In its own internal reviews the World Bank has come to the same conclusion – “ownership,” or strong domestic support of reforms, is essential for adjustment lending to succeed. Before 1990 about a third of adjustment loans failed to achieve expected reforms, and the lack of borrower ownership or commitment was a key factor in the failures.
World Bank. Assessing Aid: What Works, What Doesn’t and Why, 1998.1
One of the principal lessons that the International Monetary Fund (IMF), the World Bank and donors drew from past failures of aid effectiveness was a belief that borrowers’ lack of political will was a crucial part of the problem. In response, after considerable debate and disagreement, institutional actors developed the strategy of fostering country ownership. In contrast to earlier eras, in which development policies sought to separate politics from economics or to redefine political questions as economic, the ownership strategy treats the political support for development programs as a legitimate object of international financial institutions (IFIs) and donor action. Four assumptions underpin this strategy: politics (or at least political economy, defined in public choice terms) is relevant to economic development; policies should be tailored to local contexts; borrowing countries must take more responsibility for their own progress; and they must participate more actively in IFI and donor programs. In combining these four assumptions, the strategy of ownership brought together a number of concepts that had existed before, but separately, such as “courageous political leadership,” “self-sufficiency,” and participation.2 It was only once institutional actors and critics began to problematize the role of politics in development and to identify the political sources of policy failure that these problems were brought together and made governable.
In this chapter, I will focus my attention on two key practices that are part of the ownership strategy: the streamlining of conditionality and the introduction of Poverty Reduction Strategy Papers (PRSPs). Although these practices exist in part as specific policies in certain institutions like the IMF, the World Bank and various donor agencies, they all exceed any one particular organizational context. I will therefore examine each policy as it functions both within and across several different institutional contexts, and then consider the combined effects of these new practices as they interact and intersect with one another to make up a more provisional kind of governance.
This chapter seeks to answer two questions about the strategy of ownership: why it emerged, and how it works. I will first answer the why question by examining the various pressures that led to the adoption of the ownership strategy, before considering how the strategy of ownership works by applying a meso-level analysis and focusing on the principal governance factors. I will use a number of concepts derived from actor network theory (ANT), examining how actors sought to operationalize ownership by developing inscriptions that could translate a broader range of voices and concerns into a single powerful document. This analysis reveals some significant shifts in how these new policies do the work of governing, including changes in several key factors of governance: the application of small “i” ideas, the development of increasingly symbolic and informal techniques, the enrolment of new civil society actors, the expansion of the basis of institutional authority and the increasing reliance on productive and indirect kinds of power.
I will conclude by suggesting that efforts to foster ownership provide considerable evidence of the emergence of a new, more provisional style of governance in which governing is done indirectly, often in the gaps in official policy, while governance strategies are increasingly proactive and performative, relying on symbolic techniques for their effectiveness, yet always aware of the possibility of failure.
Before considering the institutional evolution of the two specific policies that I will examine here, it is worth considering how IFIs and donors ended up with the practice of fostering “ownership” itself. Why this new emphasis on explicitly political strategies, given the donor institutions’ long history of trying to appear objective and apolitical? It is tempting to see the strategy of ownership as part of a linear trajectory in the evolution of development policy. Yet the evolution of the strategy and policy was both contingent and contested.
By emphasizing the importance of country ownership, IFIs and donors have sought to respond to two criticisms: to non-governmental organizations (NGOs) and borrowers who charged them with heavy-handedness, ownership promised to shift more control to local communities; and to taxpayers who were suffering from aid fatigue, it was part of the “aid effectiveness” agenda that promised to deliver better results.3 Much of the impetus behind the ownership push revolved around emerging concerns about what counted as success and failure, as the debate on aid effectiveness began to erode earlier metrics and IFIs and donors found themselves under pressure to demonstrate achievements on the ground. Drawing on public choice theory, researchers at the World Bank argued that one of the crucial reasons for the poor performance of many Bank programs was the lack of a good “policy environment” in borrowing countries, which they defined as poor macroeconomic performance and distorted institutions. This meant that IFIs could not resolve countries’ economic problems without also addressing their political context, and that a willingness to reform was essential if programs were to succeed.4 Policymakers thus began to view politics as a central part of the problem of development finance, one that was crucially linked to rates of success and failure.
One strand of the ownership strategy’s development began in a technical part of the Bank, its Operations Evaluation Department (OED), responsible for in-house evaluation. OED staff first became interested in the promotion of country ownership in the 1990s, as part of their efforts to find new ways of assessing programs in the face of what appeared to be declining success rates: by their 1994 report, success rates as measured by the OED had dropped from 80 per cent (common throughout most of the 1980s) to 65 per cent.5 Of course, such figures are somewhat misleading, as the metrics used also changed over time, as I discussed in Chapter 4. Nonetheless, this vastly increased rate of failure became a matter of concern and debate within the organization. By 1998, Dollar and others in the Development Research Group identified ownership as key to aid effectiveness, and the OED also integrated it as a metric of program success, making it a crucial part of evaluation practices.6
Concerned about the apparent decline in success rates, the newly appointed World Bank President, James Wolfensohn, made the improvement of Bank performance a priority.7 In January of 1999 he launched his comprehensive development framework (CDF) in a bid to improve aid effectiveness, in part by fostering country ownership. As I will discuss below, NGOs had been pressing for policies that gave borrowing governments more flexibility to adapt policies to their own priorities throughout the 1990s. There was also pressure for reform from major shareholder governments, including the British and the Americans.8 Actors within the World Bank sought to use the strategy of ownership to translate critics’ concerns about responsiveness and flexibility in a way that would tackle what had become an embarrassing rate of failure in their programs.
