How and why the shift occurred

The first chapters of this book are concerned with uncovering what has changed since the structural adjustment era, and understanding how and why this change occurred. There are those who argue that there is in fact very little new in the global governance of development finance, and that any apparent changes are only at the level of rhetoric and not practice.¹¹ Yet, as I show in Chapter 3, if we compare the earlier structural adjustment-era practices to those of the past decade and a half, it is evident that there have been significant shifts in how development finance is undertaken.

The structural adjustment era stands out even now as the high point of the power of the IFIs and Western donors, when their capacity to exert influence over low- and middle-income countries appeared incontestable. Why then did it not last? Ironically, those very aspects of structural adjustment policies that made them seem so stable, such as their consistent reliance on universal economic principles and efforts to separate or subordinate politics to economics, ultimately proved to be unable to address the increasingly complex problems that institutions were faced with. Of course, there were significant sources of conflict between donor organizations and borrowing states and civil society organizations, all of which helped erode the structural adjustment policies. But these conflicts combined with tensions that began to emerge within the practices of governance themselves. As the IMF and World Bank delved deeper into the structural aspects of borrower countries’ economies, they found their policy tools ill-suited for the task and began to experiment with new criteria for evaluating success and failure. The difficult events of the 1990s, including the Mexican and Asian financial crises and the recognition of a failed decade of aid to sub-Saharan Africa, were viewed as signs of profound failure in the governance of development and finance. Debates about “aid effectiveness” in the 1990s not only sought to resolve the problem of failure, but, more significantly, to develop a new consensus on what constituted success and failure.

These organizations thus came face to face with what the political theorist Sheldon Wolin, in his interpretation of Max Weber’s political and methodological writings, describes as one of the central paradoxes of expert authority: the need for expertise to ground itself on methodological foundations which themselves are fragile and prone to contestation.¹² As I will discuss in later chapters, such moments of contestation often occur when the gap between a system of measurement and the complexity of its objects becomes too big – as the fluidity of the world overtakes our capacity to translate it.¹³ In the case examined here, key international organizations (IOs), and state and non-governmental organization (NGO) actors, challenged the grounds of governance expertise and sought to redefine it through a process of problematization – debating and developing new techniques and practices. What emerged over time were several new governance strategies.

How the new practices work

How do we go about understanding this transformation, and mapping the contours of these emerging practices of governing in the context of failure? In other words, how do we study the how of global governance? One of the challenges of investigating the changes discussed in this book is that they cannot be readily witnessed through the study of any one individual institution, such as the IMF or the World Bank. Although IO scholars focusing on an individual institution gain crucial insights into the complexities of internal bureaucratic politics and the dynamics between internal and external pressures, they run the risk of ignoring the ways in which policies pursued at one institution are connected to and dependent on processes at others and within a broader community of practice including donor agencies, NGOs and IOs.¹⁴ At the same time, focusing only on the broadest level of analysis, examining macro-trends in global governance – in the transformations of advanced capitalism, for example, or in neoliberalism – runs the risk of over-generalizing the changes taking place and missing the complex particularities that are involved in each institution and policy.¹⁵

Many of the important changes taking place in global governance – including the emerging strategies discussed in this book – occur at a meso-level that is between these two more common levels of analysis. In Chapter 2, I develop an analytic framework for studying these meso-level processes – a “how to” guide of sorts – to assist those who are interested in understanding these messy intermediary processes of global governance but are uncertain of how to go about doing so.

This framework focuses on three interrelated meso-levels of practice. The first level of analysis is made up of governance strategies such as managing risk and vulnerability or fostering country ownership. These are broad clusters of governance practices organized around a particular problem: how, for example, to address the political sources of policy failure (by fostering ownership). These strategies cut across a range of different institutions. They are developed, often piecemeal, by various policymakers, politicians, economists and critics through a process of debate and problematization, in which a new set of issues or concerns is defined and new techniques developed for making them governable.

Although there has been a myriad of individual policy initiatives, this book argues that it is possible to identify four broad trends in policy that most key development financing organizations and many NGOs have participated in over the past decade and a half. Put simply, these are strategies of fostering ownership, developing global standards, managing risk and vulnerability, and measuring results. The first of these strategies, most apparent in IFI efforts to streamline conditionality and to replace structural adjustment lending with Poverty Reduction Strategy Papers (PRSPs), places new emphasis on ensuring that policies are responsive to local contexts, and seeks to build local ownership of IFI and donor programs. The second strategy of standardization seeks to develop universal standards of good governance and best economic practice, and to disseminate them to developing and emerging market countries. The third strategy of managing risk and vulnerability reconceptualizes the objects of development assistance – such as poverty reduction or project success – as more contingent and prone to failure, and works to develop pre-emptive measures in response. The final strategy of results-based measurement seeks to catch up with the increasing complexity of finance and development policies by creating ever more sophisticated methods for measuring policy success and failure, and integrating the measurement and evaluation of results deeply into the process of policy management.

The second meso-level of analysis drills down to the building blocks, or factors of governance, that make up these governance strategies: these include the actors who govern, the techniques and knowledge that they use, and the forms of power and authority involved. By mapping shifts and continuities in these key factors, we can gain a nuanced appreciation of how the work of governance is being done.

