5   IMF Stewardship of Global Finance

For over three decades, the IMF served as the chief institution for managing financial crises in EMDEs. In this domain, it had no competitors, leading a prominent student of the Fund to remark that it is “the most powerful international institution in history” (Stone 2002, 1). This was an apt description—certainly up to and through the Asian crisis, when the IMF exerted extraordinary influence over policy and institutional formation over a large swath of the globe.

But the Asian crisis marked a fundamental turning point for the IMF. For better or worse, the declining influence of the IMF after the Asian crisis was not associated with the rise of alternative institutions with sufficient resources, expertise, and legitimacy to respond to national, regional, or global financial crises. It is therefore unsurprising that the global crisis led to the resurgence of the IMF, restoring it to its leading position in crisis response—directing assistance to ailing economies, conducting surveillance, and disciplining clients.

Those arguing on the side of continuity in the global financial governance architecture marshal substantial additional evidence to support a univocal reading. They claim that the crisis has induced only trivial adjustments in IMF governance and practice. But continuity is only half of the story—and by no means the most interesting half. A more adequate reading concludes that the crisis has had complex, contradictory, and uneven effects on the institution. If we look carefully inside the Fund and at the external environment in which it operates, we can glean evidence of substantive discontinuities, inconsistencies, and, in deference to the best analyses of the matter, what I will refer to as ambiguities across various dimensions of the institution.¹

Continuities at the IMF during the global crisis entail replenishment of its financial resources and reinvigoration of its central mission; stability in its formal governance practices, leadership selection, and overrepresentation of key member states; and continued reliance on politically intrusive austerity policies in client economies despite a fleeting embrace of countercyclical policy during the earliest days of the crisis.² Discontinuities entail substantial shrinkage in and transformation of the IMF’s geography of influence; the nature of its partnership with European institutions; financial support to the institution from EMDEs, which extends opportunities for informal influence over IMF practices; the institution’s practical and rhetorical support for capital controls; and the new architectural context within which the Fund is entering into new kinds of complementary and competitive institutional relationships. Key ambiguities entail inter alia a widening gap between IMF rhetoric and research on the one hand and its operations on the other. Ambiguity is most apparent in the scope of IMF conditionality; its attention to social spending targets and vulnerable groups during crises; and its practices regarding austerity, inequality, and fiscal consolidation. These assertions are summarized in table 5.1 and elaborated in what follows.

Table 5.1 Continuities, Discontinuities, and Ambiguities at the IMF during the Global Crisis
Institutional Attributes		Continuity		Discontinuity		Ambiguity: New Rhetoric, New Research, Old Practice
IMF identity and influence		• Resurrection of its role in crisis management and restoration of its financial vitality		• New and constrained geography of influence • IMF in the Troika • Institutional complementarity or competition on the horizon?
IMF governance		• Gridlock on formal governance reform, 2010–2015 • Resilience of the “gentleman’s agreement” on leadership selection • Lack of independence from key member states		• New lenders, rising powers, new (informal) channels of influence
IMF practice		• Austerity and politically intrusive policy adjustment in support packages • Restoration of traditional discomfort with countercyclical policy		• Support for capital controls, particularly on inflows but also on outflows		• Narrower scope of conditionality • Attention to social spending targets, the poor, and the vulnerable in support packages • Growing inconsistency on austerity, inequality, and fiscal consolidation
Source: Author’s analysis.

A Conceptual Note for Economists: Constructivist Accounts of Ambiguity, International Organizations, and Institutional Change

A word is in order, especially for an economics audience, on the matter of institutional practice and change. Within political science, constructivism refers to a broad school of thought that refuses to reduce institutional behavior to geopolitical or economic imperatives or to self-interest (see Abdelal, Blyth, and Parsons 2010). In the words of leading constructivists, “The central insight of constructivism is that collectively held ideas shape the social, economic, and political world in which we live” (ibid., 2). The constructivist approach examines existing and emerging tensions among stakeholders and the myriad fluid factors that bear on their influence over institutional practice. Constructivist accounts emphasize the centrality of shared ideas and norms not because these are seen to cancel economic and material power but because those imperatives are always mediated by the ideational framing that agents engage to make sense of the world. Critical questions for constructivists, then, are how the diversity of ideas and norms that permeate institutions and epistemic communities, and that are held by clients and other external actors, affect institutional practices; how, when, and why ideas and norms change; how actors within institutions and in epistemic communities demarcate standards of appropriateness and legitimize particular understandings of rights and obligations; and how changes in these domains ramify at the level of institutional and policymaker behavior. As this discussion suggests, constructivist accounts often stress processes that are complex, slow, and incremental and that involve layering of new ideas and practices over old.³

Constructivism peers into the black box by examining the internal culture of international organizations with an eye toward illuminating the diverse influences on their identities and changes in practices over time. The concept of ambiguity emerges within this context and is deployed in diverse ways. Jacqueline Best (2005; 2012) advances a sophisticated set of arguments about the fuzziness of rules within (complex) international organizations. Rather than impeding institutional performance, ambiguity is theorized as functional and even necessary to institutional viability and success. Fuzziness promotes cooperation among actors with diverse and even conflicting understandings and interests. It facilitates institutional expansion and the exercise of staff discretion, rather than the paralysis and turmoil that would arise in an organization with unambiguous rules. Ambiguity also helps to explain the divergences that arise routinely between institutional mission statements and actual practice. In Best’s account, ambiguity arises not as a deliberate organizational strategy (even though internal actors may prefer it) but rather as a consequence of unresolved conflicts among major principals.⁴

Alternative constructivist accounts of the BWIs emphasize “strategic ambiguity” (Van Gunten 2015) or “organized hypocrisy” (Weaver 2008).⁵ Strategic ambiguity arises as a form of productive avoidance in the presence of conflict among and between societal norms, multiple political masters with heterogeneous preferences, divergent interests and norms of internal staff, and in the absence of professional consensus. Under these circumstances, the organization may purposely enact vague policy and permit ceremonial rather than literal conformity to adopted policy as ways of managing. Organizational hypocrisy is apt to deepen when the interests of critical stakeholders diverge dramatically and/or when there is a breakdown in the professional consensus among an organization’s staff members and among outside professional experts. Such moments generate “escalating hypocrisy” (as per Kentikelenis, Stubbs, and King 2016)—a dynamic layering process that can sustain the myth of organizational cohesion in the face of division, or organizational reform in the face of continuity. In their account, Kentikelenis, Stubbs, and King argue that, as the global crisis progressed, the IMF came to be marked by institutional schizophrenia. In this view, the Fund has come to cloak itself in multiple and increasing layers of rhetorical appeasement, becoming increasingly adept at using ceremonial reforms in order to mask the growing gap between its new rhetoric and the retention of its standard practices.

As the foregoing suggests, constructivism has little difficulty accounting for gaps between institutional rhetoric, research, and practice. Ambiguity speaks directly to this issue, of course, and serves as an important counterbalance to analyses that reduce inconsistencies in IMF practice to alleged dishonesty, callousness, failures of management oversight, or institutional self-promotion.⁶ To be sure, these motivations figure into gaps between IMF policy and rhetoric. But this is just part of the story. The use of a blunt, reductionist approach to examine international organizations (as some analysts are wont to do) diverts attention from internal policy debates, tensions, pressures by diverse stakeholders, subtle changes, and contradictions, all of which can ultimately culminate in fundamental change even in the most hidebound institutions.⁷

Research on an organization’s staff members (as opposed to its leadership) illustrates the ways in which space emerges between top-level decision making and ground-level application.⁸ Recent analyses of the IMF, for instance, have demonstrated the degree to which staff economists shared a powerful worldview over the past several decades. The consensus around the virtues of market liberalization was sustained by recruitment and hiring practices that emphasized shared educational backgrounds, expertise, and normative commitments, socialization practices within the IMF, and a technocratic culture that yields professional authority to the IMF’s economists. These factors contributed to a shared vision in which applied economics is seen to be driven by solid objective analysis. The technocratic presumption provided IMF economists considerable leeway to operate in ways that ultimately frustrated directives or contradicted rhetoric from the organization’s leadership (Barnett and Finnemore 2004, 22–27). Momani (2010) explains what she sees as the IMF’s “split institutional personality” in these terms—as the consequence of the institution’s rigid organizational and technocratic culture and the narrow disciplinary and paradigmatic training of IMF staff economists. In this account, IMF staff economists possess neither the intellectual breadth nor the will to change their practices, even when the institution’s leadership articulates new commitments that may even be embodied in policy directives.⁹

Institutional Change

Constructivism also has much explanatory power with respect to institutional change. The emphasis on the relative influence of shared ideas and norms helps to round out alternative materialist accounts that reduce institutional behavior and change to economic or other “objective” factors. Constructivist accounts combine the subjective and objective domains. Hence, exogenous shocks like economic crises may induce—though not dictate the content of—ideational transformations, which in turn bear on institutional behavior (see Blyth 2002). Crises may permit the flourishing of previously latent or subordinate ideas and norms, alter the balance of power within an organization among advocates of contending ideas, or create the ideological space for the generation of new ideas altogether.¹⁰ Ambiguity facilitates normative adjustment when the stakes are high for internal actors. But since in the constructivist view ideational change is underdetermined by objective factors, change is theorized as messy, contingent, and contested, reversals are likely and perhaps even inevitable, and there is no stable ideational equilibrium.¹¹

In recent work, Best (2014; 2016) has explored postcrisis transformations at the IMF. She finds that the overconfidence of staff members during the neoliberal era has been shaken by the series of financial crises, and especially by the global crisis. Today, IMF staff members operate with an appreciation of the risk of failure. This is to the good, since hubris combined with authority can be dangerous (DeMartino 2011; 2013a). In Best’s view, there has been a pronounced shift at the IMF and other international organizations toward tentative policy advising, experimentation, and new metrics for assessing success and failure. But the pursuit of fail-safe policy entails a paradox, especially to the degree that decisionmakers seek to insulate themselves from the consequences of failure. As Best (2016, 6) puts it, “Ironically, this move towards more cautious policy responses may actually produce more failures in the longer-term: a development program that is limited to those initiatives whose outcomes can be quantified will avoid tackling many crucial challenges; and a system of financial regulation that continues to act as though systemic uncertainties are reducible to micro-level risks will be more [failure] prone.”

We now turn to the primary focus of this chapter, which is to chart the degree to which the IMF evolved during and since the global crisis. As will be apparent in what follows, constructivist (and other) insights inform the analysis and sustain the major conclusions, which are read through a Hirschmanian lens.

IMF Identity and Influence

As with G-20 leaders, officials at the BWIs recognized the need and opportunity for fundamental reinvention in the early days of the crisis. As World Bank president Robert Zoellick put it in 2009: “The old order is gone.… Today we must build anew” (Zoellick 2009, 19). IMF managing director Dominique Strauss-Kahn echoed that, calling “2010 … a year of transformation” (Strauss-Kahn 2010a). But reality was far more complicated than the claim of wholesale reinvention suggests.¹²

Continuity

The global crisis has been good to the IMF. It rescued the institution from the irrelevance that followed the Asian crisis. The demands of the crisis reversed the short-lived downsizing of the institution during spring 2008 and spring 2009. More than one hundred economists were hired by the end of April 2009, and the IMF’s workforce rose in size by about 20% between the end of fiscal year 2009 and 2012 (IEO 2014, 42).

The Crisis and the Resurrection of the IMF

The IMF’s loan portfolio had shrunk dramatically by the time Strauss-Kahn was appointed managing director in November 2007. Total outstanding loans were then US$17 billion, compared with US$110 billion four years earlier (Weisbrot 2015, 156). By the time Strauss-Kahn resigned in May 2011, outstanding loans had risen to US$125 billion, and agreed-upon loans were at least two times greater (ibid.). Nonconcessional lending by the IMF was almost nil before the crisis; it stood at US$400 billion in 2008–2013 (IEO 2014, 2). The institution approved 17 SBAs for more than SDR50 billion in the first year of the crisis, and an additional 20 SBAs for SDR50 billion between September 2009 and the end of 2013 (ibid., 26).

Relative to borrowers’ quotas, the size of loans during the global crisis was the largest in the institution’s history (Nelson 2014a, 161). The average size of the 84 loans that the IMF extended between 2002 and 2007 was 94.6% of country quotas; the 92 loans extended after 2008 and through 2012 averaged 398.3% of quota (ibid.). The largest of the IMF’s programs after 2008 ranged from 10% to 16% of the recipient country’s GDP (Reinhart and Trebesch 2016, 12). By contrast, the median IMF program during the EMDE debt crises of the 1980s reached just 4% of GDP (ibid.). The Mexican and Indonesian loan programs of 1995 and 1997, respectively, made headlines at the time because the loans were almost six to seven times their quota, but they amounted to just over 5% of GDP (ibid., 11, table 1).

The maximum amount of cumulative access for IMF members was raised from 300% to 600% of quota in March 2009 (Boughton 2012, 752). The IMF also provided what it terms “exceptional access” to borrowers whose needs exceeded the official limit. It had done so in the past for “systematically important” countries, such as Mexico, Brazil, Argentina, Turkey, Indonesia, and Korea (Nelson 2014a, 161). From March 2009 to 2011, over half of approved arrangements exceeded 300% of quota (Edwards and Hsieh 2011, 80). Between 2008 and 2012, there were six IMF programs above 1000% of member quota (Reinhart and Trebesch 2016, table 1). Greece’s 2010 loan amounted to 3212% of its quota (and its 2012 loan, 2159%), Ireland’s 2010 loan was 2322% of quota, and Portugal’s 2011 loan was 2306% of quota (ibid.).

The IMF’s reempowerment was driven by several factors. It enjoyed monopoly status by virtue of its expertise in responding to financial distress. As Nelson (2014a, 156) observes, “Aside from perhaps one state (the USA) and one supranational organization (the EU), there is no other governmental or intergovernmental player that can match the IMF in terms of command of material resources,” let alone social power and experience in crisis response. Events on the periphery of Europe also had a significant impact on the IMF. European officials perceived a need for the IMF’s expertise, financial resources, and authority. The IMF responded by inserting itself into the European rescues, and even in countries like Ukraine that are far from the heart of Europe (Lütz and Kranke 2014).

The IMF’s resurrection was facilitated at critical moments by key member states. In November 2008, the G-20 announced the need to ensure the adequacy of the IMF’s resources (and those of other multilateral financial institutions) and its readiness to increase them as necessary (Lesage et al. 2013). Though the matter received little attention, the Japanese government’s decision to lend US$100 billion to the IMF in November 2008 came at an important moment, helping to catalyze financial support from the United States, EU, China, and others (Grimes 2009c; Holroyd and Momani 2012; Jiang 2014, 173).

Prior to the April 2009 G-20 Leaders’ Summit in London, U.S. Treasury secretary Timothy Geithner primed the pump for IMF recapitalization by calling for a tripling of its lending capacity and for rapid progress on long-promised governance reforms (Beattie 2009). G-20 leaders responded by reconfirming IMF leadership in crisis response efforts. The significance of the G-20’s decision was not lost on Strauss-Kahn, who, at the end of the meeting, said that, “Today is the proof that the IMF is back” (Landler and Sanger 2009). The decisions not only restored the IMF’s mandate but also yielded massive new funding commitments (though a portion of what was announced as new funding turned out not to be) (Woods 2010; Helleiner 2014b, 51). G-20 leaders committed US$1.1 trillion to combat the crisis, with the bulk of the funds, US$750 billion, to be delivered through the IMF. The commitment permitted the tripling of the IMF’s lending capacity called for by Geithner. It boosted the institution’s lending resources by US$500 billion through what are called the New Arrangements to Borrow (NAB) and a new general allocation of SDRs equivalent to US$250 billion (Lesage et al. 2013).¹³

The use of the NAB as a vehicle for increased support has consequences for the IMF, as I discuss later. The NAB provides the IMF with a temporary funding source to supplement the standard mechanism of quota subscriptions paid by member nations. The NAB provides support to the IMF through bilateral credit arrangements between the Fund and 38 countries and institutions, such as central banks. Loans under the NAB were to be reaffirmed every six months, and were intended to provide temporary support until the permanent increase in IMF resources called for in a 2010 reform agreement was put in place.