At the IMF, both internal and external actors also played a role in introducing the strategy of ownership. Fund staff and directors began to discuss the issue of ownership in the mid-1990s in the context of discussions about how to ensure that adjustment policies did not negatively affect investment and growth in borrowing countries.9 The issue moved to the front of the institution’s agenda in 1997, when the IMF commissioned two different reviews of the organization’s Enhanced Structural Adjustment Fund (ESAF), one internal and one external.10 Despite its orthodox approach to the problem, the internal report identified country leaders’ “commitment to reform” and broader political factors as key determinants of program success.11 The external review went much further, concluding that one of the failings of ESAF programs was their inability to solicit country ownership.12
At about the same time, Masood Ahmed, the senior World Bank manager responsible for the highly indebted poor countries (HIPC) initiative and PRSP policies, moved to the IMF’s Policy Development and Review Department (PDR). The ESAF review and the change in personnel played important roles in highlighting the idea of ownership for Fund policy staff at a time when they were looking to reform their lending policies – partly in response to severe external criticism of their handling of the Asian financial crisis.13 Michel Camdessus, the Managing Director at the time, remembers being a strong proponent of the policy:
Personally, I felt extremely strongly about requiring ownership. A country that was not ready to publicly state its support for the policies it had agreed to, including the difficult steps it needed to take to fix the economy, didn’t deserve to be eligible for funding.14
Although IMF staff adopted the practice of fostering ownership rather later than at the Bank, some of them ultimately became quite keen proponents of the strategy – including a group of staff in PDR who played a central role in integrating it into organizational practice.15
Both institutions thus embraced the practice of fostering country ownership as a way of responding to what were perceived to be some serious failures in their policies – made visible by the contested failures of the Asian financial crisis and the “lost decade” in sub-Saharan Africa. Rather than admitting wholesale to their responsibility for these past failures, IMF and World Bank staff redefined the terms of the debate by introducing the concept of country ownership. Suggesting that the causes of failure were linked to a lack of political commitment, a subject that they had previously believed to be beyond their control, Fund and Bank staff were able to place a measure of responsibility on low-income countries (LICs) (for their lack of commitment) while promising to improve success rates (by building more ownership). In practical terms, this new attention to country ownership translated into the development of several new policies, including the introduction of general budget support, the streamlining of conditionality, and the development of the PRSP.16 These last two policies will serve as the basis for this chapter’s analysis.
Although the number and character of IMF and World Bank conditions has varied significantly over time, the trend in both institutions has been a persistent increase in both their number and scope – at least until recently.17 Long gone are the days when the IMF contented itself with credit ceilings and a few monetary targets, as in the 1950s and 1960s.18 Instead, by the mid-1990s, IMF stand-by agreements and extended facility arrangements often included a host of structural conditions on trade policy, privatization of state-owned industries, and financial sector reforms. With the advent of structural adjustment loans (SALs) in the 1980s, the World Bank had got into the business of imposing conditions and had also gradually expanded those conditions to cover a similar range of areas. Donors followed suit for the most part, either imposing their own conditions or linking their aid to a country’s successful negotiation of an IMF agreement.
This trend reached its peak in the IMF response to the Asian financial crisis in 1998, as the Fund negotiated some of its most intrusive financing programs. Yet this financial crisis was also a turning point, as it produced a backlash against the continued expansion of conditionality not only among the IMF’s critics, but also, to a lesser extent, within the organization itself.19 As one IMF staff member put it to me, you only had to look at the reaction caused by the Fund’s rigid attitude to Latin America to understand why the various regional departments at the Fund were treading more softly in recent years: no one wanted to end up facing that much criticism again, or losing so many clients. Although the IMF was perhaps the most self-conscious and public in its efforts to reconsider and revise its practices, it was not the only organization engaging in soul-searching about the nature of its conditions. The World Bank and several donor agencies also revised their practice of conditionality in the early part of the new century.
It was just a few years after the Asian financial crisis that Horst Köhler began his leadership of the Fund and revised the IMF’s guidelines on conditionality as one of his first major initiatives.20 Arguing for the importance of “streamlining” conditionality, Köhler sought to reduce the number and scope of IMF conditions.
There are several different stories about what motivated him to call for these revisions. One senior IMF official has suggested that Köhler’s earlier role in negotiations on Indonesia’s post-crisis financing package had led him to conclude “that the Fund had gone overboard” in demanding conditions.21 A former member of the Executive Board suggested that Köhler’s Africa tour first prompted his concern with the scope of IMF conditionality, after he witnessed the level of Fund conditions on the Mozambique cashew industry.22 What is clear is that he acted quickly after his appointment as Managing Director, issuing an interim notice on streamlining conditionality in September 2000, followed by formal guidelines that were approved by the Executive Board in September 2002.23
Why did the Fund embark on these reforms? Although Köhler himself clearly played a central role in making them a priority, he was also responding to broader pressures: growing uncertainty and debate about the effectiveness of the IMF’s conditions and increasing criticism of its policies. The Fund’s concern with achieving country ownership was driven in part by concerns about program effectiveness, particularly after condition-laden programs negotiated in the context of the Asian crisis were deemed by many to be failures. This was a highly contested issue: as one IMF staff paper put it, “The relationship between conditionality, ownership and the implementation of Fund supported programs has been the subject of an extensive debate, both inside and outside the Fund.”24 This paper and several others around this time nonetheless concluded that political support was crucial to program success.25 There was also a growing sense that the ever-expanding number of conditions was making things worse. As one senior IMF official noted: “If conditionality is excessive, it leads to unnecessary political fights that are to the detriment of the programs.”26 Fund staff thus began to use the concept of ownership to identify the causes of program success and failure in both political and economic terms. They also began to explore ways of translating the idea of ownership into a measurable practice – creating a new kind of expertise capable of governing these new challenges.27
The bid to reform conditionality and increase ownership was not entirely driven by concerns about re-establishing expert authority: it was also a rather effective way of responding to some of the Fund’s critics. On the one side, NGOs critical of the IMF’s activities had become increasingly numerous, organized and vocal in the course of the 1990s. They saw a rise in the number of structural conditions as the imposition of free-market ideology on developing countries. On the other side were major shareholder governments, who had also been calling for reforms to the IMF. British officials had been pushing for reforms to the IFIs since the election of Tony Blair’s government; the United Kingdom government’s first White Paper on development in 1997 argued for a shift to a partnership-based approach to lending that “moves beyond the old conditionalities of development assistance and requires political commitment to poverty elimination on both sides.”28 Their Executive Director at the time, Stephen Pickford, was also an advocate for reform.29 In the United States, Treasury officials were under pressure from both left and right to push for IMF reform and sought to respond to concerns that they knew the Meltzer Commission report was going to articulate.30 In late 1999 and early 2000, Treasury Secretary Lawrence Summers called for more focused conditionality as part of the Fund’s overall need to return to its “core competencies” in several speeches. Yet throughout these debates, US representatives remained quite reluctant to reduce the number or scope of conditions.31
After a vigorous debate on the Executive Board, the ultimate form that the new guidelines took was based on five new guiding principles: parsimony, tailoring, coordination, clarity and ownership. Fund staff were to use parsimony in determining what conditions are necessary to achieve program objectives, limiting structural conditions to those that are deemed “macro-critical,” a formulation introduced by Canadian Executive Director, Thomas Bernes.32 Staff must also now carefully tailor programs to the specific contexts of individual countries. The move to redesign conditionality around the idea of country ownership was an effective way for the Fund to both re-establish its expert authority by providing a technical solution to its perceived failures, and to respond to its critics by promising to be more responsive to local contexts.
