Power Analysis: Encyclopedia Entries

DIIS WORKING PAPER WORKING PAPER DIIS WORKING PAPER 2010:34 Power analysis: Encyclopedia entries Stefano Guzzini DIIS Working Paper 2010:34 1 DIIS WORKING PAPER 2010:34 STEFANO GUZZINI PhD, Senior Researcher, DIIS. www.diis.dk/sgu DIIS Working Papers make available DIIS researchers’ and DIIS project partners’ work in progress towards proper publishing. They may include important documentation which is not necessarily published elsewhere. DIIS Working Papers are published under the responsibility of the author alone. DIIS Working Papers should not be quoted without the express permission of the author. DIIS WORKING PAPER 2010:34 © The author and DIIS, Copenhagen 2010 Danish Institute for International Studies, DIIS Strandgade 56, DK-1401 Copenhagen, Denmark Ph: +45 32 69 87 87 Fax: +45 32 69 87 00 E-mail: diis@diis.dk Web: www.diis.dk Cover Design: Carsten Schiøler Layout: Allan Lind Jørgensen Printed in Denmark by Vesterkopi AS ISBN: 978-87-7605-415-1 Price: DKK 25.00 (VAT included) DIIS publications can be downloaded free of charge from www.diis.dk 2 DIIS WORKING PAPER 2010:34 CONTENTS Introduction 4 Power and International Politics 5 The ‘lump’ concept of power in realist theories Neo-institutionalism: redefining the link between resources and outcomes Structural power in the global political economy Rule in world politics: the social construction of legitimacy and order Conclusion: The politics of power analysis 5 6 7 9 10 Constructivist view of power in International Relations 11 Constructivism Constructivist conceptualisations of power Constructivism-inspired analyses of power in IR 11 12 13 Relational power A relational conceptualisation Causality and aggregation 15 16 17 Fungibility of power resources 19 References and further reading 22 3 DIIS WORKING PAPER 2010:34 INTRODUCTION This Working Paper makes accessible four interconnected commissioned encyclopedia entries on the analysis of power. The first one gives an overview of the different concepts of power in International Relations in their theoretical and historical development. The second treats more specifically with the constructivist view of power in International Relations. That approach has developed out of the critique of conceiving power in terms of mere material properties. In contrast, constructivists stress the constitutive character of ideas for the self-understanding and interests of actors. They also rely usually on a relational understanding of power, which is the topic of the third entry. In a relational understanding of power, power is not the possession of a person, nor does it correspond to a mere production of effects; it is constituted within a social relation. Only by knowing the respective value systems and beliefs specific to the relationship can the analyst attribute power. The fourth and final (shorter) entry is on the phenomenon of fungibility of power resources. In the analysis of power, the problem of fungibility refers to the issue whether or not different types of resources (e.g. military, economic, cultural, diplomatic) have the characteristic of being freely exchangeable or replaceable. This issue has gained prominence for two reasons. If resources are highly fungible, i.e. can be mutually substituted without losing much of their value, then this allows them to be aggregated, so as to permit the construction of overall power resource indexes. This is fundamental for balance of power analyses and for the explanation of behaviour in terms of power maximisation, both typical for realist approaches in IR. The entry shows how lacking fungibility can be seen to undermine such theorising in International Relations. 4 DIIS WORKING PAPER 2010:34 POWER AND INTERNATIONAL POLITICS1 Different aims inform the study of power. Simplifying somewhat, power plays a role in two distinct if related domains. In the field of political theory as understood here, the purpose of analyses of power is to capture the nature of the ‘polity’ in which questions of the organisation of (organised) violence, the common good and freedom are paramount. In these studies, power stands for ‘government’ or ‘governance’ and political ‘order’, as well as personal ‘autonomy’. The logic in the field of explanatory theories, on the other hand, is to think of power in terms of a theory of action first and a theory of domination subsequently. Here, power is central to the explanation of behaviour and the outcomes of social action. It refers to ‘agency’ and ‘influence’ if not ‘cause’, and to rule or domination. For classical realist thinkers in International Relations (IR), the particular context of world politics suggested that these domains could be fruitfully merged. With no world government, the international system seemed to miss not only an ordering authority, but a polity altogether. In a sense, therefore, it was possible to think of world politics as the simple aggregation, and balance, of agent capacities to influence, such that a theory of international politics was not needed. In the explanatory domain, power then became a central variable in a double causal link. Power understood as resources or ‘capabilities’ was an indicator of the strength of actors, and consequently of the capacity to affect or control events. Likewise, a general capacity to control outcomes has been used as an indicator for the ruling of Encyclopedia entry forthcoming in Bertrand Badie, Dirk Berg-Schlosser & Leonardo A. Morlino, eds, International Encyclopedia of Political Science, London et al.: Sage Publications, 2011. 1 the international system. Rather than seeing the two domains as separate, the special nature of world politics could combine them in an explanatory sequence: by knowing who can be expected to win conflicts, we would also know who or what governs international politics, which, given the absence of a world polity, was all there was to know about power as order and government. Power analyses in recent decades have challenged these tacitly assumed links and ultimately also the neglect of the concerns of political theory. Criticising the ‘lump’ concept of power that is typical of realism, neo-institutionalism has tried to redefine the link between resources and outcomes. Similarly, with regard to the understanding of ‘rule’ and ‘governance’, different ‘structural power’ approaches have demonstrated the need to conceive of more encompassing power concepts so as to capture important, but otherwise neglected facets of international rule. Poststructuralist and constructivist approaches focus on power as authority and legitimacy, not through the establishment of an open social contract, but in the habitual working of discourses and practices which dis/empower agents. When aiming at an understanding of the world polity, they also refer back to the domain of political theory. But by starting from a historical reconceptualisation of politics and order, they do this by stepping altogether outside an analysis in terms of these two tacitly assumed links. The ‘lump’ concept of power in realist theories Although classical realism does have strong political assumptions about human nature and the role of power in politics, the two-step analysis of power mentioned above, which is driven by the explanatory domain, has be- 5 DIIS WORKING PAPER 2010:34 come dominant. There, on the macro-level, realist theory relies on the concept of the balance of power. This presupposes a common denominator for power in which all its aspects can be coherently aggregated. On the micro-level, realist theory relies on the idea that states are interested in relative gains in power. For both statements to work, power needs to be measurable. Indeed, such theories require a concept of power akin to the concept of money in economic theory. In this analogy, the striving for utility maximisation expressed and measured in terms of money parallels the national interest (i.e. security) expressed in terms of (relative) power. This central assumption has been challenged by early realist critiques and more recent institutionalist approaches. In an early argument which also anticipates and implicitly criticises the economic analogy in neorealist theory, Raymond Aron