140
THE GLOBALIZERS
the material incentives they can offer, as well as persuasive power deriving from the expertise of their staff. In any one case the influence of the IMF or World Bank may be subject to overriding political imperatives, resource constraints, and other institutional exigencies, and politics within borrowing countries. However, experts within the institutions have a power to define what is possible and to present and probe the implications of any given set of policies. In so doing they cannot only influence choices made within the abovementioned constraints, but also inform those who impose the constraints.
Chapter 6
MISSION UNACCOMPLISHED IN AFRICA
Sub-Saharan Africa came to rely heavily on the IMF and the World Bank during the 1980s. The 1970s oil price rises, the raising of U.S. interest rates and subsequent contraction in the global economy in 1979, the appreciation of the U.S. dollar, and the ongoing volatility of commodity prices rocked the continent. Deep domestic policy weaknesses and poorly aimed interventions by Cold War rivals, former colonial powers, and aid donors further enfeebled African countries who tried to deal with these problems.
In the face of a continentwide crisis, the IMF and the World Bank became frontline purveyors of advice and conditional resources for Africa. The stakes were high for both institutions, as expressed by World Bank president Barber Conable in April 1986:
The role and reputation of the Bank Group is at stake in Africa. . . . We have said publicly . . . that we are giving Africa the highest priority. . . . We have been telling Africa how to reform, sometimes in terms of great detail. ... If these programmes fail, for whatever reasons, our policies will be seen widely to have failed, the ideas themselves will be set back for a long time in Africa and elsewhere. (Kapur et al. 1997, 73°)
Africa was recognized as a serious test for both the IMF and the World Bank.
Africa was potentially a showcase for the technical expertise of the institutions because unlike Mexico and Russia, the country-level work of each international institution was not overridden by threats to international financial stability or the need to stabilize a nuclear arsenal. Nowhere was good quality economic advice more needed. Many African governments had limited capacity to analyze global economic trends and shocks, yet their economies are hugely influenced by such forces. The IMF and World Bank also had a very strong bargaining position in sub-Saharan Africa. Borrowing governments faced a disastrous external position
142
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
143
and most had few other sources of finance. The Fund and Bank were not only lenders in their own right but gatekeepers to all other aid since individual donor governments followed behind their accreditation, loans, and programs.
In 2002 both the Fund and Bank published evaluations as to why their loans, advice, and conditionality seem to have failed on the continent (World Bank 2002c, Independent Evaluation Office 2002). Scattered through their reports we find evidence of the core factors discussed in this book so far. The advice they offered to governments was not always right. Politics within borrowing countries, and a lack of sympathetic interlocutors and propitious political institutions, made their jobs difficult. And the preferences of their major shareholder sometimes eroded their bargaining power or interfered in other ways. This chapter teases out these factors and explains what drove the mission of the IMF and World Bank in sub-Saharan Africa.
Defining the Mission
In the early 1980s the IMF and World Bank plunged into a widespread debate about what kind of economic reform would work in Africa. Up until the late 1970s most developing countries had favored a statist approach to development, using economic planning, import-substitution-industrialization, price controls, credit rationing, state-owned enterprises, and government control of agricultural marketing (Van de Walle 2001, Lofchie 1994, Killick 1989, Waterbury 1999). In Africa the approach was reiterated in the Lagos Plan of Action set out by the Organization for African Unity in 1980. The concern of African leaders advancing the plan was to shift the continent away from its dependence on the export of basic raw materials, which "had made African economies highly susceptible to external developments" (Economic Commission for Africa 1980, Preamble). To this end, the plan focused on increasing Africa's self-reliance, promoting industrialization, and building up regional and subregional cooperation and integration.
The Lagos approach to development faced two severe challenges in the 1980s. First, it required resources and by the early 1980s most African countries were in economic crisis. Hit by the increase in oil prices in 1973 -74 as well as a slump in commodity prices, many had increased their borrowing in the 1970s so that by 1980 they faced a world economic downturn with a huge debt burden on their backs. There was a huge gap between the resources required for a renewed push toward industrialization and what was available. External donors were unlikely to come forward, in part because industrialized countries faced problems of inflation and a downturn in their own economies. Also skepticism had grown among governments in several industrialized countries about the statist approach to development. This was the second challenge faced by the Lagos approach.
The ideological climate in donor countries changed dramatically in the early 1980s. In the United States, the United Kingdom, and Germany, President Reagan, Prime Minister Thatcher, and Chancellor Kohl espoused a new antistate,
antigovernment, free-market rhetoric. Their hostility to government spending, industrial policy, and the welfare state soon spread into their view of aid. Suddenly the focus was on the failures of development policy in the 1970s (Bauer 1984, Tucker 1977). In the worst cases in Africa the state-owned, state-driven economic model had created and sustained a kleptocratic state. Across the continent as a whole, economic development seemed at the time to have failed. In the twenty years from i960 to 1980 the average annual rate of growth for Africa was about 4.8 percent, dropping to 2.9 percent for the least developed countries (Economic Commission for Africa 1980). At the time these figures were treated as disastrous, although in retrospect they look like a golden age of development on the continent. For example, over the period 1990-2001 Africa suffered a negative 0.2 percent average annual percentage decline in gross national income (World Bank 2003, chap. 1).
Against the background of scarce aid resources and skepticism about state-centered development, the IMF and the World Bank defined conditionality for Africa in the 19 80s. Two important choices underpinned the approach they took. First, they treated the primary cause of the 1980s crisis in sub-Saharan African countries as internal rather than external to each country. Eschewing African leaders' concerns about external shocks and constraints and how these might be mitigated (a central theme of the Lagos Plan), the institutions focused their attention on actions indebted governments needed to take. They chose to turn away radically from the state-centered industrialization model, which had prevailed until the end of the 1970s, and to focus on reducing the state in the hope that this would enhance the role of the private sector.
The IMF's analysis began first and foremost as a requirement that governments undertake stabilization policies reducing the budget deficit and stemming inflation. This was evident in the conditions attached to loans during the 1970s. The Fund's largest loan at the time was to Zambia, which took out its first standby arrangement with the IMF in 1973 when its border with Rhodesia was closed by that country's white-controlled minority government of Ian Smith, who was trying to suppress the majority struggle for control in that country. Among many other effects, the border closure severely disrupted Zambia's commercial transportation system, decimating the country's trade (Boughton 2001, 787). In 1976 and 1978 Zambia took out two further IMF loans, this time as its economy, heavily dependent on copper exports, was rocked by shifts in the world copper price. In each program the Fund required the Zambian government to take measures to reduce inflation and trim the deficit. In these terms Zambia succeeded and indeed this spurred further IMF offers of assistance (IMF External Evaluation 1998, 95; Callaghy 1990, 290; Boughton 2001, 291). However, a 50 percent reduction in the deficit in 1976-79 was essentially achieved by cutting recurrent and capital expenditure, and this policy soon caused a political backlash that wiped out the gains of reform (Callaghy 1990, 290).