The G-20 gave the IMF a second tranche of new funds in June 2012 as the Eurozone crisis expanded. At the G-20 Leaders’ Summit in Los Cabos, it was announced that the IMF would receive a new infusion of US$456 billion, to come from 37 EMDEs and AEs, though the United States would not be contributing. Geithner argued that the Eurozone’s problems should be solved in-house. He articulated this view as early as mid-March 2012, when he said that the Treasury Department had “no intention to seek additional US resources for the IMF” (quoted in BWP 2012), a position that was important for the administration of President Barack Obama leading up to the 2012 election insofar as opponents were eager to target him as wasting taxpayer resources on Europe.

The global crisis reinvigorated not only the IMF but also other multilateral institutions, such as the World Bank and the Inter-American Development Bank (IADB).¹⁴ The activism of these institutions was facilitated by the G-20’s April 2009 decision to devolve a portion of the new financial commitments to the IMF to regional/multilateral institutions following a proposal by Indonesia (Chin 2010).

Did the United States Outgun the IMF?

Even while U.S. credibility was initially damaged by the crisis, its unilateral monetary power was reaffirmed early on as the Federal Reserve acted quickly to open currency swap lines with the central banks in countries of geostrategic importance. A currency swap is an agreement between two or more central banks that enables a central bank in one country to provide (via an exchange or swap) a certain amount of foreign currency liquidity to another central bank in the event of a sudden liquidity shortage. Swaps enhance financial stability by relieving pressures in short-term money markets. Swaps may also be used to stabilize a currency or maintain trade patterns with key trading partners.

During 2007 and 2008, the Fed opened vast, ad hoc, temporary swap arrangements with 14 central banks. The value of the swaps peaked at almost US$600 billion in November and December 2008; all of the swaps expired in February 2010 (Helleiner 2014b). In September and October 2008, the Fed extended swaps to major AE central banks, namely, the European Central Bank (ECB), Bank of Japan, Swiss National Bank, and the Bank of England, and also to the central banks of Canada, Australia, Sweden, Denmark, Norway, and New Zealand. These were quickly followed by swaps with “diplomatically and economically” important EMDEs, namely, Mexico, Brazil, Singapore, and South Korea, each of which signed a US$30 billion agreement with the Fed. Mexico already had a standing swap agreement with the Fed, but the 2008 agreement extended the amount to which it had access. Notably, the global crisis occasioned the Fed’s first swaps with EMDEs outside of Mexico. Swaps were “highly sought after and highly welcomed” since they were seen as an alternative to IMF support (Gallagher 2014, 75). Most of the central banks that had swap agreements with the Fed drew on them extensively. U.S. interests, including the exposure of U.S. banks to certain foreign markets, critically shaped decisions to extend swaps.¹⁵

Some IMF observers have concluded that the breadth and depth of U.S. bilateral swaps signaled to the BRICs and other EMDEs the necessity of committing new funds to the IMF as a means of offsetting U.S. influence. At the same time, however, the size and speed with which the United States opened swaps early on moderated the bargaining power that BRICs contributions to the IMF might otherwise have facilitated, and which the BRICs might have been able to parlay into IMF governance reform (Helleiner 2014b; James 2014; Lesage et al. 2013).

U.S. swaps raise the following question: does this exercise of monetary power imply that the IMF played a subordinate role during the global crisis? For some analysts, the answer to this question is largely “yes.” Helleiner (2014b) argues that the Fed’s swaps were the single most important type of stabilizing monetary cooperation during the global crisis, dwarfing the IMF’s support. In his view, the swaps reflect and reinforce the vast unilateral monetary power of the United States (see also James 2014; Prasad 2014). A more nuanced response to the question suggests that IMF impact and agency cannot be so easily dismissed. While it would be wrongheaded to deny the continuity of U.S. monetary power, the IMF undertook critical stabilization efforts that U.S. officials could not have undertaken on their own. IMF assistance packages (especially relative to quotas and GDP) and the number of countries with Fund programs after 2009 enhanced the institution’s influence during and following the crisis. The crisis also restored the IMF to its role in enforcing policy change in clients. It is best, then, to recognize the interdependence of U.S. and IMF power and influence during and since the crisis. The crisis response by the United States and other leading countries helped to restore the IMF to a central position in global financial governance.

It is important to note in this context that there were political limits on the willingness of the United States to serve as unilateral international lender of last resort and to increase its support to the IMF. The U.S. Treasury began to distance itself from Europe in late 2009. It also refused to join others in the G-20 in a second round of IMF funding in 2012. In this context, contributions to the IMF by the BRICS and other EMDEs became more important. These shifts help to explain why demands for IMF governance reform by the BRICS became more assertive in 2012—as was apparent in the war of words waged by BRICS representatives, particularly Brazil and India (Nogueira Batista 2012).¹⁶

The Fed was not alone in undertaking bilateral swaps. The global crisis induced the rise of ad hoc “swap diplomacy.” Since 2008, more than 80 swap agreements have been signed, involving over 50 countries in every region of the world. At the start of 2015, swaps represented collectively US$1 trillion in lendable resources (McDowell 2017, 2, table 1). Swaps constituted a large, broad, ad hoc international liquidity network (ibid.). In April 2009, the Fed itself accepted swaps from the ECB, the Swiss National Bank, the Bank of England, and the Bank of Japan. The ECB also created small swaps with Poland, Denmark, Switzerland, Sweden, and Hungary. The PBOC and the Bank of Japan each expanded existing swaps to South Korea. The PBOC ultimately established swaps with more than 20 countries. The Danish and the Swedish central banks extended a small swap to Latvia in December 2008 as a bridge to its IMF loan. And the central banks of Denmark, Sweden, and Norway each extended small swaps to Iceland in May 2008 before the country received an IMF loan (Helleiner 2014b, 41–42, fn63; Henning 2016, 129).¹⁷

For advocates of continuity, the creation of an international swap network has had limited implications for the international monetary system (McDowell 2017). But this view employs a narrow metric for gauging significance. The cross-cutting swap network reflects the disorderly nature of the initiatives that emerged during the crisis. Central banks and the IMF scrambled to respond to a global crisis of uncertain proportions and borders. That central banks played a critical stabilizing role is undeniable. For now, it bears repeating that this melee provided the IMF with the means and opportunities to reestablish itself as a leading, if not the leading, institution of crisis management.

Discontinuities

The global crisis marked a sharp transformation in the IMF’s geography of influence, with the institution no longer enjoying wall-to-wall influence across EMDEs. Most of the institution’s clients during the global crisis—especially those that received its largest support packages—came from regions that it had rarely frequented in the past. The rise of relatively autonomous EMDEs, such as Brazil, China, South Korea, and India, has had dramatic impacts on the IMF’s sphere of influence. Many EMDEs enjoyed the fiscal space to provide financial support to the IMF for the first time in its history. Equally important, their behavior served as an example for less powerful countries that were able to respond to the crisis in ways that would have been inconceivable during previous crises.

New Geography of Influence and the IMF in the Troika

The IMF has provided crucial assistance to EMDEs during the global crisis via SBAs and other programs. But Europe has been the IMF’s chief preoccupation. The IMF has made loans to countries on the periphery of Europe, including Iceland, Ireland, Greece, Portugal, Cyprus, Central and Eastern European countries, and Ukraine. Iceland’s 2008 SBA was the first for an AE in decades, and in 2010 Greece became the first country in the Eurozone to sign an SBA. Greece continues to be the most challenging of the IMF’s new clients and its biggest failure (Mody 2016; Moschella 2016). IMF loans to Eurozone countries peaked in 2014 at US$92 billion (IMF 2016d). As of March 23, 2015, the IMF had arrangements with eight countries in Europe (including precautionary lines of credit, as with Poland), with commitments totaling about US$78.8 billion (IMF 2015c). By March 15, 2016, the number of loans to European countries had fallen. By then, five countries in what the IMF terms emerging Europe had outstanding loans (with commitments totaling around US$37.4 billion), and total outstanding credit to European members was around US$56 billion (IMF 2016d). Programs in Greece (2010 and 2012), Ireland (2010), and Portugal (2011) were record breaking in terms of loan size in dollars and loan size relative to GDP and quota (Reinhart and Trebesch 2016, 19, table 1).

IMF loans had been regionally concentrated during prior crises. But the extent of the concentration and involvement in Europe was unprecedented in the organization’s history. The IMF lent over US$200 billion from 2008 to 2015, of which around two-thirds went to AEs in Europe (Weisbrot 2015, 157; Reinhart and Trebesch 2016, 10). Greece, Ireland, and Portugal accounted for around 80% of total IMF lending during 2011–2014 (IEO 2016, 2). As of mid-February 2015, almost 85% of outstanding credit (plus a prospective US$10 billion disbursement to Ukraine in 2015) was extended to Greece, Ireland, Portugal, and Ukraine (Schadler 2015, 5). Programs in the European periphery place the IMF in a familiar role in a new neighborhood. The IMF is again promoting procyclical austerity policies, though mostly in high-income Eurozone countries and some middle-income countries in Eastern Europe and the former Soviet Union (Weisbrot 2015, 157).

The crisis in the European periphery brought the IMF into a partnership with two European institutions—the EC and the ECB—forming the “Troika.” The Troika was formally launched on March 25, 2010, in connection with Greece, when euro area heads of state declared “their willingness to take determined and coordinated action, if needed, to safeguard financial stability in the euro area.” They proclaimed themselves ready “to contribute to coordinated bilateral loans” as part of “a package involving substantial IMF financing and a majority of European financing” (quoted in Moschella 2016, 5). The Troika was a partnership of necessity. The IMF alone could not fill the borrowing gaps faced by European clients, and European institutions, while (at least initially) possessing sufficient financial resources for the assistance packages, needed to import the institution’s expertise, authority, and credibility (Nelson 2014a).¹⁸ The treaty establishing the European Stability Mechanism (ESM) does not legally require IMF involvement when precautionary financing is extended to member nations, though it does state that assessments “shall be conducted together with the IMF,” “wherever appropriate and possible” (quoted in Henning 2016, 124).¹⁹ As recently as winter 2016–2017, European creditor nations affirmed their strong preference to include the IMF in its programs (ibid.).

During earlier crises, the IMF worked with the G-7 and other multilateral institutions, especially the World Bank. The IMF had also previously been involved in discussions with European institutions—for example, around the formation of the European Monetary Union after the signing of the Maastricht Treaty (Moschella 2016, 5). Prior to the launch of the Troika, the IMF also worked with the EC on programs for EU members that were not part of the Eurozone, and worked with European actors in the case of (non–EU member) Iceland. But the IMF had not been involved as regularly and deeply with other institutions until the Troika. The arrangement has constrained the IMF in particular ways since “each participant was bound by the decisions of the group and the expectation that any concerns or disagreements would not be raised in public” (Bernes 2014, 12; IEO 2014, 7). As we will see, these expectations were not always fulfilled.

The IMF’s Subordinate Position within the Troika

There is substantial evidence that the IMF has occupied a subordinate position in the Troika and that the relationship between the IMF and its European partners has been fraught with tension over fundamental policy decisions. In the case of Latvia and Romania, the IMF attempted unsuccessfully to temper European demands for severe austerity (Lütz and Kranke 2014, table 3). In Estonia, Latvia, and Lithuania, the IMF initially favored external devaluation of domestic currencies, while European Troika members and Baltic governments themselves favored the painful strategy of “internal devaluation” involving downward adjustment of nominal wages and fiscal contraction (Kattel and Raudla 2013, fn13).²⁰ The IMF’s IEO also takes note of tensions with the EU in the Baltics. In the summer of 2009, IMF staff apparently had doubts about the realism of Latvia’s fiscal targets and the decision to maintain the currency peg, and mission staff reported that its negotiating position was weakened by EU action (IEO 2014, fns43, 46).

More broadly, European Troika members have generally been far more “hawkish” than the IMF on fiscal policy (Clift 2014). For example, in negotiations over Ireland’s program, the IMF favored a fiscal adjustment that was half the size demanded by the EC and ECB—ultimately the Troika demanded a fiscal adjustment that represented a compromise between their respective proposals (ibid.).²¹ Leaked reports on Cyprus reveal disagreement within the Troika. European Troika members in May 2013 pushed to impose a levy on insured and uninsured depositors in the country’s solvent and insolvent banks, over the objections of the IMF (Evripidou 2013).²²

Tensions within the Troika surfaced in negotiations around all three of Greece’s assistance packages. Conflicts between the German government and the IMF also flared on several occasions. A leaked internal (German) Bundesbank report written after Greece’s first SBA refers to the IMF as the “Inflation Maximizing Fund” (Evans-Pritchard 2010; Vary 2011). The report illustrates the depth of German fabulism on Greece and the IMF. Equally revealing is German finance minister Wolfgang Schäubel’s April 2014 remark that “Greece could serve as a model for Ukraine” (Weisbrot 2015, 2).

The IMF’s evaluation of Greece’s first SBA reveals its relatively weak position in the Troika. In 2013, the IMF issued an “Ex Post Evaluation” of the first SBA in Greece (IMF 2013a). The assessment notes that European partners ruled out debt restructuring despite the fact that the IMF flagged the country’s public debt as unsustainable (ibid., 29);²³ that Europeans demanded conditions that were too detailed and broad ranging, and were slow to make decisions; and that it was difficult to coordinate with European Troika partners (IMF 2013a; Whelan 2013).²⁴ Officials of the EU were disquieted by the IMF report, and an EU spokesperson announced that it “fundamentally disagrees” with its substance (New York Times 2013).²⁵

The first SBA with Greece was replaced by another in March 2012 for €130 billion (not all of which was disbursed). Intra-Troika tensions flared again during the negotiations that culminated in the signing of the agreement. Conflicts within the Troika in the summer of 2011 centered on bank recapitalization, debt relief and sustainability, and fiscal policy (Moschella 2016). The IMF advocated swift bank recapitalization, whereas European lenders were hostile to the idea (Eichengreen 2012, 3; Moschella 2016). Greece was ultimately permitted to restructure its debts in early 2012 after European leaders found the case for a partial write-down too obvious to ignore (Moschella 2016). Banks and investors accepted a deal that paid them about half the face value of their bond holdings.²⁶

In July 2015, conflicts between European and especially German officials and the IMF were apparent in a pitched battle over a third loan to Greece.²⁷ The IMF issued a “debt sustainability analysis” on July 2, just three days before a referendum that Greek prime minister Tsipras called to force a popular vote on the Eurogroup’s austerity package.²⁸ Senior IMF officials (including IMF managing director Christine Lagarde and then chief economist Olivier Blanchard) announced that the institution could not participate in a new Greek program in which debt was unsustainable. Greek debt was then at 175% of GDP, above the 110% of GDP that the IMF had set as a sustainable target. The analysis indicated that Europe’s officials needed to develop a plan for easing Greece’s debt burdens via restructuring or sizable debt relief, something that the German government found unacceptable. U.S. Treasury secretary Jacob Lew reinforced the IMF’s message. In commenting on the sustainability analysis, an IMF staff member said that “the EU has to understand that not everything can be decided based on their own imperatives” (Taylor 2015). Not surprisingly, Prime Minister Tsipras viewed the IMF’s intervention favorably, and cited it in a televised appeal to voters before the referendum. European members of the IMF’s Executive Board reportedly tried to stop publication of the sustainability analysis, and the EC responded to the IMF’s analysis by producing its own, less pessimistic analysis. Undaunted, the IMF’s message on debt sustainability was reinforced in a memo to EU authorities ahead of a summit on Greece. The memo stated that “Greece’s debt can now only be made sustainable through debt relief measures that go far beyond what Europe has been willing to consider” (IMF 2015b, 1). The memo highlighted other points of contention between the Fund and European Troika members, such as the feasibility of a medium-term primary budget surplus target of 3.5% of GDP and the assumption that Greece would be able to borrow at AAA rates on international markets by 2018.²⁹

In August 2015, Greece ultimately reached agreement with European officials on a third loan of €86 billion (US$94 billion). European negotiators pressed for and received from the government commitments to new, more severe austerity policies and a primary budget surplus of 3.5% of GDP by 2018. In addition, a Eurosummit statement that preceded the agreement with Greece included the provision that “the government needs to consult and agree with the Institutions [as European Troika members started to refer to themselves at that time] on all draft legislation in relevant areas with adequate time before submitting it for public consultation or to [the country’s] Parliament” (EU 2015, 5).