opposed this aggregated concept of power and the underlying power–money analogy. The different degrees of the fungibility of money and power resources make this impossible. The term ‘fungibility’ refers to the idea of a moveable good that can be freely substituted by another of the same class. Fungible goods are universally applicable or convertible, in contrast to those that retain value only in a specific context. Whereas fungibility seems a plausible assumption in monetarised economies, it is not so in world politics: even apparently ultimate power resources like weapons of mass destruction might not necessarily be of great help in getting another state to change its monetary policies. Aron recognised that economic theory can be used to model behaviour on the basis of a variety of conflicting preferences. But for him, with the advent of money as a general standard of value within which these competing preferences can be situated on the same 6 scale, compared and traded-off, economists were able to reduce the variety of preferences to one utility function. In world politics, for reasons of its lack of a real-world fungibility, power cannot play a corresponding role as a standard of value. And not being a standard of value means that power cannot be the currency of great power politics, and national security in terms of power is not equivalent to utility. In response, realists insisted that diplomats had repeatedly been able to find a measure of power and hence the difference is just one of degree, not of kind. Yet even if actors can agree on some approximations for carrying out exchanges or establishing power rankings, this is a social convention which by definition can be challenged and exists only to the extent that it is agreed upon. Power resources do not come with a standardised price tag. With the link between resources and outcomes foregone, the realist chain of causes for understanding the international structure is broken. For a single international power structure relies either on the assumption of a single dominant issue area or on a high fungibility of power resources – neither of which are realistic . Neo-institutionalism: redefining the link between resources and outcomes When the US lost the war in Vietnam, some scholars tried to explain this power paradox away by identifying the lack of ‘will’ on the side of the US to use its resources, i.e. socalled ‘conversion failures’. In such an explanation, the war did not indicate the relative weakness of the US (in spite of its military capabilities), but simply its unperformed strength. Obviously, such an explanation can re-interpret any outcome ex post to suit any DIIS WORKING PAPER 2010:34 power distribution. As so often, the trouble with this type of power analysis is not that it is wrong, but that it cannot go wrong. Neoinstitutionalist analysis offers two responses. One conceptual way out consisted in accepting the apparent lesson of the Vietnam War. Consequently, control over resources, even issue-area specific ones, does not necessarily translate into control over outcomes. Power no longer functions as a determining cause. In Robert Keohane’s analysis, for instance, determinacy in the explanation shifts from interests defined in terms of the distribution of power to rational choice made on the basis of given interests defined in terms of power, expectations, values and conventions. Hence, only predictions of a very limited kind are possible – with a secondary role for power. Another solution to the paradox of unrealised power has been proposed by David Baldwin, who has taken the issue of power fungibility most seriously. His approach keeps a strong causal role for power by further specifying the relational and situational context that defines which policy instruments can count as actual power resources in the first place. Baldwin’s conception is shaped by his relational understanding of power. If power is about the capacity to get someone else to do what he/she would not have otherwise done, then threatening a suicide candidate with a gun implies that the person holding the gun has no power. In other words, power resources have no intrinsic value or effect but depend on the actual value systems of human beings in their relations with each other. Hence, the major difference from utilitarian action theories is that personal value systems cannot be simply assumed in the empirical power analysis. Instead, the researcher has first to analyse the value systems of the interacting parties in order to establish wheth- er there are any power resources in the first place. For this reason, Baldwin insists that one can only study power, if understood as a causal variable, in well-circumscribed ‘policycontingency frameworks’. Any assessment of power independent of such situational factors is erroneous (and there goes realism); any generalisation beyond such cases is contingent and has to be established separately (and there goes behaviouralism). The price for this, however, is that power analysis must potentially become very narrowly circumscribed to particular instances, where no prediction is possible. Whereas Keohane’s institutionalist move retained, however limited, the predictive capacity of a theory which is based on rational choice and not on power, this second move saves a central causal role for power at the price of predictability in IR/IPE. Keohane’s solution points to the direction of a rationalist neo-institutionalism, Baldwin’s less generalisable, contingent and situational solution to historical institutionalism. Structural power in the global political economy With the link between resources and control weakened, the micro–macro link between control over outcomes and international rule might not be worthwhile studying at all. And yet this is where international political economy (and constructivism and poststructuralism; see below) have made their most important contributions to the analyses of power. In fact, concepts of structural power redefine the context within which strategic interaction takes place, the resources considered important for assessing capabilities in the first place, and the outcomes that should be included in power analysis. Their common claim is that the sole reference to the first link, as made by 7 DIIS WORKING PAPER 2010:34 neo-institutionalists, is insufficient, if not biased, for understanding rule in the international system. It is the second link between outcomes and rule that becomes the starting point in analyses of power. A first version of structural power might be called indirect institutional power. This refers to the conscious manipulation of the institutional setting within which bargaining relations take place. Many important issues are decided before they reach the bargaining stage – indeed, often because they never reach it. For understanding the distribution of power, it is as important to see who prevails in decision-making as it is to analyse which ‘non-decisions’ were made. Despite occasional claims to the contrary, this version is perfectly compatible with neo-institutionalist approaches. Structural power has also been conceptualised as non-intentional power. Susan Strange’s concept of structural power stresses both the diffusion of the origins of power (and the variety of power resources) and the diffusion of its effects. Here, there is no reason to exclude from power analysis all those crucial effects that might not have been intended. As an old Chinese saying has it, it makes little difference to the trampled grass beneath whether the elephants above it are making love or war. This analytical shift from intentions to effects diminishes the importance of the neo-institutionalist approach for understanding power based on resources, interests and rationality. It focuses on the systematic and structural aspects of power, not on chosen ones. Thirdly and finally, structural power can also be understood as systematic bias or impersonal power. This refers to an impersonal ‘mobilisation of bias’ whereby social structures systematically favour certain agents. Such an understanding of power is common cur- 8 rency in dependency writings, both Marxist and non-Marxist, as well as in neo-Gramscian