What the IMF soon recognized was that stabilization measures worked only as a short-term measure. In and of itself stabilization did not enhance a country's capacity to repay the Fund. Indeed, even as Zambia met its core program condi-
144 THE GLOBALIZERS
tions, its debts mounted alarmingly, and by the early 1980s Zambia could no longer repay the IMF in a timely fashion (Boughton 2001, 787). For the IMF this spelled out the need for deeper measures of "structural adjustment," while critics argue that the case of Zambia in the 1970s underlined the extenuating impact of external factors—political, strategic, and economic.
The World Bank's approach was very similar to that of the IMF. In 1981 in a report named after its coordinator Elliot Berg, the Bank set out a tough critique of African governments for failing to provide incentives for agricultural growth, discouraging the private sector, poor public sector management and investment, and poor exchange rate and trade policies. The Berg report underlined the need for the countries of the region to "adjust" (World Bank 1981). Many have treated this as a statement of the "technical consensus" of the time. That consensus, however, was highly contested outside of Washington, D.C.
The World Bank's diagnosis of Africa's economic position in 1981 created a storm of controversy. As historians of the Bank later recorded, "Never before had the Bank been as publicly critical of such a large group of borrowers" (Kapur et al. 1997, 719). At the April 1982 meeting of the Economic Commission for Africa the report was declared to be "in fundamental contradiction with the political, economic and social aspirations of Africa" (Economic Commission for Africa 1982). Of course, one would expect this response from those most heavily criticized in the report. However, African countries were not alone in arguing that the multilateral institutions were taking insufficient account of factors beyond their control such as terms of trade, international economic conditions, and climatic and regional security problems. Nor were they the only ones to reject the consensus expressed in the Berg report. Strong critiques were also expressed at the 1982 meetings of the OECD Development Assistance Committee, by Arab/ OPEC countries, the European Economic Community, the United Nations Development Program, and by the United Nations Children's Fund (UN1CEF).
It was not strictly true to say that the IMF and World Bank were ignoring external factors or exogenous shocks. The Berg report recognized compounding factors beyond governments' economic policy such as the rise in oil prices, slow growth in industrialized countries, adverse climatic conditions, civil and military strife, and donor policies that supported and even encouraged inappropriate domestic strategies and institutions (Kapur et al. 1997, citing internal World Bank memoranda, 716-17). However, these factors were not emphasized in the prescriptions of the report. The importance of exogenous shocks was recognized in the IMF in 1963 when a Compensatory Financing Facility was established for countries affected by commodity price shifts. But such a facility could only ever provide short-term alleviation of the problem. Furthermore, limited funding and limited shareholder support rendered it very difficult to build on that approach in the 1980s.
The alternative to the tough stabilization approach taken by the IMF and World Bank was a more explicitly gradualist approach to reform as advocated by many development economists at the time and through the 1980s. The Economic Commission for Africa produced an African Alternative Framework as a
MISSION UNACCOMPLISHED IN AFRICA 145
conceptual starting point, although this did not include specific program designs (Economic Commission for Africa 1989). A more specific alternative was drawn up by an independent team of advisers to Uganda, sponsored by the Canadian International Development Research Centre, who advocated a program of economic stabilization and reform while retaining several key elements of the existing system of centralized planning and control (Uganda Economic Study Team 1987). At the core of gradualist alternatives was an attention to attenuating the vulnerability of African economies to world markets, exogenous economic shocks, and their reliance on exporting primary commodities—in the case of Uganda 90 percent of its export earnings came from global coffee markets (Lox-ley 1986).
In Tanzania in 1980-81, Robert McNamara arranged, with the agreement of the government of Tanzania, a three-person "wise-men's group" to attempt to find an accommodation between the IMF and Tanzania. After about a year's work by expatriate and local staff an alternative adjustment program was developed. It placed much greater emphasis on supply-side expansion rather than demand-side restraint, took much greater care with the income distributional implications of the required macroeconomic adjustments, and more gradual implementation. That said, in the end, both the IMF and Tanzania turned it down (McDonald and Sahle 2002).
Commodity exports lay at the core of the problem for many low-income developing economies. Their reliance on exporting commodities laid a vicious economic trap for three reasons. First, access to markets for commodities was (and still is) tightly controlled by industrialized countries who instead of opening their markets, operate tight discretionary policies. Second, the price and demand for primary commodities is in a long-term decline, which means that even if the volatility in world prices for commodities is alleviated, an alternative long-term strategy is still required. Finally, the possibilities for poor countries to pursue a longer-term strategy of moving away from raw commodities into semiprocessed and processed goods are blocked by industrialized countries who apply higher and higher barriers to these goods, effectively kicking away the development ladder from any countries trying to move up it: a 1988 United Nations Conference on Trade and Development (UNCTAD) study showed industrialized countries were applying twice the level of nontariff barriers to manufactured goods from developing countries compared to what they were applying to manufactured trade with each other (UNCTAD 1988, Chakravarthi 1989).
An alternative approach to Africa's crisis in the 1980s would recognize that all small, low-income economies were being buffeted by factors beyond their control, including shifts in terms of trade, in capital flows, and in world interest rates. Calling on small, low-income economies to adjust their own economies was like exhorting passengers in a lifeboat to paddle faster when their raft is in the middle of the Atlantic Ocean in a hurricane. No matter how impressive the efforts of the passengers, it is unlikely that their paddling will bring them to safety. Without a coherent approach to international conditions, it was clear to some economists that the "adjustment" programs being foisted on one country at a time
146 THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
would not work. The fallacy in the Fund and Bank's approach was, as Tony Kil-lick expressed in 1990, that adjustment "has come to be viewed primarily as something to be undertaken by deficit countries, with no equivalent pressure for action on surplus countries" (Killick 1989, 1990).
The problem for Fund or Bank staff, even if sympathetic to this approach, was twofold. A different approach required resources that did not seem to be available; and it countered the new ideological predilections of their most powerful shareholders. The support and influence of major shareholders in the Fund and Bank was a critical feature of the institutions' work in Africa in the early 1980s. Having extended loans to African countries throughout the 1960s and 1970s for a variety of geostrategic, postcolonial, economic, and domestic political reasons, the industrialized countries found themselves in relationships with aid-dependent states that could not repay even the most concessional loans. They turned to the IMF and World Bank for help and the institutions duly became more active in Africa.
Loans from the IMF and World Bank in the 1980s reflected new stringent constraints: a squeeze on resources as their industrialized country members responded to general economic downturn; and a new ideological imprimatur imposed very rapidly and forcefully in each institution when the Reagan administration took office (Boughton 2001, Kapur et al. 1997, interview with former U.S. IMF Executive Director Charles Dallara 1995). These constraints meant that it was easier for the IMF and World Bank to call on borrowers to tighten their belts than it was to extract more resources from industrialized country members, or indeed even their cooperation in macroeconomic coordination. Further reinforcing this approach was the fact that as the institutions became more involved in lending to Africa, their priority became to ensure that short-term repayment schedules were met and hence their own resources assured.