The past two decades have witnessed significant shifts in the various factors involved in the work of governance. New, more engaged actors have become implicated in the processes of governance, most notably through the integration of various kinds of civil society actors as the source of “demand” for particular kinds of government policies and market services. Forms of knowledge have also evolved, as practical, small “i” ideas, such as new public management and new institutionalist economics, have become the drivers of institutional change, replacing the more ambitious big “I” Ideas like the Keynesian and Neoclassical paradigms. The techniques have also shifted accordingly, relying on new forms of participation and the production of different kinds of documents, or inscriptions, to coordinate action.¹⁶ The forms of power and authority involved in the governance of finance and development have also undergone a transformation, as IFIs and donors have begun to rely on more popular and moral forms of authority, and as their expert authority has become more provisional in character. In the process, they have also begun to replace some of the more overt, instrumental forms of power used in the structural adjustment era with less direct, more productive (but still exclusionary) forms, such as scoring and ranking processes that sort countries based on their performance.

The third and final level of analysis that I am undertaking in this book considers whether there are any broader underlying patterns apparent in the strategies and factors of governance at a given moment in time. As I will elaborate in the next chapter, some historical moments are characterized by a particular style of governance. Such styles are defined by the particular ways that institutional actors have found to resolve the tensions facing governance efforts – in particular, the methodological dilemmas that I discussed above, as they seek to maintain expert authority in the face of a slippery world that resists full comprehension. In Chapters 3 and 4, I suggest that the structural adjustment era and the present day are each defined by a different style of governance – the earlier era being characterized by a far more confident and direct style than the present-day provisional form of governance.

Implications

What are the implications of such changes in how governance is done? This is a potentially vast question, which could be answered on many different levels – focusing on the effects on domestic communities, on interstate dynamics, or on the IFIs and donor organizations themselves. This book seeks to answer the question of implications in the final chapter by focusing primarily on the last of these questions – examining the effects of these changes on organizations by asking what their implications are for the politics of global governance, and considering how sustainable these new strategies ultimately are.

What is the future of this provisional style of governance? If we look more closely at the different patterns that constitute it – the shift towards more proactive and indirect approaches to governance, the reliance on symbolic techniques, and the increasing awareness of the possibility of failure – we do not find a single coherent telos but rather two possible paths. On the one hand, many of the practices involved in these strategies are open-ended and even experimental.¹⁷ They respond to the uncertainty of the world through a trial-and-error approach and bring new actors, particularly local ones, together with local forms of knowledge into the process to better respond to the unknown and learn from past failures. Yet this more open-ended and inclusive form of expertise coexists with, and is often trumped by, a much more risk-averse one that responds to those same uncertainties by relying on the security of more traditional forms of expertise, trying to reduce everything to numbers – an approach best captured by the new emphasis on measurable results.

Each of these paths also has significant political implications. More experimental approaches to governance often cede some authority to a wider range of actors, such as civil society organizations, poor people and local governments. Yet, when caution wins out, these messy and less reliable forms of input have to be translated into traditional expert categories, often reducing genuine debate and deliberation with thin proceduralist forms of consultation.¹⁸ The repoliticization of these governance processes paradoxically turns into a kind of depoliticization, as various forms of political action are read through the lens of economic expertise and then reduced to quantitative indicators.

The effects of this approach to governing failure are paradoxical. Policymakers’ caution is one of the key ways they attempt to hedge against the possibility of failure. Yet, despite such efforts, failures persist. These new strategies continually confront the limits of their efforts to make ownership and governance measurable, to draw tidy lines between policies and results, or to reduce the uncertainties of finance and development to algorithms of risk. These failures of performance can lead to failures of consensus. Although one might expect that IFI and donor staff would embrace these new techniques of governance and the forms of power and authority that they afford, my interviews reveal that many of them are ambivalent about these reforms, precisely because of their continued messiness and refusal to fit within bureaucratic norms of neutral and apolitical expertise.¹⁹

But do these failures actually matter? After all, as scholars like James Ferguson and Timothy Mitchell have noted, although global development policies frequently fail, such failures seem to have a negligible effect on the development machine.²⁰ Indeed, I will suggest, some of these failures are benign or even constructive, doing no damage to the institutions involved in development governance. Yet some of these failures are destructive to them: when failures of performance combine with failures of consensus, the ground is fertile for further erosion of governance authority.

Empirical contributions

In empirical terms, this study contributes to our understanding of some key changes in the governance of finance for development, speaking to scholars and policymakers interested in global governance, international organizations and international development. The book is the culmination of seven years of research into the changes taking place in the policies of the IMF, the World Bank and several key donors. Most of the book’s empirical material is drawn from the IMF and the World Bank, given their dominant role in governing development finance. I do, however, also examine the policies of certain donor agencies, particularly where their influence has been important in shaping the direction of development policy – for example DFID’s movement to eliminate economic conditionality, the Canadian International Development Agency (CIDA) pioneering adoption of results-based measurement, and the American Millennium Challenge Corporation’s (MCC) pass–fail approach to conditions.²¹ The research is based on extensive document analysis, archival research at the IMF, World Bank, Canadian and British National Archives, and over fifty interviews with staff and management at the IMF, the World Bank, the Organization for Economic Co-operation and Development (OECD), NGOs, and in certain donor countries.