approaches, but it also applies to constructivist and poststructuralist approaches which emphasise non-materialist structural biases (see below). Such a conceptualisation has been criticised for deducing power from rewards, the so-called ‘benefit fallacy’ of power. We usually do not call a free-rider powerful who certainly profits from a certain systemic arrangement, but who basically remains at its mercy. But the benefit fallacy exists only within a causal framework itself. To say that a system benefits certain people does not mean that they have created that benefit or that they control it. It just means that in understanding power in a social system, it seems odd not to take into account the effects of that system which can systematically advantage some actors. In other words, in terms of the second link between rule and outcome, systematic benefits are relevant. Structural power analysis in international political economy tries to overcome the difficulty of conceiving power along the resource-outcome-rule line by starting from the other end. These approaches run into two types of problem, however. First, they tend to overplay the causal strength of their analysis. Moving backwards from rule to outcomes faces similar problems as moving from resources to outcomes. ‘The US won because of its structural power’ faces the same translation or conversion questions as classical resource-based analysis. It often appears to offer an answer when in reality it begs the question: power cannot be just substituted for cause. The second risk is related to this. IPE approaches tend to understate the non-materialist aspects of rule or governance, indeed the extent to which structures affect events only though the meaning given to them. DIIS WORKING PAPER 2010:34 Rule in world politics: the social construction of legitimacy and order Still staying with IR’s emphasis on the explanatory domain of power analysis, constructivism redefines power at the systemic and agent levels in IR. Its systemic analysis of power often looks at the origins of consent in terms of practices of tacit legitimacy. It is therefore close to power concepts of the family of ‘authority’. But rather than looking at formal or institutional authority, constructivism is interested in the intersubjective practices of power – not in the position of authority, therefore, but rather in what ‘authorises’, ‘legitimates’ or ‘empowers’. Moreover, it is not necessarily looking at intentional or agent power, but at the impersonal effects of discourses and/or habits for the production and reproduction of order, in particular in cases where practices go without saying, appear natural and are therefore perhaps the most effective power relations there are. At the actor level, such a view implies an emphasis on the process of interest formation as a primary locus for power relations. For constructivists, interests cannot be understood outside of their intersubjective contexts in terms of shared constitutive norms, of shared knowledge and understandings, and also through the effects that practices have on self-understandings or identity. For constructivists, what we want follows from who we are (or want to be). One larger power research agenda therefore concerns the background knowledge or constitutive ‘rules of the game’ which mobilise certain biases and which define the competent players and their effective moves. Naturalised understandings evoke certain actions and empower certain agents. If an event is understood as analogous to ‘Munich’, a collective memory is mobilised that authorises some action and undermines the legitimacy of others. Whether or not the end of the Cold War has ushered in a ‘clash of civilisation’, such an understanding mobilises and is empowered by pre-existing Cold War scripts in which totalitarianism was replaced by fundamentalism in security discourse. It gives it the potential to become a self-fulfilling prophecy. Just when the security sector started to be de-militarised, the increased use of Private Military Companies gave them ‘epistemic power’ because their practices shape the understanding of security and the self-understanding of its actors so as to authorise an increasingly technical and military understanding of the field while being, in turn, authorised through it. With regard to their focus on identity, constructivist scholars have not only looked at the impersonal effect of discourses and/or practices on self-understandings, but also on the ‘power politics of identity’. If identity is crucial for interest formation, then it is only a small step to analysing how diplomatic practices, sometimes intended, can try to blackmail actors by taking profit from contradictions in another actor’s self-understandings or between its action and self-representation. As a result, some of the constructivist research agendas converge with Foucauldian approaches in their understandings of order as diffused practices of rule, rather than as clear and/or formal hierarchies. Such convergences can be seen, for instance, in the analysis of how international standards, which are often established by private actors, are practices of rule once they become accepted convention and interact with the actors and issues they were supposedly only neutrally measuring (e.g. credit rating). But at the same time, such analyses link power in explanatory theory back to political theory, or, particularly in the post-Foucaultian vein, actually puts the latter first. The un- 9 DIIS WORKING PAPER 2010:34 derstanding of power is not primarily about cause and effect, not about influence and outcomes. Instead it embeds power into a historical analysis of the changing nature of the modern and liberal order, which, so the thesis goes, increasingly works by making the subjects of order active subjects of their own ordering. Ole Jakob Sending and Iver Neumann, for instance, analyse the role of nonstate actors in this vein, understanding them not so much as civil society in opposition to the political power of states, but as part and parcel of a decentralised and self-disciplining logic of a global order (governmentality). Conclusion: The politics of power analysis Even if careful scholarly discussion can discard some conceptualisations of power, there is no one root concept which we can unravel simply by digging deeper, neither in the domain of political, nor explanatory theory. Power concepts derive their meanings from the theories in which they are embedded and meet there the meta-theoretical or normative divides that plague and enrich our theorising. At the same time, the debate has come full circle. Initially, realist writings combined the domains of political theory, centred on the understanding of order in the polity, with the domain of explanatory theory by assuming that, in the absence of a genuine world polity, the analysis of capabilities and influence was all there could be. By attacking the double link between agent influence and the balance of power, later studies redefined a more or less causal role for power, be it at the agent or the structural level. Hence they stressed the explanatory domain of power at the expense of the political theory of power. But this can work only so far, since the two domains intrude into each other: structural power is 10 necessary to understand not only outcomes, but also autonomy in a polity; inversely, the analysis of the changing nature of global governance and order provides the background against which the very processes of power can be understood in the first place. And so, to close the circle with the post-structuralist and constructivist turn, the analysis of power returns again to show the links between the two domains. But contrary to early realism, it does so by assuming the existence of a genuine world polity within which power has to be understood. Yet so far it cannot pretend to have a theory of power capable of combining the two domains. And finally: power analysis is not only tied to the understanding of politics; it is itself political. The reason is that some concepts, like power, have a special status in our political discourse. They are used for a variety of purposes. For power, two are particularly important. Power