Implementing the Mission in the 1980s
By the end of the 1980s both the IMF and the World Bank had each staked significant material and intellectual resources in their work in Africa. They coordinated the region's relations with creditors, setting down the conditions debtors needed to meet in order to continue borrowing not just from the institutions themselves but from all donors. This position gave the international financial institutions significant bargaining power since sub-Saharan African countries became massively indebted throughout the decade. As illustrated below, the total debt of countries on the continent doubled between 1979 and 1985 and doubled again by the early 1990s. The value of their external debt as a share of gross national product (GNP) rose from around 25 percent in 1980 to more than 80 percent in 1994. As the IMF and World Bank became more involved in Africa, indebted countries began to use bilateral loans from individual donor agencies to repay the IMF and the World Bank who were necessarily their "preferred creditors." The result was both to create reverse flows of funds to the IMF (see
250
200
D 100
50
0 -I—■—1—■—I—■—1—■—1—■—1—■—1—■—1—■—1—■—1—i—1—■-1—■—I
1971 1973  1975 1977 1979   1981  1983 1985   1987 1989 1991   1993 1995
Figure 6.1 Sub-Saharan Africa, total debt, 1971-1995
figure 6.3 below) and to create strong political pressure for a change in the debt strategy.
The IMF was at the heart of the rescheduling of African debt. Any country needing to reschedule its debts to governments had first to conclude a deal with the IMF and then present itself to the "Paris Club" to negotiate a new repayment schedule. The Paris Club was (and still is) a forum in which creditor governments could gang up on individual debtor countries, demanding concessions defined by the IMF. The process has been described by participants as "a deliberately complex obstacle course, full of chicanery" (James 1996, 523) and as a necessarily "unpleasant affair" (Rieffel 1985, 15).
The reschedulings of the 1980s led to a vast increase in the debt burden of African countries. As debt-service payments were postponed outstanding debt was increased as debt-servicing obligations were added to the capital sum. While
90 -1--1
0 H—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—1—'—1—1—1—1—
1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 Figure 6.2 Total sub-Saharan Africa debts as percentage of GNP, 1971-1995
148
THE GLOBALIZEES
1971 1973   1975 1977 1979   1981   1983  1985   1987  1989   1991 1993 1995 Source: World Bank, World Debt Tables Figure 6.3 Sub-Saharan Africa, net flows of lending from international agencies, 1991 -1995
the IMF lay at the heart of the rescheduling process, the World Bank attempted to coordinate donors more generally through consultative group meetings for donors on a country-by-country basis, and on specific sectors. This would later be described as a particularly thankless task (Kapur et al. 1997, 739) but it included the creation in 1988 of the Special Program of Assistance for Africa, which was supposed to act as a focal point for coordinating the balance of payments portion of external assistance to sub-Saharan Africa with all major official donors.
A second role the IMF and World Bank played in respect of Africa was as lenders to debt-ridden African countries. But IMF creditor nations seriously limited the resources the institutions were willing to lend to Africa. In March 1986 the Structural Adjustment Facility (SAF) was created in the IMF with $3.2 billion to provide loans to the poorest countries (defined as those eligible for assistance from the Bank's International Development Association) with balance of payments difficulties. However, after strong U.S. opposition to new or easy money, the facility was meagerly funded from repayments on previous loans to the IMF's Trust Fund, and accompanied by particularly stringent conditionality (Boughton 2001, 646).
A second attempt to increase IMF lending was made in late 1987 when the Enhanced Structural Adjustment Facility (ESAF) was created, which had larger funding and offered a longer support framework (IMF 1988, 120). Again however, the United States was very reluctant to contribute to a new facility. The U.S. administration argued that it needed to concentrate on securing appropriations for the International Development Association (IDA) from Congress and refused to countenance selling some of the IMF's gold stock in order to finance the new facility (and U.S. approval was a sine qua non since such as sale required 85 per-
MISSION UNACCOMPLISHED IN AFRICA
149
cent of total voting power on the Board of the Fund). Eventually the United States made a very modest contribution of about 4 percent of the total grant commitments of the ESAF, leaving it to the IMF to establish a trust fund negotiated with ad hoc contributions from other countries, among whom Japan became by far the largest contributor.
The Enhanced Structural Adjustment Facility magnified the bargaining power of the IMF vis-a-vis Africa. It combined much-needed loans with particularly far ranging and high-level conditionality covering medium-term policy changes and short-term monetary and fiscal management. It was a prerequisite for loans from all other bilateral donors and other international funding programs. Fund conditions were thus "at the top of the hierarchy of donor conditionality" not because of the amount of resources that the Fund transferred but because the Fund was the lead coordinator (IMF External Evaluation 1998, 26).
The World Bank's role in the adjustment process was a complementary one to that of the IMF. The Bank's agenda was to reshape the role of the state and increase the role of markets and the private sector in African economies. In practice, however, the Bank soon found that its most feasible goals were to liberalize trade policies and to devalue overvalued exchange rates. These goals were much easier to achieve than deeper institutional reform within borrowing countries; furthermore liberalization and currency reform were prerequisites for ESAF lending. Within this framework, the World Bank increased its lending to Africa through its concessional arm in the 1980s (see figure 6.3 above) and increased its overall stake in the continent. During the early 1980s the Bank came to deploy the largest share of staff and budgetary resources to sub-Saharan Africa: a third of its regional staff resources, an increasing percentage of research time producing numerous special regional reports, and a plethora of special initiatives and programs launched (Kapur et al. 1997, 731-72). The Bank's concessional lending to Africa increased from less than a quarter of IDA from 1977 to 1979 to nearly half of IDA from 1988 to 1990 and was further increased by a Special Facility for Africa created in January 1985 based principally on contributions from France, Italy, the Netherlands, Germany, Japan, the United Kingdom, and a transfer from the Bank's net income (Kapur et al. 1997, 733). Much of the Bank's new lending was aimed at bringing about policy reforms within African borrowing members. Obviously, to be effective in this role more than monetary incentives were needed.
A third role, played by the IMF and World Bank in Africa in the 1980s, and by far the most contentious, was their attempt jointly to induce particular economic reforms. Their roles in coordinating assistance and lending to African countries gave them some bargaining power. However, as we saw in Mexico and Russia analyzed in earlier chapters of this book, the IMF and World Bank depend on sympathetic national policymakers to bring about policy change. Their interlocutors need to be interested in pursuing policies prescribed by the IMF and World Bank. Furthermore, they must be situated within institutional arrangements, which permit them to implement such measures. In Africa, the Fund and Bank attempted to shape economic policy in what they saw as a hostile political
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
151
context—they were somewhat weak and lost in an alien terrain. By contrast, critics saw them as tremendously powerful and arrogant—blind to the political needs and constraints of even the most well-intentioned policymakers. So how powerful were they?