Although the Bank only undertook a formal review in 2004, it had introduced some significant changes in its conditionality policy over the previous few years, driven by similar concerns to the Fund about compliance and ownership as well as external pressure for reform.
In a 2004 retrospective on adjustment lending, Bank staff suggested that both technical and political considerations played an important role in driving changes in conditionality policy at the institution.33 The Bank’s own research and analysis of the successes and failures of the 1980s and 1990s indicated that overly detailed and complex conditions had little chance of success, that program ownership was key to any adjustment program, and that selectivity played a central role in improving program completion.34 At the same time, Bank staff were aware of the criticisms that had been directed towards them of late by “the development community,” acknowledging that external criticisms helped shape changes in Bank policy.35 One of the more influential critiques came from Paul Collier, who announced the “Failure of Conditionality” in a widely-cited essay.36 The paper was published in 1997, just one year after he became Director of the Development Research Group at the World Bank.
Although most of these critical voices were from NGOs and academics, some of the more influential critics were actually within the British government, and included the British Executive Director to the World Bank, two successive Secretaries of State for International Development, Claire Short and Hilary Benn, and the Chancellor himself, Gordon Brown. Since the election of the Labour Party under Tony Blair in 1997 and the subsequent creation of the Department for International Development (DFID), the British government had begun to stake out a more critical approach to international development financing, as I will discuss below. The British Executive Director, Tom Scholar, along with Tony Benn, pushed hard for the Bank to undertake a formal review of its conditionality policy.37
One important shift in the Bank’s conditionality policy came in 2004 with the Board’s adoption of a new operational policy and procedure for lending: the World Bank decided to replace adjustment lending with development policy lending.38 This involved not simply a change in name – although the symbolic importance of that change should not be underestimated – but also a number of shifts in policy: the new policy document eliminates all reference to privatization as an explicit objective of such programs and places more emphasis on poverty, participation, the environment, fiduciary responsibility and other related governance issues.39,40 The World Bank also undertook a formal review of its conditionality policy in 2004, producing a report on its findings in 2005.41 The report, dismissed by critics as more of a justification than a critical review of current policy, ultimately argued for a series of “good practice principles” similar to those adopted by the IMF in its streamlining exercise: criticality, transparency and predictability, customization, harmonization and ownership.
International organizations (IOs) are not the only players that have made significant changes in their approaches to conditionality. Both the UK and the US made more radical changes in the 2000s – although they did so in different ways.
In its 2005 policy paper, Partnerships for Poverty Reduction: Rethinking Conditionality, the British government declared that it would no longer impose economic policy conditionality on those states with which it had a bilateral aid relationship.42 While the British retained certain conditions to ensure fiduciary responsibility – i.e. ensure that the money is spent transparently – and also introduced some new governance conditions, they committed themselves to stop imposing formal economic conditions.43 This placed them in a relatively small minority of donor states, including the northern Europeans, who believe that conditions undermine rather than support development efforts, and demonstrate this conviction in their aid programs.44 Although, in the words of one NGO representative, the British were sometimes viewed as “pot smokers” by other donor government for their radical position on conditionality, they did quietly persuade other nations to pay attention to their ideas.45
Why this significant shift in development philosophy and policy by the British government? Nearly all of those I interviewed on the subject suggested that it was a logical outcome of the previous twelve years of Labour government policy, consistent with the creation of DFID as a separate department and with a succession of White Papers by that department. Some argued that the first Secretary of State for Development, Clare Short, had been instrumental in changing the overall tone and philosophy of the UK’s aid to one based on the principle of partnership.46 Others suggested that the more radical shift to eliminate economic conditionality owed more to the second Development Secretary, Hilary Benn, who was willing to take bolder action,47 while others suggested that it was the movement of so many NGO staff into DFID offices that explained the shift in policy. Underlying these individual personalities and dynamics was a broad preoccupation with “the failures of existing aid,” as one former assistant to Clare Short put it.48 As they noted in their first White Paper on development, the British were keenly aware of the costs of the lost decade in sub-Saharan Africa.49 Not unlike the World Bank, they traced the source of many of the failures of aid to “two flaws” in thinking about the relationship between politics and economics: the belief that the state is the only solution, and the belief that it is the only problem.50
This emphasis on the political economic dimensions of aid failures at both the World Bank and DFID was not a coincidence: DFID, together with the UK think tank, the Overseas Development Institute (ODI), were major actors in the debate on aid effectiveness. Together, they produced a significant amount of research in the late 1990s and early 2000s indicating that excessive conditions were often part of the problem, including Tony Killick’s influential work on “the failings of conditionality.”51 Finally, the British government was no more immune than other institutions from political pressure: this took the form of an organized and vocal community of NGOs in the UK, as well as the recognition that, in the words of one Treasury staff member, it is politically very difficult as donors to impose conditions on other nations.52
On the other side of the Atlantic, the Americans also made a significant change to their conditionality policy, through the creation of a new donor agency, the Millennium Challenge Corporation (MCC), in 2004. The MCC seeks to promote sustainable economic growth through a different kind of process than that used by the IMF, the World Bank or DFID. Rather than imposing ex-post conditions on countries that must be followed once a timetable of financial assistance is agreed upon, the MCC reverses the process by pre-selecting countries that have already attained certain basic criteria in three broad areas: “good governance, economic freedom and investments in people.”53
The criteria are interesting in the way that they seek to provide a quantitative measure of compliance, even when the objects being measured are extremely complex (like the level of political rights). I will further discuss this effort to measure development results in Chapter 8. For the present, it is worth noting that the MCC transforms these criteria into a “scorecard” for each country, and determines whether the state has passed or failed in each of these different areas, drawing more attention to the question of failure and creating a particularly stark distinction between those eligible and ineligible for assistance. Although in many ways the MCC approach to conditionality is significantly different from those being adopted by the IMF, World Bank and DFID, it does share one important attribute: the MCC places significant emphasis on ensuring – and measuring – the level of country ownership.54
Analysing conditionality reform
In the context of debates about what counts as program success and the move to begin to problematize the political dimensions of policy failure, IOs and donors developed a range of strategies for redesigning conditionality. Despite their differences, they share some significant commonalities in their underlying style of governance, similarly conceptualizing the problem of politics, developing more symbolic techniques to engage the problem, and broadening the basis of authority and informalizing power relations in the process.
