional norms and culture.

This shared expertise has also come to include a range of other military activities in which states are occupied, such as crisis management, civil-military relations, and humanitarian intervention. This has been especially evident in recent years in Europe where there is intense transnational activity with the advent of the Common Security and Defence Policy. Naturally, this does not mean that the whole military profession constitutes an epistemic community. Rather, I suggest that specific groups of high-ranking military officials who interact transnationally have the potential to form epistemic communities by virtue of their shared professional norms and expertise, as long as they seek collective policy goals as a result of these qualities. Military expertise has not traditionally been understood as scientific, but a broader understanding of knowledge reveals that groups of generals may indeed comprise epistemic communities. For example, permanent representatives in the European Union Military Committee (EUMC), who are three-star generals or admirals, report that reaching consensus on their collective military advice to the political decision-makers in Europe is unproblematic. Their similar training and overlapping career experiences give them a body of shared knowledge that is virtually taken for granted within the group.⁸⁷ They are not simply following instructions, but exercising collective agency. As an example of this, they were instrumental in crafting the 2005 ‘Long-Term Vision for European Defence Capability and Capacity Needs’, which outlines the need for security integration among EU member states in light of the demographic, economic, political, and security challenges that the EU will face in the next two decades.⁸⁸ This strategy has been accepted and implemented by the member states. EU military generals thus operate as a relatively strong epistemic community, demonstrating that the emergence of military epistemic communities is a possibility in other contexts and circumstances.

Similarly, high-level networks of diplomats may also qualify as epistemic communities.⁸⁹ For Adler, diplomats are part of the epistemic community story, but he sees their role as limited to communicating the ideas of domestic epistemic communities.⁹⁰ However, there is much to suggest that in some policy domains, diplomats actually constitute epistemic communities in their own right. There are multiple diplomatic epistemic communities based in regions and cities around the world that vary in size and strength. Of course, not all diplomats are members of epistemic communities, but those that are at the highest levels of the professional hierarchy – ambassadors in particular – and those that are career diplomats are more likely to have such membership.

The European experience with its historical evolution of transnational activity affords an example that is at the foundation of the contemporary international order. Epistemic communities of diplomats have existed across time, with their origins dating back at least to the time of the Renaissance in Europe. They have not simply become stronger over time in a linear fashion; rather, they have varied in strength as their internal dynamics have changed.⁹¹ Their shared expertise, norms, and worldviews have enabled them to find areas of agreement that would not have otherwise been found. Over time, diplomatic expertise has evolved to include: in-depth knowledge of the substance of the specific policy areas, a thorough understanding of different countries’ interests and concerns as well as the personalities involved, and an overarching expertise in the processes of diplomacy itself.

Two reasons that diplomats have not typically been considered epistemic communities is that they are obliged to represent national interests and that they seem to be generalists rather than experts. Nevertheless, there is much to suggest that formal groupings of diplomats often transcend their prescribed role, and routinely operate as an epistemic community. Diplomats are experts at the art of negotiation, persuasion, and compromise. Their internal processes of deliberation and the professional norms that govern these processes are so important that they determine the success or failure of potential international agreements. Moreover, the epistemic community framework better captures their role than that of Slaughter's transgovernmental networks because the latter are more concerned with improving their own national-level work through transnational sharing of ideas.⁹² For example, judges cite each other's cases and regulators convey best practices so that they can do their jobs better at home. By contrast, epistemic communities are ultimately concerned with outward policy outcomes that benefit society beyond the nation. For very cohesive epistemic communities of diplomats, such as those in the context of the EU, success is not defined as maximising one's bargaining position, but rather as a compromise solution that brings states together on a policy issue of collective concern.⁹³

Nukhet Sandal draws upon one of the oldest professions in history – the priesthood – to argue that religious leaders can comprise epistemic communities. One type of knowledge is ‘hermeneutics’ – expertise in understanding a particular text, like the Koran or Bible. Sandal notes that religious leaders are not the same as scientific communities, but that they meet the conditions for epistemic communities. They share norms and causal beliefs derived from their expert knowledge of the same biblical texts. These include ‘respect for life, equality, a belief in a transcendent being and need for a just economic system sensitive to the environment’.⁹⁴ They benefit from a common and rigorous technical education that makes them into experts on ‘accepted methods of interpretation’, and enables them to devise new laws of interpretation. These faith leaders also publish on these issues in professional journals and attend conferences where they debate the nature of their work using similar frames of reference. In Sandal's two case studies – Protestant leaders in South Africa and Catholic leaders in Northern Ireland – she shows that these religious epistemic communities are able to influence political outcomes. In South Africa, Protestant leaders were influential in ending apartheid and establishing racial equality. In Northern Ireland, Catholic leaders were eventually able to ease the intensification of conflict with the Protestants. In both cases, it was because of their shared authoritative claim to knowledge and the legitimacy, authority, and influence that this conveyed.