During 2015 and the first half of 2016, the IMF continued to flag its refusal to participate in the third loan to Greece until Europeans agreed to significant debt relief (Lagarde 2015b). Lagarde said that the budget surpluses that were a part of the August 2015 agreement would require “heroic” efforts by the Greek population, and that maintaining such surpluses well into the future is “highly unrealistic” (Donnan and Giles 2016). The IMF’s shift to persistent advocate of Greek debt relief rather than “persistent scold” surprised many observers (Thomas 2015), particularly since it had hardly played a positive role in earlier rounds of the Greek crisis (see Galbraith 2016).

In May 2016, the IMF walked back from confrontation with European officials. It signed off on a May 2016 deal to unlock $US11.5 billion in loans in a deal that commits Greece to greater austerity while putting off further discussion of IMF participation until late 2016, following a new analysis by the Fund of what the Eurogroup’s vague commitments mean for Greece’s debt (Varoufakis 2016). Prior to release of the formal analysis, IMF research director Maurice Obstfeld and director of the European Department Poul Thomsen noted on the institution’s blog in December 2016 that assessment of the European plan revealed that it would deliver a primary budget surplus of 1.5% of GDP by 2018 and that additional austerity required to achieve a 3.5% surplus would prevent “the nascent recovery from taking hold” (Obstfeld and Thomsen 2016). Notwithstanding the absence of any signs of recovery in Greece (nascent or otherwise), the blog post triggered a new round of dueling public statements. Eurozone officials expressed disappointment that IMF staff published the blog post during negotiations with Greece, and they chided the institution for its breach of confidentiality norms. They stated that they “hope that we can return to the practice of conducting program negotiations … in private” (Reuters 2016a).³⁰

The simple arithmetic of the assistance packages helps to explain the IMF’s junior position.³¹ There was a great deal of variability in the degree to which the IMF cofinanced European programs. It provided a relatively small portion of the financing toward many of the region’s SBAs. Many of these programs had a complex financing geometry involving a number of European governments and support mechanisms. For instance, Greece’s first package in 2010 involved an IMF commitment of €30 billion, in comparison with €80 billion in European assistance (though this amount was later reduced). The IMF contributed just €1 billion to a €10 billion combined financing initiative with the ESM in a May 2013 agreement with Cyprus and US$2.1 billion to Iceland’s November 2008 SBA, while Denmark, Finland, Norway, and Sweden contributed US$2.5 billion, Poland US$200 million, and the Faroe Islands US$50 million. The IMF contributed US$17 billion to Ukraine’s April 2014 program, while the United States, Europe, and World Bank contributed US$15 billion in additional lending. In May 2011, the IMF provided financing of €12.6 billion to Portugal, while the EU provided €25.2 billion in a three-year package. In the region’s most complicated cofinancing arrangement, the IMF provided €22.5 billion of the total of €67.5 billion in assistance to Ireland in December 2010.³² In the case of Latvia (2008) and Greece (2012), the IMF provided 20% or less of the total financing, whereas in Hungary (2008) and Romania (2009 and 2011) the IMF provided more than 60% of the support and the EU provided the balance (Miyoshi 2013, 19–21).

Notwithstanding intra-Troika conflicts, the IMF should not be viewed as the failed savior of Greece or other peripheral European countries. The IMF ultimately supported many European demands for greater austerity in Greece and elsewhere, and continues to do so. The IMF equivocated in its position on debt sustainability in Greece. The IMF also charges a much higher interest rate than European creditors (up to 3.9%, while Europe charges slightly above 1% on average); IMF loans have to be repaid on average in five to seven years, compared with up to fifty years for European loans; and the IMF has extracted more than €86 million annually in profits on European loans since 2013 (Gros 2016). Indeed, some observers emphasize cooperation rather than conflict within the Troika (Weisbrot 2015). A balanced assessment must conclude that there is ample evidence of both conflict and cooperation.

The IMF’s decisions and recommendations in Europe are deeply influenced by the European governments on its Board of Governors and Executive Board, especially the larger and more powerful countries among them (Lütz and Kranke 2014; Weisbrot and Jorgensen 2013). European chairs have significant influence in program approvals, not least because they hold almost a third of IMF quotas (Mohan and Kapur 2015, 51). European overrepresentation facilitated not only IMF decisions to provide large amounts of support to the region but also decisions on exceptional access and systemic risk waivers (discussed later). Hence, Vestergaard and Wade (2015, 3) could argue that overrepresentation was “invaluable in securing the IMF’s willingness to act as a junior partner of the eurozone politicians.”

Increasing Complementarity or Competition on the Horizon?

Despite massive increases in lending capacity, the IMF by itself is nevertheless inadequate to the task of responding to crises in internationally integrated, highly liberalized, liquid, and increasingly opaque financial markets. This is apparent when we compare the IMF’s still limited resources against the massive pools of capital held by money managers, and the resources in shadow financial markets, derivatives, and private equity and hedge funds, or even when we consider its resources against the collective size of the SBAs in relatively small countries on the European periphery (see Henning 2009b, 3).³³ Liquidity support from the IMF during crises will increasingly need to be augmented, and in some cases it may be supplanted by other institutions (see chapter 6 and Grabel 2013b). The IMF is now, and will continue to be, pressed into relationships of convenience during crises, in which it complements the resources of existing and new institutions, even when such arrangements dilute IMF autonomy.³⁴ The IMF’s strained partnership with the Troika illustrates the importance of thinking carefully about the nature of any partnerships between the institution and liquidity support arrangements in EMDEs, and the importance of doing so outside the context of a crisis (see chapter 6).

Institutional substitution, or forum shopping, is emerging as a live possibility for countries in distress (see chapter 6). Today the IMF faces competition from China and even the World Bank. For instance, writing early in the crisis, Wade (2010, fn10) noted that the IMF lost new business to the World Bank outside of the European rescues. Even in Europe, Turkey broke off negotiations with the Fund in early March 2010 because of the severity of its conditions. A few weeks later, the country negotiated a US$1.3 billion loan with the World Bank (ibid.).

The case of Nigeria illustrates the new opportunities for forum shopping. In February 2016, the Fund called on policymakers to remove its currency peg and capital controls, a recommendation that the country’s president rejected. Soon after, Finance Minister Kemi Adeosun announced that the country would not apply for IMF loans (Nwachukwu 2016). Instead, Adeosun announced that Nigeria would receive US$6 billion in FDI from China (Premium Times 2016), while also receiving support from the World Bank through its International Development Association (IDA), the unit at the World Bank that works with the poorest countries. In 2016, the government also applied to the World Bank and the AfDB for additional, larger loans (of US$2.5 billion and US$1 billion, respectively) (Financial Nigeria 2016).³⁵

The government of Azerbaijan has also engaged in successful forum shopping. In July 2016, the country’s finance minister highlighted both the continuing stigma on borrowing from the IMF and the high interest rates on its emergency assistance (Farchy 2016a). The government announced that it was pursuing lower-cost loans from the World Bank and possibly the Asian Development Bank (AsDB) and European Bank for Reconstruction and Development (EBRD). The same dynamic played out in a number of other country contexts during 2016. Indeed, as a consequence of increased demand, the International Bank for Reconstruction and Development (IBRD, the World Bank’s main lending arm) lent US$61 billion in fiscal year 2016 (Donnan 2016c). By comparison, the previous peak of IBRD lending during the crisis was US$44.2 billion in 2010 (ibid.).

The increase in emergency budgetary support by the World Bank not only suggests a growing trend toward forum shopping. It also suggests the messiness in global financial governance as countries turn to other multilaterals for emergency support that would traditionally have come from the IMF. By now, IMF staff are certainly aware that institutional competition or opportunities for forum shopping are broadening. A key question going forward is how this awareness will affect IMF practices.

IMF Governance

Many IMF observers, myself included, hoped that the global crisis would create space for long-awaited governance reforms that would give EMDEs more voice and influence at the institution. Governance changes at the Fund could involve formal and informal adjustments. On the formal level, voting rights and decision-making procedures at the institution can be altered. On the informal level, there is the possibility of changes that bear on the relative influence of individual members over decisions, practices, and norms at the institution even in the absence of reform of formal procedures.

Continuities

Formal IMF governance processes have long been a point of contention among EMDE member states and civil society organizations. To date, progress on even modest formal governance reforms at the BWIs has been glacial.

After nearly twelve years of pressure, the Singapore reforms of 2006 embodied largely inconsequential changes in the voice and vote of EMDEs at the IMF. Under the agreement, the voting share held by the United States fell from 17% to 16.7% and that of high-income countries fell from 52.7% to 52.3%. China secured an increase from 2.9% to 3.6% in its voting share, and the shares held by the remaining 163 of the IMF’s 185 member countries dropped from 37.1% to 36.6% (Weisbrot and Johnston 2009).

In 2007, Brazil’s then finance minister Guido Mantega expressed frustration with stubbornly slow and insignificant progress on IMF governance reform. At the IMF’s annual meeting in October 2007, Mantega maintained that absent serious governance reforms EMDEs “would go their own way.… We will seek self insurance by building up high levels of international reserves, and we will participate in regional reserve-sharing pools and regional monetary institutions. The fragmentation of the multilateral financial system, which is already emerging, will accelerate” (Mantega 2007, 3). These sentiments continued to resonate and deepen during the global crisis.³⁶

Governance Reform and the Global Crisis

IMF quota subscriptions (hereafter quotas) are the funds that each member nation hands over to the IMF. They constitute a significant part of the Fund’s capital and provide a secure permanent basis for its lending. Quotas largely determine how much a member can borrow and its share of voting rights, which translate into member countries’ voting weights on the Fund’s Board of Governors and Executive Board. Each country gets a “basic” vote, plus additional votes in proportion to its quota. The United States has always enjoyed unilateral veto power at the IMF since it holds over 17% of total quotas and over 16% of total votes. It is the only country holding veto power over important decisions, such as changes to the organization’s Articles of Agreement, which require an 85% supermajority of member votes. U.S. veto power has been termed the “most important ‘frozen asymmetry’ of the IMF governance structure” (Lesage et al. 2013, 20). Given the voting shares held by the United States, it is not surprising that the institution “reflects US economic policy preferences more faithfully than perhaps any other international organization” (Henning 2009b, 1).

In late 2008, new quota negotiations began at the BWIs and in five regional development banks. In mid-March 2009 (in advance of the first BRICs Leaders’ Summit in June 2009), the BRICs issued a communiqué that explicitly linked quota reform to an increase in their contributions to the BWIs (Armijo and Roberts 2014, 516). The first EMDE funding commitments to the IMF in April 2009 were not conditioned on implementation of specific governance reforms. Nevertheless, the institution’s new funders, particularly Brazil and China, used the moment to emphasize the need to reallocate IMF quotas (and hence voting shares) to reflect more accurately the economic significance of large EMDEs in the global economy (Woods 2010). During the April 2009 discussions, senior Chinese officials announced that Beijing would be willing to increase its contributions if its quota reflected its economic weight; namely, by basing its quota on output per person (Landler 2009). In the fall of 2009, the BRICs countries and the Group of 24 (G-24) proposed a 7% increase in quotas in favor of EMDEs (Wroughton 2009; Reuters 2009a).³⁷ China proposed a far more ambitious goal: in September 2009, a finance ministry official called for establishment of a timeline to transfer 50% of the voting rights at the BWIs to EMDEs (Reuters 2009a; 2009b).³⁸

At the time of the negotiations, the BRICs did not consistently present a united front on governance reform. The Russian government was particularly ambivalent. On various occasions, it aligned itself with critics of IMF governance, and it was among the most outspoken critics of the “exorbitant privilege” enjoyed by the United States (Johnson 2008). But Russian officials occasionally distanced themselves from positions taken by fellow BRICs countries. For example, during the fall 2009 meeting of the IMF and World Bank in Istanbul, a Russian central bank official emphasized that its purchase of IMF bonds was not conditioned on IMF governance reform (Reuters 2009b). Russia’s ambivalence exemplifies what Jones (2011) notes is the fluid, uncertain relationship among the BRICs and between them and other EMDEs (see Ban and Blyth 2013, fns1, 2; Mittelman 2013).

In 2009, the G-20 announced with much fanfare that it had moved forward on its commitment to reform formal governance at the BWIs. Representatives agreed to redistribute quota shares by at least 5% from “over-represented” members to what it termed “dynamic” EMDEs at the IMF, and by at least 3% at the World Bank by January 2011 (G-20 2009b). The language of the agreement was opaque, making it difficult to determine exactly which countries would gain votes and which would lose them. The lack of linguistic clarity reflected conflict within the G-20 (Wade 2011, 363). The G-20 also agreed in 2009 to reallocate some seats on the IMF’s Executive Board to EMDEs by consolidating some European seats, and to increase the IMF’s general quota so as to triple its lending capacity. The IMF’s International Monetary and Financial Committee (IMFC) endorsed the G-20 agreement the following month (Lesage et al. 2013), continuing the vague language on quota redistribution. Negotiations on the governance reform components of the agreement then stalled for a year owing to conflicts between the United States and Europe and among European countries.

In October 2010, the Finance G-20 regained momentum on IMF governance reform when its members agreed to a number of proposals, including a shift of 6.2% in quota shares to dynamic EMDEs by October 2012 (G-20 2010). In November 2010, the IMF’s Executive Board approved this measure, which was included in a broader package known as “The 2010 IMF Quota and Governance Reform Agreement” (hereafter the 2010 agreement). Specifically, the 2010 agreement involved several measures: a vast increase in the lending power of the institution by doubling IMF quota resources; an amendment of the IMF’s Articles of Agreement to provide for an all-elected Executive Board;³⁹ and a shift of 6% in quota and voting shares away from AEs (principally in Europe) toward rapidly growing EMDEs, with an understanding that, in order to increase the representation of EMDEs, “advanced European” countries would reduce their representation on the 24-person Executive Board by two seats via a complicated arrangement involving sharing and rotation of seats and chairs within groups (Truman 2015; IMF 2010f; Lesage et al. 2013, 560–561; Wade 2011, 364–365).⁴⁰ Under the agreement, China would become the IMF’s third-largest member country behind the United States and Japan. Moreover, the top ten shareholders at the Fund would henceforth represent the ten largest economies in the world, which now include China, Brazil, India, and the Russian Federation (Eichengreen and Woods 2016; IMF 2016b). (We will see that the headline figure of 6% overstates the degree of change.)

The IMF’s Board of Governors ratified the 2010 agreement in November of that year. It was to be implemented in October 2012, following ratification by member countries. The majority of members approved the agreement. But approval by an 85% supermajority was required. Implementation of the agreement then stalled for five years because the U.S. Congress failed to approve it in late 2012 and again in January 2014, even though the government’s executive branch had done so and despite repeated exhortations by the IMF.⁴¹

In the face of gridlock, the IMF was unable to secure the anticipated increase in its secure quota-based lending resources. It remained dependent on the willingness of its members to lend to it on a temporary bilateral basis through the NAB. The IMF was forced to rely on temporary borrowing for 70% of its credit capacity, with only 30% of lending capacity derived from stable country quotas (IEO 2014, 25). This left it vulnerable to sudden demands on its resources. The degree of dependence on temporary borrowed loans had no historical precedent: prior to the global crisis, member quotas provided more than 80% of the IMF’s lending capacity (Mohan and Kapur 2015, 47; IEO 2014, 25).