is used in practical contexts in which we are interested in what we can do to others and what others can do to us. It is also important in moral and legal contexts where it functions as an indicator of effective responsibility: if actors could not have done an act (if they did not have the capacity to do so), they cannot be found guilty of it. The first indicates the realm of action; power becomes an indicator of politics as the ‘art of the possible’. The second assesses possible blame. Since power is often conceived as a counterfactual, that is, about things which could have been otherwise, invoking power is to call for a justification of why things were done the way they were. As a result, choosing concepts of power which are relatively narrow diminishes the realm where ‘something can be done’ and in which action needs to be justified; unintended effects, for instance, are handled as regretful but unavoidable collateral damage. Inversely, wider concepts of power suggest DIIS WORKING PAPER 2010:34 realms for action, even where there may be none. The fundamental point for such a performative analysis of power – not ‘what does it mean’, but ‘what does it ‘do’ – is that invoking the presence of power politicises issues. This also explains a curious paradox. Scholars and practitioners often engage in debates about where power ‘really’ lies, for example, whether it is hard or rather ‘soft’. By doing this, they must appeal to an underlying idea that we can know this is in a somewhat objective manner. If power were measurable, however, such debates would be quite pointless. Precisely because power is not as fungible as money and its understanding is to some extent conventional, observers try to shape the common understanding and fix the meaning of what power is and where the power ‘really’ lies. For such understandings have authoritative effects on national security and foreign policy doctrines when used to define the national interest, as well as on actual political rank and standing when an actor’s main potential power resource comes to be considered insignificant. CONSTRUCTIVIST VIEW OF POWER IN INTERNATIONAL RELATIONS2 In the 1980s, constructivism has appeared as a new turn in the theorising of international relations (IR). Its success was helped by the unexpected end of the Cold War. Although the Soviet Union was militarily not less powerful than before – let alone if compared with the early post-1945 period up until the 1960s – it decided to peacefully retrench from its positions in Eastern Europe. If the balance Encyclopedia entry forthcoming in Keith Dowding, ed., Encyclopedia of Power, London et al: Sage Publications, 2011. 2 of power was to be the main theory of IR, it met here an anomaly, not because it did not predict the event, but because, according to its tenets, such an event was not to happen in the first place. For constructivists, the end of the Cold War showed that a materialist understanding of power, and balance of power theories with it, were woefully insufficient, since outcomes in international politics could not be explained by some shifting balances of capabilities. By criticising the explanatory role of power, constructivism aimed at the core of established IR theories In order to more precisely establish the constructivist view of power, it is necessary first to introduce into constructivism and develop its implications for understanding and conceptualising power. Constructivism Constructivism can be understood as a metatheoretical commitment which is based on three characteristics. First, it makes the epistemological claim that meaning, and hence knowledge, is socially constructed. It is constructed, since concepts are the condition for the possibility of knowledge. Our senses are not passive receptors of ‘given’ facts. The very identification of facts out of the ongoing noise is dependent on pre-existing notions that guide our view of the world. This knowledge is moreover socially or intersubjectively constructed. Concepts are part of language. Language can neither be reduced to something subjective nor objective. It is not subjective, since it exists independently of us to the extent that language is always more than its individual usages and prior to them. It is not objective, since it does not exist independently of our minds and our usage (language exists and changes through our use). It is intersubjective. 11 DIIS WORKING PAPER 2010:34 Second, constructivism makes the ontological claim that the social world is constructed. As in John Searle’s famous example about a money bill, it is only for our shared beliefs that this piece of paper is ‘money’. As all people who have had to go through periods of hyperinflation would recognise, the moment that this shared belief ceases to exist, the bill is literally no more than a piece of paper. This assumption does not entail that everything is constructed, but it covers that part of reality in which the social sciences are usually interested. Hence, the physical type of support for money (paper, plastic, etc.) is usually not the most relevant for social analysis. What is most relevant is the social or institutional fact; the ontological result of ‘our making’. Third, since constructivism clearly distinguishes and problematises the relationship between the levels of observation and action, it is finally defined by stressing the reflexive relationship between the social construction of knowledge and the construction of social reality. In other words, it focuses on reflexivity. On the micro-level, reflexivity has to do with what Ian Hacking calls the ‘looping-effect’. Categories we use for classifying/naming people interact with the self-conception of those people. Whereas it makes no difference to stones how we classify them, it can make a difference to people and affect their self-understanding and behaviour. Identity thus becomes a crucial term for constructivism. On the macro-level, reflexivity refers to ‘self-fulfilling prophecies’. The concern in the response to Samuel Huntington’s ‘Clash of Civilisation’ thesis had much to do with this reflexive relationship between knowledge and the social world. Whether or not the main fault lines of conflict really have to be thought in this way, if all people assume they do, and act accordingly, the world would indeed become one of inevitable clashes of civ- 12 ilisations. Assuming the claim to be true, our actions would tend to produce the very reality the claim was only supposed to describe. But the relationship between social reality and the social construction of knowledge also works from social facts to knowledge, a component perhaps less touched upon in constructivist writings. Constructivist conceptualisations of power This meta-theoretical commitment has implications for the type of social theories which would be compatible with constructivism. And those theories, in turn, have implications for the types of power which can be conceived therein. Constructivism is part of the interpretivist family of social theories. As such, it cannot conceive of power in terms of resources alone. People act towards objects on the basis of the meaning they give to them: objects themselves do not determine their meaning. Nuclear missiles might be mighty weapons; small Luxemburg does not fear its huge French neighbour for them. A fortiori, constructivism is not prone to repeat what Robert Dahl once called the lump fallacy of power, where all possible power resources would be mixed and added. Such an aggregate power (resource) assessment, independent of the actor’s understandings and the contingent situational setting, would not only be wrong, but conceptually impossible. This makes constructivism more receptive to a relational understanding of power. Often confused with a relative understanding of power – one’s power resources are always to be seen in relation with the other’s power resources – such an understanding sees power defined by the specific relation between actors. Here, power lies not with given (re)sources, DIIS WORKING PAPER 2010:34 but can be established only once we know the precise scope and domain of the relation, that is, one must state who is being influenced and in what