These three examples of military officials, diplomats, and faith leaders illustrate how knowledge can be understood more broadly, and how the line between scientific and non-scientific knowledge is itself often blurred. As Adler points out, sometimes the basis of an epistemic community's knowledge can be purely ‘imaginary’ because there is no way for the group to actually test the theory, as in the case of the arms control epistemic community. Nuclear war scenarios were only hypothetical and so the agreed-upon policy goals were based on a theory of international behaviour that was not based on scientific fact.⁹⁵ However, policymakers, persuaded by the epistemic community, treated the theory like it was indeed proven, turning it into a self-fulfilling prophecy. Antoniades argues that authoritative knowledge is a product of social context.⁹⁶ The main point is not whether the knowledge has been definitely proven or not, but rather whether it is socially recognised. As Haas argues in a more recent article, expertise is socially constructed in that it is most powerful when epistemic communities are seen to have integrity and to be free from political interference.⁹⁷ In effect, society must confer the authority and expertise that make epistemic communities influential.⁹⁸ This is where its power to influence lies. If an epistemic community is socially recognised – regardless of professional field – it can persuade others, get them to use particular language in defining their aims, and come to shape their worldviews. The more cohesive they are in presenting their knowledge, the more likely their authority and expertise will be treated as legitimate.

At the same time, one should not go too far in the opposite direction and assume that nearly any transnational network whose members share an interest in something constitutes an epistemic community. The parameters of shared professional expertise still hold.⁹⁹ A gathering of ‘Free Tibet’ activists is not likely an epistemic community because they are motivated by moral imperative rather than professional expertise, though they may seek advice from a community of human rights or legal experts who are.¹⁰⁰ Neither are political scientists meeting at annual conferences like the International Studies Association because there is little overarching agreement on policy, though smaller subsections of political scientists with a high degree of cohesion and coherent policy positions may possibly constitute an epistemic community. An epistemic community is rarely so broad as to include an entire discipline because its members must be able to understand the issues at stake, interpret the information similarly, and then form the same goals about what should be done. Disagreements about what constitutes knowledge within a profession can lead to the formation of multiple or even competing epistemic communities. For example, Youde describes competing epistemic communities surrounding the causes of AIDS, and Adler finds evidence for them with respect to nuclear weapons and deterrence.¹⁰¹

All together, the innovations put forward here suggest a far more significant role for epistemic communities in the field of IR. Through revisiting and reconceptualising the various components of the framework put forward twenty years ago, it is much easier to imagine how epistemic communities comprise an integral part of the fabric of transnationalism. The utility and explanatory power of the concept has been seriously under-recognised.

Conclusion

This article has suggested that epistemic communities can be strong or weak in terms of their persuasive abilities, and more comparative case study research would go a long way in identifying important differences in the influence of epistemic communities. I have argued that internal cohesion within an epistemic community is central to explaining these outcomes. Furthermore, the primary element that captures these internal dynamics is professionalism, which includes norms, standards, training, socialisation, status, and jurisdiction over a certain area of knowledge and work. Only specific groups within professions are ‘activated’ into forming epistemic communities, but a deeper understanding of their internal dynamics is greatly enhanced through an examination of its members’ professional underpinnings. On the other side of the coin, these same internal qualities would also be useful in pinpointing unsuccessful efforts or failed cases.

The epistemic community literature thus far has focused too narrowly on scientists because of the misguided notion that scientific knowledge is somehow superior to other forms of knowledge. Instead, I have suggested that the recognition and legitimation of expert knowledge is socially constructed. This is based in part on the integrity of the experts themselves, and also on whether or not they have reached consensus among themselves. A consensus among high-status experts who are viewed as free from political interference is often quite powerful.

Why should we care about the role of epistemic communities? Again, there is an important and growing linkage between globalisation and epistemic communities. As global processes become increasingly complex, ushering in conditions of ongoing uncertainty over a variety of issues, the need for specialisation is increasing. Epistemic communities are dealing with tangible realities that require policy solutions. In addition, new professions are emerging that deal with the advances in technology, uncertainties of the virtual world, threats to security, and rapidity with which global processes take place, making the likelihood of epistemic community emergence even greater. Policymakers will likely increasingly rely on the expertise of these networks to devise transnational solutions to global problems. Transnational epistemic communities have the advantage of projecting shared knowledge at the same time as understanding the particular circumstances of the various countries they represent, and being seen as a legitimate voice domestically. As transnational interaction grows, uncertainty abounds, and the role of non-state actors becomes ever more prominent, the intersection of global governance and expert knowledge should be a significant part of the next generation of epistemic community research.