Michael Barnett and Martha Finnemore, Rules for the World: International Organizations in Global Politics, Cornell University Press, 2004.
|
RATING: 25
|
Buy this book?
|
No
|
This book is bad, despite or perhaps because of its sophistication, because it applies a general theory of the tendency of bureaucracy towards dysfunction to international organisations (IOs) without having any theory of the social world, and therefore without means of deciding what actually counts as dysfunctional. Instead, its point of reference is normative - the 'liberal values that are the foundations of a global liberal culture' (173). The theoretical impasse at the heart of the book becomes clear right at the start. The authors' goal is to understand 'why IOs behave as they do', and they rightly reject the idea that they just 'do what states want', arguing instead that they tend to 'develop their own ideas and pursue their own agendas' (2). But things immediately go horribly wrong, in a textbook example of the weakness of 'middle-range' theory:
'Scholarship on organizations generally (not just IOs) has made it abundantly clear that organizations routinely behave in ways unanticipated by their creators and not formally sanctioned by their members. Organizations that start with one mission routinely acquire others. Organizations adapt to changing circumstances in unanticipated ways and adopt new routines and functions without getting approval from their "stakeholders". Organizations are notoriously resistant to reform or redirection because change threatens entrenched organizational culture and interests. International organizations evidence all these familiar traits. They exhibit mission creep. They wander far from their original mandates and into new terrains and territories. They develop new rules and routines in response to new problems that they identify. They formulate rules that are politically safe and comfortably routine rather than efficient or effective. We want to understand these behaviors in IOs' (2).
The problem here is one that I have noted previously in relation to theories of 'institutional persistence' and path dependency (Paul Cammack, 'New institutionalism: predatory rule, institutional persistence and social change', Economy and Society, 21, 4, 1992, pp. 397-429). If institutions stay the same, have they have ossified and become anachronistic, or just kept doing the job they were set up to do? If they change, is it because the world has changed and they have responded, or because something has gone wrong? There is no way of telling, if your theory of organisations is not backed up by a theory of change in the world at large. In this case, Barnett and Finnemore flip between possibilities: organisations may take on new missions in response to new problems they identify; or they may fail to adapt because of entrenched cultures and interests, drift off course, settle for safe and comfortable routines, and lose the capacity to carry out their original mission. Their 'constructivist' approach, they say, will 'help explain the power they exercise in world politics, their propensity toward dysfunctional, even pathological, behavior, and the way they change over time' (3, see also 29-44). Unfortunately, it won't tell them which situation is which. In practice, IOs are brought under a particular theory of bureaucracy, which sees it as having its own 'internal logic and behavioral proclivities', prior to any analysis of the substance of what particular international organisations do and why. The 'constructivist' part of the argument itself does not require this, as it simply and reasonably says that IOs exercise power through their ability to make rules that 'not only regulate but also .. constitute and construct the social world' (3). The title of the book - Rules for the World - reflects this, but the substantive analysis - as reflected in the titles of the first two chapters, Bureaucratizing World Politics and International Organizations as Bureaucracies - pursues the 'propensity toward dysfunctional, even pathological, behavior' (3). This hybrid analytical machinery is then turned onto the issue of the relationship between IOs and states: Barnett and Finnemore, rejecting as noted the statist claim that IO behavior follows directly from state demands, look for 'consistent patterns of IO action that cannot be traced to state pressures' (10), and attempt to explain the five kinds of autonomous IO behavior they see (independent action consistent with state interests; operations in areas to which states are indifferent; failure to carry out state interests; opposition to state interests; and action that changes state interests - see also 27-29), taking care in their examination of IO decisions to 'go beyond the immediate to consider how environmental forces shaped the rules that IO staff use to see the world and make decisions' (12). This, they claim, 'allows us to determine how, when, and why IOs are able to act in ways that are not demanded by states and thus to exhibit different kinds of autonomy' (12). True, but as noted above it doesn't allow them to decide whether the resulting behaviour is 'pathological' or not.