The case of Senegal, a leading recipient of aid per capita in Africa from 1980 to 1987, illustrates the way politics, economics, and conditionality were intertwined. In the late 1970s economic crisis and a collapse in revenue from peanut exports on which Senegal depended brought reformer Abdou Diouf to power, first as prime minister and then as president (Mbodji 1991). In a first flurry of reform, Prime Minister Diouf launched an adjustment program with the World Bank supported by a $60 million structural adjustment loan approved on 18 December 1980 (World Bank 2004a) and a loan from the IMF's Extended Fund Facility. The IMF loan carried tough conditions requiring the government to cut its current account deficit by more than half, almost double net public savings by 1985, increase overall investment from i6percentin 1981 to i8percentin 1985, and achieve a 4 percent annual growth rate of GDP (World Bank 1989c; Ka and Van de Walle 1994, 309).
Both multilateral loans soon ran into difficulties. Bad weather affected exports and necessitated greater food imports, public debt was higher than originally admitted, and fiscal revenues actually declined from 1981 to 1984 (Ka and Van de Walle 1994, 311). The IMF loan was discontinued in January 1981 and replaced by a one-year standby arrangement. The World Bank canceled the second tranche of its structural adjustment loan in June 1983 because of noncompliance. For a government facing a sharp drop in the export price of peanuts and in the run-up to an election, it was increasingly difficult to sustain unpopular, contractionary reforms (Landell-Mills and Ngo 1991, 48; Mbodji 1991, 124-25). For some analysts this demonstrated that Diouf's political base was too narrow and technocratic with insufficient grounding in political parties, the political process, and electoral politics of Senegal—a constraint that soon began to change (Ka and Van de Walle 1994).
Immediately after the 1983 elections in Senegal, Diouf began to consolidate his political power. He eliminated the post of prime minister and limited the power of the National Assembly, strongly reinforcing his position as president. He also began to usher a new breed of technocrats into positions of authority across all ministries, enhancing and streamlining the capacity of the government to negotiate with external aid and lending agencies and to undertake new economic policies. Principal among the new breed of officials was Mamoudou Toure, a former IMF official who was to lead Senegal's structural adjustment effort from 1985.
By mid 1984 Senegal enjoyed three newly approved World Bank loans and a new IMF loan (IMF-Senegal 2004, World Bank 2004a). The government embarked on a program of economic reform that was approved by a World Bank-organized consultative group meeting in December 1984 (Landell-Mills and Ngo 1991 )• Subsequently, government expenditure was slashed, credit was controlled, and fiscal and current account deficits were both cut. As Senegal struggled with
an exchange rate fixed within the CFA franc zone and fluctuating against the dollar, it relied heavily in the period 1980-87 on foreign aid flows, which grew by about 18 percent per year, totaling about one fifth of Senegal's GDP. Much of this aid was coordinated with IMF and World Bank lending, further enhancing the potential leverage of the organizations.
By 1987 the president's reform agenda faced powerful opposition. Although Senegal's public finance situation had improved by this point, as World Bank economists have written, "The bulk of the program was achieved through containment of expenditures and reliance on extraordinary revenues generated from petroleum and rice imports" (Landell-Mills and Ngo 1991, 50). Austerity and cuts in government spending soon led to student boycotts, school closures, strikes, and union opposition to the government. In the aftermath of the 1988 election a state of emergency was called by the government as opponents of the government went on a rampage, and even once order had been restored, public demonstrations against reform continued. In the spring of 1989 riots took on an ethnic dimension as tensions with neighboring Mauritania spilled over in the streets of Dakar, forcing Mauritanian shopkeepers out.
The IMF and World Bank had succeeded in supporting the government to undertake stabilization, but longer-term reforms seemed to be slipping rapidly out of reach. The key technocrats in charge of structural adjustment—Mamoudou Toure and Cheikh Hamidou Kane—both left government in March 1988. Meanwhile, key structural adjustment policies were reversed in the face of the need to shore up political support and the government's lack of revenue. For example, the government had removed trade protective tariffs as a core part of a relatively successful new industrial policy (Boone 1991). By 1988, the policy was reversed because the government needed the revenues that tariffs produced and a small number of large, powerful businesses lobbied against it (Ka and Van de Walle 1994). While outside commentators accuse the IMF, World Bank, and donors of having imposed conditions that were too detailed and copious to be implemented and too seldom enforced (Ka and Van de Walle 1994, 329), Senegalese critics of structural adjustment in that country argue that it imposed unsustainable and unacceptable costs in health, sanitation, education, and literacy (Ndiaye 2003).
The IMF and World Bank had enjoyed some key preconditions for their success. They had incentives to offer Senegal and sympathetic interlocutors within the government with whom to work. They had shown themselves able to suspend, cancel, and defer loans when conditionality was not met. Yet after the first phase of stabilization and structural adjustment, further reform seemed virtually impossible. In retrospect, a survey of the assumptions underpinning reform and the evidence of impact makes this finding unsurprising.
During the 1980s, the IMF and World Bank justification for their programs in Senegal was that once the government undertook stabilization and a first phase of adjustment, it would achieve an annual growth rate of around 3.8 percent. This prediction was based on some extraordinary premises. For example, it was assumed that liberalization in agriculture and industry would produce an imme-
152
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
diate "supply response." In other words, farmers could and would rapidly increase production in response to greater market freedom. Similarly, industry would expand as privatization and liberalization attracted new credit and permitted new export sectors to flourish. Unsurprisingly (given all other cases of stabilization and structural adjustment) new policies would take much longer to produce change, and in Senegal there were technical and environmental factors along with wide fluctuations in world market prices of exports and low international peanut prices that prevented an expansion of food production and exports (Landell-Mills and Ngo 1991, 52.). In respect of industry, the establishment of new private sector activity and increased investment would require at the very least a more developed banking system. More generally, in the words of one scholar examining the evidence in the textile industry, "Senegal's Structural Adjustment programs offered no economically viable or politically acceptable means of restructuring the existing textile industry" (Boone 1991, 146). What does the failure of IMF and World Bank predictions tell us?
Hemmed in by their own resource constraints, yet desiring to play a role in a large number of countries across the world, the Fund and Bank had their own reasons for adopting policy prescriptions that cast an onus on developing country borrowers to adjust and to keep adjusting even in the absence of any evidence of economic growth. Conditionality had to be premised on a prediction of growth or the institutions would be explicitly trying to persuade patients to take medicine that was bad for them. At the same time neither the Fund nor the Bank could lend or catalyze lending that would directly fund growth-inducing investment. Furthermore, both the Fund and the Bank had to ensure that borrowers repaid them for previous loans and this put a stringent priority on stabilization.