If the experiences of the 1990s had produced some significant debates about the expert authority of the IFIs and donor states, then these new conditionality policies can be understood in part as efforts to respond to those contested failures.55 All four institutions’ strategies are shaped by their interpretation of the causes of those earlier failures – whether it is the problem of excessive conditions for the World Bank, the IMF and DFID, the importance of selectivity at the Bank and the MCC, or the importance of country ownership at all four.
What all of these policy innovations have in common is the problematization of the “political economy” dimensions of development, using a cluster of public choice assumptions to conceptualize and engage with their objects. A handful of studies are cited over and over again in the policy literature on the subject: analyses by David Dollar, Jakob Svensson, Paul Collier, Tony Killick and Craig Burnside, all of whom draw heavily on public choice theory (and most of whom worked for the World Bank at some point).56 What does politics look like from this perspective? It takes two principal forms: first, as the “environment” within which conditions are implemented, an environment made up of existing policies and institutions.57 Second, the political appears in somewhat more active guise as the leaders of borrowing countries, who are deemed to be either “reformers” – keen to implement “sound” economic policies – or not.58 Dollar and Svensson argue that these “political economy variables” play a crucial role in determining whether or not an adjustment loan achieves its intended outcomes, regardless of World Bank efforts.59
From the perspective of public choice theory, development financing should therefore be more selective, directed towards “reformers,” since “adding more conditions to loans or devoting more resources to manage them does not increase the probability of reform” for poor performers.60 What then is the role of conditions, if they are not needed to “buy” or even encourage reform? They become a kind of “commitment technology,” in the language of public choice and credibility theory: they are a way for governments to (voluntarily) bind their own hands, in order to show markets and donors that they are credible reformers.61 Yet this new logic for conditionality contains a dilemma for donors: markets will only find governments credible if they trust that they will follow through on their promises. Research by Collier and others indicates that many governments state their willingness to undertake reforms in return for financing, and then fail to follow through.62 In public choice terms, IFIs and donors thus face the problems of adverse selection and moral hazard: because most World Bank loans are fully disbursed to borrowers even when conditions are not met, governments have considerable incentives to misrepresent themselves as reformers. Hence the crucial importance of selectivity: IFIs and donors must not lend to poor candidates, not only because aid is unlikely to foster reforms, but also because they will actually undermine the credibility of good candidates’ conditions.63
The logic of public choice theory points towards a different kind of conditionality policy – one where the number and scope of conditions matters much less than the credibility of governments’ commitment to implement them. Commitment, or ownership, as it eventually came to be called, becomes the key to addressing the “failure of conditionality,” as Collier bluntly called it.
How was this new public choice conception of conditionality translated into practice? There is little question that there has been a culture shift in the IFIs, particularly at the World Bank, around staff’s perception of the role of conditionality. A number of staff who I spoke with suggested that conditions were far less important than they had been, one former senior staff member suggesting that in the region that he was involved in, “we do not have conditionality any more.” Instead, “government comes up with a plan which we discuss; the plan is put down on paper and discussed, and if it seems to be something worthy of support, we give them the money to help them implement it.”64 A closer examination of the evolving practices of conditionality, however, suggest that rather than disappearing, conditions have changed in form, becoming increasingly symbolic in character, and informal in application.
As I discussed above, public choice theory proposes a much thinner conception of conditions as commitment technology. Conditions become less about what they are (required reforms) and more about what they represent (signals of “sound” intentions). Their relationship with concrete policy changes therefore becomes increasingly distant and hypothetical – more symbolic than real. Although all conditions today potentially play this symbolic, signalling role, the IMF has gone so far as to introduce a kind of conditionality that is purely about signalling commitment. The policy support instrument (PSI) does not include any financing whatsoever: it is conditionality without the money. To date, seven countries, all of them African, have negotiated a PSI with the IMF. Why would a government agree to conditions without receiving any financing in return? It would be in order to signal to donors and financial markets their willingness to stick to the sound policies that the IMF supports.65
At the same time as formal conditions have been pared down and become increasingly symbolic, informal conditions have proliferated. At the World Bank, the number of formal conditions has dropped from above thirty-five on average in the late 1980s, to twelve in 2005.66 Yet even as these formal conditions have declined, there has been an increase in those conditions that the Bank does not officially define as such: these include benchmarks and triggers, which are used in programmatic lending, as well as the country policy and institutional assessment (CPIA), which all play a role in determining whether funding will be provided, and at what level.67 Thus, while official conditions may have decreased, by 2001 the average number of benchmarks increased from fifteen in the early 1990s to twenty-three or more, and as many as thirty-five for the poorest countries.68 Even at the IMF, where one of the objectives of the 2002 streamlining exercise was to rely less on subjective structural benchmarks in order to increase the “clarity” of conditions, the Board reversed course after the 2008 financial crisis and replaced quantitative performance criteria with more subjective reviews, again blurring the boundaries of what counts as conditionality.69 Paradoxically, even as IFIs and donors seek to make conditions more transparent, they are becoming less visible.