There are three chapter-length case studies, of the IMF's use of its professional expertise to regulate and reconstitute the economies of Third World states, the emergence of a 'repatriation culture' at the UNHCR, and the pathology of UN peacekeeping in Rwanda, all of which reflect these problems in specific ways. I focus primarily on the IMF case study, but comment briefly on the other two as well.
Remarking that analysis of dysfunction 'requires that we develop criteria so that we can identify it when we see it,' and clarifying that they care most about 'those occasions when behavior significantly exceeds the inevitable minimum of noise and leads the organization to act in ways that are inconsistent with its social purpose' (35), Barnett and Finnemore go on to offer five pathology-inducing mechanisms that flow from 'the defining features of bureaucracy itself' (39): irrationality of rationalization, bureaucratic universalism, normalization of deviance, insulation, and cultural contestation (39-41). No studies of international organisations are cited in the exposition of these five mechanisms. However, it is preceded by a brief discussion of dysfunction as a matter of perspective, and preceded and followed by thumbnail applications of the concept to the three case studies, with devastating consequences for the argument in the case of the IMF in particular. On perspective, the authors say, echoing Cox, that behaviour 'is dysfunctional only for something or someone; behavior by itself is neither functional nor dysfunctional,' and offer in response to this that they 'analyze potential pathologies from several different angles, showing how behaviors that seem self-defeating or undesirable from one perspective might make perfect sense from another' (36). This is fine in itself, although it is completely at odds with the previous appeal to stated social purpose as a criterion. But the manner in which it is applied is crucial. First:
We study a UN Secretariat that presents itself as a guardian of human rights but nevertheless turns a blind eye to crimes against humanity. We study financial stabilization programs that produce one failure after another. We study a UNHCR that violates refugee rights (35).
Second:
A UNHCR that fails to challenge states every time they violate refugee rights might be functional for interested governments and even for the organization, which must pick its battles wisely to preserve state support. IMF policies may create widespread misery in deficit countries but be extremely functional for global capitalism (emphasis mine). The decision by the UN Secretariat not to intervene in the Rwandan genocide might have been completely dysfunctional from a human rights perspective but extremely functional for strong states that wanted to stay out of the conflict and for a Secretariat that believed that the likely failure of the operation would fatally damage the organization (36).
And third:
We examine two cases of pathologies in detail: UNHCR's repatriation culture and its sanctioning of involuntary repatriation of the Rohingyan refugees from Bangladesh to Burma in 1994; and the Secretariat's institutional ideology of impartiality and its preference not to intervene in the opening weeks of the Rwandan genocide (41).
Commentary is almost superfluous. The cases of refugees and peace-keeping may not speak as much to loss of social purpose as to the pragmatics of limited power - and studies of only two conspicuously hard cases may not help us to decide on the general implications anyway. But the case of the IMF speaks directly to function and social purpose, and Barnett and Finnemore offer casually and in passing the thought that its purpose may be tied up with the health of global capitalism rather than with the 'success' or 'failure' of specific instances of financial stabilization. Its founding articles certainly suggest as much. This does not stop them harping on financial stabilization as a defining criterion for IMF action throughout the book (alongside intermittent broader characterisations of its role). Not surprisingly, the brief case study that follows is inconclusive. For all their theoretical huffing and puffing in earlier chapters, the authors don't identify a single instance of pathological behaviour (66), and they come down fairly and squarely on the view that by and large the Fund does what powerful states (the G7) want (68). They do argue that its ability to do so hinges on its claim to expertise, showing that it has devised increasingly wide-ranging rules in its work with member countries, and that the data it deploys contribute to remaking as much as describing the world. However, the development of the argument completely undermines the narrow identification of the IMF mission with financial stabilization, and points directly instead to a consistent concern for 'global capitalism'. They show that from the start the Fund saw short-term adjustment as a necessary prelude to medium- or longer term increases in productivity, thereby invalidating any claim of 'mission creep'. As they document, the means it used changed as the focus of its activity