In Senegal the harsh effects of adjustment were magnified by the country's inability to devalue its currency. As a member of the West African Monetary Union, Senegal was locked into the CFA franc zone arrangements. In essence this left the government with only two real instruments of adjustment: cutting government expenditure, and controlling exports and imports. The overvalued CFA franc made the latter extremely difficult.
Why did the IMF (and World Bank) accept and support Senegal's currency arrangement? In economic terms a permanently fixed and externally guaranteed exchange rate coupled with a supranational central bank should promote low inflation and encourage savings, investment, and growth. These benefits have been reviewed by several IMF and World Bank economists (Bhatia 1985, Devarajan and de Melo 1987, Elbadawi and Majd 1992.). Certainly low inflation was achieved within the franc zone and some scholars go further and positively correlate the currency arrangement with growth (Devarajan and de Melo 1987, Guillaumont et al. 1988). However, these studies also show that members did not benefit equally. Indeed, smaller countries such as Senegal did much worse than the larger members (Medhora 2000). Furthermore, the most obvious benefit of the currency arrangement—exchange rate stability-—may well have been illusory for Senegal since the real effective exchange rate was more unstable than the nominal effective exchange rate (de Macedo 1986). In economic terms there was (and
still is) genuine debate and disagreement as to the merits and demerits of Senegal's currency arrangement through the 1980s.
For the IMF and World Bank there was a further political reason underpinning support for Senegal's currency arrangement. This highlights an already-mentioned structural constraint at work facing the institutions. As one of France's former colonies and largest aid recipients, decisions about Senegal are led by France's preferences, with other powerful shareholders in the international institutions loath to intervene in respect of what they recognize as a special sphere of influence. Senegal's currency arrangements in the 1980s were part of France's CFA-franc zone encompassing the West African Monetary Union and a currency union among the central African states across which France guaranteed the convertibility of the common currency—the CFA franc (Medhora 1992). France vigorously opposed CFA franc devaluation and fought the Bank's and Fund's recommendations in this respect.
At a more general level the structural constraint of special spheres of influence has at times permitted major shareholders to pursue geostrategic goals in the context of the Cold War, to reinforce former colonial ties, or to bolster narrow economic interests, sometimes with catastrophic consequences for development. The extreme cases of this were the support provided to Nicaragua under Somoza, to the Philippines under Marcos, and to Zaire under Mobutu. In these cases the IMF and World Bank were not lending on technical economic or developmental grounds. Rather they were following the directions of their major shareholders, who permitted dictators to amass vast personal fortunes leaving behind a crippling debt burden which these impoverished countries have been forced subsequently to service (Kremer and Jayachandran 2003). That said, even in cases where the structural constraint was not a determining factor, the results of the IMF and World Bank's loans to Africa in the 1980s were extremely disappointing.
Conditionality and structural adjustment simply did not work in the 1980s. The large number of evaluations undertaken by the IMF and the World Bank themselves provides ample proof. Combing through their studies, which use a variety of methodologies, it is difficult to find any evidence that countries that entered into programs of structural adjustment with the IMF and World Bank did any better than countries that did not.1 Their critics argue that this was at least in part because their prescription was both wrong and in itself damaging.
Was the Prescription Wrong?
Independent analysts have argued that the Bank and Fund misdiagnosed the problem in African economies in the 1980s, making inappropriate forecasts for recovery and applying the wrong policy conditions. Far from facilitating nec-
1 See the excellent summary of the IMF work from Khan 1988 onward in Boughton 2001 and the ESAF review (IMF External Evaluation T998). The World Bank reviews include World Bank 1989a (Adjustment Lending); World Bank 19896 (Africa's Adjustment), and World Bank I989f (Sub-Saharan Africa).
154 THE GLOBALIZERS
essary adjustments and reform, the conditionality pushed by the international financial institutions drove countries into a vicious circle of stagnation and poverty.
What was needed for effective structural adjustment was a boost in low domestic savings so that countries could fund the investments necessary for structural change and growth. It required increased imports of raw materials and spare parts, which necessitated additional foreign exchange. And for structural adjustment to work, there needed to be political support and a sense of confidence and sustainability in policies undertaken. The converse, the "import compression" or "import strangulation" phenomenon, resulted in serious underutilization of existing capacity (due to the shortage of critical inputs), not just a limited ability to invest in order to expand it.
Instead, a narrow set of structural adjustment targets were imposed by the Fund and Bank in the context of increasingly onerous debt repayments schedules. A vicious cycle was created. Governments forced to meet enormous debt repayments obligations did not have foreign exchange resources to finance imports and without necessary imports, exports could not be increased, thus further reducing the capacity to purchase imports. Debt servicing also claimed domestic savings needed for investment and the maintenance of capital stock (Killick 1989 calculates some 20-25 percent of domestic savings being absorbed in debt repayments). The lack of investment was exacerbated by the uncertainties introduced by "debt overhang," which further discouraged investment and diverted governments away from longer-term problems of structural reform (Killick 1989, 1990). The result was to grind economies to a halt rather than to permit restructuring that would bring about growth.
The impact of IMF conditionality (on which World Bank lending hinged) was rigorously analyzed by an expert group commissioned by the board of the IMF in 1996. Their brief was to analyze the most far-ranging and high-level conditionality applied by the Fund—the ESAF, or Enhanced Structural Adjustment Facility. Designed to deliver concessional financing to low-income countries, ESAF required medium-term policy changes across the economy as well as shorter-term monetary and fiscal management targets. Although the amounts lent from ESAF were small, the associated conditionality was highly leveraged because compliance with the IMF's ESAF conditionality was a prerequisite for most aid and lending (in technical jargon, there is "nonreciprocal cross-conditionality"), particularly program assistance.
The goal of ESAF was to enhance investment and growth in low-income countries by channeling funds not to governments (who complied with conditionality) but mostly to a country's central bank to bolster reserves and thereby to promote confidence and greater investment in a country's economy. The key to the success of ESAF was enhancing investment.2
2 The goal of ESAF is set out in an IMF Board decision of 15 December 1987: "To promote in a balanced manner, both balance of payments viability and growth, through mobilization of domestic and external resources." There were only three situations in which ESAF loans actually provided budget support to governments (its best effect according to the External Evaluation): (1) where the extra reserves produce
MISSION UNACCOMPLISHED IN AFRICA 155
The External Evaluation of ESAF substantiated several powerful criticisms of IMF conditionality, some of which had previously also been leveled at the World Bank's structural adjustment lending.
A first problem with ESAF conditionality was that it simply did not seem to work. In many countries ESAF targets were not met. The review found that three-quarters of ESAF programs collapsed or were interrupted (IMF External Evaluation 1998, 32). Perhaps more seriously, the evaluation cited the evidence that where ESAF programs were being followed, they seemed to have no impact on investment flows (IMF External Evaluation 1998, Rodrik 1995). Finally, the review found that conditionality can be counterproductive in the sense that "one of the IMF's most valuable functions is the signal of credibility that it provides to private investors by approving a program. This signal becomes noisy as its recipients become aware that the design of approved programs may be faulty and that program interruptions are indeed common" (IMF External Evaluation 1998, 32).