Scholars of international political economy (IPE) and finance tend to study the IMF and to focus on the interactions of major industrialized states, while development scholars study the World Bank and donors and tend to ignore the IMF. This book cuts across these two solitudes and provides a synthetic analysis of the changes taking place in these various organizations, while at the same time remaining attuned to the important differences among them. In fact, as the evidence in this book makes clear, the common claim that the IMF is “not a development organization” is untrue: even if development is not a formal part of its Articles of Agreement, the organization’s actions have profound developmental effects.²² In choosing to focus on the institutional side of recent changes in finance for development, I have of course downplayed the other side of the equation: the impact of these changes in developing countries. This book is ambitious enough without attempting to do justice to these important questions. However, as these changes in policy have begun to take hold, other scholars have begun to tackle these issues.²³

Methodological innovations

This volume’s approach and structure also constitute an important methodological innovation. How do we go about studying the how of global governance? Much of the literature to date has tended to focus either on individual IOs or on broad-level governance trends and patterns. Yet many of the important changes taking place in global governance – including the emerging strategies discussed in this book – occur at a meso-level that is between these two more common levels of analysis.

This book argues for the value of a methodological approach that begins in the middle, focusing on the concrete policies, strategies and techniques through which various actors do the work of global governance.²⁴ This kind of analysis is “meso” for several reasons. It is a kind of analysis that starts in the middle: looking at what is going on in the form of concrete policy practices, like the consultations to produce PRSPs or efforts to streamline conditionality, and seeking to understand them. The objects of this analysis also exist somewhere in a middle ground between materiality and discourse, linking the two without being resolved into either one: the documents and consultation processes that are key to the PRSP, for example, are both material and discursive – their power, in effect, derives from their capacity to translate ideas into material form. This book thus undertakes an analysis focused primarily on processes – how ownership is fostered, for example – rather than on outcomes or interests. Finally, this is a meso-level analysis because its level of analysis exists between and connects the macro, more structural level of global governance and the micro level of individual state, NGO, academic and bureaucratic actors: to understand the strategy of ownership, for example, we must look at how certain practices emerged in and circulate among these different actors and institutions.

Theoretical insights

In focusing on the “how” of global governance, this research seeks to make theoretical contributions to several key academic debates. My principal inspirations and interlocutors can be found in the literatures on global governance and IOs, critical IPE and social theory. My goal is to bring some of the underappreciated insights of social theory, particularly certain concepts from actor-network theory (ANT) and science and technology studies (STS), into the global governance and IPE literature. In so doing, I hope to enrich the sociological turn in international relations (IR) through a contribution to our understanding of how global governance works.

More specifically, this book makes four key contributions to theoretical debates. The book focuses on strategies and techniques that link the material and the discursive, thus contributing to the practice turn in social theory and IR. The book also seeks to provide an account of change not only of norms but also of governance practices. It seeks to advance our understanding of the centrality of expertise and its limits, in part by examining the politics of failure. Finally, my attention to the rise of provisional governance contributes to but also moves beyond existing work on risk in social theory.

The plan of the book

This book is organized into four sections. Chapter 2 continues the discussion initiated in this Introduction on how we might go about studying the “how” of global governance. After a discussion of this book’s relationship with the broader practice turn in IR and social theory, I provide a more substantial account of the main categories of analysis used in this book – governance strategies, governance factors and styles of governance – followed by a brief overview of how I will put them together to understand the transformation of global governance practices.

Chapters 3 and 4 then consider the historical context of the recent changes in IFI policy, tracking the changes underway in governance factors and tracing the reasons for the emergence of new governance strategies. In order to establish whether policy strategies such as ownership and risk management are in fact new, it is important to show how they differ from earlier governance practices. Chapter 3 does just that, taking a careful look at how the IMF and World Bank sought to govern development financing in the 1970s and 1980s, revealing a far more confident and direct style of governance. Chapter 4 traces the gradual erosion of that governance style, a process driven by debates about contested failures in finance and development and the problematization of new issues. Throughout this period, staff, critics and leaders sought to re-establish the basis of IFI authority, not just by developing new policies such as the PRSP and good governance agenda, but also by developing entirely new governance strategies and definitions of success and failure.

In Chapters 5 through 8, I examine the four new governance strategies that have emerged in response to this erosion. In Chapter 5, I begin by examining the strategy of country ownership, the chief means by which IFI and donor actors have sought to govern the political dimensions of economic policy. Through their development of the PRSP and their efforts to streamline conditionality, the IMF, the World Bank and many donors have begun to pay more attention to the local dynamics of adjustment and development, as well as to the importance of political will. Yet even as they have touted the increased transparency of these new policies, in practice these institutions have been gradually informalizing and obscuring power relations. In Chapter 6, I move from the particulars of country ownership to the universals of global standards, tracing the ways in which IMF and World Bank staff members have transformed the rigid economic universals of the structural adjustment era into more flexible and ambitious global standards covering everything from accounting practices to maternal health. As they have moved into this more contested terrain, standards have become increasingly preoccupied with fostering credibility, making them both more symbolic and more performative.