switched from the developed to the developing world and again as the legitimacy of structural adjustment was called into question. Its increasing focus on the domestic economy, its development of conditionality tools and its engagement with 'poverty alleviation' fit this perspective perfectly well: while the means applied may have expanded, the social purpose remained unchanged. Yet Barnett and Finnemore never explore their own suggestion that the IMF might be 'functional for the development of global capitalism' - nor do they explore in any detail the complementarity of its work with that of the OECD and the World Bank, among other organisations - an essential step if one wishes to evaluate its role. At the end of the chapter and of the book they stick to the narrow view of the Fund's purpose, claiming that its application of its 'objective models has failed to stabilize borrower economies in a great many cases' (71) and associating it again with 'the pursuit of financial stability through markets' (157). Their assessment of its success in pursuing its original mission, in the words of its original 1944 Articles, 'to facilitate the expansion and balanced growth of trade, and to contribute thereby to the promotion and maintenance of high levels of employment and real income and to the development of the productive resources of all members as primary objectives of economic policy', is therefore fatally flawed throughout; and of course they don't have the concepts (these being part of the toolkit of historical materialism) to understand how failure in such a mission might be explained without recourse to general theories of bureaucratic pathologies.
The other two case studies are more substantial, but do little to justify the overarching theme of bureaucratic dysfunction. In the first, the UNHCR gave increasing emphasis to repatriation as a durable solution from the 1980s, and in the case of the Rohingyas repatriated from Bangladesh in the mid-1990s it bent the concept of voluntary repatriation virtually out of recognition. But as the case was a hard one, and no evidence is given either of the overall evolution of numbers UNHCR was dealing with over time or the extent to which similar practice and interpretation applied elsewhere, it is impossible to tell whether this was (a) evidence of the 'normalization of deviance' (as invoked, 119), or alternatively (b) an isolated instance, or perhaps (c) a reflection of the limited powers and options of the organisation. In any case, repatriation was always one of the prescribed possible options, and it was little used in the early years, as we are repeatedly told, only because in the context of the Cold War it seemed unthinkable to return refugees to areas under Soviet control. So the shift towards repatriation is better described as moving from a temporary deviance to the full regime originally envisaged.
In the second case, the UN Secretariat failed to recommend intervention in the Rwandan genocide in 1994. It had had difficulty setting up the peace-keeping force, UNAMIR, in the first place (it was approved by the Security Council with 2,500 lightly armed troops, half the level requested), had it renewed in March 1994 only over initial US objections, and once the mass killing started, was working 'with New Zealand ambassador Colin Keating, the President of the Security Council during the month of April 1994, to oppose the U.S. desire to withdraw the operation completely in favor of a token presence' (145 and 203, ft. 98). The authors attribute the failure of the Secretariat to portray the violence to the Security Council as organised genocide and request intervention to the 'peacekeeping culture', the 'tendency to cling to preexisting beliefs and to reach premature closure' (148), and the fact that because its mandate was to anticipate and help stop a civil war from returning, 'that was all it could see' (149). This is entirely implausible, principally because the commitment to 'consent, neutrality and impartiality' was not a case of embedded and powerful rules that determined goal definition, as the 'irrationality of rationalization' is described (39) - on the contrary, it was a very recent and considered reversion to that approach to peacekeeping after it was abandoned in 1992 in favour of a more active practice of 'peace-making', leading to setbacks in both Bosnia and Somalia, as Barnett and Finnemore demonstrate at length. In any case, there is not a shred of evidence that the Security Council would have countenanced the tripling of forces on the ground as requested by General Dallaire, and in view of Dallaire's misgivings about the quality of his troops (especially in light of the scheduled withdrawal of the Belgian contingent), and the scale of the civil war and genocide, the likelihood that a larger force would be effective must have seemed low. None of this excuses the failure of the international community as a whole to challenge the organised genocide that unfolded. But to attribute the outcome to the UNHCR and particularly the Secretariat's pathological construction of the crisis is arbitrary and unconvincing.