way. It includes hence an interaction and person-specific component. If, for instance, power is defined as the capacity to get B to do something he/she had not planned to do, then this implies knowing the specific plans of B before being able to assess whether A’s action had any effect. Yet, constructivist theorising would give a communicative twist to this, insisting on the role of open or tacit recognition which, in turn, relies on a wider social or cultural context. Such recognition is typically based on conventions, since, as mentioned above, resources are given weight not by themselves, but by shared understandings in social relations, and also since the recognition of a general power status is social. Just as individual communications are part of and make sense within the context of a language at large, the relational aspect of power is conceived in this wider manner so as to allow social norms to become visible in their role for the assessment of power as authority. As a corollary of the interpretivist and communicative setting, constructivists will not use power in terms of an efficient cause. Power is part of constitutive relations and effects (see also below): a master does not ‘cause’ a slave, but both, and their respective powers, are constituted through this masterslave relations. For the same reason, constructivism will view power in an often impersonal and hence also not necessarily intentional way. Invoking certain metaphors or historical analogies can be very influential, whether intended or not, since they mobilise a pre-given understanding. The particular way issues are framed empowers certain arguments and actors at the expense of others. If a situation is understood in terms of the ‘lessons of Mu- nich’, pleading for negotiations becomes an indefensible act of ‘appeasement’; an understanding in terms of the lessons of the First World War would make negotiations an act of prudence to pre-empt a further escalation nobody wanted. This power of existing biases is ‘impersonal’ to the extent that it is done through a set of common understandings or discourses, rather than reducible to the interpretation of one person; it is intersubjective not subjective. But, just as language, to be effective it requires persons mobilising it. Finally, constructivism is interested in the power aspects of performativity, where it relies mainly on speech act theories and Foucauldian approaches. If the categories with which we order the world are themselves part of, and can significantly affect, the order in the (social!) world, then they are a crucial element to understand power in any society. So does, for instance, the category ‘failed states’ interact with some states in their self-understanding and subjectivity and therefore change the social world and do not just describe it. It also prompts and legitimates certain actions, which would not have been legitimated by other categorisations, such as international interventions which overrule the otherwise fundamental norm of sovereignty. Applied to the concept of power itself, such a performative analysis can also look at the way the analysis of power affects power: the ‘power politics of power analysis’. Constructivism-inspired analyses of power in IR Constructivism redefines power at the systemic and at the agent level in IR. Its systemic analysis of power is often looking at the origins of consent in terms of power relations, i.e. at issues of tacit legitimacy. It is therefore close to power concepts of the family 13 DIIS WORKING PAPER 2010:34 of authority. But rather than looking at formal or institutional authority, constructivism is interested in the intersubjective practices of power, not in the position of authority, but rather in what ‘authorises’, ‘legitimates’ or ‘empowers’. Moreover, it is not necessarily looking at intentional or agent power, but at the impersonal effects of discourses and/or habits for the production and re-production of order, in particular there were practices go without saying, appear natural and are therefore the perhaps most effective power relations there are. At the actor level such a view implies an emphasis on the process of interest formation as a primary locus for power relations. Constructivism insists in making this interest formation part of the analysis – and not just simply assumed – something which cannot be derived outside of the specific interaction and the wider culture or shared understandings within which it takes place (from relational to communicative, see above). And for constructivists, interests cannot be understood outside of such cultures in terms of shared constitutive norms, of shared knowledge and understandings, and also through the effects practices have on self-understandings or identity. For constructivists, what we want follows from who we are. Such a view informs constructivist views of power, although they may not always be openly framed as power analyses. One larger research agenda is about the background knowledge or constitutive ‘rules of the game’ which define the competent player and the effective moves. This has been applied both to the world of diplomats, but also to the world of experts. Richard Ashley in particular has analysed how the ‘authorised’ expertise most often defined through tenets of the realist school in IR 14 systematically enacts conceptual blackmails and biases that marginalise other practices. Anna Leander, to cite another Bourdieu-inspired example, has analysed the ‘epistemic power’ of Private Military Companies when they shape the understanding of security and the self-understanding of its actors, as well as their ‘structural power’ in reproducing a field of security characterised by experts which authorise an increasingly technical and military understanding of the field – just when the security sector had started to be de-militarised – and are, in turn, authorised through it. With regard to their focus on identity, constructivist scholars have not only looked at the impersonal effect of discourses and/ or practices on self-understandings, but also on the ‘power politics of identity’. If identity is crucial for interest formation, then it is only a small step to analyse how diplomatic practices, sometimes intended, can try to blackmail actors by taking profit from contradictions in an another actor’s self understandings or between its action and selfrepresentation. Janice Bially Mattern calls this process one of ‘representational force’. As her study on US-UK relations during the Suez crisis in 1956 shows, the US could exploit such tensions to make the British government change its behaviour such as to conform to a certain self-understanding of what it stood for. Finally, performative or reflexive analyses of power study the conventions of power definitions, the definitional struggles and their effects on the social world. As mentioned earlier, constructivists, and not only, would reject any power index based on some resource aggregate as basically meaningless. Since different types of power resources are not commensurable (how much does one Bio. people weigh compared to DIIS WORKING PAPER 2010:34 running a world reserve currency?) and depend for their actual value on the meanings attached to them, their ranking and measure is but the effect of a shared convention which establishes their efficacy, their status and the status of actors. This convention informs the type of interests and hence most rational policies. Before diplomats can start counting, they must first agree in what counts. If all diplomatic actors come to agree that the authority given by cultural attraction, but not by military resources, weigh much in our times of globalisation, this shared idea will strongly influence the status and hence privileges of particular actors in world affairs. Less ambitiously and applied to one country only: if a certain understanding of power becomes predominant in one country, it redirects the foreign policy of that country, as can be seen in definitions which stress ‘soft power’ (thus de-militarising foreign policy) and other attempts to resist it. Given their conventional status, there is a power politics of power analysis. As a result, some of the