As donors and IFIs have scaled back formal conditions, they have also been increasing their technical assistance, upping the role of consultants and other sources of policy advice. For example, in the UK, as both DFID staff and NGO critics point out, even as formal conditionality has declined, the agency’s budget for technical assistance remains “huge” and has become a crucial means for applying less formal pressure on borrowers.70 As ODI researcher, Ruth Driscoll, put it, “In practice, what you get is a bit of a fudge where DFID staff engage in a lot of ‘policy influencing’.”71 Finally, in their studies of “new” or “post-conditionality” in Africa, Graham Harrison and Jeremy Gould have both demonstrated an increase in perhaps the most invisible kind of development conditionality: the growing internalization of donor and IFI norms by developing country leaders, particularly by finance ministry officials who have come to accept the imperatives of domestic adjustment.72 The combination of this kind of internalization with the persistence of significant aid dependence in many countries makes it difficult to identify the degree of external influence.73
As public choice theory would have it, the new conditionality avoids power dynamics altogether, since borrowing countries use conditions to tie their own hands and to signal to markets and donors their ownership of policy reforms. Not too surprisingly, this account turns out to be something of a fantasy. Instead of eliminating power dynamics, the informalization of conditionality techniques discussed above produces a certain informalization of power relations. Even where formal conditionality has been reduced or eliminated, its ghost lives on at the edges of or in the spaces between formal lending programs – its invisibility lending it a paradoxical kind of power.
There is no question that the reduction in the number of formal conditions by the IMF, World Bank and DFID alters the ways in which the lender can influence the behaviour of the borrower. At the IMF, the attempt to limit step-by-step conditions means that staff have less ability to control how borrowing states comply with conditions, while efforts to focus on their core areas of expertise limit the issues they can address. Yet there is little doubt that the institution retains significant power in its relationship with borrowing states. The very fact that poor African countries would choose to accept the constraints of the PSI without financing suggests that the IMF wields considerable authority through its capacity to decide what counts as sound economic policy.
The fact that the World Bank and DFID have moved even further away from formal conditions suggests that the diminution of lender control will be even more noticeable; yet their increased emphasis on technical assistance, consultants and more flexible benchmarks does not so much reduce as transform their power to influence borrowers’ policies. One former senior World Bank staff member I spoke to suggested that conditions have given way to conversations at the institution. In cases where there is considerable evidence that the country is moving in the right direction, that conversation can be reasonably open-ended. On the other hand, “if you have a situation where there is no progress on the poverty reduction front, then you need to ask tougher questions, and it means a tougher dialogue – which can be interpreted as conditions.”74 This shift towards a policy of conversations is a double-edged sword as it can both complicate and conceal the power relations at work in the aid relationship, evidenced by the fact that surveys have indicated that borrowers do not see much difference between formal and informal conditions.75 To put it more succinctly, the difference between a condition and a conversation is not always clear to those on the borrowing side of the relationship.
Even at the MCC, where the practice of relying on quantitative pass-fail indices appears to be the least ambiguous approach possible, the dynamics of selectivity also work to informalize power. As the then Chief Executive Officer of the MCC, John Danilovich, noted with pride, poorer states that have yet to qualify for MCC assistance have begun to proactively make changes to their governance practices and business regulations in hopes of eventually qualifying.76 These changes are no doubt in large measure aimed at achieving a better score in the MCC indices; yet they are also a kind of performance of good faith – a demonstration of country ownership. When conditions become increasingly symbolic, power relations also take a representational turn, as borrowing countries seek to signal their willingness to be good political economic players.
This willingness to conform demonstrates the expert authority of the IFIs and donors as they set themselves up as the arbiters of “sound” policies that the markets (and other donors) will view as credible. This expert authority is a particularly productive form of power. Ilene Grabel points out that the discourse of credibility is highly political and performative: the act of declaring a policy as credible (generally by an external, authoritative actor) seeks to create rather than to simply signal credibility.77 If the initial performance of the statement of credibility is accepted as authoritative, donors and market actors will follow through with funds and further endorsements that effectively reinforce the initial statement about credibility. Moreover, those policies that are deemed “sound” remain strikingly familiar, focusing on low inflation, financial and trade liberalization, and the creation of a friendly business environment – the stock-in-trade of neoliberalism.
Paradoxically, as conditions become more symbolic, they also become more performative: conditions are designed to do a lot more than simply indicate what policies should be changed; they now also communicate a particular kind of political will. Similar to the standards and codes that I will discuss in the next chapter, these seemingly simple techniques are thus delegated a new kind of power, as conditions take on a life of their own.
These new subtler forms of power can nonetheless be devastatingly effective. We should not forget that underpinning the new conditionality is a belief in the value of selectivity – of only providing financing to those countries with good policy environments and leaders who demonstrate their commitment to reform. The aid effectiveness studies discussed above suggested that those not deemed “genuine reformers” should receive advice and technical assistance but no money; in practice, as one former senior World Bank staff member noted, not only has the allocation of dollars been reduced to countries deemed poor candidates, but so has the time spent on them, which is a serious mistake given that “the whole point was that these countries needed more engagement.”78 Moreover, in the wake of these changes, Bank aid in particular has become more selective.79 The cost to poor countries of not complying with informal conditions – of not convincing lenders of their commitment – is therefore very high indeed. Selectivity and the threat of selectivity act as particularly potent forms of exclusionary power, sorting countries into the saved or the potentially savable, and the damned.