The final chapter reflects on IO action from a cosmopolitan liberal democratic perspective, with a focus on issues of legitimacy and accountability. It builds on comments at the end of the first chapter, where Barnett and Finnemore claim that 'one consistent finding is that IOs tend to expand,' and hold out the prospect of an 'increasingly bureaucratized world' that is potentially worrying because 'the liberal norms embodied and promoted by these organizations are generally not matched with accountability or participation procedures that liberalism favours' (15). No attempt is made to provide empirical evidence to support this claim for IOs in general, and the theme then disappears until we come to the final chapter, 'The Legitimacy of an Expanding Global Bureaucracy', which asks again whether 'this expansion that is creating an increasingly bureaucratized world' is a good thing. At the same time, though, the claim is qualified: 'While the overall trend line suggests that IOs are expanding in size and scope, there are moments when IOs resist opportunities and even explicit invitations to expand' (158). One could go further - as they report, the UN briefly expanded its role into 'peace-making', but quickly drew back to its original mandate when it 'learned' that it had over-reached. This does not deter them from adducing 'learning from failure' as a mechanism promoting expansion (160), and three pages later they so forget themselves as to cite the UN Secretariat's reversion to an 'institutional ideology of impartiality' in a section on the 'pathology of expansion'. On their own account, on the admittedly slender evidence of two case studies, it seems eccentric to describe the UNHCR or UN peace-keeping in terms of expanding power and reach (170). On the authors' own account, they have been starved of both resources and authority, and caught between conflicting pressures, as reflected in a more realistic statement in the closing pages: 'As UNHCR action has created (and constituted) new constituencies, particularly advocates for refugees, contradictions between the demands of these groups and those of member states have made crafting policies increasingly difficult' (171). In any case, even if demonstrated, the expansion of IOs does not equate to an increasingly bureaucratized world, let alone to the emergence of 'increasingly powerful and autonomous bureaucrats' (173). Even on the limited and theoretically impoverished canvas deployed, Barnett and Finnemore give ample evidence of the limits to IO action set by states, and the tendency for the challenges to be met to outrun the resources available.
One persistent problem is that there are obvious deficiencies, from a critical political economy perspective, to an approach that treats all international organisations as generically the same despite the different arenas of their activity, and does so from a perspective drawn from Weber's analysis of bureaucracy. Such an approach is bound to miss the specific issues raised by the governance of the global economy, and the relationship between states, markets, international organizations and populations it entails. In this case, Barnett and Finnemore are able to identify some important empirical phenomena, but unable to provide a satisfactory analysis of them. Four successive statements from the final chapter pinpoint the scope and limits of their approach:
'As IO governance activities expand, they are not only helping to create a new structure to world politics but are also working to create and constitute new domestic societies. IOs appear to be steadily expanding their mandates in a convergent direction: all are increasingly involved in the domestic affairs of states, and, specifically, all are trying to create durable, modern nation-states that are organized around democracy and markets' (164).
'IO expansion, in short, entails creating particular kinds of states with particular kinds of interests' (165)
'The bureaucratization of world politics means that international organizations have more authority than ever before and therefore have more power over other actors than ever before' (165)
'This does not mean that authority is zero-sum' (ft. 27, p. 205)
It is not that Barnett and Finnemore do not identify something of importance, but rather that they are unable to develop their insight effectively. Some IOs, if not all, certainly are 'working to create and constitute new domestic societies'. But it is a stretch to apply the idea to the UNHCR, and in terms of the division of labour, more applicable to the OECD, with its relentless focus on competition in labour and product markets and advocacy of market-supporting social protection over universal welfare, and the World Bank, with its sights trained on forging ideal neoliberal citizens, than on the IMF. It is a mistake to focus so much on expansion as opposed to the core missions of IOs in this respect though, another to rely so heavily and with such self-defeating consequences on theories of bureaucracy, and another again to take constructivism as a starting point instead of attending first to the material and social forces driving change in the world. Class struggle in the context of the evolving world market creates particular kinds of states in which particular interests are dominant though always contested; IOs concerned with the governance of the world market work to reconcile states to it, and to build the hegemony of capital over labour on a global scale. So, as the third statement above and its immediate qualification in a footnote show, to address this in terms of the relative power of IOs and states (whether in zero-sum terms or not) in the absence of a theory of change in the social world leads to a dead end.