constructivist research agendas converge with Foucauldian approaches in their understanding of order as diffused practices of rule, rather than as clear and/or formal hierarchies. Such convergence can be seen, for instance, in the analysis of how international standards, invented often by private actors, are practices of rule, once they become accepted convention and interact with the actors and issues they were supposedly only neutrally measuring (e.g. credit rating). Similarly, to give a last example, Ole Jakob Sending and Iver Neumann analyse the role of non-state actors not that much in opposition to the political power of states, but as part and parcel of a decentralised and self-disciplining logic of global order. RELATIONAL POWER 3 For its defendants, relational approach to power is not just one type of power – power as applied in a relationship – it is a basic characteristic of all power. In a relational understanding of power, power is not the possession of a person, nor does it correspond to a mere production of effects; it is constituted within a social relation. Only by knowing the respective value systems and beliefs specific to the relationship can the analyst attribute power. Power is here explicitly understood within the social world (as opposed to, for example, electric power). Such a view has important consequences for the analysis of power, in that its conceptualisation significantly differs from approaches that regard resources as either the locus of power or a sufficient proxy for it. But the conceptual solution also produces, in turn, several internal problems. One difficulty derives from linking power to causality. Doing so tends to look for power as a master-cause in the analysis of behaviour and outcomes. To assure that role, the very assessment of power has to factor in all situational qualifications of the respective social relationship under analysis, with the result of over-blowing the role of power. A second and related problem has to do with the possibility of giving an overall picture of power in a society. Initially, the conceptual move to relational power was meant to criticise simple and aggregate ‘lump’ concepts of power, justifiably requiring a domains and situation-specific analysis of power instead. But pushing this research based on relational specificity to its conclusion, this analysis eventually risks undermining any aggregate view of power within a polity. Encyclopedia entry forthcoming in Keith Dowding, ed., Encyclopedia of Power, London et al: Sage Publications, 2011. 3 15 DIIS WORKING PAPER 2010:34 A relational conceptualisation The relational conceptualisation of power takes place in the context of post-Weberian definitions of power. Weber had defined power as any chance (and not ‘probability’, as often translated) ‘within a social relation to impose one’s will also against the resistance of others, regardless of what gives rise to this chance’. For Dahl, A has power over B to the extent that he can get B to do something that B would not otherwise do. Hence, the main characteristic of a relational approach is that it locates power in a human relationship, thus distinguishing it from the sheer production of effects (power in nature). At the same time, ‘relational’ is not to be confused with ‘relative’ in that it means something different from the bottom of a balance sheet where power corresponds to one’s net ‘amount’ when the power (or worse, the resources) of others has been taken into account. Such relational concepts of power take issue with a vision of power in terms of its resources or instruments: power exists in and through a relation, it is not the possession of any agent. In a famous example, Bachrach and Baratz illustrate this with reference to a situation in which a sentry levels his gun at an unarmed intruder, whom he orders to halt or else he will shoot. If the intruder stops, it seems the threat has worked: the sentry has exercised power. Not necessarily, they say. If the intruder was himself a soldier, he may obey because that is what a soldier does when receiving an order from a sentry. The alleged power resource was ineffectual here, since it was the intruder’s value system that made him obey, not the gun. Inversely, if the intruder does not obey and gets himself killed, we may again not be seeing a power relationship. Strictly speaking, the killing of the intruder is not power, since the intruder 16 apparently valued entering the base more than his own life; the killing only shows the ultimate powerlessness of force (violence) in the face of a suicide attack. (In a more strictly Weberian reading, however, it would be fair to say that the sentry exercised power in imposing his will – not to allow anyone unauthorised to enter the base – against the resistance of the intruder.) Pushing the example to its extreme, the intruder may have wanted to commit suicide but gets the sentry to do it for him. In this case, the intruder, by being shot, exercises power over the sentry. The central point is that no analysis of power can be made without knowing the relative importance of conflicting values in the mind of the power recipient, if not also of the supposed power-holder. The capacity to sanction and the resources on which the sanctions are based are a part of power analysis, but in themselves insufficient to attribute power, since what counts as a sanction in the specific power relation is itself dependent on the specific values in the minds of the people involved. Thus, a relational conception reads power relations through the eyes of the recipient, or, more precisely, looks at all the actors involved as potential recipients of power relations. For this reason, Carl Friedrich’s ‘rule of anticipated reactions’ has retained its prominence in relational analyses of power. Politicians winning most of their political battles may be seen as having much power; but such victories may also express powerlessness in so far as they only come about because politicians carefully chose the few insignificant fights in which they anticipated standing a chance of winning. Hence, by concentrating on the recipient of power exercises, relational approaches also stress the ‘latent power’ of actors (‘having power’) who do not necessarily need to act (‘exercis- DIIS WORKING PAPER 2010:34 ing power’) – all it takes is that the recipients adjust their behaviour in pre-empting real or even imagined (negative) sanctions, whether or not the power wielder is even aware of this in every case. The causal link from sanctions to the altered behaviour of the recipient is upheld, but the explanation is shifted from the recipient of power to its wielder, not the other way round. Causality and aggregation Opposing a concept of power that is possessive, i.e. the reduction of power to its instruments or resources, can be done in a number of different ways. To demonstrate its implications for the understanding of causality and its use in analyses, a comparison with a strict dispositional conceptualisation is useful. Peter Morriss’s cogent conceptualisation of a dispositional approach defines power neither as a resource or vehicle, nor as an event or its exercise, but as the capacity to effect an action. In such a conceptualisation, the stress is on the effecting of outputs, not the full analysis of outcomes. The relational component of a social analysis is not denied, but postponed until after the assessment of power as such. Hence, this dispositional conceptualisation does not tie power to resources, nor does it tie it causally to the explanation of social outcomes. Causality is to be established elsewhere. Indeed, for Morriss, power itself does not explain events. This is quite different from the original intuition of those who have proposed a relational approach to power. For Dahl, Oppenheim and Baldwin, all defenders of a relational approach, the interest in power and its synonymous use with influence stems from the possibility that power can be used as a kind of core causal variable. For Oppenhe- im, saying that B was influenced by A is no mere description, but also a partial explanation of B’s conduct: actor B did something because of A’s power. Hence, the analysis moves to the