Congress finally ratified the 2010 agreement in mid-December 2015, as part of an “Omnibus” spending bill. Congressional Republicans were ultimately won over because the language of the bill required the United States to take all necessary steps to repeal a systemic risk exemption that the IMF had introduced in 2010 (to be discussed) and that the U.S. executive director at the IMF vote to repeal this exemption (U.S. Congress 2015, section 9004-5). The Treasury Department agreed to push to repeal the exemption to appease congressional Republicans (Calmes 2016).

Strauss-Kahn wrapped the 2010 agreement in hyperbole, calling it “historic” and calling the “landmark reforms” “the most fundamental governance overhaul in the IMF’s 65-year history and the biggest ever shift of influence in favor of EMDEs to recognize their growing role in the global economy” (IMF 2010f). In the view of critics, what Strauss-Kahn hailed as landmark change was better described as “microscopic” and even “misleading” (Vestergaard and Wade 2014; 2015). The reforms shifted just 2.6% of voting shares from AEs to EMDEs, with the rest of the increase in dynamic EMDE shares coming from what were deemed to be overrepresented EMDEs. The G-7 countries as a group sacrificed just 1.8% of their aggregate voting power (Vestergaard and Wade 2013b). The 2010 agreement made the shift in voting shares to EMDEs appear more significant by virtue of a new classification system introduced at the time, under which South Korea and Singapore were reclassified as EMDEs even though they continued to be grouped with AEs in the IMF’s flagship publication, World Economic Outlook (Lesage et al. 2013). Civil society groups noted that the agreement did little to address Africa’s effective disenfranchisement at the Fund (BWP 2010b). The agreement also maintained the unilateral veto power of the United States since its voting share fell (trivially) from 16.7% to 16.5%, whereas BRICs countries as a bloc hold 13.51% of voting shares and are therefore deprived of veto power (Lesage et al. 2013).⁴²

Why Did the BRICs Governments Approve the 2010 Reforms?

It may seem puzzling that BRICs country governments ratified the 2010 agreement, given its serious shortcomings and the financial contributions to which their governments committed themselves.⁴³ After all, at the beginning of 2009, the BRICs together held more than ten times the amount of currency reserves that the IMF had at its disposal. Despite these holdings, the group was in a relatively weak and sometimes divided position at the IMF. The bloc was not in a position in the short run to force governance changes that were more significant than those embodied in the 2010 agreement.

Approval of the 2010 agreement by the BRICs reflects several factors. Its members preferred to minimize default risk and to outsource conditionality by contributing to the IMF rather than engaging in bilateral assistance. Notwithstanding solidaristic rhetoric, BRICs leaders privileged increasing their own influence at the Fund over the broader matter of increasing EMDE influence. The incremental nature of the 2010 agreement was seen by BRICs governments to open the door for increasing their influence at the Fund over time. The BRICs also saw it as being in their interest to contribute to the IMF to reduce the likelihood that the United States would unilaterally inject massive amounts of liquidity into the global financial system, thereby undermining the value of the dollar-denominated investments that they held.

The “Gentleman’s Agreement” on IMF (and World Bank) Leadership

Early in the crisis, critics raised another aspect of governance reform that centered on the process by which leaders of the BWIs are selected. EMDEs had long chafed at the Bretton Woods–era gentleman’s agreement that has ensured that the IMF’s managing director is a European and the World Bank’s president an American. G-20 finance ministers agreed to end this practice at their March 2009 meeting in Sussex, England (Eichengreen 2009). In reality, the practice continued.

In April 2011, BRICS⁴⁴ representatives announced that they sought an end to the gentleman’s agreement (Armijo and Roberts 2014, 514–515). It appeared that the first chance to act on this would be at the conclusion of Robert Zoellick’s term as World Bank president in June 2012 (ibid.). But the arrest and resignation of Strauss-Kahn in May 2011 provided an earlier opportunity. IMF directors representing the BRICS quickly issued a statement calling for an open, merit-based selection process (ibid.). Some—though by no means all—EMDE leaders and representatives at the IMF used the opening in 2011 to press publicly for Mexico’s Agustín Carstens. Lagarde and, very briefly, Stanley Fischer also emerged as candidates. Russia briefly promoted the candidacy of Kazakhstan’s Grigori Marchenko, but quickly endorsed Lagarde (Armijo and Roberts 2014, 515). Old appointment patterns prevailed in July 2011 when the IMF’s Executive Board, using the usual weighted voting that reflects country quotas, replaced Strauss-Kahn with Lagarde, a fellow European. Supporters of Lagarde’s appointment justified the decision on grounds that a European was needed at the helm because of the Eurozone crisis. Critics rightly identified the hypocrisy of this view insofar as Europeans led the IMF through decades of crises in EMDEs. Lagarde was quietly reappointed to a second IMF term in February 2016, when she was the sole candidate for the position.

The commitment to the gentleman’s agreement on leadership was also reaffirmed at the World Bank. In April 2012, it was announced that an American, Dr. Jim Yong Kim, would serve as World Bank president beginning in July 2012. Two highly qualified candidates had emerged early on as leading candidates for the World Bank post. Nigeria’s former finance minister (and former World Bank managing director), Dr. Ngozi Okonjo-Iweala, was nominated for the post by South Africa. Colombia’s former finance minister and former UN undersecretary-general for economic and social affairs, Dr. José Antonio Ocampo, was nominated by Brazil. World Bank governors in Europe interviewed candidates Kim, Okonjo-Iweala, and Ocampo and reportedly deemed the latter the best prepared, possessing the clearest ideas about where to take the Bank, and the most knowledgeable on economics (Wade 2012). Ocampo withdrew on April 13, 2012, to make space for Okonjo-Iweala’s candidacy once it became clear that it was politically important that the World Bank’s executive directors representing EMDEs—a group known informally as the G-11 inside the Bank—present a unanimous vote for one candidate (Wade 2012; Wroughton 2012b). Okonjo-Iwela withdrew shortly thereafter, once other EMDEs threw their support behind Kim and when it became clear that politics and bilateral relations with the United States rather than merit were driving the process.⁴⁵

The fluid and sometimes conflictual relationships within the BRICS and between the BRICS and other large EMDEs created space for Kim’s appointment.⁴⁶ The BRICS and other EMDEs did not offer unified support for either Ocampo or Okonjo-Iweala. Kim’s candidacy was supported by Mexico and Russia; Ocampo by Brazil and not by the government of his country of origin, Colombia; and Okonjo-Iweala by South Africa. In addition, Kim’s appointment was facilitated by the ability of the United States to exploit these fractures by offering leadership carrots to EMDEs, such as when Jin-Yong Cai, a Chinese national, was appointed chief executive officer of the International Finance Corporation (the private sector lending arm of the World Bank Group) in August 2012 and Kaushik Basu, an Indian national, was appointed chief economist at the World Bank in September 2012 (see Wade 2012).⁴⁷ Following the pattern with Lagarde, Kim was the only candidate put forward at the end of his first term, and he was unanimously reappointed in July 2016.

Lack of Independence from Key Member States

The circumstances surrounding the IMF’s decision to grant exceptional access to Greece (and others in the Eurozone) and the introduction of the systemic risk waiver provide a window into the overrepresentation of key member states in IMF decision making.

In 2010, it was clear that Greece would either need to default on its debt, which largely comprised bonds held by private investors and banks in France, Germany, and Britain, or would need exceptional access to IMF loans beyond its quota to support debt service (Ellmers 2016, 11). French and German banks blocked debt restructuring (Blustein 2015, 11). U.S. banks were indirectly exposed to Greek debt through credit default swaps that insured European bank loans (Ellmers 2016). With restructuring off the table, the IMF granted exceptional access to Greece in 2010. The IMF’s exceptional access policy allows nations to access funds provided that four criteria are met—namely, that there is a high probability that public debt is sustainable in the medium term; the country faces exceptional balance of payments pressures; it has good prospects for regaining market access; and it has a policy program likely to succeed. The Fund acknowledged that it violated the first of these criteria when lending to Greece in 2010. In order to approve the 2010 loan, the IMF introduced the systemic risk waiver. The waiver provides exceptional access to borrowers when there is “high risk of international systemic spillover effects,” even if public debt is deemed unsustainable in the medium term (IMF 2010c, 20). The waiver did not require an upfront debt restructuring. Waivers were also invoked to justify IMF loans to Ireland and Portugal.

The decision to create the systemic risk waiver tarnished the IMF’s credibility (Schadler 2014; Robertson 2015). As the IMF itself later acknowledged, the waiver aggravated long-standing internal concerns about governance. EMDE and other representatives criticized the exceptional treatment of Europe as another instance of the IMF’s lack of evenhandedness, the imprint of Europe’s overrepresentation on the Executive Board and managing directorship, and the specter of U.S. influence (IEO 2013, 25–26, 29; 2014, 7, 25; Blustein 2015, 2; Ellmers 2016). Representatives from EMDEs questioned whether the exceptional access granted to Europe would be available to them in future crises (IEO 2013, 25; 2014, 7, 25). The waiver was also controversial among the IMF’s rank and file (Ellmers 2016). In January 2016, the IMF dropped the systemic risk waiver following horse trading in the U.S. Congress in late 2015 around the 2010 reform agreement.⁴⁸

The IMF’s lack of independence from European members is also apparent in the decision to support so many European rescues, to join Greek programs absent debt write-downs or extensions of payment periods, and to serve as a junior partner to European institutions. Mody puts the matter bluntly: “The IMF’s subordination to its major shareholders became painfully clear when it parachuted into the euro crisis” (Mody 2016). In a similar vein, Rogoff argues that the IMF “sycophantically supported each new European initiative to rescue the over-indebted eurozone periphery, committing more than $100 billion … risking not only its members’ money but ultimately its … credibility” (Rogoff 2011). Mody (2016) also sees the IMF’s shift to a fiscal consolidation message in 2010 as consistent with (and reflective of) the sentiments of the United States and powerful European members.

The contents of IMF Article IV surveillance reports also reflect the substantial influence of key states.⁴⁹ They are often given a pass on macroeconomic conditions that induce a stern warning for less powerful states (IEO 2011, 25, 44; 2013, 25–26, 29). Powerful members exercise influence not just over what is in surveillance reports but also over what is removed. In a study of Article IV and other reports from 2012 to 2014, Eichengreen and Woods (2016, table 1) find that “deletion rates” are generally higher for advanced and emerging economies than for developing countries (see also IEO 2013, 26).

Discontinuities

The preceding discussion of IMF governance offers substantial evidence of continuity. But concurrent and continuing changes at the IMF (and World Bank) are beginning to reconfigure channels of informal and even formal influence of member states. These changes will no doubt accelerate in the coming years. One channel of evolution stems from the new roles that leading EMDEs are assuming at the institution.

New Lenders, Rising Powers, New Influences?

During the crisis, China was initially reluctant to contribute additional funds to the IMF. At the November 2008 G-20 meeting, the country’s representatives argued that it had already fulfilled its quota to the institution. It emphasized that its voting power should increase if other countries wished it to contribute more. It also insisted that were additional funding to be forthcoming, it would be through modalities that minimized its risk (Jiang 2014, 172–173). As the crisis worsened, however, China saw fit to reverse its position.

Some of the Fund’s former clients took on a new role at the April 2009 G-20 meeting that gave the IMF its first tranche of funding related to the global crisis. For the first time in IMF history, the institution issued its own bonds. This provided the vehicle for unprecedented financial support from EMDEs. The bonds were SDR-denominated, tradable in the public sector, and limited to a one-year maturity—conditions demanded by new EMDE lenders that were willing to commit funds only through instruments that would maintain pressure for governance reform while also offering them a safe alternative to bilateral support to distressed economies (Holroyd and Momani 2012, 208; Jiang 2014, 174).

Negotiations over new financial commitments and the means through which they were to be committed to the IMF were contentious. China was among the first at the 2009 G-20 meeting to commit new resources to the IMF (Chin 2014b, 194). In doing so, the country’s policymakers broke new ground in emphasizing that their contributions would help to stabilize the global economy and protect poorer countries from the fallout of the crisis (ibid.). In Chin’s view, equally notable is the fact that the country worked closely with the BRICs (and later the BRICS) and other East Asian governments to press for international monetary and IMF governance reform (ibid.).⁵⁰ China worked closely with Brazil and Russia to insist that new contributions would take place through purchase of the new SDR-denominated bonds and not through the NAB, since contributions through this channel do not translate into greater voting shares (ibid., 195). Indeed, all of the BRICs countries and South Korea made their contributions through the IMF’s SDR bonds rather than the NAB, though most other countries committed new resources to the IMF through the NAB.⁵¹

China committed to purchasing US$50 billion of the new bonds, while Brazil, Russia, South Korea, and India each committed to purchasing US$10 billion. Thus, US$90 billion in new resources for IMF lending in 2009 came from EMDEs, some of which were past clients. The historical significance of this decision was not lost on Brazil’s then president, Lula da Silva, who quipped, “Brazil is now a creditor of the IMF!” (quoted in Armijo, Katada, and Roberts 2015, 37).

In late 2011, Lagarde called on EMDEs for a second tranche of commitments while the IMF also sought new resources from AEs. Among the former, Brazil’s government was initially most receptive. However, Brazilian president Dilma Rousseff refused to announce the precise nature and dollar amount of the country’s second contribution until she was apprised of plans for IMF governance reform and until Eurozone leaders articulated a credible rescue plan. She also made it clear that the country’s decision on new funding would be tied to a later meeting among BRICS leaders, an announcement that signaled solidarity among the often fractious group. Never one to miss a chance to note historical ironies, Mantega stated during Lagarde’s visit to the country, “It’s a great satisfaction to us that this time the IMF did not come to Brazil to bring money like in the past but to ask us to lend money to developed nations” (quoted in Leahy 2011).

At the same time as Lagarde sought a second round of support from the G-20 in 2011, European policymakers appealed directly to the BRICS and other large EMDEs for contributions to the European Financial Stability Facility (EFSF). China, in particular, was targeted for support. The country bought EFSF bonds after French president Nicholas Sarkozy made an emergency call to China’s president, Hu Jintao, in October 2011 (Jiang 2014, 167). About 25% of China’s currency reserves might have been used to purchase European sovereign debt, which China viewed as advancing its commercial and political interests within the EU (ibid., 166–167). The other BRICS declined European requests for bilateral support. In September 2011, Mantega announced a BRICS bailout fund for the Eurozone but then dropped the initiative when China did not support it (Armijo, Katada, and Roberts 2015, 37).

Lagarde continued to seek additional IMF funding in 2012. G-20 finance ministers announced new commitments at their April 2012 meeting. As expected, the United States was unwilling to contribute. The specific contributions to be shouldered by EMDE members were not announced until BRICS representatives met informally on the eve of the June 2012 G-20 Leaders’ Summit in Los Cabos (Chowla 2012). The BRICS pledged US$75 billion of the US$456 billion in new IMF funding committed by G-20 countries. China’s share was US$43 billion, Brazil, Russia, and India each committed US$10 billion, and South Africa pledged US$2 billion.⁵² The new BRICS pledges were made via temporary, bilateral standby credit lines under the NAB to be activated as needed, with the understanding that they were to be drawn on only after the IMF’s existing resources had been substantially utilized (Wroughton 2012a). BRICS governments explicitly conditioned release of this second tranche of funds upon implementation of the stalled 2010 agreement (ibid.).⁵³ Mantega clearly articulated the quid pro quo involved in the BRICS position when he stated that the promise of additional funding was tied to “an understanding that the reforms of the Fund’s quotas … will be implemented according to the timetable agreed by the G20 in 2010” (quoted in Giles 2012). China’s President Hu also linked the announcement of new funding to implementation of the 2010 agreement (Jiang 2014, 174). Even after this second recapitalization by the BRICS, Lagarde continued to seek support from EMDEs. During a visit to Colombia in December 2012, she noted “that (the country) is in a situation where it can offer support” (quoted in Stringer 2012).