'Scholarship on organizations generally (not just IOs) has made it abundantly clear that organizations routinely behave in ways unanticipated by their creators and not formally sanctioned by their members. Organizations that start with one mission routinely acquire others. Organizations adapt to changing circumstances in unanticipated ways and adopt new routines and functions without getting approval from their "stakeholders". Organizations are notoriously resistant to reform or redirection because change threatens entrenched organizational culture and interests. International organizations evidence all these familiar traits. They exhibit mission creep. They wander far from their original mandates and into new terrains and territories. They develop new rules and routines in response to new problems that they identify. They formulate rules that are politically safe and comfortably routine rather than efficient or effective. We want to understand these behaviors in IOs' (2).
The problem here is one that I have noted previously in relation to theories of 'institutional persistence' and path dependency (Paul Cammack, 'New institutionalism: predatory rule, institutional persistence and social change', Economy and Society, 21, 4, 1992, pp. 397-429). If institutions stay the same, have they have ossified and become anachronistic, or just kept doing the job they were set up to do? If they change, is it because the world has changed and they have responded, or because something has gone wrong? There is no way of telling, if your theory of organisations is not backed up by a theory of change in the world at large. In this case, Barnett and Finnemore flip between possibilities: organisations may take on new missions in response to new problems they identify; or they may fail to adapt because of entrenched cultures and interests, drift off course, settle for safe and comfortable routines, and lose the capacity to carry out their original mission. Their 'constructivist' approach, they say, will 'help explain the power they exercise in world politics, their propensity toward dysfunctional, even pathological, behavior, and the way they change over time' (3, see also 29-44). Unfortunately, it won't tell them which situation is which. In practice, IOs are brought under a particular theory of bureaucracy, which sees it as having its own 'internal logic and behavioral proclivities', prior to any analysis of the substance of what particular international organisations do and why. The 'constructivist' part of the argument itself does not require this, as it simply and reasonably says that IOs exercise power through their ability to make rules that 'not only regulate but also .. constitute and construct the social world' (3). The title of the book - Rules for the World - reflects this, but the substantive analysis - as reflected in the titles of the first two chapters, Bureaucratizing World Politics and International Organizations as Bureaucracies - pursues the 'propensity toward dysfunctional, even pathological, behavior' (3). This hybrid analytical machinery is then turned onto the issue of the relationship between IOs and states: Barnett and Finnemore, rejecting as noted the statist claim that IO behavior follows directly from state demands, look for 'consistent patterns of IO action that cannot be traced to state pressures' (10), and attempt to explain the five kinds of autonomous IO behavior they see (independent action consistent with state interests; operations in areas to which states are indifferent; failure to carry out state interests; opposition to state interests; and action that changes state interests - see also 27-29), taking care in their examination of IO decisions to 'go beyond the immediate to consider how environmental forces shaped the rules that IO staff use to see the world and make decisions' (12). This, they claim, 'allows us to determine how, when, and why IOs are able to act in ways that are not demanded by states and thus to exhibit different kinds of autonomy' (12). True, but as noted above it doesn't allow them to decide whether the resulting behaviour is 'pathological' or not.
There are three chapter-length case studies, of the IMF's use of its professional expertise to regulate and reconstitute the economies of Third World states, the emergence of a 'repatriation culture' at the UNHCR, and the pathology of UN peacekeeping in Rwanda, all of which reflect these problems in specific ways. I focus primarily on the IMF case study, but comment briefly on the other two as well.