After the shocks of the Asian crisis, the AIDS crisis and the most recent global financial crisis, both IFIs and donors have begun to focus more on risk and vulnerability, the subject of Chapter 7. At the World Bank, key units have re-defined poverty as social risk, while the IMF has developed a ranking system to assess borrowing countries’ vulnerability to excessive debt. As agencies have begun to conceptualize the objects of their governance through the lenses of risk and vulnerability, they have also developed new tools for pre-empting the things that might go wrong. At the same time, as decisions increasingly get filtered through a risk–reward matrix at these institutions, poor countries find themselves ranked and sorted in ways that significantly affect their capacity to borrow. How should these increasingly complex and dynamic objects and techniques of governance be measured and evaluated? This is the challenge at the heart of the fourth and final policy strategy, examined in Chapter 8: that of results measurement. In various ways and with varying degrees of success, the World Bank and donors have sought to define a new category of knowledge, called “results.” By demonstrating results, IFIs and donors hope to re-establish some of their lost authority. Although this turn to demonstrable results appears to be the exception to the turn towards more provisional forms of expertise, I suggest that the often-heroic assumptions that make such claims about possible results leave considerable room for hedging against failure.

What then is the future of provisional governance? This is the central question examined in the Conclusion. As a mode of governance that is unusually preoccupied with avoiding failure, it is ironic (if perhaps unsurprising) that efforts to pursue these new more provisional strategies nonetheless face resistance, limits and failure. After assessing the implications of these failures, I examine the two possible directions that provisional governance might take – more open-ended and experimental, or cautious and risk-averse. I conclude by considering which is the more likely future path for global governance.

In the wake of the 2008 financial crisis, we have witnessed both a decline in the volume of traditional donor assistance and the return of more explicitly political kinds of aid, particularly among those donors with new conservative governments. These two shifts have precipitated calls for hard, quantitative results-based forms of expertise to demonstrate the “value for money” of various aid initiatives. Such manoeuvres reinforce the trend towards a cautious, even cynical kind of provisional governance in which expertise is increasingly tied to political convenience. At the same time, efforts to develop ever more standardized forms of evaluation only intensify the difficulties of translating the complexity of development into tidy forms of expert knowledge. Recent trends thus only exacerbate the tensions faced by those trying to manage development finance, accentuating the fragility of their expert authority and the persistence of the politics of failure.

Understanding governance as practice

Over the past decades, a growing number of social theorists have begun to use the concept of practice in their work. They are an eclectic bunch, ranging from Pierre Bourdieu (who developed a “practice theory” based on the concepts of practice, habitus and field), to Michel Foucault (who focused on discursive practices and, in his later work, on embodiment), and Michel Callon and Bruno Latour (who have studied the practice of knowledge-creation in the sciences through actor network theory (ANT)). Most IR scholars who identify themselves as part of the practice turn have drawn primarily on the work of Bourdieu.² My framework, in contrast, owes more to the insights of Callon, Latour and other scholars of science and technology studies (STS), as well as to some of the ideas of Foucault.

What then is a practice, and why is it useful for understanding global governance? In his introduction to a field-defining book on the subject, Theodore Schatzki defines practices as “embodied, materially mediated arrays of human activity organized around shared practical understandings.”³ It is worth spending a moment unpacking some of the implications of this definition. Practice theorists’ emphasis on the material character of human action differentiates them from constructivist and certain post-structuralist approaches, which tend to focus more narrowly on its discursive or ideational dimension. Yet practices are not just activities (e.g. whirling around in a circle), but meaningful ones, organized around common understandings (e.g. about the pirouette as a kind of dance movement). Practices are therefore both material and discursive, combining an action with a frame of reference.

Because practices are socially situated but enacted by individuals and groups, a focus on practice provides one possible answer to the conundrum of the relationship between structure and agency. Agency is constituted in part through practice: we are defined in part by what we do (as dancers or paper-pushers), and in part by the social context that makes this action possible. At the same time, practices are modified through individual and collective action, and change over time. There is a wide array of meaningful activities that we might define as practices, ranging from pronouncing words, to writing a memo, to negotiating a loan with a low-income country.⁴ As these examples suggest, practices can be thought of as connected and nested in one another, with more complex practices relying on a whole range of more basic and often unnoticed ones.

In IR circles, a number of scholars have begun to make use of the idea of practice in their work.⁵ Emmanuel Adler and Vincent Pouliot, in particular, have provided an elegant and coherent pitch for the importance of practice in IR, building on Bourdieu’s work to do so. Although my use of practice in this book shares much with their contribution, it also differs in several important respects – partly on theoretical grounds, but also on practical ones. Much of the recent work on Bourdieu in IR, Adler and Pouliot’s included, has focused on empirical cases in the realm of security and diplomacy. In this particular realm, Bourdieu’s ideas have proven to be very fruitful. Bourdieu’s concepts of field, doxa and habitus are particularly useful for explaining the persistence of logics of practice. Pouliot, for example, examines the operation of the field of diplomacy – which he defines as a relatively autonomous community of practice in which everyone agrees on the stakes, knows the rules (or doxa) and plays the same game.⁶ In this context, Pouliot puts considerable emphasis on the role of habitus, the tacit know-how and assumptions that various actors learn through their position in the wider social structure, and bring to the game of diplomacy.⁷

Although my theoretical framework shares much with these earlier contributions to the practice turn in IR, there are also a number of key areas in which it differs. In order to make the practice turn work for the study of global governance, I argue that we need to place more emphasis on knowledge-making practices, shift from fields to problems as the basis for communities of practice, and pay greater attention to the dynamics of change.