A second set of problems detailed in the report might be summarized (although the experts did not summarize in this way) as the IMF's overly doctrinaire and short-term focus on reducing budget deficits in ESAF countries. This had several very negative effects. Three effects mentioned in the evaluation are particularly worth elaborating.
By putting such priority on balancing the budget, the Fund supported policies that had adverse long-term effects. For example, the Fund resisted lowering import tariffs, without analyzing the longer-term consequences on growth. In respect of privatization, the Fund was so keen to use the sale of assets to improve the budget deficit that it paid little heed to the way privatization was undertaken and the consequent longer run efficiency implications (or social implications). Among other cases, the evaluators cite the privatization of the public telephone company in Cote dTvoire, which resulted in a highly profitable monopoly charging much higher prices and setting back the development of access to infrastructure necessary for development. According to the ESAF report, the IMF simply did not adequately trade off the short-run fiscal benefits and the long-run social costs in such cases. It left this work to the World Bank to pursue in an entirely separate way, which did not work.
A further effect of the Fund's single-minded focus on balancing the budget was that the institution was too quick to assume an end to external aid to countries and this hindered the prospects for growth in the poststabilization phase. The Fund's emphasis has been to plan radically to reduce "aid dependency." Yet, as the ESAF report argues, poststabilization low-income countries need more money not less—so as to begin to invest. Instead the IMF's approach (exemplified by Uganda) was to force a reliance on trade and petroleum taxes that were very costly in terms of growth.
income which could be transferred to the governments; (2.) where the exchange rate appreciated, facilitating imports but penalizing exports; (3) in the CFA franc zone where ESAF loans were used entirely for budget support because France's support of the common currency made reserves unnecessary.
i56
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA 157
Compounding the damage of obsession with budgetary balance was the Fund's exaggeration of countries' fiscal deficits. The independent evaluation details the way the Fund included only pure grants in its calculations of fiscal balance, excluding the grant element in all other loans and treating them instead as commercial loans. In respect of IDA loans, the external evaluation team argued that some 70 percent should be treated as grant aid. By not treating IDA loans in this way, the IMF probably discouraged investment and pushed for too stringent a budget contraction (IMF External Evaluation 1998, 33).
A third set of problems with IMF conditionality lay in their design. Several countries were encouraged to undertake financial and exchange rate liberalization before they had stabilized their economies. The results were disastrous, not only making stabilization unnecessarily difficult but leading to broader economic collapse. As the ESAF report details, in Zimbabwe and Zambia the policy sequence led to economic crisis as the government lost fiscal control and interest rates rose, deterring any investment. In Zambia the share of public expenditure in GDP halved in a two-year period. In Zimbabwe deep cuts in health and education spending could have been avoided.
A related problem in conditionality—which affected both the IMF and World Bank—was a reliance on unleashing market effects to bring about structural reforms. In reality, specific structural reforms needed to complement liberalizing measures. For example, in Zambia privatization in agriculture needed to be complemented with early reforms to improve rural transport, extension, and storage. Without these, farmers simply "got stuck" (IMF External Evaluation 1998).
The ESAF report criticisms of the IMF were not new and some had already been applied to the World Bank. In the 1980s and early 1990s both the IMF and the World Bank had been criticized for basing their structural adjustment programs on overly optimistic projections (Helleiner 1987, Van der Hoeven and Van der Kraaij 1994). Key assumptions regarding the demand in industrial countries for primary commodities, terms of trade, private flows, and costs of servicing commercial debt all went in the opposite direction to that assumed in the Fund's programs. The Fund itself acknowledged this at a very early point (IMF 1982, 96) and similarly the World Bank would later conclude that the external economic environment "turned out to be substantially worse than was assumed at the start of the 1980s" making "adjustment slower and more difficult than initially expected" (World Bank 1989c). In 2002. a report of the IMF's Independent Evaluation Office detailed that the IMF had projected a mean export growth rate of 10.5 percent in countries making prolonged use of Fund resources. In reality the mean export growth rate in these countries was 7.4 percent. Similarly, real GDP growth in prolonged-use countries had been projected at 4.1 percent (again as a mean) whereas the actual growth rate was 3.5 percent (Independent Evaluation Office 2002).
Over-optimism in the IMF has frequently reflected a desire on the part of staff and the country to ensure that the board would accept a loan program. But this does not obviate the subsequent problem that over-optimistic projections have knock-on effects for program design, funding needs, and expectations about meeting conditions.
IMF missions in countries making prolonged use of IMF resources have typically been staffed by more junior personnel, those less willing and less able to challenge head office orthodoxy. In sub-Saharan Africa IMF missions were often more rigid, interacting with local officials on the details of programs that had already been broadly constructed in Washington (IMF External Evaluation 1998, Independent Evaluation Office 2002).
The content of World Bank conditionality had been criticized from a number of sources. As mentioned above, the Bank relied heavily on reforms in prices and market signals yet as one analysis of reform in rural Africa puts it, "reliance on markers may not necessarily ensure competitive processing or marketing of crops, where monopolies exist, or where historical factors explain oligopolistic tendencies" (Lele 1988, 204). The Bank had been overly optimistic about the prospects for traditional exports, particularly where several countries were being simultaneously advised to expand their exports of a particular commodity (Cassen 1994, Koester et al. 1987). The push for rapid privatization—as per the Washington consensus—was misguided (Adam 1994) and displayed too little regard for the ways it could be instrumentalized by politicians to consolidate power and direct profits toward favored groups and sectors as has been documented in the case of Cameroon (Van de Walle 1989, Konings 1989).
Too often, Bank conditionality was based on simplified but incorrect presumptions about both the situation on the ground and the likely impact of adjustment policies. For example, Bank staff in the 1980s worked on the premise that all over Africa government employees were overpaid and overemployed. This assumption reflected an idea popular at the Bank that the rapid and excessive expansion of government service in the immediate aftermath of independence had produced too many public sector employees (Goldsmith 1999). Critics argue that the Bank's perception was based on out-of-date and faulty evidence. The assumption of too many overpaid civil servants was contradicted by the Bank's own subsequent data and analysis (Lindauer, Meesook, and Suebsaen 1986; Dipak Muzumdar cited in Kapur et al. 1997, 737). In Anglophone African countries over the 1970s and 1980s civil service salaries had in fact collapsed by more than 80 percent of their real value (Robinson 1990). In many countries, an increase in the number of civil servants was accompanied by a dramatic decline in quality and remuneration (Van de Walle 2002). A better analysis of the problem, as the World Bank would later admit following consultations with African leaders, institutions, and donors, was not that Africa needed "less government" but that it needed "better government best pursued through technical assistance, institution-building, public expenditure reviews and the like" (Agarwala et al. 1994)-
In fashioning reforms, Bank and Fund officials discounted the realities of formal and informal financial markets and structures (Johnson 1994). In Nigeria for example, financial sector liberalization hugely increased corruption within the banking sector (Lewis and Stein 1997). Similarly, trade liberalization often resulted in an exploitation of new opportunities for fraud and rent-seeking behavior (Van de Walle 2002, Hibou 1996). Finally, the World Bank was heavily criticized for paying insufficient attention to areas where greater public invest-
158 THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA 159
ment was needed and most especially for inadequate efforts being made to protect the poor and public programs beneficial to them (Stewart 1994).