Remarking that analysis of dysfunction 'requires that we develop criteria so that we can identify it when we see it,' and clarifying that they care most about 'those occasions when behavior significantly exceeds the inevitable minimum of noise and leads the organization to act in ways that are inconsistent with its social purpose' (35), Barnett and Finnemore go on to offer five pathology-inducing mechanisms that flow from 'the defining features of bureaucracy itself' (39): irrationality of rationalization, bureaucratic universalism, normalization of deviance, insulation, and cultural contestation (39-41). No studies of international organisations are cited in the exposition of these five mechanisms. However, it is preceded by a brief discussion of dysfunction as a matter of perspective, and preceded and followed by thumbnail applications of the concept to the three case studies, with devastating consequences for the argument in the case of the IMF in particular. On perspective, the authors say, echoing Cox, that behaviour 'is dysfunctional only for something or someone; behavior by itself is neither functional nor dysfunctional,' and offer in response to this that they 'analyze potential pathologies from several different angles, showing how behaviors that seem self-defeating or undesirable from one perspective might make perfect sense from another' (36). This is fine in itself, although it is completely at odds with the previous appeal to stated social purpose as a criterion. But the manner in which it is applied is crucial. First:
We study a UN Secretariat that presents itself as a guardian of human rights but nevertheless turns a blind eye to crimes against humanity. We study financial stabilization programs that produce one failure after another. We study a UNHCR that violates refugee rights (35).
Second:
A UNHCR that fails to challenge states every time they violate refugee rights might be functional for interested governments and even for the organization, which must pick its battles wisely to preserve state support. IMF policies may create widespread misery in deficit countries but be extremely functional for global capitalism (emphasis mine). The decision by the UN Secretariat not to intervene in the Rwandan genocide might have been completely dysfunctional from a human rights perspective but extremely functional for strong states that wanted to stay out of the conflict and for a Secretariat that believed that the likely failure of the operation would fatally damage the organization (36).
And third:
We examine two cases of pathologies in detail: UNHCR's repatriation culture and its sanctioning of involuntary repatriation of the Rohingyan refugees from Bangladesh to Burma in 1994; and the Secretariat's institutional ideology of impartiality and its preference not to intervene in the opening weeks of the Rwandan genocide (41).
Commentary is almost superfluous. The cases of refugees and peace-keeping may not speak as much to loss of social purpose as to the pragmatics of limited power - and studies of only two conspicuously hard cases may not help us to decide on the general implications anyway. But the case of the IMF speaks directly to function and social purpose, and Barnett and Finnemore offer casually and in passing the thought that its purpose may be tied up with the health of global capitalism rather than with the 'success' or 'failure' of specific instances of financial stabilization. Its founding articles certainly suggest as much. This does not stop them harping on financial stabilization as a defining criterion for IMF action throughout the book (alongside intermittent broader characterisations of its role). Not surprisingly, the brief case study that follows is inconclusive. For all their theoretical huffing and puffing in earlier chapters, the authors don't identify a single instance of pathological behaviour (66), and they come down fairly and squarely on the view that by and large the Fund does what powerful states (the G7) want (68). They do argue that its ability to do so hinges on its claim to expertise, showing that it has devised increasingly wide-ranging rules in its work with member countries, and that the data it deploys contribute to remaking as much as describing the world. However, the development of the argument completely undermines the narrow identification of the IMF mission with financial stabilization, and points directly instead to a consistent concern for 'global capitalism'. They show that from the start the Fund saw short-term adjustment as a necessary prelude to medium- or longer term increases in productivity, thereby invalidating any claim of 'mission creep'. As they document, the means it used changed as the focus of its activity switched from the developed to the developing world and again as the legitimacy of structural adjustment was called into question. Its increasing focus on the domestic economy, its development of conditionality tools and its engagement with 'poverty alleviation' fit this perspective perfectly well: while the means applied may have expanded, the social purpose remained unchanged. Yet Barnett and Finnemore never explore their own suggestion that the IMF might be 'functional for the development of global capitalism' - nor do they explore in any detail the complementarity of its work with that of the OECD and the World Bank, among other organisations - an essential step if one wishes to evaluate its role. At the end of the chapter and of the book they stick to the narrow view of the Fund's purpose, claiming that its application of its 'objective models has failed to stabilize borrower economies in a great many cases' (71) and associating it again with 'the pursuit of financial stability through markets' (157). Their assessment of its success in pursuing its original mission, in the words of its original 1944 Articles, 'to facilitate the expansion and balanced growth of trade, and to contribute thereby to the promotion and maintenance of high levels of employment and real income and to the development of the productive resources of all members as primary objectives of economic policy', is therefore fatally flawed throughout; and of course they don't have the concepts (these being part of the toolkit of historical materialism) to understand how failure in such a mission might be explained without recourse to general theories of bureaucratic pathologies.