Practice-oriented IR scholars have tended to emphasize the importance of tacit or practical knowledge in international politics, and to differentiate it from more reflexive, self-conscious forms. Pouliot for example seeks to contrast “the abstract schemes produced by technocrats and social scientists” with the tacit, unverbalized knowledge that informs practice: it is this tacit, practical knowledge that he sees as crucial to international practice.⁸ Yet, those who study global governance would certainly want to question such a tidy distinction between the abstractions of bureaucrats and the concrete practices of governance.⁹ The main practitioners of global governance are in fact technocrats, and many of them are also social scientists (particularly economists). Their practical work involves translating the messiness of the world into useful abstractions (reports, tables, matrices, scores, indexes) that can then be deployed to govern their unruly objects. To grasp the dynamics of global governance, we therefore need to understand the production of expert knowledge as a kind of practice – a task, I will suggest below, that is particularly suited to the insights of ANT scholars like Callon and Latour.

The concept of field used by many IR practice theorists also needs to be used with caution when considering the processes of global governance. Fields have a kind of coherent logic that enables those operating within them to know the rules and to agree on the stakes involved, rather like playing a game.¹⁰ While it is possible to identify a number of fields within the realm of development finance, the closer we look at current practices of governance, the less clear it becomes where the fields begin and end. Actors working at the IMF, the World Bank, NGOs and aid agencies share many assumptions about the tacit rules of the game and the relative hierarchy of economics over other forms of intellectual capital. Yet they also have quite different cultures: the IMF is a centralized institution that focuses tightly on “hard” financial issues and concerns, whereas the World Bank is known for its diffuse structure and more heterogeneous intellectual culture. If we focus on who is actually engaged in the practices of governance, we see a very loose network of actors (including international organizations (IOs), NGOs, governments, the private sector and academia) playing a multitude of games, often using different rules, and seeking different stakes.¹¹ Moreover, many of the recent changes in development finance are expanding the community of practice by including an ever-wider range of actors in the processes of governance, making the boundaries of that community subject to change and contestation. As I will discuss below, I have therefore found it more useful in this study to look at how actors and practices become connected around concrete problems and strategies rather than through predefined fields.¹²

It is also important that we pay attention to changes in governance practice. As social theorists like William Sewell, David Stern and Anthony King have pointed out, Bourdieu’s concepts of field and habitus tend to push his analysis towards the structuralist, or objectivist, side of the balance, making it easier to explain the stability of practices than to understand their changes.¹³ Although it is important to be able to understand what does not change in IOs and other institutions over time, we also need conceptual tools to help us understand shifts in practices.¹⁴ Even mundane bureaucratic practices change significantly over time. Why is it that everyone working at a development agency or government-funded development NGO nowadays (at least in certain countries) knows how to prepare a results matrix when proposing or evaluating a program, whereas they had not even heard of the practice fifteen years ago? Why did various practices designed to foster ownership become ubiquitous in the early 2000s, but have become less so more recently? To answer these questions, we need an approach to practices that is attentive to their contingency as well as their sedimentation.

In the remainder of this chapter, I will elaborate the analytic framework that I propose to use for the rest of this book. This meso-level approach draws on some of the insights of other practice theorists in IR while being more attentive to the role of knowledge-making practices, focusing on problems rather than fields as the glue that links governance practices, and seeking to understand what drives the changes in governance strategies. As this brief discussion has probably already made clear, the concept of practice is a slippery one: because it includes everything from basic coping practices like pronouncing words to highly sophisticated ones like managing an IO, the concept can be difficult to use with precision. To avoid conceptual muddiness, it is useful to use more specific terms to designate the different kinds of practice that are involved. It is for this reason that I have chosen different terms – strategies, factors and styles, rather than practices – for my key conceptual categories.

Focusing on governance strategies

Chapters 5 through 8 each examine one key governance strategy: creating global standards, fostering ownership, managing risk and vulnerability, and measuring results. These strategies are constellations of practices that are linked by their connection to a concrete problem and a way of defining and tackling it, rather than by their situation in a common field. Focusing on strategies is a particularly effective way of understanding institutional practices because they are in many ways problem-driven machines.