The critiques of the Fund and Bank prescriptions for African economic reform suggest that conditionality was too often aimed at narrow, measurable, short-term targets. There also seem to have been too few incentives for the organization and staff to achieve the longer-term aims of each institution. By contrast the incentives facing officials of each organization weighed heavily in favor of setting and achieving short-term targets. The result was too little attention paid to analyzing or taking into account the clear trade-offs arising between short-term and long-term goals.
Political pressures also pushed and shaped the conditionality of the IMF and World Bank. Most obviously, politically powerful members imposed a resource constraint. Each institution has to cut a robe to fit its available cloth and to some degree this explains the incentive on both the IMF and the World Bank to stay blind to the (obvious) fact that the debt position of most African borrowing members in the 1980s was unsustainable and that stabilization and adjustment was not producing the effects necessary to reverse that position .
This seeming blindness of the Fund and Bank to the failure of their approach to sub-Saharan Africa persisted even as the studies of the IMF and the World Bank themselves demonstrated that stabilization and adjustment failed to elicit positive investment effects (World Bank 1989a, 1989c, 1989^ Khan 1990; Killick 1990; Corbo and Rojas 1992; Elbadawi and Majd 1992; Bird 1995; Killick 1995). It permitted structural adjustment to be offered as the most convenient diagnosis and prescription for agencies needing to ensure that repayments were made in a timely fashion without catalyzing accelerating needs for further financial assistance.
The most politically mobilizing criticism of structural adjustment as pursued by both the IMF and the World Bank was that it had an unacceptably harsh impact on the poor and vulnerable in economies across Africa. In 1986, UNICEF launched a report on The State of the World's Children, calling for "adjustment with a human face" (Cornia et al. 1987). Study after study of the impact of adjustment in Africa and elsewhere pronounced adverse effects on the poor or at the very least highlighted how little attention had been paid to protecting the poor (Havnevik 1987, Bassett 1988, Hodges 1988, Helleiner 1987, Van der Hoeven and Van der Kraaij 1994). Public sector job retrenchments, job losses in other areas, cutbacks in food subsidies and other welfare provisions, as well as a loss in the quality of welfare provision, the effects of the general economic slowdown, and the lack of any political voice in the process of adjustment all exacted a high price on the poor in sub-Saharan Africa. An internal Bank memorandum reflecting on the impact of adjustment noted that "adjustment through further economic contraction is not a feasible alternative in a continent where per capita income levels are no higher than they were twenty years ago" (internal memorandum written in 1986, cited in Kapur et al. 1997, 732).
If the Bank and Fund had wanted to prioritize protecting the poor in the 1980s, they needed to build into stabilization and structural adjustment pro-
grams protections of five core aspects of the lives and opportunities of the poor: access to productive assets such as land; the quality and availability of extension services which increased the returns of the poor from the assets they did have; employment opportunities; access to education and health services; and supplementary resources, such as food subsidies (see UNICEF 1986 and the World Bank staff paper published in 1987: Demery and Addison 1987).
To plan economic adjustment with a human face would require the IMF and World Bank to work differently. To protect the poor, they would need to use local information about who was poor in any country and how they might be protected. Such information is unlikely to be held by Fund and Bank interlocutors in the Ministry of Finance or Central Bank. As the External Evaluation of ESAF noted: "It is not possible to devise, a priori, safety net interventions that will work across ESAF programs" in different countries. There is no substitute for detailed country-level work using socioeconomic survey data. However, to the extent that information was available, the Bank and Fund failed to use it to build safety nets into the design of ESAF programs (IMF External Evaluation 1998, 18). The IMF could make better use of the household poverty expertise of the World Bank in integrating projections of social impact into program design and monitoring the outcomes. At the very least the IMF and World Bank could work together to share information and their respective expertise. This had not happened in the 1980s.
The Bank and Fund Modify Their Approach
By the late 1980s a growing wave of criticism of structural adjustment and of the Bank and Fund washed over donor countries. Indeed some donor countries broke the link of bilateral aid to the region with IMF programs (James 1996, 525). Nongovernmental organizations (NGOs) in Canada, Scandinavia, and in the United States took up a vociferous role urging their governments to address the hardship being suffered by people in heavily indebted countries. In September 1987 American NGOs held a press conference delivering a letter calling for greater World Bank efforts on poverty and signed by 153 members of the U.S. Congress and 40 senators (Kapur et al. 1997, 368).
External pressures on the World Bank were leveraged by two big funding struggles. First the Bank had begun negotiations to replenish its concessional lending fund—called in Bank jargon IDA 8. This occurred in the context of the disastrous previous negotiations on IDA 7, in which the United States had cut funding. Second, the Bank was also negotiating a general capital increase, and in that context the whole adjustment with a human face issue was raised, with the U.S. Treasury committing to report to Congress on the Bank's involvement with NGOs, poverty programs, women's programs, micro-enterprises, and other issues (Kapur et al. 1997, 368).
The Bank and Fund considered some measures to alleviate the impact of adjustment on the poor. For example, the World Bank called for more external fi-
i6o
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
161
nancing to reduce the social costs of adjustment and gradually began to accept that compensatory services and public works projects might usefully ensure some protection for the poor. At the same time, staff began to focus on how government social services might be channeled more directly to the poor through targeted, needs-based benefits, funded by charges and user-fees paid by the better-off (Nelson 1995, 23). Similarly the IMF began an internal debate about how to monitor poverty impacts and protect programs from external shocks. This led to the creation of a new Compensatory and Contingency Financing Facility (CCFF) m 1988, which integrated the preexisting Compensatory Financing Facility with a new external contingency mechanism.
In practice, there was little modification to the overall approach of the Fund and Bank. Both institutions held to their existing paradigm, which assumed that stabilization and adjustment were prerequisites to alleviating poverty. Official Bank and Fund documents all robustly promulgated the view that adjustment was a necessary step toward poverty alleviation. For example, in the World Bank's 1992 review of structural adjustment we find several assertions that adjustment reduced the incidence of poverty, that the "distributional effects of well-designed policies often favour the poor," and that "adjustment is much better for the poor than non-adjustment" (World Bank 1992, 19-20). But the evidence does not back up these claims. One central assumption was that adjustment would improve the rural/urban terms of trade and therefore, because poverty in rural areas was far greater than in urban areas, reduce poverty overall. However, critics combing the actual available figures on poverty have found no hard evidence to back this claim. For example, one study found that poverty was high and increasing in most of sub-Saharan Africa in the 1980s in both adjusting and nonadjusting countries, belying the claim that adjustment was better for the poor (Stewart 1995, 138-70).