What should we make of the BRICS and their potential to exercise their voice at the BWIs? Armijo and Roberts (2014, 519–520) put the matter well: “Instead of dismissing the BRICS as trivial for their limited achievements thus far, it makes more sense to conceptualize them as in the process of building capacity, adjusting to China’s looming presence within the club itself, and working through common positions …; the BRICS organization itself functions most readily as an ‘outside option’ for China to employ to exercise leverage within the major existing global governance institutions …; BRICS’ preferences, singly and jointly, for global governance turn on reform and evolution, not revolution.”

What is most important about the commitments to the IMF by the BRICS and EMDE members of the G-20 is that they reflect the emergent power and autonomy of countries that weathered the global crisis relatively successfully. Long-standing frustrations with the BWIs and gridlock around the 2010 agreement contributed to the decision by the BRICS in July 2014 to make good on their promise-cum-threat to launch a new development bank and reserve pooling arrangement (Giles 2012).⁵⁴ The same frustrations contributed to new China-led initiatives as well (see chapter 6). There is every reason to expect that in the coming years continued frustrations with the IMF by increasingly efficacious actors will generate additional pressure on the institution to reform its formal internal governance procedures. Absent such reforms, we should expect increased forum shopping by a larger set of EMDEs.

Informal Channels of Influence

An examination of formal governance procedures—whether it concerns voting shares or leadership selection processes—does not address the fact that institutional decisions and practices are also influenced by myriad informal processes. As with any extensive, complex organization, the IMF comprises diverse constituencies that often disagree among themselves about fundamental matters pertaining to the institution’s mission and strategies. Scholars whose research looks inside international organizations have documented the myriad fissures within these organizations, the diverse channels of formal and informal influence within them, and the complicated processes by which meaningful change within them may be facilitated or stymied.

There is by now an extensive, insightful body of literature on the various means by which countries seek to achieve greater influence in IMF affairs than is afforded by their formal voting rights. At the same time, internal staff members can sometimes exert greater influence than is codified within formal IMF procedures.

Nelson (2014a; 2017) finds that formal governance mechanisms at the IMF, including voting by the IMF’s Executive Board, can be less important to the institution’s activities than some outsiders recognize. On paper, many important decisions require a supermajority, which gives the United States effective veto power. But Nelson notes that Executive Board voting on lending programs is informal and is recorded on an up or down basis, with the board almost always unanimously approving staff proposals. In practice, the procedures provide space for internal and external actors to influence the terms of IMF agreements through informal back channels. Moreover, even when powerful countries use formal channels of influence at the IMF, staff may well reach decisions and take actions that stray from positions that key states communicate through formal voting mechanisms. For example, IMF staff extended a loan to Sri Lanka in July 2009 despite U.S. and U.K. abstentions from an Executive Board vote, by which they conveyed reservations about the loan (Nelson 2014a, 167).⁵⁵

While the EMDE loans to the IMF and increasing EMDE assertiveness are certainly contributing to a gradual process that can be expected to result in significant changes in IMF practices over time, we must keep in mind that it is not generally possible to overhaul both formal and informal practices all at once. Stickiness in either domain may act as a temporary brake on institutional reforms that would seem to be warranted. Alternatively, dynamism in one domain might allow actors to circumvent enduring roadblocks in the other domain. Adjustments in voting shares, then, might be less important than many observers appreciate (as per Nelson 2017). Stagnation in formal governance processes may be overcome by new informal institutional norms and practices that result from changes in funding sources and instruments, personnel, ideologies, recognition of complementarities and/or competitive threats from other institutions, and the changing economic fortunes of diverse members. In short, we should expect evolution in IMF governance to be messy and uneven, with at least some consequential changes occurring out of the public eye.

From this perspective, the support for the Fund coming from EMDEs represents a landmark event for the institution. It is implausible to assume that these changes in financial flows between member nations and the IMF will not ultimately bear on the ability of EMDEs to exert influence at the institution in ways that decisively alter its informal and even its formal practices. But it is easy to be skeptical about the influence of new lenders on the IMF. It took five years to implement the modest 2010 agreement, which was the most high-profile objective that they have sought. But the skeptical view fails to recognize that what we might call offensive influence, securing change that is detrimental to the interests of one’s opponents, represents the highest level of influence. Lower down on the ladder is defensive influence, or the ability to block, derail, or veto initiatives by other parties that one views as detrimental to oneself. If we are interested in searching for evidence of changes in informal governance processes at the IMF, then we might look first for changes in EMDEs’ defensive capacity at the IMF.

China’s visibility on the global stage and its demonstrated commitment to creating new financial institutions exemplifies the potential that new lenders have to influence informal decision making at the IMF, even only gradually and primarily via defensive influence. The visibility and defensive assertiveness of the Chinese government is reflected in many other ways. For example, the country’s policymakers have always pushed back aggressively when the U.S. Treasury has branded it a “currency manipulator.”⁵⁶ China has always resisted U.S. efforts to enlist the IMF in containing or criticizing it.⁵⁷ For instance, the IMF announced in August 2007 that it would monitor global imbalances as per a plan advanced by the United States. Identification of imbalances would trigger an automatic audit and IMF staff recommendations to correct the problem. In response, China suspended Article IV consultations and presented a competing analysis that centered on the contribution of AEs, especially the United States, to persistent trade imbalances. China found support for this position among its BRICs partners in early 2009 as they were preparing for their first Leaders’ Summit. In March 2009, BRICs finance ministers signed a formal statement that called on the IMF to extend surveillance to AEs. In June of that year, the IMF’s new Article IV guidelines abandoned the language on exchange rate levels to which China had objected.

BRICS leaders pushed back a second time in April 2011, when the IMF again took up the U.S.-driven idea of providing data on trade imbalances. The communiqué of the fourth BRICS Leaders’ Summit, in late March 2012, argued that it was “critical for AEs to adopt responsible macroeconomic and financial policies [and] avoid creating excess global liquidity … [while also noting that the BRICS were] concerned about the slow pace of quota and governance reforms in the IMF … [and calling] upon the IMF to make its surveillance framework more integrated and evenhanded” (BRICS 2012). These examples demonstrate the ways in which the BRICS, as per Armijo and Roberts (2014, 520), can amplify China’s leverage and deflect criticisms advanced through the BWIs, G-20, and other arenas of global governance.

China’s assertiveness is also reflected in the government’s indictment of the role of the United States in the global crisis, its questioning of the continued viability of a dollar-denominated international monetary system, and the downgrade of U.S. government debt by China’s credit rating agency in August 2011 and October 2013. The appointment in July 2011 of Min Zhu, former deputy governor of the PBOC, as deputy managing director of the IMF provides a channel by which China has already enhanced its influence over informal decision making at the IMF. It remains to be seen how China’s growing influence will affect the voice of other EMDEs. Equally important, Zhu’s succession by Tao Zhang (in August 2016), also a former deputy governor of China’s PBOC, might indicate the emergence of a new norm in which the position is reserved for a Chinese official. In addition, in March 2012, Jianhai Lin was appointed secretary of the IMF and of its IMFC, and in 2008 “Justin” Yifu Lin was appointed chief economist and senior vice president of the World Bank. As Henning and Walter (2016, 9) argue, China’s representation among senior staff at the BWIs demonstrates a substantial increase in its informal influence at the BWIs since 2008 (despite continued setbacks in formal governance reform). In November 2015, China also achieved a long-sought (though largely symbolic) goal of having the IMF agree to include its currency in the SDR basket alongside other currencies that the institution had long designated as having “global reserve currency” status; namely, the dollar, euro, yen, and British pound. The decision was operationalized in October 2016.

The now regular (once to twice a month) meetings of the BRICS executive directors at the BWIs have created an informal network of influence that facilitates broader changes at these institutions, notwithstanding tensions within the group (Wade 2011). These and other emergent networks among EMDEs are now laying the groundwork for more significant changes at the BWIs in the future (as per Woods and Martinez-Diaz 2009).

These developments are occurring at a time when the Trump administration in the United States is sending contradictory messages about its role in global institutions. Trump appointees at the Treasury Department are at once indicating that they intend to press to reduce the reach of the IMF in Europe and beyond and to revive failed U.S. efforts to involve the IMF in exchange rate surveillance (Donnan 2017; Mnuchin 2017). At the same time, however, they intend to reduce U.S. funding commitments to the IMF and World Bank, without recognizing that decreased funding will almost certainly reduce U.S. leverage over these institutions (Donnan 2017; Economist 2017b). China appears to be seizing the resulting opportunities for increased influence within the IMF (Economist 2017b).

The IMF discovered new vitality as first responder to economic distress at the same time as it has faced a diminished terrain over which it can dictate economic policy. Equally important, the institution became dependent on raising new resources from vibrant EMDEs while facing a closed door on contributions by the United States in 2012. Even if this new role for EMDEs does not translate into meaningful formal changes in the institution’s governance in the near term, it cannot be dismissed as inconsequential since it reflects broader changes in economic power that are expressed within the IMF in other ways. It is too early to gauge the extent to which the EMDEs are ultimately able to use the emerging networks among their representatives to enhance their formal and informal influence at the Fund. It is also too early to know the degree to which the IMF must change course more than it has to date to maintain its absolute and relative influence in relation to competing institutions and increasingly autonomous EMDEs. But the fact that these are now pertinent questions, even with uncertain answers, indicates the degree to which the IMF confronts a dramatically altered landscape that poses significant challenges to its internal governance and external influence.

IMF Practice

We now consider the IMF’s conduct during the crisis. We find evidence of both important continuity with and significant departures from past practice. The emphasis on practice and rhetoric allows us to identify important ambiguities and inconsistencies and, especially, notable gaps between the institution’s public statements, research, and operations.

Continuity

As the global crisis emerged, the IMF’s leadership initially seemed to renounce its traditional embrace of procyclical (austerity) policy responses. In 2008 and 2009, key officials publicly advocated the use of countercyclical macroeconomic policies (involving lower interest rates and increased government spending) to boost growth. Strauss-Kahn and Blanchard championed the dramatic expansion in government expenditures, in November 2008 going so far as to propose to the G-20 that AEs and EMDEs with fiscal space should participate in a global fiscal stimulus equal to 2% of world GDP (IEO 2014, 9; IMF 2008). In May 2009, Strauss-Kahn argued that “fiscal policies should counteract the crisis, not make it worse.… Many African countries have room on this front” (quoted in Reichmann 2013, 37; IMF 2009b). Farrell and Quiggin (2017, 16–17) note that what was remarkable was not so much Strauss-Kahn’s proposal as the nearly complete “absence of dissent within the IMF, an institution which had until recently been associated with very different economic ideas.” The IMF’s rhetoric of countercyclicality tracked and reinforced the same powerful Keynesian resurrection in which the G-20 and many prominent and formerly anti-Keynesian economists suddenly found themselves (see Farrell and Quiggin 2017; Skidelsky 2010; 2011).

The early rhetorical embrace of countercyclical policy reflected the sensibilities of key IMF officials and foxhole Keynesianism. But it also served instrumental purposes. Strauss-Kahn and others saw in the early moments of the crisis not just the need but also an opportunity to regain the resources and standing that the institution had previously lost. The IMF’s early embrace of Keynesian responses also distracted attention from the degree to which it shared responsibility for the crisis, not least by having reified light touch regulation for several decades and failing to flag fragilities that culminated in crisis.⁵⁸

Austerity and Politically Intrusive Adjustment in Support Packages

Rhetoric and traditional IMF practice were realigned in June 2010 when the Fund began to advise major AEs to pursue “fiscal consolidation” while maintaining monetary expansion if needed.⁵⁹ The shift mirrored the pivot back to orthodox ways of thinking about fiscal policy on the part of most G-20 leaders and many prominent economists. Strauss-Kahn returned to the script of his predecessor, Michel Camdessus, when he said in June 2010 that he was “totally comfortable” with deficit cuts “even if it has some bad effect on growth” (BWP 2010a).

In reality, the conditionality programs advanced during and since the global crisis are in important respects similar in scope and content to those advanced during the Asian, Mexican, and other EMDE crises (see Grabel 2011). Nelson (2014a, 162–163) puts the matter well, arguing that “changes in the design of lending arrangements have been subtle (some might say nearly imperceptible) … [and that] the crisis of 2008 was not a breaking point in either the scope or content of conditionality” despite the IMF’s public advocacy or use of countercyclical policy. Continuity should not be surprising since, in the view of many Fund staff, “many of its patients were suffering from the same disease” (ibid., 163), a disease that it had long treated in other patients.

Support Early in the Crisis

There is by now a vast amount of literature that shows that IMF assistance programs negotiated early in the global crisis featured its traditional austerity prescription and the associated portfolio of macroeconomic policy adjustments. IMF programs in Iceland, Latvia, Hungary, Romania, Greece, Portugal, Pakistan, Ukraine, and El Salvador included limits on or large cuts in fiscal outlays and tax increases (Grabel 2011, 821–822). A review by UNICEF (2010) of IMF programs in 86 countries found that the Fund advised two-thirds of their governments to reduce total public expenditures in 2010 and to further their fiscal contraction in 2011. Weisbrot et al. (2009) studied 41 countries with IMF agreements and found that 31 of these agreements involved tightening fiscal or monetary policy, or both. A study by Islam et al. (2012) of Article IV consultations from 2009 to 2010 in 30 low-income countries and 20 middle-income countries found that fiscal adjustment was a feature in 48 of the 50 cases.⁶⁰

Assistance packages involving the Troika early in the crisis parallel IMF programs elsewhere. SBAs in Hungary in November 2008, Romania in April 2009, and Latvia in December 2008 looked a great deal like earlier programs in EMDEs (Gabor 2010, 822). An examination by Weisbrot and Jorgensen (2013) found that the 67 Article IV consultation reports prepared for 27 countries in the EU from 2008 to 2011 emphasize fiscal consolidation, reductions in social expenditures, measures that weaken the bargaining power and income of labor, and measures that make it more difficult for governments to promote growth and employment or reduce poverty and social exclusion. They also observe that a number of Article IV reports refer to the Eurozone crisis as an opportunity to make changes that would in other circumstances be politically difficult (see also Weisbrot 2015, chap. 1, especially 46–47).⁶¹ The Asian crisis had previously provided similar opportunities for the IMF, the U.S. government and U.S. firms, and domestic liberalizing forces, such as opponents of Korean chaebols (see chapter 3 and also Weisbrot 2015, 132–134; Crotty and Lee 2002; 2009).

Support Later in the Crisis

One might argue that the sudden onset of the crisis left the IMF with little choice but to dust off the old playbook. Continuity might therefore have been expected early in the crisis. But IMF programs later in the crisis are also marked by continuity. In fact, the number of conditions per loan has increased in recent years, with widespread use of expansive, politically sensitive conditionality involving reductions in welfare programs (Griffiths and Todoulous 2014). A vast review of IMF “Country Reports” for 174 countries from January 2010 to February 2013 by Ortiz and Cummins (2013) concludes that the IMF has contributed significantly to decisions to cut fiscal expenditures in EMDEs.⁶² In their comprehensive examination of IMF conditionality in loan agreements involving 131 countries from 1985 to 2014, Kentikelenis, Stubbs, and King (2016) find unambiguous evidence of continuity in IMF conditionality across much of this period. They conclude that, despite its claims to the contrary, the IMF continues to advocate labor market liberalization and reductions in public sector employment and government wage spending. On this basis, the authors argue that the IMF has an “escalating commitment to hypocrisy.”