establishment of the role of power in a causal chain for the explanation of behaviour. By being so closely connected to causality, relational power analysis needs to avoid the usual tautologies which conflate power with either poles of the causal chain, i.e. which equate power with either outcomes or resources. The proposed solution is twofold. First, the very assessment of what has to count as a power resource or power base (and not just an instrument or vehicle) must be defined and qualified by the particular type of power relation under analysis. Relational analyses do not assume a high fungibility of power resources. (A good is fungible if it is of such a nature as to be freely replaceable by another of like nature or kind without decreasing in value.) As a result, the analysis of power has to delimit carefully the domain within which a policy instrument can be assumed to have effect, that is, to count as an actual power resource in the first place. Secondly, since the analysis is based on the values of the actors in the power relation, especially if individual actors are analysed, the actual values of the actors need to be factored in before we can identify what can count as a power resource or base in this particular relation, and not just in a general domain. In addition, skill and motivation must also be added before we can establish the power base. All this is necessary to maintain the causal link between power base and outcome. This avoids, almost by definition, the risk of non-falsifiability prominent in those studies that are marred by what Baldwin called the ‘paradox of unrealised power’. In such an 17 DIIS WORKING PAPER 2010:34 explanation, even powerful actors can lose against less powerful ones because they did not, for one reason or the other, use all their power. When allegedly overriding power does not to translate into influence, it is not because an actor lacks sufficient power, but because of conversion failures (a lack of political will, for example). Some power gets lost on the causal path, and power analysis degenerates into finding ‘conversion failures’ which end up making the same outcome explicable by opposite causes (power or powerlessness). Therefore, whatever the outcome, it does not undermine the initial power assessment. The value of resources is ultimately objectified, and everything that does not fit in terms of influence is explained away with reference to incompetent agency: power resources never fail, only politicians. However, in a relational conceptualisation of power, power is causally connected to outcomes, and hence this cannot be: power bases do not fail, otherwise they could not to be considered such to start with. And so the very definition of what can count as a power resource needs to be made far more comprehensive and situation-specific in order to maintain a close causal link. It was not that power failed, but that the analyst did not qualify it correctly. This tendency to heavily qualify the understanding of the very power resources or power base is further exacerbated by the need to analyse power relations from the receiving side. As noted above, the analysis does not start from the power wielder, but the power recipient. In understanding the behaviour of that recipient, power is said to play a role. By tying the concept of power to influence in a causal scheme, the tendency must be to look for all possible factors that may have influenced the recipient’s behaviour. But then, rather than focusing on the 18 power wielder’s capacity to get someone else to change behaviour – i.e. the manipulative and agency-oriented view prominent in the underlying relational definition of power – the power analysis strives for a fairly comprehensive view of all the factors that influence the recipient’s behaviour. By understanding power as causal influence, the temptation is then great to incorporate ever more of those causal (or permissive) factors into the very assessment of power. It is therefore no coincidence that Baldwin ends up adding even social norms to his catalogue of factors which define the specific environment within which power analysis takes place. Pushed to its extreme, such analysis will embed so much into the qualification of the power base that only purely contingent factors – ‘luck’ in Dowding’s approach – are left as causes outside of power. Besides causality, a second important issue raised by the relational understanding of power is the issue of aggregation or generalisation. Ever since his initial salvo against power-elite approaches, Dahl has castigated what he called the ‘lump’ concept of power, in which all possible resources are added up to establish a single measure for an actor. This argument is of particular significance in International Relations, where it undermines the widely used balance of power theories. Such theories invoke the distribution of power as the main cause in understanding the behaviour of the state and its effects on the international system. To establish the distribution of power in the international system, the analyst must locate the general poles of power. These poles of power, in turn, can be defined either in purely military terms or in terms of the combination of their different resources. In the first case, a relational approach would show that the mil- DIIS WORKING PAPER 2010:34 itary is not necessarily the most significant resource in all domains, and hence it is not possible to reduce the international power structure to a single one, with one single index or measure. And once one opens up that Pandora’s box, given the problem of fungibility, a simple addition can no longer work independently of domain, scope, the values of the recipient or the other factors that are now to be included in the power assessment. As a result, the ubiquitous balance of power analyses must assume a unified and general international power structure, which a relational approach can show not to be workable in principle. Without a ‘lump’ concept, general balance of power theories cannot work. But pushed to its logical conclusion, such a relational understanding then seems to imply that only very circumscribed analyses of power are possible in which all the factors have been included, as noted above. In other words, as Dahl points out, this means that there could be as many power structures in any political system as there are individuals who impute different intentions to other actors. However, having made it so demanding to define power in the first place and having broken down power analysis into many autonomous small power analyses, a relational approach risks losing the overall picture from sight. It is not just that the generalisation of power bases across domains cannot be taken for granted: the aggregation of powers that is necessary for qualifying the polity as such may no longer be possible. Hence, despite having been conceptualised within and for the analysis of ‘community power’, later theoretical developments of relational power seem to make it increasingly difficult to answer the question ‘Who governs?’ for a polity in general. FUNGIBILITY OF POWER RESOURCES4 A good is fungible, if it is of such a nature as to be freely replaceable for another of like nature or kind without decreasing in value. Convertibility and mutual substitutability are hence forms of fungibility. Prime examples of fungible goods are money bills or currencies. In the analysis of power, and in a looser sense, the problem of fungibility refers to the issue whether or not different types of resources (e.g. military, economic, cultural, diplomatic) have the characteristic of being freely exchangeable or replaceable. This issue has gained prominence for two reasons. If resources are highly fungible, i.e. can be mutually substituted without losing much of their value, then this allows them to be aggregated, so as to permit the construction of overall power resource indexes. This is crucial for balance of power theories, since they rely on such an understanding of aggregate power for explaining the dynamics of the international system and/or the options (and behaviour) of individual states. Also, this time on the level of agency, if resources are highly fungible, then power could become the equivalent of money in economic theory through which different aims can be weighed on a common scale. This allows a unique value of utility as the maximisation of different aims can all be converted into power. Yet, research has shown that the assumption of high fungibility is highly problematic, if not mistaken. Balance of power theories used to focus mainly on military resources. The greatest stability of the system, and hence the least violent resolution of conflicts is expected Encyclopedia entry forthcoming in Keith Dowding, ed., Encyclopedia of Power, London et al: Sage Publications, 2011. 