Conditionality policy is not the only policy through which IFIs and donors have sought to foster country ownership: the PRSPs are also key to that strategy. In fact, the PRSPs can be understood as a more active process for generating the country ownership that is needed for streamlined conditionality to work. First introduced in 1999, PRSPs replaced the poverty framework papers (PFPs) that the IMF and World Bank developed in the late 1980s to coordinate their programs.80 By the late 1990s, institutional actors recognized that “the PFP process was broken”: the plans were drafted in Washington, largely by the IMF, with countries signing off but not terribly committed to the program.81 PRSPs were first developed by staff at the World Bank and IMF in the context of the debt relief initiatives for “highly indebted poor countries” (HIPC). Staff developed the idea of requiring governments to draft a PRSP before receiving debt relief – developing a “big picture” plan for how the funds would be used. The practice of creating the PRSP was also intended to give borrowing governments more responsibility for setting their own priorities, and at the same time to nudge them towards a more responsive relationship with their population.
Like efforts to streamline conditionality, the development of the PRSP was driven in part by political pressure from states and NGOs, and in part by expert concerns about past policy failures and the belief that country ownership was essential if development finance was to succeed. With the PRSP these institutions went considerably further, however, in their efforts to re-establish their flagging authority by developing techniques designed to foster public involvement in the creation of development plans and thus actively build local ownership. Although its development was contested, the PRSP ultimately became the key mechanism for putting country ownership into practice.
Karin Christiansen and Ingie Hovland, in their excellent analysis of the dynamics underpinning the development of the PRSP, have described its adoption as the consequence of a “tipping point” in the international development field.82 As both internal and external actors began to problematize the political dimensions of policy failure, debates about aid effectiveness, new concerns about poverty and participation, and internal bureaucratic dynamics converged to make the PRSP the crucial practice for operationalizing ownership.
External actors played an important role in pressuring the IMF and World Bank to adopt a policy that would enable country ownership and poverty reduction. As in the case of conditionality, the UK, under the leadership of Clare Short and Gordon Brown, argued forcefully for a more participatory and country-owned process.83 The Clinton administration was less enthusiastic but ultimately supported an instrument that they saw as ensuring that the neediest people and states would benefit from debt relief and additional assistance.84 The Development Assistance Committee (DAC) of the Organization for Economic Co-operation and Development (OECD), an organization bringing together major donors, also played an important role in championing the new emphasis on ownership.85 NGOs like Jubilee 2000, Oxfam, Christian Aid and ActionAid also pushed for debt reduction and argued for a poverty-focused approach to aid. Finally, although southern actors played a smaller formal role in the development and adoption of the PRSP process, the strategy was modelled in part on several developing states’ own earlier poverty reduction plans, including those in Tanzania, Bolivia, Mozambique and, above all, Uganda.86
It is impossible to neatly separate out political from expert pressures, however, since all of the main “political” actors, like donors and NGOs, participated in the debates about policy success and failure, and articulated their concerns in these terms. Nevertheless, the public choice-informed analyses of aid effectiveness discussed above tended to tackle the problem of reform in relatively narrow terms, emphasizing the need for orthodox economic reform, and pointing to political commitment as a crucial determinant of success. These concerns about how to reduce policy failures combined with a growing emphasis on poverty reduction as a central goal of development finance, though there was little consensus on how to best tackle the problem.87
These emerging concerns about political commitment and poverty reduction combined with a third debate around participation. The practice of integrating participation into development had been circulating for some time, appearing in a number of different contexts with different meanings. As I will discuss in the next chapter, in the public choice-informed debates on aid effectiveness, local participation was defined as a means to better service delivery. The idea of participation was also popular among the more “critical” units of the World Bank, like Social Development, where it was framed as a way of getting civil society actors to hold government accountable – a conception that would eventually come to underpin the “demand side” of good governance. More broadly, increased emphasis on development finance as a “partnership,” in which developing country governments must take some responsibility for reform, relied on more active participation from borrowing country actors.88
By the late 1990s, there was therefore an emerging consensus about the importance of increasing aid effectiveness, focusing on poverty reduction, and encouraging greater borrower participation and responsibility – even if the actual meaning of the term “ownership” remained contested. This process of debate and problematization helped create a tipping point that meant that some kind of change in policy was needed; what form that policy was to take – the question of how to operationalize ownership – was, however, far from inevitable. Several different possibilities were on the table, including World Bank President James Wolfensohn’s proposed CDF.
Although the PRSP has been popularly represented as a brainchild of Wolfensohn and a logical continuation of his CDF, the World Bank President was not actually pleased with the PRSP when it was first developed, seeing it as competing with his framework.89 His fears were at least partly vindicated, since the PRSP quickly overshadowed the CDF as the dominant practice for tackling ownership within the broader aid community. Wolfensohn’s CDF was defined by four major principles, emphasizing long-term thinking, citizen participation, country ownership and measureable results.90 It was therefore very much in tune with the broader shifts in developing thinking that I have discussed, and also consistent with the principles underpinning the PRSP. Yet the CDF remained in the realm of ideas – it was a framework, not a policy – and in an organization as vast and decentralized as the World Bank, frameworks need to be put into practice to have much effect. The PRSP, in contrast, promised a set of concrete practices for achieving these objectives: it was practical, and as such, quickly eclipsed the CDF. The PRSP was eventually represented as a way of putting the CDF into practice.91
The IMF’s own adoption of the PRSP was actually a matter of chance. Facing pressures from the Asian crisis, and NGO and developing country criticism of conditionality policy, PDR staff sought to restructure their lending relationship with poor countries. The PRSP, which in its earliest form had been developed partly by senior IMF staff member Jack Boorman,92 seemed like a policy that might just do the job.93 The Fund’s transformation of the old ESAF into the poverty reduction and growth fund (PRGF) (which was then tied to the PRSP) was far from automatic, as the very question of whether the Fund should have a presence in LICs was contested at the institution; yet the IMF Managing Director at the time, Michel Camdessus, was a deeply religious man who saw the Fund as having a moral responsibility towards the poor, and ultimately pushed hard to ensure that the institution included poverty reduction as a central part of its mandate.94 In my interview with him, Camdessus himself noted that focusing on poverty as well as macroeconomic stability required a difficult shift in the culture of the institution – one that was resisted by some staff members and stakeholders.95 The significance of this shift for the Fund’s internal culture should not be underestimated. As one senior IMF staff member put it: “It’s now accepted in this building that the IMF . . . [is] an instrument of the international community to end poverty. You would have been laughed out of the building if you had said that twenty years ago.”96
Over time, the PRSP’s role was expanded, as it came to replace the PFP as the planning document for IMF and World Bank concessional lending. Donor governments also began to use the PRSP for their own aid strategies, making it a lynchpin in the governance of development finance.