Other research reaches similar conclusions. Ban (2013) found that in 2013 the IMF continued to demand fiscal constraint in Bulgaria, Romania, Hungary, Portugal, Lithuania, and Estonia. A September 2013 loan of US$6.7 billion to Pakistan includes provisions for privatization and reductions in the fiscal deficit, principally through extending taxation. Hanieh (2015) and Mossallem (2015) find strong evidence of continuity in loans to several countries in the Middle East and North Africa despite a shift in IMF rhetoric toward social inclusion after the 2011 Arab Spring uprising.⁶³ Loans extended in 2016 continue the pattern. A US$1.5 billion IMF loan to Sri Lanka in April 2016 highlights the need to curtail state-owned enterprises (SOEs) and subsidies and to liberalize trade; a US$613 million loan to Bosnia in May 2016 calls on policymakers to address SOEs, cut taxes, and improve the business environment; a US$2.9 billion loan to Tunisia in May 2016 emphasizes reduction of the fiscal deficit and subsidies, exchange rate flexibility, and SOE reform; and a US$478 million loan to Suriname in May 2016 targets the elimination of electricity subsidies and increased tax revenues. A loan package of US$12 billion to Egypt in November 2016 was conditioned on liberalization of the exchange rate, reduction in energy subsidies, introduction of a value-added tax, fiscal consolidation, reductions in the public sector wage bill, and a tightening of monetary policy.⁶⁴

Several African countries have sought IMF assistance since 2014 as they grappled with commodity price weakness, capital outflows, and a slowdown of global and especially Chinese growth, FDI, and loans.⁶⁵ Some observers have suggested that African countries have returned to the IMF not just out of desperation but also out of a perception that it is less neoliberal than in the past (Pilling 2016). But the few recent agreements with African countries retain the IMF’s usual focus on fiscal restraint and privatization. A US$918 million loan to Ghana in April 2015 followed a 2014 Article IV report highlighting control of the wage bill and enforcement of a “no subsidy policy” on fuel and utilities (IMF 2014b, 17).⁶⁶ A US$1.5 billion precautionary loan to Kenya in March 2016 followed a 2014 Article IV consultation that urged the government to contain the wage bill, widen the tax base, privatize SOEs, and reduce the number of parastatals (IMF 2014c).

Discontinuity

The most prominent discontinuity in IMF practices during the global crisis concerns its rhetoric and practice on capital controls. This matter warrants substantial attention and is explored at length in chapter 7. The IMF came to embrace controls on inflows but also on outflows. The change in IMF practice has contributed to the normalization of capital controls as a tool of prudential macroeconomic management. The revival of capital controls began haltingly after the Asian crisis. But it became far more significant and consistent during the global crisis. These changes are by no means irreversible, and indeed are a site of contestation. But, as we will see, the case of capital controls provides a window into the complicated dynamic of an IMF that has been racing to catch up with the policy decisions of increasingly autonomous EMDEs; changes in the material power of member states that hold diverse views on this instrument, some of which have worked through the IMF and/or the G-20 to press for greater policy autonomy in regards to capital controls; and changes in the ideas of prominent mainstream academic economists and some of the Fund’s research staff.

Despite the studies cited regarding continuity in IMF conditionality, some observers find evidence of subtle discontinuities in its lending programs.⁶⁷ Early in the crisis, Broome (2010) found that while IMF loans to Mexico, Belarus, and Iceland in late 2008 and early 2009 featured the usual procyclical policy targets, he argues that the institution demonstrated a more flexible approach to crisis management, as evidenced by its differentiated treatment of borrowers, the provision of precautionary financing to Mexico, and its tolerance for short-term capital controls. This conclusion is cautiously echoed by Bird (2009). Later in the crisis, Broome (2015) finds evidence of both continuity and discontinuity across 93 SBAs from 1985 to 2012. He concludes that SBAs negotiated during 2008–2012 are characterized by the IMF’s usual mix of procyclical adjustments, but he argues that the institution moved away from promoting the structural reforms that were features of earlier SBAs. Broome speculates that discontinuities in Fund practices may stem from endogenous change at the IMF and exogenous change on the country level insofar as structural reforms such as trade liberalization have already been “locked in.”

Ambiguities: Discontinuities in Rhetoric and Research, Continuities in Practice

Though the design of IMF assistance packages during the crisis reveals strong continuity with past practice, there are subtle departures in IMF rhetoric and less subtle departures in IMF research that bear noting. These gaps between practice on the one hand and rhetoric and research on the other can be dismissed as subterfuges designed to disarm critics. But a more nuanced reading highlights the tensions and ambiguities that are revealed by new rhetoric and research, and the potential that these gaps hold to expand policy space.

Narrowed Conditionality: Rhetoric versus Practice

Throughout the global crisis, the IMF intensively marketed the idea that it had narrowed the scope of conditionality. Toward this end, Fund reports and statements by key officials herald the “end of conditionality” and emphasize country ownership and national policymaker involvement in reform programs.⁶⁸ The IMF’s Lagarde distanced the institution from conditionality in a 2011 press briefing in which she said, “Structural adjustments? That was before my time.… We don’t do that anymore” (quoted in Kentikelenis, Stubbs, and King 2016, 2).

What the IMF signaled as a radical rethinking of conditionality in fact predated the global crisis. Disdain for the IMF after the Asian crisis led staff to review conditionality in 2001 and then to adopt streamlined conditionality guidelines in September 2002 (IMF 2014a). Real change proved illusory, however. A 2007 study of conditionality by the IEO found that despite the adoption of streamlining guidelines in 2002, the number of structural conditions on Fund programs remained steady, and the Fund continued to promote conditions that were not necessary to achieve program goals. Moreover, the IEO found evidence of inconsistency in IMF conditionality (IEO 2007; 2009). Scholarly studies of the scope and terms of conditionality prior to the global crisis corroborate the IEO’s findings (e.g., Best 2007; Bird 2009; Momani 2005; Steinwand and Stone 2008).⁶⁹

The global crisis provided the IMF with another opportunity to inaugurate and then trumpet changes in its approach to conditionality. The purported end of conditionality was marked by the announcement in March 2009 that the Fund had eliminated “structural conditions” in lending programs as “performance criteria” and demoted them to the more flexible status of “benchmarks.” Structural conditions are generally nonquantifiable measures, such as privatization, that alter the economy’s underlying institutions and practices in ways deemed necessary for achieving program goals. They can be binding or nonbinding. Performance criteria are specific, measurable quantifiable conditions, often relating to macroeconomic variables (Edwards and Hsieh 2011). They must be met in order for a borrower to receive installments of a loan beyond the first installment (Bird 2009). Violation of performance criteria automatically triggers program suspension unless the IMF’s Executive Board approves a waiver. By contrast, benchmarks are enforced at staff discretion.

As the first wave of SBAs was being signed in Europe, it seemed that the IMF’s forceful new rhetoric around conditionality would be matched by its practice. But the same package of structural reforms that has long been at the heart of IMF practice were essential features of European SBAs. The Greek programs of 2010 and 2012 contained benchmarks involving privatization and restructuring of state-owned assets. For instance, the 2010 program was conditioned on privatization and restructuring of the national railway, state holdings in road transportation, airports, ports, utilities, real estate, and gaming (IMF 2010d, 50, 69, 80). In fact, the number of structural conditions increased over the course of Greece’s first program (IEO 2016, 31). The 2012 program was conditioned on reductions in minimum wages, changes in national collective bargaining arrangements, and measures that eliminate tenure and reduce employer contributions to social security programs (IMF 2012c, 21–23; Traynor 2012). In March 2014, the Troika insisted that the Greek government lower salaries for new public employees, reduce salaries for other civil servants, and retain the previously lowered minimum wage (Dabilis 2014). Many EMDE and European SBAs call for pension system and public sector reforms and elimination of subsidy programs. At the same time, SBAs continued to feature privatization as a structural condition in those cases where prior programs left some SOEs standing. Jamaica was one such case, and privatization of its national airline featured in its February 2010 SBA.

By now it is clear that the scope and content of conditionality in IMF assistance packages during the global crisis are consistent with those of previous crises (Grabel 2011, 821; Kentikelenis, Stubbs, and King 2016). Structural conditions remain central to the IMF’s toolkit, even if some structural conditions have been downgraded to benchmarks. Indeed, the use of benchmarks has increased in recent years.⁷⁰ The IMF reincorporated and widened the scope of many reforms, including structural reforms, as the crisis progressed after 2008. Agreements in 2014 had the highest number of benchmarks since this type of condition was introduced in the 1980s, and the number of conditions on loans rose during the global crisis. The numbers contradict the IMF’s long-standing claims regarding its streamlined approach to conditionality, let alone its more dramatic claims about the end of conditionality.

More broadly, borrower experience with IMF conditionality has not changed in any meaningful way since the crisis (Nelson 2014a, 165). The IMF’s own review of conditionality in programs signed between 2002 and September 2011 shows that “the number of conditions per program has fallen since peaking in 2004—but only back to the 2002 level” (Nelson 2014a, 162–163; IMF 2012b). Nevertheless, recent reports by the institution’s staff emphasize discontinuity by highlighting the reduction in the number of structural conditions in the SBAs of the global crisis (e.g., IMF 2012a; 2012b).

The shift in rhetoric regarding conditionality hung over one of the new instruments that the IMF introduced during the global crisis. In March 2009, the IMF introduced the FCL. The FCL is essentially a precautionary (one- to two-year) line of credit designed for countries that meet a demanding set of preconditions.⁷¹ To qualify, a country must demonstrate international capital market access and a sustainable external position; sound monetary, fiscal, and exchange rate fundamentals; a track record of sound policies and a credible commitment to their retention; the absence of bank solvency problems; and effective financial sector supervision. The criteria for policy soundness are so demanding as to render most EMDEs ineligible. Not surprisingly, few countries (namely, Mexico, Poland, and Colombia) have applied for and received FCLs.⁷² The IMF was apparently disappointed that more countries did not apply (Henning 2016, 123). One observer characterizes the FCL as a “large insurance policy one never wishes to claim” (Güven 2012, 874). The IMF has cited the FCL as an example of its new, modernized conditionality. But the FCL transforms traditional structural ex post conditionality into a demanding ex ante conditionality that elevates precisely the same neoliberal policy and institutional agenda that the Fund has been promoting over the last three decades (Grabel 2011).

Social Spending Targets and Attention to Vulnerable Groups in Support Packages

The IMF made much of its new commitment to social and pro-poor spending targets during the global crisis. Indeed, the support packages of the global crisis can be distinguished on paper from those of the Asian and prior crises by the IMF’s new emphasis on social protection. This emphasis appears in many IMF reports—for example, one that surveyed 15 SBAs between July 2008 and September 2009 (IMF 2009d) and another that surveyed program packages in 19 low-income countries in the same period (IMF 2009b, especially Annex 3; see also IMF 2014e). A 2009 report argues that social goals were supported in Fund programs and that social spending rose in some countries, such as Costa Rica (IMF 2009d). A 2011 IMF report goes further. It concludes that social spending as a percentage of total public spending rose in almost all program countries. More recently, the IMF has highlighted the ways in which its programs helped governments in EMDEs (including in low-income countries and countries in the Middle East) and in Greece to maintain or increase social spending and social assistance during the crisis (IMF 2015a; Clements, Gupta, and Nozaki 2013). The IMF has also claimed that program countries enjoyed increased spending on public health as a share of GDP (Clements, Gupta, and Nozaki 2014).⁷³

One example of the IMF’s increased attention to the most vulnerable during the crisis involves its stipulation that rather than eliminate the “13th month pension” in Hungary in response to fiscal difficulties, the government impose an income test to insulate poorer pensioners from the harshest effects of the crisis. A similar feature was incorporated into the SBAs with Latvia and Romania (Lütz and Kranke 2014).⁷⁴ In Greece, the IMF claims that it supported employment programs that target youth and jobless households and enhanced healthcare access for the uninsured via a voucher system, and that pension cuts targeted recipients of the highest pensions and supplemental pensions (IMF 2015a).

Critics contend that the ameliorative social policies that the IMF purports to have introduced have been inadequately or superficially incorporated into program design. For example, data from social expenditure targets in sub-Saharan Africa show that they have not been met half of the time, and that concerns regarding these targets remain secondary to the IMF’s preoccupation with macroeconomic targets (Kentikelenis, Stubbs, and King 2016).

It is difficult to square the IMF’s emphasis on protection of socially vulnerable groups with the severe fiscal constraints that are a key feature of so many of its recent programs. It is true that the decline in tax revenues and ODA has complicated the matter of financing social protection during the crisis. On the practical matter of where the funding for social protection is to come from, especially in the short run, the IMF has been largely silent. And in the case of Greece, for example, it is clear that IMF claims do not square with myriad indicators of severe social misery in the country during the crisis (Foy 2015; Galbraith 2016; Smith 2015). Its inconsistent opposition to the Troika’s demands for Greek austerity pale in comparison with the extent of the dislocation that austerity has induced. Making matters worse, even when the IMF addressed the matter of financing social protection, its claims strained credibility. For example, a 2016 program in Egypt calls for a strengthened social safety net that includes increased spending on food subsidies and cash transfers to be financed by savings from fiscal consolidation (IMF 2016e).

The gap between IMF rhetoric and practice on protecting the vulnerable in recent programs tracks prior failures to change IMF practices on poverty and inequality (on prior failures, see Momani 2010). IMF critics are right to demand much more of the institution, but it is important to recognize that the new rhetoric on pro-poor spending is not strictly reducible to a public relations offensive, ceremonialism, or outright misstatements. The IMF’s failure to incorporate social spending targets reflects in part the challenges of changing organizational cultures in institutions comprising diverse stakeholders, embedded interests, familiar scripts, and continuity in staff. In that kind of environment, the incorporation of new principles, norms, goals, and strategies cannot be orderly or seamless (Vetterlein 2010).⁷⁵ Changing practice to incorporate new aspirations is complicated and difficult, and can be frustrated by staff that are ill prepared, resistant, and that operate with some degree of autonomy, and by constraints imposed by external actors (as per Park and Vetterlein 2010).

Growing Inconsistency on Austerity, Inequality, and Fiscal Consolidation

In 2010, and especially after the fall of 2012, the IMF’s key officials, reports, and research began to highlight the costs of extreme austerity, inequality, and fiscal consolidation. These messages reflected self-assessments by IMF staff of the institution’s work, new staff research, and intra-Troika tensions. The IMF’s gentler rhetorical stance after 2012 also reflected proactive efforts to protect the institution’s restored franchise, as well as broader zeitgeist concerns about inequality.

On the matter of austerity, in late 2012, Lagarde nudged European Troika counterparts and especially the German government toward a more flexible position. She called them to task for what she termed “wishful thinking” in the Greek program (Reuters 2012a), while also pressing to curtail the austerity program at the annual IMF–World Bank meeting in Tokyo in October 2012. At the same meeting, Germany’s Schäuble publicly rejected this challenge. A senior IMF official involved in the negotiation of Ireland’s November 2010 program highlighted intra-Troika tensions over austerity in remarks in 2013 (Smyth 2013).

IMF economists expressed unease with radical austerity and severe inequality during the crisis, arguing that they were counterproductive to growth. In 2010, Daniel Leigh and his colleagues provided empirical refutation of the “expansionary austerity” thesis for AEs, as offered by Alesina and Ardagna (2009) and former ECB chair Jean-Claude Trichet, which holds that slashing spending during crises creates jobs. The research of Leigh et al. (2010) was featured in the IMF’s October 2010 World Economic Outlook. It reflected early evidence of the costs of austerity in some of the first countries caught up in the global crisis.

Research by Blanchard and Leigh in 2012 and 2013 concluded that the fiscal multipliers that the IMF (and other researchers) used had seriously and consistently underestimated the negative effect of fiscal tightening on European growth early in the crisis (see, e.g., Blanchard and Leigh 2013).⁷⁶ This work first appeared in the October 2012 World Economic Outlook (IMF 2012g, especially box 1.1), and led to tense exchanges between Lagarde and European critics at the IMF–World Bank meeting in Tokyo. The EC shot back quickly and released a report that downplayed the effect of austerity on European growth (Spiegel 2012; EC 2012). But despite the high profile of Blanchard and Leigh’s work, it had no noticeable effect on the IMF’s practices in Europe or elsewhere. And even the authors themselves seemed disquieted by their own findings. They inexplicably defended the Fund’s work in Europe as “mostly right” (Blanchard and Leigh 2013; IMF 2012g, especially box 1.1). In 2015, Blanchard defended the IMF’s work in Greece, saying that “fiscal consolidation explains only a small fraction of the output decline” (IEO 2016, fn5). And in statements to the press, Blanchard spoke against stimulus policies and reversals of austerity (Weisbrot 2015, 163), and he supported the IMF’s insistence on further fiscal consolidation (Mody 2016).