The other two case studies are more substantial, but do little to justify the overarching theme of bureaucratic dysfunction. In the first, the UNHCR gave increasing emphasis to repatriation as a durable solution from the 1980s, and in the case of the Rohingyas repatriated from Bangladesh in the mid-1990s it bent the concept of voluntary repatriation virtually out of recognition. But as the case was a hard one, and no evidence is given either of the overall evolution of numbers UNHCR was dealing with over time or the extent to which similar practice and interpretation applied elsewhere, it is impossible to tell whether this was (a) evidence of the 'normalization of deviance' (as invoked, 119), or alternatively (b) an isolated instance, or perhaps (c) a reflection of the limited powers and options of the organisation. In any case, repatriation was always one of the prescribed possible options, and it was little used in the early years, as we are repeatedly told, only because in the context of the Cold War it seemed unthinkable to return refugees to areas under Soviet control. So the shift towards repatriation is better described as moving from a temporary deviance to the full regime originally envisaged.
In the second case, the UN Secretariat failed to recommend intervention in the Rwandan genocide in 1994. It had had difficulty setting up the peace-keeping force, UNAMIR, in the first place (it was approved by the Security Council with 2,500 lightly armed troops, half the level requested), had it renewed in March 1994 only over initial US objections, and once the mass killing started, was working 'with New Zealand ambassador Colin Keating, the President of the Security Council during the month of April 1994, to oppose the U.S. desire to withdraw the operation completely in favor of a token presence' (145 and 203, ft. 98). The authors attribute the failure of the Secretariat to portray the violence to the Security Council as organised genocide and request intervention to the 'peacekeeping culture', the 'tendency to cling to preexisting beliefs and to reach premature closure' (148), and the fact that because its mandate was to anticipate and help stop a civil war from returning, 'that was all it could see' (149). This is entirely implausible, principally because the commitment to 'consent, neutrality and impartiality' was not a case of embedded and powerful rules that determined goal definition, as the 'irrationality of rationalization' is described (39) - on the contrary, it was a very recent and considered reversion to that approach to peacekeeping after it was abandoned in 1992 in favour of a more active practice of 'peace-making', leading to setbacks in both Bosnia and Somalia, as Barnett and Finnemore demonstrate at length. In any case, there is not a shred of evidence that the Security Council would have countenanced the tripling of forces on the ground as requested by General Dallaire, and in view of Dallaire's misgivings about the quality of his troops (especially in light of the scheduled withdrawal of the Belgian contingent), and the scale of the civil war and genocide, the likelihood that a larger force would be effective must have seemed low. None of this excuses the failure of the international community as a whole to challenge the organised genocide that unfolded. But to attribute the outcome to the UNHCR and particularly the Secretariat's pathological construction of the crisis is arbitrary and unconvincing.