What kinds of problems am I talking about? When we look at recent policies adopted by various organizations and governments involved in financing development, it is clear that many share similar concerns. For example, in the past few years there have been numerous policies that identify risk and vulnerability as key challenges in a more uncertain global environment – including social risk policy at the World Bank, the Organisation for Economic Co-operation and Development (OECD) and DFID, and renewed attention to low-income countries’ vulnerability to external shocks at the IMF following the recent financial crisis.¹⁵ Similarly, a wide range of institutional actors became preoccupied with developing and implementing new global standards in the late 1990s and early 2000s, and sought to achieve them through a range of policies including the good governance agenda, the standards and codes initiative, and the MDGs. Standardization, risk and vulnerability are all concepts, but they only really become effective when they are translated into concrete practices that seek to foster or control them. They are thus examples of what I am calling governance strategies.

The term “strategy” has its origin in military planning. In global governance as in warfare, a strategy looks at the medium to long term, while tactics are the more immediate means through which those objectives are pursued. Yet, unlike wartime strategies, governance strategies are not always explicitly articulated, nor are they the source of a central will or intention. My use of “strategy” is therefore quite different from traditional IR use, in which “strategic” interaction refers to individual rationalist action, often in a game-theoretic context.¹⁶ Governance strategies are social rather than narrowly individual, and are embedded in the day-to-day work of governance.

Governance strategies are defined by a set of assumptions, goals and ways of doing things. Strategies link together several policies, often across more than one institution. In so doing, they work to problematize certain aspects of social, political and economic life: to draw a line between one issue and another, to make an issue visible, to suggest a direction or a point of attack – in brief, to make things governable.¹⁷ My concept of governance strategy therefore resembles Foucault’s concept of problematization, which he uses in his later work. In the Use of Pleasure, for example, Foucault’s goal is to understand how certain sexual practices came to be problematized at a certain moment in history, “becoming an object of concern, an element for reflection, and a material for stylization.”¹⁸ This book examines how such reflexive moments of questioning and contestation – or problematization – emerge from, and are translated into, routine everyday forms of practice. In so doing, the concept of strategy links discursive or ideational approaches, like constructivism and post-structuralism, and materialist and pragmatic ones, like Bourdieusian practice theory.

The various strategies that I am studying emerged as certain issues came to be viewed as matters of concern, either for the first time or in new ways. For example, in the debates prior to the development of the strategy of country ownership, key actors began to see the challenge of dealing with a country’s domestic politics as both relevant and problematic in new ways. As I will discuss in the coming chapters, this problematization of politics was precipitated by contested failures in development and finance, particularly in Africa and Asia. After decades of denying or downplaying the political dimensions of their policies, development policymakers began to see political institutions and a lack of political will, or ownership, as a key determinant of program failure, and thus a legitimate object of policy action. This process of problematization was not limited to elite policy-makers or economists but was a wide-ranging debate that included NGOs, critics, state leaders and institutional staff. In the process, these actors brought background assumptions about what counted as success and failure into the foreground, forcing the IFIs to defend and adapt not just their policies but the expert authority on which they were based.

Key IFI and donor actors began to see fostering country ownership as a way of addressing the problem of politics and re-establishing their authority. Yet this was a far from coherent process: there was no singular individual or group responsible, nor a general commanding troops to ensure ownership. While many institutional, intellectual and governmental actors championed the ownership strategy, others resisted it. Nonetheless, over time, the practice of fostering ownership has become an explicitly articulated and generally accepted strategy, one that has been adopted by a whole host of organizations ranging from the IMF to many donors and NGOs and that has had profound effects across a multitude of issue areas.

Examining factors of governance

Studying specific governance strategies may be more manageable than trying to make sense of broad practices like “the governance of development finance,” but strategies are still very complex things. If we want to understand how strategies do the work of governance, we need to break them down further into their constituent parts or governance factors. Even the simplest of practices are complex phenomena made up of many different dimensions.¹⁹ I want to focus on five dimensions of practice here that are central for understanding governance strategies: the roles of actors, techniques, knowledge, authority and power. The concept of “governance factors” encourages us to look both at and beyond day-to-day practices: to look not just at what is done, but who is doing it, how they conceptualize their work, what specific techniques they use, how they are authorized, and what kinds of power relations are implicated. By studying these factors we can compare past and present practices, determine whether patterns exist that link policies through a common strategy, and assess whether changes are occurring.

Recognizing governance styles

So far, I have focused on the meso-level strategies that give shape to global governance practices as well as the more basic factors, or building blocks, that make them up. I have suggested that this kind of analysis helps us to overcome the limits of focusing narrowly on individual organizations or policies, as well as to avoid the overly broad generalizations of certain global governance approaches. Beginning in the middle, however, does not mean that we cannot examine broader patterns of governance: rather, it means that any such analysis must be inductive – developed through careful study of concrete governance practices. If it becomes apparent that similar kinds of actors, techniques, forms of knowledge and kinds of authority and power keep reappearing across a range of different strategies, this tells us that a broader process is underway.

During my research for this book, I was struck by the ways in which each of the strategies that I identified seemed, despite their differences, to share some marked similarities. As I will discuss at much greater length in the coming chapters, I saw in each strategy a tendency to integrate more local actors in the processes of governance while also attempting to constitute them as more active participants; to use increasingly symbolic and performative inscription techniques; to draw on a similar set of small “i” ideas as a way of making politics amenable to economic governance; to deploy more productive and indirect forms of power in order to sort and rank states and individuals; and to redistribute authority while expanding its basis to include popular as well as expert forms.