Research and public statements by key IMF staff in 2012 raised questions about the strategy of “fiscal consolidation,” especially “front-loaded” consolidation, which the G-20 and the IMF promoted in 2010 and 2011. In the foreword to the April 2012 World Economic Outlook, Blanchard argued that financial markets do not react to fiscal consolidation as conventional economic theory assumes. Blanchard wrote, “Markets appear somewhat schizophrenic—they ask for fiscal consolidation but react badly when consolidation leads to lower growth” (IMF 2012h, xiv). Clift (2014) notes that the IMF’s growing skepticism about extreme fiscal consolidation was a point of contention within the Troika. But he also observes that from 2010 onward the IMF advocated only a slower, somewhat more mindful procyclical fiscal policy than the EU and ECB had recommended (ibid.).⁷⁷

Internal assessment of the IMF’s work in Greece and broader research by the institution’s staff in 2013 and 2014 continued to reflect tensions within the Troika and discomfort at the IMF with radical austerity and inequality. The April 2013 World Economic Outlook took a sidelong glance at the obsession with fiscal rectitude and inflation by the European Troika and some G-20 members (IMF 2013e). The World Economic Outlook refers to inflation as the “dog that didn’t bark” (ibid., chap. 3). Other Fund research published around this time evidences unease with extreme austerity. One study concludes that Northern Europe would need to accept more inflation if Southern Europe is expected to adjust (Blanchard, Jaumotte, and Loungani 2013), a theme that is obliquely echoed in the April 2014 World Economic Outlook (IMF 2014d, 54–55). In addition, a widely cited study by Ostry, Berg, and Tsangarides (2014) examined the costs of inequality for growth, political and economic stability, social cohesion, and progress in health and educational outcomes. It concludes that some redistributive policies can have benign or even beneficial effects on growth.⁷⁸

The IMF’s views on austerity, inequality, and fiscal consolidation evolved inconsistently during the crisis. Research by the institution’s staff and public statements by key officials remained largely orthogonal to the institution’s practice with client states (Ban 2013; Gabor 2010; Güven 2012; Ortiz and Cummins 2013). Ban (2015) characterizes the IMF’s stance on fiscal matters as “hybridized … editing … a very modest recalibration” in which precrisis orthodoxy is overlaid with mildly Keynesian themes in its research and public statements in 2008–2009 and again after 2012. The IEO takes note of the perception that the IMF’s advice on fiscal policy matters is evolving modestly and gradually, even while it continues to emphasize the primacy of fiscal space when considering the viability of short-term fiscal stimulus (IEO 2013; Reichmann 2013). In summing up interviews and surveys, especially among EMDE representatives, the IEO reports that there is “a shift in perceptions.… [A] common view … was that the IMF … abandoned its emphasis on fiscal adjustment and was … more attuned to … social and economic development … [and that] the IMF became a strong proponent of short-term fiscal stimulus, wherever fiscal space was available, further dampening its preoccupation with fiscal discipline” (IEO 2013, 11). In mid-2016, Lagarde continued to speak of her desire to have the institution go “deeper and further” in its work on inequality, but at the same time she acknowledged skepticism about the matter by outsiders and some within the institution (Donnan 2016a).⁷⁹

For some observers, the new framing of outdated policies represents a desperate attempt by the IMF to burnish its image. But it is better to see the gap between research, rhetoric, and practice as one among several ambiguities—resulting in part from the IMF’s “silo-like” organizational culture (as per Nelson 2017, 205)—that reflect a contested, uneven, and ultimately uncertain process of change being negotiated within an extraordinarily complex international organization.⁸⁰

This discussion sustains the following tentative conclusions. The IMF’s response to the global crisis diverges in some respects from, and is less coherent than, its response to the Asian crisis and prior crises. IMF practice concerning macroeconomic adjustment evidences strong policy continuity with that of previous crises, despite new research by the institution’s economists, a multifaceted rhetoric of change, and Keynesian-inflected concerns during the early moments of the crisis. The IMF continues to apply pressure to secure compliance with stringent fiscal policy targets and expansive conditionality.⁸¹ To a limited extent, the Fund now emphasizes the need to protect social spending targets and the dispossessed while moderating the pace of fiscal consolidation and austerity. Although it has failed to move much beyond aspirations in these areas, the new rhetoric and research mark important breaks with the past that legitimize a concern for the poor and for the social costs of inequality and austerity. Though it is certainly not the IMF’s intent to do so, the new rhetoric is empowering external actors to hold the institution accountable to IMF-sanctioned normative criteria and concerns that had largely been missing from the IMF platform. In addition, the rhetorical turn exposes the long-held neoliberal script to contestation from IMF insiders, such as IEO researchers who have proven themselves willing and able to indict the institution. Indeed, for the first time in decades, there is growing overlap between the insights of IMF insiders and outside researchers on IMF failures.

Conclusion

The question is not “do we have a new IMF?” as some observers, such as Weisbrot (2015, 160), have bluntly put the matter. For him, the question can easily be answered in the negative. Proof of concept is found in the scope and severity of conditionality, the influence of key member states and the failed ideas they advance, and IMF hypocrisy (ibid., 160, 163). Weisbrot presses the matter further in asserting, with equal certainty, that there will be no significant change in the future (ibid., 239). He predicts that the IMF will be less important to EMDEs as China gives countries more and better options (ibid.), but he doesn’t recognize that this shifting landscape will alter IMF governance and practice.⁸²

In my view, the matter is far more complex than other analysts will allow. The IMF has been and is evolving in ways that bear centrally on global financial flows and governance. In some domains, we find clear and compelling evidence of continuity. In others, especially as concerns capital controls, we find consequential discontinuities (see chapter 7). And in still others, we find mounting evidence of inconsistencies and ambiguities that involve substantive gaps between the rhetoric, research, and policy practice. As is clear by now, I treat change at the IMF in a manner that is in keeping with subtler treatments that emphasize unevenness, discontinuity, and incrementalism.⁸³ The conclusion to be drawn is that there has been partial and inconsistent though meaningful change at the IMF, despite evidence of continuity.⁸⁴ This finding is driven by the empirical record when read through a Hirschmanian lens—one that recognizes that institutional evolution is typically contingent, messy, and contested, and that though historical developments are path dependent, the future is fundamentally unscripted.

In seeking to come to some conclusion about whether the IMF has actually changed, many observers scrutinized a widely circulated 2016 essay in the IMF’s Finance and Development magazine. The essay “Neoliberalism Oversold?” by IMF economists Jonathan Ostry, Prakash Loungani, and Davide Furceri (Ostry, Loungani, and Furceri 2016) reflects on the implications of their own and their colleagues’ recent research on austerity, inequality, and fiscal consolidation. Some IMF watchers seized on the essay to support the case that the IMF had changed fundamentally. Others emphasized its affirmation of central doctrines of neoliberal policy as evidence that the change is more apparent than real. In reality, the essay exemplifies an internally contested, inconsistent, and still evolving process of change at the IMF during the global crisis. On the one hand, Ostry, Loungani, and Furceri state that “there is much to cheer in the neoliberal agenda” (ibid., 38). They affirm the IMF’s traditional view in stating that many countries have little choice but to engage in fiscal consolidation because markets will not let them borrow, and they emphasize the benefits of privatization, trade expansion, liberalization, and FDI. But they also argue that aspects of the neoliberal agenda “have not delivered as expected” (ibid.). In making this case, they cite recent research that shows that countries with “ample fiscal space” would do better to live with debt; that expansionary fiscal consolidation has been discredited; and that the short-run costs of fiscal consolidation have been underplayed.⁸⁵ In a subsequent Financial Times article, Ostry notes the simultaneous wariness and new openness at the IMF, stating that the essay “did not reflect ‘mainstream culture’ at the IMF and would not have made it into a fund publication as recently as five years ago” (Donnan 2016b). In the same article, Rodrik quips, “What the hell is going on?” He notes that “ ‘there is definitely a gap’ between the IMF’s research arm and other parts of the institution. ‘The operational side … is typically more orthodox.… There the change is slower and is lagging behind the thinking’ ” (quoted in Donnan 2016b). Rodrik also observes that the IMF’s push for debt relief in Greece suggests that research by Ostry and others is influencing IMF policy. In an interview on the Ostry, Loungani, and Furceri essay, the IMF’s Obstfeld describes the process of change at the IMF as “evolution not revolution” (IMF 2016b, 1). This seems a reasonable interpretation of how change occurs in a complex organization comprising diverse internal and external stakeholders with conflicting interests and competing views about prudent economic policy.

As events in Europe demonstrate, for the first time in decades, the IMF’s role as the central driver of crisis response activities is in question. The centrality of the IMF will depend on its continued ability to retain the financial vitality and authority it regained during the global crisis, and on whether EMDEs are willing or are forced to turn to the IMF in the coming years. The latter, in turn, depends very much on the ability of EMDEs to navigate future global turbulence, and on whether emerging institutions in these countries develop the capacity to compete with the IMF or evolve in ways that create linkages with the institution that are distinct from those that characterize the Troika. For now, the IMF has many “unmet challenges,” as Eichengreen and Woods (2016) have recently argued. These include addressing the content and lack of evenhandedness of conditionality and surveillance, addressing the long-identified absence of a sovereign debt restructuring mechanism, and making significant progress on governance (ibid.). Failure to take on these challenges will reinforce deep concerns about its competence, credibility, and legitimacy (Eichengreen and Woods 2016; Nelson 2014a).

That said, crisis-induced transformations in global financial governance provide the IMF with yet another opportunity for reinvention. One path entails an effort to reassert itself as the driver of institutional and policy coherence across the EMDEs (and those AEs that are unfortunate enough to face IMF conditionality). That path will be difficult to tread given increased EMDE autonomy and assertiveness during and since the crisis, and the erosion in confidence in the simplistic neoliberal script. The other path leads toward recognition of the legitimacy and necessity of policy autonomy, where the IMF coordinates activities with established and emerging EMDE institutions of financial governance to mediate institutional and policy diversity rather than imposing conformity (see chapters 6–8). This task is perhaps less appealing to economists trained during the latter decades of the twentieth century, who were socialized to theorize and engineer a globally coherent regime that infused national economies while spanning national borders. It may require a Hirschmanian shift in weltanschauung that encompasses a new vision of the role of economists in the development process.

If we widen our gaze beyond the IMF to consider the EMDEs themselves, we find widespread experimentation and an increased density in the landscape of institutions that provide liquidity support and development finance. Existing institutions have been given new life by the global crisis, and entirely new institutions have been created. Some of these initiatives, especially the largest, which feature China as the key driver, are already having significant impacts on the BWIs. In this context, some observers, such as Ocampo (2010c), have called for the IMF to take on new roles in coordinating or supporting the evolving, dense geometry of global financial governance. In the next chapter, we turn to the new emerging landscape.

Notes

1. See Grabel (2011) for an early elaboration of some of these arguments.

2. The continuity view of the IMF is reflected in Gabor (2010; 2015), Güven (2012), Islam et al. (2012), Kentikelenis, Stubbs, and King (2016), Muchhala (2011), Nelson (2014a; 2017), Ortiz and Cummins (2013), Rowden (2009), UNICEF (2010), Van Waeyenberge, Bargawi, and McKinley (2011), and Vernengo and Ford (2014). Reinold (2016) argues for continuity in one dimension of IMF practice; Güven (2012) and Vestergaard and Wade (2013a) find continuity at the World Bank; and Hanieh (2015) concludes that continuity marks BWI practice in Tunisia, Morocco, and Egypt.

3. On the BWIs, see Park and Vetterlein (2010). On the IMF, see Abdelal (2007, chap. 6), Babb (2003), Ban (2015), Barnett and Finnemore (2004, chap. 3), Best (2005; 2006; 2012), Broome and Seabrooke (2007; 2012), Chwieroth (2010), Lütz and Kranke (2014), Momani (2005; 2010), Moschella (2010), Nelson (2017), and Woods (2006). On the World Bank, see Weaver (2008). On the role of local interlocutors, see Ban (2016) and Woods (2006). Another tradition in political science, historical institutionalism, also emphasizes the complex interaction of the economic and social realms and focuses on the mutually constitutive relationship between agents and the ideational, material, and political conditions that open and restrict their fields of action (see, e.g., Mahoney and Thelen 2010, 6–7).

4. See related discussions in Abdelal (2007, chap. 6), Broome and Seabrooke (2007; 2012), Moschella (2012), and Weaver (2008). Mahoney and Thelen’s (2010, 8) treatment of “ambiguous compromises” echoes key constructivist themes.

5. Wade’s (1996; 2002) and Weaver’s (2008) work on the World Bank are taken as seminal. Wade advances the idea of “paradigm maintenance”; Weaver’s concept of organized hypocrisy is an extension of Brunsson (1989; 2003; see Weaver 2008, chap. 2). See Kentikelenis, Stubbs, and King (2016) on the related idea of “organizational facades.”

6. Broome and Seabrooke (2012) offer Fine (2001) as an example of analysis that fails to resist this temptation. Another example can be found in Vernengo and Ford’s (2014) treatment of the IMF. Relatedly, Palley (2013) focuses on the public relations value and political cover provided by economists’ strategy of asserting change in economic ideas when none has actually occurred. See also Vollmann (2015) on image and political cover in recent IMF research on inequality and labor unions.

7. Historical institutionalist analysis has tended to emphasize institutional continuity over long periods. New contributions push the boundaries of the approach, taking account instead of the ways in which institutions evolve endogenously, incrementally, and in subtle ways that can add up to consequential shifts in identity and practice. Piecemeal adjustments tend to be overlooked and underappreciated by those who focus on momentous exogenous shocks leading to radical, abrupt institutional transformation. See Mahoney and Thelen (2010, chap. 1) (and essays therein) on modalities of gradual institutional change.

8. Seabrooke (2010, 139) focuses on another type of gap at the IMF—the gap between the IMF’s “reform ‘talk’ and borrowers’ reform ‘walk.’ ” The gap between IMF dictate and borrower practice has grown since the Asian crisis, as more EMDEs achieve greater relative autonomy from the IMF (discussed later).

9. See Barnett and Finnemore (2004, chap. 3), Chwieroth (2007a; 2007b; 2010), and Nelson (2014b; 2017) on the shared neoliberal worldview, selection, training, and socialization of IMF staff, and related work by Best (2007) and Momani (2005).

10. Several key contributions focus on the way that exogenous shocks undermine the capacity of predominant ideas to explain events while creating opportunities for new ideas to gain traction, either rapidly or incrementally, and stick (Best 2003; Blyth 2002; Broome and Seabrooke 2007; Chwieroth 2014; 2015; Moschella 2010; 2014; Widmaier, Blyth, and Seabrooke 2007), while others focus on the interaction between ideas and external interests in driving ideational change (Blyth 2003; Kirshner 2003; Moschella 2010). McNamara (1999) illustrates how ideas influence economic policy formation, Widmaier (2007) shows that similarities in crisis conditions facing policymakers can yield dramatically divergent interpretations of interests and crisis responses, owing to their respective understandings of monetary trends (see also Widmaier 2014), and Widmaier (2016) demonstrates how ideas that initially reduce uncertainty can ultimately breed misplaced certainty and crisis.

11. See citations in note 3 of this chapter.

12. See Wade (2010; 2013a) on the World Bank.

13. There was also a very modest increase in IMF resources through gold sales and other means. The NAB was established in 1997 and activated in 1998 to finance an SBA for Brazil. The NAB was not utilized again until April 2011. It has been activated ten times since its 2011 enlargement.