Analysing the PRSP
Perhaps even more clearly than efforts to streamline conditionality, the introduction of the PRSP brought with it a number of important changes to how the IFIs and donors did the work of financing development, including the introduction of new inscriptions and technologies of community, the integration of new actors into the work of governance, and the application of subtler but more productive forms of power.
Engaging new actors
Those involved in developing and implementing the PRSP have sought to use this policy strategy to engage new actors in the practices of governance. Who has historically been responsible for the day-to-day work of governing international development finance? If we look back to World Bank President Robert McNamara’s “war on poverty” in the 1970s, McNamara himself was not only a force behind the adoption of new policies, but also played dual roles as a diplomat working behind the scenes to influence borrowers, and as a technocrat whose measurement techniques ensured that staff were achieving their targets.97 As the Bank and Fund became actively involved in structural adjustment in the 1980s, more work was delegated to technical staff. With the shift from one-off projects to more program-based lending, domestic government actors became more important; yet they were conceptualized in passive terms as consumers of policy advice rather than as active participants, as IFI staff often bypassed domestic institutions and actors.
With the introduction of policies like the PRSP, Bank and Fund staff have continued to be key actors. Yet the PRSP’s architects hoped that the day-to-day practice of governing economic growth, adjustment and poverty reduction would increasingly be undertaken by a new set of actors. Domestic governments were seen as much more active participants in their own governance, preparing their own strategies and taking more responsibility for their successes and failures. The PRSP’s creators also sought to integrate civil society actors as active players, participating in the creation of the PRSP and then using it to hold the government to account. In practice, the range of actors involved in the formulation of the PRSPs has varied considerably, but has included religious groups, parliamentarians, worker and peasant organizations, indigenous peoples, women’s groups, local government officials and others. Even the poor themselves, who were once deemed unable to organize themselves into an effective political force, are now seen as a group capable of having an effect on government policies.98 In the process, developing countries’ civil society has come to be viewed as a necessary quasi-political space – a third space, to use Nikolas Rose’s term – in between the rent-seeking realm of the state and the self-interested logic of the market.99
In order to understand what role these new actors play, we need to consider the techniques through which the PRSP does its work. Two kinds of techniques in particular are crucial: inscription techniques, and what Rose calls “technologies of community.”100 The technique of putting things down on paper has long played an important role in IFI lending to poor countries: the PFP was also designed to translate complex domestic economic factors into practical goals. Both the PRSP and the PFP it replaced can usefully be understood as inscriptions: they each seek to translate the complexities of a country’s economic, social and political context, their aspirations and objectives, into a single document capable of enrolling actors and orienting action. Yet they are very different kinds of inscription.
If we compare them, the most obvious difference between the two is the length and scope of the documents. PFPs were relatively short (twenty- to thirty-page) documents that covered a range of areas, from social to monetary policy, viewing them all through the lens of economic efficiency.101 The object of the PFPs, and the structural adjustment programs that they facilitated, was economic transformation. The PRSPs, in contrast, are not merely much longer documents – often close to 200 pages long – but also much more ambitious in their objectives.102
What is perhaps most interesting about the PRSP is the way that it is produced – a key selling point for its advocates. This is where technologies of community combine with inscription techniques to produce a more dynamic and performative document. Whereas the PFP was a report created by Bank and Fund staff in order to coordinate their respective development programs, the PRSP is, at least in theory, prepared by the borrowing government. It is not just the Treasury staff who are involved: many of the affected ministries now play a role, together with a range of civil society groups and individuals. The goal of the PRSP is to engage a myriad of actors in discussing the shape of the country’s development policy. By translating some of the past experiences with micro and project-level participation to a macro level, staff involved in implementing the PRSPs seek to engage a range of stakeholders in the process of formulating, implementing and monitoring a country’s poverty reduction strategy. In practical terms, techniques include information dissemination strategies, various kinds of consultations, workshops and focus groups, citizen surveys and report cards.103
By linking PRSPs to a wide range of different development practices, the IFIs have sought to both re-establish and expand their institutional authority. They have sought to re-found their economic expertise by putting into practice many of the insights of public choice theory and institutionalist economics. At the same time, IFIs’ use of various “technologies of community” help to enhance their claims to popular authority – allowing them to argue that their programs are based on a wider, more robust kind of ownership than in the past. These more popular technologies also delegate a certain amount of governance authority to a wider range of actors – giving domestic governments and local populations more control over their own economic development.
Given its popular, participatory dynamic and its effort to redistribute authority, the PRSP appears to be less subject to power imbalances than the PFP. Yet this delegation of authority is far from unconditional. Although the instrumental power that was so often visible in SALs has become less prominent, power relations remain in a different guise.
The PRSP has been designed not simply to engage government and civil society actors in governing, but also to help shape them. As Joseph Stiglitz, then Chief Economist at the Bank, put it, “At the heart of development is a change in the way of thinking and individuals cannot be forced to change how they think.”104 Rather than operating through coercion, the PRSP is designed to work more subtly, fostering change through the document’s production and reception. Even after the document is produced, its publication is to have performative effects. One of the “wagers” of this strategy, as one development think-tank member put it, is that by giving civil society actors new capacities and tools, they will put pressure on governments to become more responsive.105 In the process, PRSPs, as inscriptions, have become increasingly performative, as institutional actors have delegated considerable powers to them – powers that continue to operate after their original formulation.