As I moved back in time and looked at the recent history of financing development (as I do in the next chapter), I found some markedly different patterns at work. During the structural adjustment era of the 1980s and early 1990s, the actors involved were restricted, the techniques of governance more direct and straightforward, the ideas more narrowly economic, the forms of power more coercive, and the authority exclusively expert. By focusing in on these factors, I had uncovered some broad differences in how the work of governance was done then versus now. These differences are not random, nor are they driven by a kind of monolithic logic. Instead, they are subtler, but important, differences in the style of governance. The structural adjustment style was confident and direct: certain of its authority, direct in its techniques, seeing the future as a continuation of the present. More recent governance strategies are defined by a more provisional style: far less certain about its authority or the future, and relying on more complex and indirect techniques and forms of power to achieve its ends.⁴⁵

I have chosen to talk about “styles” rather than “logics” of governance in order to avoid overstating their coherence.⁴⁶ A style colours but does not determine governance practices. The two historical moments that I discuss in this book are characterized by the predominance of a particular style of governance. However, this is not true for all periods: sometimes there is more than one style in operation. Today, for example, the more confident style of the structural adjustment era still defines certain aspects of development finance – certain issue areas, institutional units and individual economists – even if the more provisional style has gained considerable influence.

What do different styles of governance actually do? If strategies emerge out of the problematization of given issues and questions and seek to provide particular solutions, then governance styles do the same at a more fundamental level. As I will elaborate in Chapter 4, modern governance practices have had to come to terms with a very basic problem: the inherent fragility of their claims to expert authority. This fragility has methodological, ontological and epistemological aspects. IFI and donor actors are faced with significant methodological challenges: how do they measure and translate the complexities of real development problems in narrowly economic terms? In fact, as I have discussed elsewhere, institutional staff will often spend considerable energy debating just how qualitative or quantitative their metrics should be, and how objective or subjective their findings are.⁴⁷ These difficulties often have ontological underpinnings: time passes and situations change, making past methods obsolete; more fundamentally, the world is messy and unpredictable, and exceeds our capacity to represent it. Underpinning these debates about metrics is a more profound epistemological dilemma: given such complexities and contingencies, how can institutional actors know what they know, and act on a given problem with authority? Modern governance practitioners must therefore come to terms with the ever-present possibility of a very basic kind of failure – the failure to know what counts as a failure.

A style of governance is a particular way of defining and resolving these dilemmas of governance. During the structural adjustment era, institutional actors relied on a simplistic ontology, treating the world of development as if it were amenable to straightforward methods of measurement and intervention. This allowed them to govern through a narrowly economistic methodology and a highly confident set of practices. More recent governance strategies reveal a far more provisional attitude to the world and its governability. The ontology of recent governance practices is messier and more fluid as the world seems more volatile and complex. New strategies like ownership and risk management are underpinned in turn by more cautious epistemological claims about what is actually known.

As I will discuss in the chapters ahead, this caution is driven by a greater awareness of and preoccupation with the possibility of failure – failure to tackle the political roots of development pathologies, to measure the achievements of complex programs, to predict the next big shock. This preoccupation has led IFI and donor actors to develop proactive governance strategies as a way of inoculating development programs against failure. Such strategies may mean playing the long game, by fostering a results culture, for example, or promoting certain risk-management capacities among poor people, which will only pay dividends over the long term. IFI actors are also relying increasingly on indirect techniques that operate in the gaps of formal policies, through the proliferation of informal conditions, for example, or through policies that rely on pressure from peers or civil society groups. The new strategies also make use of increasingly symbolic practices in their efforts to govern development – practices that depend more on what they represent than on what they are. The value of global standards, for example, lies increasingly in their capacity to signal a country’s commitment to reform, making it visible and open to evaluation. Such symbolic practices are also performative, as they seek to mobilize market and civil society actors to act on the basis of these signals provided. Finally, many of those engaged in these strategies are not only preoccupied with failure, but also seek to hedge against future failures.

We are therefore witnessing the rise of the kind of “provisional expertise” that Niklas Luhmann pointed towards, in which experts sought to avoid failure by minimizing risks and deflecting responsibility; yet this is a kind of provisional approach to governance that is not limited to risk-management techniques, as he suggested, but that is present in a wide range of different governance strategies.⁴⁸ In the coming chapters, I will examine just how this provisional style colours each of the new strategies, and I will reflect on the sustainability of this shift and its implications in the Conclusion.

Understanding change

My discussion of the strategies, factors and styles of governance all emphasize their dynamic character. The next two chapters provide a more detailed account of the forces driving the recent shift in how IFIs and donors manage the financing of international development. For now, I will just provide a basic outline of how attending to the styles and strategies of governance and the factors that make them possible enables us to appreciate not just how governance is practised, but also how those practices change over time.