14. See McElhiny (2009) for a description of how the IADB’s holiday party in 2008 celebrated the increased demand for infrastructure packages. See Kulish (2009) on reinvigoration of the European Bank for Reconstruction and Development (EBRD). See chapter 6 for a broader discussion.

15. See Helleiner (2014b, 40, fn58), Henning (2016), and Prasad (2014). On bank exposure, see Broz (2014).

16. See also Harding (2014), Reuters (2013b), and Yukhananov (2013).

17. After the Fed, the PBOC provided the second-largest set of swaps to EMDEs (Gallagher 2014, 75). Together, the two provided over US$200 billion in swaps to EMDEs during the global crisis (ibid., 76). McDowell (2017) notes that EMDE central banks have been key actors in the spread of the international swap network, which was motivated by recognition of the risks of international capital flows and of dollar dependence in trade.

18. European officials were initially reluctant to involve the IMF, but by March 2010 they were forced to acknowledge the necessity of doing so (Eichengreen 2015, 134).

19. The ESM is a crisis-resolution mechanism that was created by euro area countries during the global crisis. The ESM can raise funds from numerous sources, including private markets, and can provide support to Eurozone countries, subject to conditions (Schadler 2014, fn8). It is to replace the temporary European Financial Stability Facility (EFSF).

20. The preference for internal devaluation over external devaluation by national authorities reflected what Kattel and Raudla (2013) identify as a hardening of neoliberal resolve in the Baltics during the crisis. Troika members apparently feared that external devaluation would induce severe instability in Baltic capital markets and capital flight, and would have negative spillovers in Central and Eastern Europe.

21. See also Ban (2016) and Moschella (2016).

22. The levy was ultimately rejected by the Cypriot Parliament.

23. Strauss-Kahn, a small group of IMF staff, and officials from the German and French finance ministries quietly explored Greek debt restructuring in 2010. Debt relief was not granted until March 2012 (Blustein 2015, 1).

24. The report also contains self-indictment: it acknowledges that the IMF underestimated the negative effects of spending cuts and tax increases, and misjudged the government’s capacity to implement structural reforms. Nevertheless, it concludes that the IMF’s “overall thrust of policies” was “broadly correct” (IMF 2013a, 32).

25. The IEO takes note of EU discomfort with the IMF’s Ex Post Evaluation (IEO 2016, 3). The IEO (ibid., 16) also notes that senior IMF officials were divided and that staff were generally skeptical regarding its involvement in Greece.

26. One of the main goals of the second SBA was to protect European banks that bought Greek bonds in the hopes of making a profit (Ewing and Alderman 2015).

27. Discussion in this paragraph relies on Taylor (2015).

28. The Eurogroup is an informal body comprising Eurozone finance ministers, the EC, and the ECB.

29. The primary budget excludes interest payments on debt.

30. As of May 2017 the IMF remains non-committal about whether it will participate in the disbursal to Greece that is scheduled for July 2017.

31. Despite much evidence to the contrary, a 2016 IEO report rejects the view that the IMF had a “muted voice” in Troika programs in Portugal, Ireland, and (the first program in) Greece (IEO 2016). The report does note that the institution lost its “characteristic agility” as a crisis manager and that the technical judgments of staff were potentially subjected to political pressure. It also faults the fragmented nature of euro-area decision making. News reports suggest that IMF staff, some executives, members of the European Department, and some European executive directors unsuccessfully tried to obstruct production of and remove passages from the IEO report (Thomas 2016).

32. Additional support to Ireland also came in the form of bilateral loans of €3.8 billion from the United Kingdom, €0.6 billion from Sweden, and €0.4 billion from Denmark; €17.7 billion from Euro-area member states/EFSF; and from the Irish government itself, which contributed €17.5 billion from its own cash reserves and liquid assets. Reporting on IMF participation in European packages is complicated by program suspensions, renewals, replacement of programs by new ones, and the rapid evolution of European support mechanisms. Figures here are drawn from IMF press releases and Miyoshi (2013, 19–21).

33. For example, at the end of December 2015, the notional amount of outstanding over-the-counter derivatives contracts was valued at US$493 trillion (BIS 2016).

34. Some analysts disagree, arguing that the IMF would likely not consent to being a junior partner in arrangements with regional financial actors outside Europe (Mohan and Kapur 2015, 51).

35. Disbursal of the larger loan by the World Bank has been delayed until at least 2017, owing to a dispute between the government and the institution over macroeconomic reforms. Nigeria did receive disbursement of the US$500 million IDA loan in 2016.

36. See Yukhananov (2013) for comments along these lines by other BRICS country leaders.

37. The Intergovernmental Group of Twenty-Four on International Monetary Affairs and Development (known as the G-24) is an informal organization established in 1971. It has a permanent secretariat in Washington and is housed within the IMF’s building, though it does not operate on the IMF board level (Woods and Lombardi 2006). The G-24 coordinates the positions of a collection of EMDEs at the BWIs and advocates on their behalf in discussions of global financial reform (G-24 2014).

38. See Woods (2010) on BRICs positions on governance reform, financial support to the Fund, and instruments through which support would be provided.

39. The agreement abolishes the right of the five largest quota holders to appoint their own members of the Executive Board. Instead, all board members are to be elected by groups of countries (Eichengreen and Woods 2016, 48).

40. Executive directors are civil servants appointed by their home governments. They sit on the IMF’s Executive Board and are apportioned voting rights on it. The Executive Board is the institutional venue where member states most directly exercise their national interests (Nelson 2014a, 167). The Executive Board is the IMF’s day-to-day governing council. Countries with more voting power have their own representatives on the board, while other countries share a representative (Eichengreen and Woods 2016).

41. Mohan and Kapur (2015, 3) note that U.S. Treasury officials had called for IMF governance reform since at least 2005. The United States was the principal architect of the 2010 agreement (Truman 2015). See Vestergaard and Wade (2015) on the politics of the gridlock, and see Lesage et al. (2013) on why the U.S. executive branch was willing to approve the 2010 agreement.

42. Vestergaard and Wade (2014; 2015) show that governance reforms at the World Bank in 2010 had an even more microscopic, perverse effect on EMDE voice than reforms at the IMF.

43. This matter is explored extensively by Lesage et al. (2013), and I draw on their discussion.

44. As noted in chapter 1, South Africa formally joined the group at its April 2011 meeting, leading to the adoption of the new acronym BRICS to replace the previous acronym BRICs.

45. See Wade (2012) on the politics of this episode.

46. So did the practice (rather than formal rule) of requiring a supermajority of votes. The norm maintains U.S. veto power (Wade 2012).

47. The divisiveness among the BRICS is proof for some analysts that they are unlikely to challenge prevailing patterns of global governance (Bond 2013).

48. Members can still enjoy exceptional access absent an upfront debt restructuring operation so long as other official creditors provide loans on sufficiently concessional terms to make the debt sustainable (Ellmers 2016).

49. Article IV consultations are carried out annually for nearly every member country, result in a paper on each country’s economy and economic policy, and provide recommendations on a range of issues. Recommendations become influential only when countries require IMF assistance (Lombardi and Woods 2008).

50. Jiang (2014) takes a different view. She sees China’s actions and rhetoric as symbolic, self-interested, and not indicative of a significant break with the past.

51. SDR-bond arrangements were to be folded into the expanded NAB, which was to be unwound once the 2010 agreement went into effect (Lesage et al. 2013, 561).

52. Some analysts suggest that China’s decision to contribute this second tranche of funds was related to Japan’s mid-April 2012 announcement that it would contribute an additional US$60 billion to the IMF (Jiang 2014, 174). Japan’s announcement apparently came as a surprise to China. Since February 2012, and in a rare instance of solidarity, China and Japan have raised the issue of increasing the EMDEs’ voice at the IMF while also pressing the Eurozone to lift a €500 billion cap on bailout funds if it hoped to secure additional funding from non-European G-20 members (ibid., 173). China’s second tranche also reflected its “responsible great power” ambitions, its commitment to provide funds through the safety of the IMF, and broader BRICS demands to increase EMDEs’ voice at the Fund (ibid.).

53. Civil society organizations protested these contributions on two grounds. First, there were far better domestic uses for these funds than helping the IMF assist European banks. Second, long-standing objections to Fund governance had not been addressed. See Bond (2013) on opposition by South African trade unions to the government’s commitment.

54. Recall that Mantega began raising this threat in 2007.

55. Abdelal (2007) highlights informal, though effective, influence exercised by France in relation to U.S. interests at the IMF prior to the global crisis.

56. The last time the Treasury Department labeled China as such was in 1994 (though it presently has the country on a “currency monitor list” along with Germany, Japan, and South Korea).

57. The discussion in the balance of this paragraph and the next draws on Armijo, Katada, and Roberts (2015, 17–18).

58. The IEO acknowledged that the IMF contributed to the crisis by promoting financial liberalization (IEO 2011) and missing signs of banking system fragility in several countries (IEO 2016, vii).

59. The IEO takes the IMF to task in noting that the “call for fiscal consolidation proved … premature,” that the policy mix was “not appropriate,” and that the negative consequences of European austerity were foreseeable in light of the institution’s precrisis research (IEO 2014, 11, 33).

60. See also findings in Molina (2010), Muchhala (2011), Rowden (2009), and Van Waeyenberge, Bargawi, and McKinley (2011).

61. More broadly, Vreeland argues that IMF conditionality provides the political leverage to advance unpopular domestic policy changes: “The moment it demands that deficits be cut … the IMF has entered into domestic politics” (Vreeland 2008, 367).

62. Country reports include Article IV consultations, reviews conducted as part of lending operations, consultations under nonlending arrangements, and publicly available IMF reports.

63. Momani and Lanz (2014) see the impact of the Arab Spring on the IMF differently. They find that there has been significant change in the cases of Tunisia, Egypt, and Morocco. In their reading, the content of IMF policy advice is “visibly different” in these cases (i.e., the IMF gave greater attention to inclusive growth, income inequality and redistribution, and health and education spending). They acknowledge, however, that the quality of the IMF’s advice on social targets is vague, particularly in comparison with advice in other policy realms. While the authors do not address the factors that drove change, their analysis suggests that political turmoil and social pressures led to increased IMF focus on the social dimensions of economic policy.

64. Egypt’s currency was devalued by 50% in anticipation of the IMF agreement.

65. Some countries, such as Nigeria, have continued to turn away from the Fund and have sought support from other multilateral institutions and investment from China.

66. The IMF continued to highlight the need to contain the wage bill around the September 2016 disbursement of a loan tranche.

67. Several self-congratulatory reports written early in the crisis by Fund staff conclude that the policy content of new IMF programs is fundamentally different from that of the Asian crisis era (IMF 2009b; 2009d; 2011a).

68. For example, in public statements on the November 2016 program with Egypt, the IMF emphasized its “homegrown” character (IMF 2016e). Best (2014) links the IMF’s emphasis on country ownership to its efforts to manage policy failure. In earlier work, she argued that the IMF’s efforts to promote borrower ownership of reforms and revise conditionality were at the heart of its struggles to reinvent and “re-legitimize” itself prior to the global crisis (Best 2007).

69. The creation of the IEO after the Asian crisis and the office’s willingness to issue reports critical of IMF performance even in areas that are central to the IMF’s mission themselves represent vitally important ruptures in IMF conduct. The reports contribute to ambiguity at the IMF. They often validate the critiques of less influential IMF members and external IMF watchers but do not dictate changes in IMF practices. The latter point was underscored in an early external evaluation of the IEO, which concluded that there is “little evidence that findings and recommendations … are being systemically taken up and followed by senior management and the Board” (Lissakers, Husain, and Woods 2006, 4). Nevertheless, IEO reports have legitimized a range of dissenting voices within the IMF and beyond, and have contributed to an environment where it is increasingly respectable for key actors to criticize and/or discount the Fund’s policy positions.

70. The remainder of this paragraph draws on Kentikelenis, Stubbs, and King (2016).

71. The Fund developed precautionary facilities after the Asian crisis. It introduced several precautionary facilities during the global crisis. It launched the Short-Term Liquidity Facility in October 2008, replaced it with the FCL in March 2009, and introduced the Precautionary Credit Line (PCL) in 2010. Macedonia was the only country that applied for and received support under the PCL (in January 2011). In November 2011, the PCL was replaced by the Precautionary and Liquidity Line, which has only been used by Morocco. The replacement of one program by another is indicative of the IMF’s inability to develop the right instrument and to appreciate the stigma and market signals associated with its support (Henning 2016, 123).

72. Mexico was the first to apply for an FCL. It was approved for US$47 billion in April 2009. In May 2009, Poland and Colombia were approved for US$20.5 billion and US$10.5 billion, respectively. These three countries have not drawn on the FCL, and each has several times requested and been approved for new, larger lines as old ones expired.

73. One scholarly study of social spending targets in IMF concessional lending programs early in the global crisis concludes that there has been nearly universal incorporation of these targets into concessional loans (Clegg 2014).

74. Lütz and Kranke (2014, 320, 324) argue that in Hungary, Latvia, and Romania the IMF was more sensitive than its European partners to social protection. See Kattel and Raudla (2013) on conflict between the IMF and EU partners over internal devaluation in the Baltics.

75. Reinold (2016) identifies the same dynamic in explaining the IMF’s failure to mainstream what it terms “social issues” (e.g., human rights and the right to food and shelter) despite much rhetoric about doing so. She also identifies this dynamic in the failure to act on Lagarde’s frequent attention to income inequality and gender equity (discussed later).

76. As Mody (2016) observes, estimation of fiscal multipliers is arguably one of the IMF’s areas of “core competency,” so this failure is particularly damaging to its credibility. The IMF has a long history of issuing overly optimistic forecasts only to downgrade them later.

77. See also Ban (2015) on modest changes at the IMF regarding fiscal consolidation, and see also the discussion earlier in this chapter.

78. Other IMF research during the crisis is indicative of a new discomfort with “big finance.” Gabor (2015) traces the emergence after 2010 of the IMF’s view that transnational banks were, in its words, “super spreaders” of systemic risk and tended to engage in regulatory and tax arbitrage. Here, too, there is a gap between IMF research and practice, as Gabor demonstrates. Another high-profile IMF study expresses reservations about the size of the financial sector in several large AEs (Sahay et al. 2015). The authors argue that these economies suffer from “too much finance” and that banks and financial institutions that assume too big a share in the economy contribute more to economic and financial volatility than to growth.

79. The IMF also toyed with new ideas on tax policy. It issued tepid support in 2010 and 2013 for taxation of large financial institutions and more general progressive taxation (Claessens, Kean, and Pazarbasioglu 2010; IMF 2010b; 2013c). Some observers saw the Fund’s Fiscal Monitor Report of October 2013 as advancing a case for a wealth tax (IMF 2013b). After the report was released, the institution quickly issued a press release to distance itself from this interpretation (IMF 2013d). The reversal illustrates the continued internal contestation over new ideas, and also the influence of key member states that sought to reframe what they saw as the IMF’s message on taxation.

80. Though an advocate of strong continuity, Nelson (2017, 205) acknowledges discontinuity around IMF research (and capital controls).

81. Monetary and exchange rate policies have not figured into the European programs since most of these countries do not have independent monetary policy and several post-communist countries chose not to break their currency peg.

82. For citations of what I have called the continuity literature, see note 2 of this chapter.

83. Scattered throughout this chapter are references to the salient literature, such as Ban (2015), Ban and Gallagher (2015), Best (2014; 2016), Broome (2015), Clegg (2014), Grabel (2011), and Moschella (2010; 2014; 2016).

84. On change in other international organizations, see, for example, Sharma (2013) on discontinuities at the World Bank and Chorev (2013) on the World Health Organization. See Babb (2013) on change surrounding the Washington Consensus.

85. The essay does not address capital flow liberalization, though this is an area where IMF research and practice has evolved most unambiguously.