Another wager contained in this policy is that if government actors participate in devising their own programs – and become “self-responsible” in the words of one IMF Managing Director – they will become better at developing “good” rather than “distorted” policy environments.106 This emphasis on fostering an internal will to reform resonates with Foucault’s concept of governmentality, in which the objective is to govern by encouraging actors to regulate themselves.107 This is also a particularly proactive strategy that seeks to create the conditions necessary for its long-term objectives, rather than simply reacting to ongoing events and challenges.
The power relations enabled by policies like the PRSP are thus not only more productive and proactive, they are also increasingly indirect. Although the goals are economic – development, adjustment, stability and poverty reduction – the means are through various intermediaries. It is only through the roundabout route of mobilizing the poor, encouraging civil society pressure, generating demand for reforms and fostering political will that the ultimate objective is attained.
Like efforts to reform conditionality, the PRSPs have the effect of informalizing power and making it less visible. This is rather ironic, given that one of the central premises of the PRSP is to make development planning more transparent. Although the preparation of the PRSP as inscriptions is a more public process, with many more participants, it is certainly not universally inclusive. In most countries, not all civil society groups are included in the consultation processes: in Uganda, for example, the unions and many other groups seen to be too “political” were excluded from the first PRSP process.108 In Bolivia, it was local government officials who played a preponderant role in consultations.109 Moreover, studies of the Bolivian, Ghanaian, Ugandan and Nicaraguan PRSPs suggest that donors continued to play a significant role in defining the ultimate form of the PRSP, often undermining the “bottom-up” accountability that the process was intended to create.110 Which actors were included in the participatory processes? Which ones were heard? Whose voices were ultimately translated into the PRSP documents? The answers to these questions say a great deal about the power relations reflected in and enabled by this particular poverty reduction strategy. Yet they are obscured as much as they are revealed by the PRSPs themselves.
When we look at the recent revisions to conditionality policy and the introduction of the PRSP together, we find that despite the somewhat different motivations and pressures behind them, they have much in common. Both have been designed to respond to the contested failures of finance and development in the 1990s and are concerned with improving the efficiency of development financing by making specific conditions and entire poverty reduction policy frameworks more responsive to local contexts. Both policies are also motivated by a concern about the flagging authority of the IFIs and donors. By reinventing structural adjustment policies as PRSPs and streamlining or redefining conditionality, IFIs and donors seek to regain expert authority and bolster it through increased popular authority. At the same time, both continue to reinforce power relations – in less formal and often less visible forms.
The turn to ownership represents a shift in how the work of governance is done, as the overall approach to governing has become more provisional. As I discussed in Chapter 2, a provisional style of governance is a particular kind of response to the problem of policy failure and to the fragility of expert authority: rather than seeking to control absolutely, those seeking to govern provisionally apply a less direct, more proactive approach to the task – one that relies crucially on increasingly symbolic techniques, and often hedges against the possibility of future failure.
In the case of the ownership strategy, we can see how governance increasingly occurs indirectly rather than through direct action. Influence is exerted in the spaces where conditions have been cut, in the form of advice, quasi-conditions, conversations or unspoken expectations that must be met to demonstrate “genuine” commitment to reform. The PRSP, in turn, is designed to solicit pressure for reform from civil society actors rather than applying it directly on government actors. This indirect form of governance relies crucially on the increasingly symbolic character of policy techniques. Both loan conditions and inscriptions like the PRSPs are valued for their capacity to signal political commitment. The symbolic nature of these policy techniques is crucial for their performative effectiveness. Paradoxically, it is because conditions and PRSPs are abstract signs of ownership that they work to foster “real” ownership. If the markets believe in the credibility of a country’s leaders’ commitment to fulfil the conditions, they will reward that country, making it easier for the leaders to stick to their commitments. Similarly, if civil society members see the PRSP as a signal of their government’s responsiveness to their concerns, they will begin to take more ownership of the strategy and related aid programs. In theory, at least, the symbolic character of these techniques produces a virtuous circle in which the appearance of a thing helps to make it real.
These policies are also highly proactive: the Bank and major donors increasingly recognize and address the temporal dimension of policies, seeking to create the conditions for longer-term reform by not only altering the incentives but also changing thinking. They are therefore seeking to pre-empt failure by changing the context in which policies are developed and implemented. It is not just this more proactive approach to governance that seeks to inoculate the policies against failure: all of these more provisional tactics of governance do the same. The indirect techniques of the PRSP and streamlined conditionality allow the IFIs and donors to step back a bit from the highly politicized fray of more direct conditions, and escape some responsibility. By delegating more responsibility for success and failure to domestic governments and their populations, they further distance themselves from potential failures. Institutional actors’ reliance on more symbolic techniques also allows them to hedge their bets: if they are interested less in “real” ownership than in its appearance, then the fact that participatory dynamics are often instrumentalized, and that ownership is not always genuine, is not a sign of failure.111
Yet despite these efforts to inoculate themselves against failure, the possibility of failure remains a continual preoccupation of IFI and donor staff – in large measure because the strategy of fostering ownership is fraught with so many of its own problems. I will take up these various failures and their implications in greater detail in the Conclusion to this book. For now, it is just worth noting that these challenges have begun to erode internal and external support for the PRSP, unravelling the initial consensus on the strategy of ownership. Although they have continued to pursue efforts to foster country ownership in various forms, IFI and donor staff have grown increasingly ambivalent about its promise, running into persistent difficulties in quantifying ownership and finding the strategy difficult to put into practice.
One of the challenges that the ownership strategy has had to come to terms with is its complex relationship with a second important governance strategy: the push to develop new global standards in everything from good governance to budgeting practice to infant mortality rates. As I will discuss in the next chapter, even as IFI and donor staff have sought to tailor their policies to particular local contexts, they have also pursued a far more universalist strategy of standardizing global practices.