The question of what drives changes in governance is one that has recently gained significant attention in IR. Constructivist scholars have followed in Martha Finnemore and Kathryn Sikkink’s steps by seeking to understand shifts in governance norms, tracing the various stages in norms’ life-cycles from their emergence to their internalization.⁴⁹ At the same time, a number of critical and post-modernist scholars have sought to explain governance changes through Foucault’s notion of governmentality, seeing them as signs of a broader, more epochal shift to a liberal form of self-government.⁵⁰ In spite of their considerable strengths and resonances with my findings, both of these different approaches tend to emphasize the discursive or ideational drivers of change, downplaying its practical dimensions. Although norms, rationalities and the ways that they are materialized are clearly connected, one cannot be reduced to the other: changes in norms can lead to changes in concrete techniques, but innovations or breakdowns in material practices can also lead to changes in governance norms and eventually to broader rationalities.

The concept of strategies lends itself particularly well to a conception of change that is attuned to both the material and discursive aspects of governance practice. Rather than being tied to a set of functions or a relatively stable field, strategies are organized around particular problems. Those problems change over time – or rather, the way that institutional actors problematize and tackle them changes. For example, although IFIs have always had to develop some kind of implicit conception of policy time and the future, it is only very recently that they have become explicitly preoccupied with the unknowability of the future, seeking to manage these uncertainties through a strategy of risk management. Such strategies come into being in the context of particular moments of problematization, in which those things that were once taken for granted are destabilized and contested: suddenly the unpredictability of the future seem very important. These moments bring part of the tacit background of governance practice into the foreground, making it the subject of a more reflexive debate.

Why do certain things that were taken for granted become open to debate? As the cases that I consider in the next chapters reveal, there can be many different reasons for settled assumptions to become the subject of problematization: particular events, key actors, a change in ideas, innovations or difficulties in techniques – any of these can precipitate these kinds of more profound debates. Yet, lurking behind these more contingent factors is often a concern about the problem of failure.

Given that the question of failure is one that has been discussed before in both global governance and international development literatures, I want to be very clear about how I am using the concept here. Susan Park and Antje Vetterlein have pointed to development policy failure as one of the triggers of change in IOs’ policy norms.⁵¹ On the other hand, James Ferguson and Timothy Mitchell have argued that many development ideas and policies persist in the face of very obvious failures in their explicit objectives.⁵² Without getting into the interesting tensions among these analyses, I would simply like to note that they all focus primarily on objective failures.⁵³ In contrast, my interest is not just in objective failures but also in the ways in which certain things are subjectively defined as failures at certain moments. The meaning of most failures is subject to contestation – but some failures in particular seem to provoke enormous debate about the very grounds of what counts as success and failure. It is these contested failures that often lead to a problematization of sedimented governance practices.

For example, the growing interest in the political dimensions of development discussed above was underpinned by a new conviction that the success and failure of aid programs hinged on domestic stability and ownership. This was a significant departure from previous conceptions of the determinants of policy success or failure. This redefinition was arguably linked to certain objective failures in development financing, notably the “lost decade” of development in sub-Saharan Africa. Yet the failure of development in sub-Saharan Africa had been going on for decades; past World Bank and other reports had acknowledged the problem, but had identified very different causes and solutions. It was only in the 1990s that this problem was defined as a particular kind of failure (of political will and of governance), opening up a wider debate about aid effectiveness, underpinned by key questions about what counted as successful or failed aid.

While changes in more specific policies or norms may sometimes hinge on certain reasonably obvious objective failures, changes in governance strategies often depend instead on more contested debates about what counts as success and failure. Such debates are not merely technical matters for the experts to work out, but are in fact highly political, and often include a wide range of actors including policymakers, academics, NGOs and political leaders. There is a lot at stake in calling a policy a failure (or a success), including potentially the survival of the organizations involved. Debates about success and failure pose significant challenges to institutional authority, since they strike at the heart of claims to expert authority. How can you be an expert on an issue if you are not sure what counts as success? As I noted above, institutional expertise is a fragile thing, depending on a whole host of methodological, ontological and epistemological compromises. Such moments of problematization can unsettle existing compromises, forcing their renegotiation.

Over time, actors develop new strategies as a way of managing these problems and re-establishing their authority: these strategies seek to move what has been problematized into the background, black-boxing their key assumptions and practices as a new kind of tacit understanding. As my discussion of the development of four new governance strategies reveals, this is not a quick or straightforward process, but rather a messy, contingent and contested affair. For, as I suggested above, the particular form taken by the strategies of fostering ownership, developing standards, managing risk and vulnerability, and measuring results is defined by a more provisional style of governance. The most recent solution to the problem of failure is to develop a preoccupation with its ever-present possibility.

The remainder of the book puts this meso-level analytic framework into practice. In the next two chapters, I will trace the path from the structural adjustment era to the present day, examining the differences between the dominant governance strategies then and now, and tracing the processes of problematization and contestation that ultimately led to their decline. Chapters 5 through 8 then examine the four new governance strategies, considering how they emerged in response to the erosion of previous practices and assumptions, and how they do the work of governing. In each case, I examine the role of each of the factors of governance and then consider whether we are in fact witnessing the emergence of a more provisional style of governance. The book concludes by pointing to a paradox of sorts: despite the cleverness of this new provisional style of governance in hedging against failure, the various strategies all face a number of their own failures. After exploring in more depth the paradoxical politics of failure, I conclude by examining the sustainability and long-term implications of this move to a more provisional style of governance.