The final chapter reflects on IO action from a cosmopolitan liberal democratic perspective, with a focus on issues of legitimacy and accountability. It builds on comments at the end of the first chapter, where Barnett and Finnemore claim that 'one consistent finding is that IOs tend to expand,' and hold out the prospect of an 'increasingly bureaucratized world' that is potentially worrying because 'the liberal norms embodied and promoted by these organizations are generally not matched with accountability or participation procedures that liberalism favours' (15). No attempt is made to provide empirical evidence to support this claim for IOs in general, and the theme then disappears until we come to the final chapter, 'The Legitimacy of an Expanding Global Bureaucracy', which asks again whether 'this expansion that is creating an increasingly bureaucratized world' is a good thing. At the same time, though, the claim is qualified: 'While the overall trend line suggests that IOs are expanding in size and scope, there are moments when IOs resist opportunities and even explicit invitations to expand' (158). One could go further - as they report, the UN briefly expanded its role into 'peace-making', but quickly drew back to its original mandate when it 'learned' that it had over-reached. This does not deter them from adducing 'learning from failure' as a mechanism promoting expansion (160), and three pages later they so forget themselves as to cite the UN Secretariat's reversion to an 'institutional ideology of impartiality' in a section on the 'pathology of expansion'. On their own account, on the admittedly slender evidence of two case studies, it seems eccentric to describe the UNHCR or UN peace-keeping in terms of expanding power and reach (170). On the authors' own account, they have been starved of both resources and authority, and caught between conflicting pressures, as reflected in a more realistic statement in the closing pages: 'As UNHCR action has created (and constituted) new constituencies, particularly advocates for refugees, contradictions between the demands of these groups and those of member states have made crafting policies increasingly difficult' (171). In any case, even if demonstrated, the expansion of IOs does not equate to an increasingly bureaucratized world, let alone to the emergence of 'increasingly powerful and autonomous bureaucrats' (173). Even on the limited and theoretically impoverished canvas deployed, Barnett and Finnemore give ample evidence of the limits to IO action set by states, and the tendency for the challenges to be met to outrun the resources available.
One persistent problem is that there are obvious deficiencies, from a critical political economy perspective, to an approach that treats all international organisations as generically the same despite the different arenas of their activity, and does so from a perspective drawn from Weber's analysis of bureaucracy. Such an approach is bound to miss the specific issues raised by the governance of the global economy, and the relationship between states, markets, international organizations and populations it entails. In this case, Barnett and Finnemore are able to identify some important empirical phenomena, but unable to provide a satisfactory analysis of them. Four successive statements from the final chapter pinpoint the scope and limits of their approach:
'As IO governance activities expand, they are not only helping to create a new structure to world politics but are also working to create and constitute new domestic societies. IOs appear to be steadily expanding their mandates in a convergent direction: all are increasingly involved in the domestic affairs of states, and, specifically, all are trying to create durable, modern nation-states that are organized around democracy and markets' (164).
'IO expansion, in short, entails creating particular kinds of states with particular kinds of interests' (165)
'The bureaucratization of world politics means that international organizations have more authority than ever before and therefore have more power over other actors than ever before' (165)
'This does not mean that authority is zero-sum' (ft. 27, p. 205)
It is not that Barnett and Finnemore do not identify something of importance, but rather that they are unable to develop their insight effectively. Some IOs, if not all, certainly are 'working to create and constitute new domestic societies'. But it is a stretch to apply the idea to the UNHCR, and in terms of the division of labour, more applicable to the OECD, with its relentless focus on competition in labour and product markets and advocacy of market-supporting social protection over universal welfare, and the World Bank, with its sights trained on forging ideal neoliberal citizens, than on the IMF. It is a mistake to focus so much on expansion as opposed to the core missions of IOs in this respect though, another to rely so heavily and with such self-defeating consequences on theories of bureaucracy, and another again to take constructivism as a starting point instead of attending first to the material and social forces driving change in the world. Class struggle in the context of the evolving world market creates particular kinds of states in which particular interests are dominant though always contested; IOs concerned with the governance of the world market work to reconcile states to it, and to build the hegemony of capital over labour on a global scale. So, as the third statement above and its immediate qualification in a footnote show, to address this in terms of the relative power of IOs and states (whether in zero-sum terms or not) in the absence of a theory of change in the social world leads to a dead end.