4 19 DIIS WORKING PAPER 2010:34 when the different states (or coalitions) have roughly equal military capabilities, since it would be irrational to wage war in such circumstances. Also, since all states wish to avoid being in an inferior position which invites aggression, their individual security policies will tend to reproduce collective power balances. As long as purely military resources are sufficient to account for such an equilibrium, the issue of fungibility does not arise too prominently, although some see the advent of nuclear weapons as a qualitative break. But the moment power theory is applied to explain more generally the outcomes of conflicts (with the expectation that the actor with more capabilities will inevitably win), fungibility becomes a crucial issue. First, it makes an ex ante assessment of capabilities difficult: how do we compare different types of resources, like, e.g. speaking the international lingua franca and running a major trade surplus (which some would dispute to be a resource in the first place) and how do they add up? Yet without an ex ante assessment of the overall power relation, any outcome can be ex post re-arranged to fit an explanation in terms of power differentials (power analysis becomes tautological). Second, such a power relation is usually taking place in different domains. While military resources may be crucial in a military dispute, how effective are they in an economic one? Finally, power relations do not necessarily take place between some interchangeable actors. Although military means would be potentially useful in a conflict with enemies, they can hardly be used with allies. Indeed, relations of amity or enmity profoundly affect the value of resources in the first place. As David Baldwin has often argued, all this leads by necessity to a power analysis, which is relational, multidimensional and highly situ- 20 ational, and hence precisely not able to rely on a fungible power resource assumption. Indeed, attempts to aggregate such resources independent of such factors are prone to the ‘lump fallacy of power’ in Robert Dahl’s felicitous phrase. But fungibility is also important for devising a power-utilitarian theory of behaviour. A basic assumption of classical realism states that actors struggle for power: they try to maximise their interest defined in terms of power. This produces a type of rationalist theory which runs parallel to economic theory: power takes the place for money, security for utility. States are maximising their security which can be expressed in terms of power. The underlying power-money analogy, and with it the attempt to directly apply economic theory to politics, has however been criticised even by some realist writers. Raymond Aron has argued that the different aims of economic actors can be made commensurable through money, whereas no such commensurability exists in politics. Aron points to the difference between an economic theory which works against the background of a monetarised economy, i.e. one where money has taken over the double role of standard of value and means of exchange, and a political theory with no such equivalent in actual politics, and where aims cannot be reduced to a common power scale. Still, as Robert Art insists, even if capabilities are not measurable in a clean way, it would be absurd to deny that state leaders are (at times) able to find some rough sense of each others’ capabilities. But this argument implies that the value of power resources and their level of fungibility is ultimately a function of interpretation and, to be effective, of a convention shared among state DIIS WORKING PAPER 2010:34 actors. Exactly because capabilities cannot just be added and weighed independent of the interaction and shared understandings, actors need to find proxies for conducting their bargaining – proxies on which they have to agree if they are to have effect. The value of resources and their fungibility do not define the political game, but vice versa (although not only). Hence, this realist answer may save power analysis only by moving it onto constructivist terrain. 21 DIIS WORKING PAPER 2010:34 REFERENCES AND FURTHER READING Aron, Raymond. (1962). Paix et guerre entre les nations [Peace and war among nations]. Paris: Calmann-Lévy. Art, Robert J., Baldwin, David A. (1999). A Debate: Force, fungibility, and influence. Security Studies, 8(4), 173-189. Ashley, Richard K. (1987). The Geopolitics of Geopolitical Space: Toward a Critical Social Theory of International Politics. Alternatives, XII (4), 403-434. Ashley, Richard K. (1989). Imposing International Purpose: Notes on a Problematique of Governance. In Ernst-Otto Czempiel and James Rosenau (eds.), Global Changes and Theoretical Challenges: Approaches to World Politics for the 1990s (pp. 251-290). Lexington, Mass.: Lexington Books. Bachrach, Peter & Baratz, Morten S. (1970). Power and Poverty: Theory and Practice. New York: Oxford University Press. Baldwin, David. (1985). Economic Statecraft. Princeton: Princeton University Press. Baldwin, David. (1989). Paradoxes of Power. Oxford: Blackwell. Barnett, Michael & Duvall, Raymond. (Eds). (2005). Power in Global Governance. Cambridge: Cambridge University Press. Berenskoetter, Felix, and Williams, Michael J. (eds.). (2007). Power in World Politics. London, New York: Routledge. Bially Mattern, Janice. (2005). Ordering International Politics: Identity, Crisis, and Representational Force. New York: Routledge. Dahl, Robert. (1968). Power. In D. L. Sills (Ed.), International Encyclopedia of the Social Sciences, vol. 12 (pp. 405-415). New York: Free Press. Dahl, Robert & Stinebrickner, Bruce. (2003). Modern Political Analysis, 6th ed. London, New York: Longman. Dowding, Keith. (1996). Power. Minneapolis: University of Minnesota Press. Gill, Stephen, and Law, David. (1989). Global Hegemony and the Structural Power of Capital. International Studies Quarterly 33(4), 475-99. Guzzini, Stefano. (1993). Structural Power: The Limits of Neorealist Power Analysis. International Organization, 47(3), 443-78. Guzzini, Stefano. (2000). A reconstruction of constructivism in International Relations. European Journal of International Relations, 6(2), 147-182. Guzzini, Stefano. (2000). The use and misuse of power analysis in international theory. In Ronen Palan (Ed.), Global Political Economy: Contemporary Theories (pp. 53-66). London, New York: Routledge. Guzzini, Stefano. (2005). The concept of power: a constructivist analysis. Millennium: Journal of International Studies, 33(3), 495-522. Keohane, Robert. (1984). After Hegemony: Cooperation and Discord in the World Political Economy. Princeton: Princeton University Press. Leander, Anna. (2005). The Power to Construct International Security: On the Significance of Private Military Companies. Millennium: Journal of International Studies, 33(3), 803-826. 22 DIIS WORKING PAPER 2010:34 Morriss, Peter. (2002 [1987]). Power: A Philosophical Analysis. Manchester: Manchester University Press. Onuf, Nicholas Greenwood. (1989). World of our Making: Rules and Rule in Social Theory and International Relations. Columbia, S. C.: University of South Carolina Press. Oppenheim, Felix E. (1981) Political Concepts: A Reconstruction, Oxford: Basil Blackwell. Sending, Ole Jakob, and Iver B. Neumann. (2006). Governance to Governmentality: Analyzing NGOs, States, and Power. International Studies Quarterly, 50(3), 651-672. Strange, Susan (1988) States and Markets: An Introduction to International Political Economy. New York: Basil Blackwell. 23 DIIS WORKING PAPER 2010:34 24