Why was it so difficult to modify Bank and Fund conditionality? Adjustment programs in sub-Saharan Africa in the 1980s demonstrated a high degree of uniformity and consistency on the part of each of the IMF and the World Bank in diagnosing the problems of these economies and in prescribing solutions (Killick et al. 1984).3 In outlining structural adjustment measures staff from both the Fund and Bank drew heavily on in-house theoretical propositions and predilections, which were not always supported by substantiating evidence. As discussed in chapter 2, there were strong incentives for staff members and their immediate interlocutors to use such a template. Originality in design would only increase the likelihood of a proposed loan being rejected. By contrast, using a template reduced the responsibility of the individuals writing the agreement for its content. Simply put, if the program turned out to be wrong but followed the institutional template, responsibility would fall more heavily on the institution than the authors of the program.
3Cf. an unpublished study by William Kingsmill which argues that there was some variation in the content of World Bank programs even though the broad thrust of the programs was the same (cited in Killick 1990, 16).
More recent evidence about the institutions' modus operandi in Africa points to another reason for a template approach to have dominated their programs on the continent. A recent IMF evaluation provides figures about how much staff time has been put into designing and monitoring programs in countries that have the longest ongoing programs with the IMF (into which category most of Africa falls). The evaluation published in 2002 highlights that far fewer staff resources are invested in the programs of these countries than in the more successful "temporary" users of IMF resources. It also shows the very high degree of turnover of staff and mission chiefs working with countries, with less than half of any mission team having been involved in the same country in the previous two years (Independent Evaluation Office 2002, Annex VI).
Similarly in the World Bank a recent report on low-income countries under stress (LICUS) notes that most of these countries "have typically not received much Senior Management attention . . . and little investment in economic and sector work, so that World Bank Group knowledge of these countries is often seriously deficient" (World Bank 2002c, vii). Under these time and staffing constraints, it is difficult to see how either the Fund or Bank might acquire expertise about the subtle and complex economic, political, and social implications of reform in any one country. Yet a final criticism emphasizes how vital such knowledge is.
The most difficult, irrefutable, and profoundly challenging critique for both the IMF and the World Bank is that their work in fostering economic reform has ignored or wished away political realities—in Africa just as much if not more as in other countries. To some degree the institutions have recognized this. To quote a working paper produced in the World Bank's evaluation department in 2000 "development constraints are structural and social, and cannot be overcome through economic stabilization and policy adjustment alone—they require a long-term and holistic vision of needs and solutions" (Branson and Hanna 2000).
A deeper critique of the institutions' policies is that political realities have turned rational policies into instruments of deeply damaging change, incurring perverse effects and hindering the prospects of positive development outcomes. The argument is not necessarily that the theory of structural adjustment is wrong. Indeed, many have questioned the extent to which stabilization and adjustment measures were ever actually implemented in Africa (Van de Walle 2001). One worldwide survey of 305 IMF programs from 1979 to 1993 found implementation failure in 53 percent of cases where failure was defined as a country not implementing 20 percent or more of the program's conditions (Killick 1996). In a different study of World Bank adjustment loans, the same author found that 75 percent of adjustment loans faced problems of noncompliance (Killick 1998).
The core political economy argument about reform in Africa is that IMF, World Bank, and other donors' conditionality made an unintended difference to politics rather than to economics. For example, Nicolas Van de Walle makes a powerful argument that conditional loans produced an entrenchment and reinforcement of patrimonial politics in Africa. He argues that two decades of economic reform have produced three key trends. First, there has been a cen-
l62 THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA 163
tralizacion of power as staffing, control of economic reform, and control of the i ]
rent-seeking opportunities have all converged on the office of the president. Sec- : ond, at the behest of the head of state, "reforms" have been used to direct benefits to specific groups in the economy, whether they are tribal, regional, or political. Third, the state has withdrawn from development, leaving nongovernmental organizations often run by the elite (who profit from them) to enter into the business of providing health, education, and so forth in an even less accountable, potentially more clientelistic way than governments (Van de Walle 2002).
Although both the IMF and the World Bank have recognized the need to take political circumstances into account, this recognition is very difficult to act on. How could the institutions acquire the kinds of knowledge required to take political, social, and institutional factors into account? At a fairly theoretical level, the IMF has probed a number of political science approaches to understanding reform feasibility and sustainability, the most recent of which is a review commissioned by the Independent Evaluation Office (Wimmer 2002). In the World Bank more practical attempts have been made to advance "reform readiness analysis" (World Bank 1999b). However, to cite World Bank researchers in the evaluation department: "This tool demands detailed knowledge of a proposed reform and of the political situation surrounding it, knowledge often unavailable to outsiders" (Branson and Hanna 2000, 6). As a result any change in either institution has been very slow and partial even though the institutions have long expounded the need for more sensitivity to political constraints. That said, other exigencies forced a change in strategy by the end of the 1980s.
A Growing Problem within the Institutions
By the early 1990s the strategy vis-a-vis the poorest, most indebted countries was not working. Debt levels and debt service payments were continuing to increase. In spite of debt rescheduling and reduction efforts, the debt stock of most ESAF-supported countries had doubled between 1985 and 1995, current account deficits had seen little reduction, and savings performance had been disappointing (IMF External Evaluation 1998, 23). Overall, to adapt the words of historian Harold James, the experience of the IMF and the World Bank with Africa had been "profoundly dispiriting, disappointing, and disillusioning" (James 1996, 543). So-called "structural adjustment with growth" was neither being consistently pursued, nor was it leading to the promised growth and recovery.
In 1996 the Fund and Bank responded to the failure of the debt strategy in Africa with the launch of the Heavily Indebted Poor Countries (HIPC) Initiative to provide exceptional assistance to heavily indebted poor countries. To quote the IMF: "For these countries, even full use of traditional mechanisms of rescheduling and debt reduction—together with continued provision of concessional financing and pursuit of sound economic policies—may not be sufficient to attain sustainable external debt levels within a reasonable period of time and without
additional external support" (IMF 200id). Eligible countries were defined as those facing an unsustainable debt burden beyond available debt-relief mechanisms, and an established track record of reform and sound policies through IMF- and World Bank-supported programs.
What brought about a new debt initiative? To some degree it was the NGOs and critics of structural adjustment who shamed and pressured the most powerful G-7 governments into action in the early and mid 1990s, however, they had an even stronger influence on the subsequent shift in policy. In the mid 1990s their calls coincided with pressing practical exigencies, which began to force a shift in the strategy toward the poorest countries.
Throughout the 1980s and early 1990s the G-7 blindly refused to accept that many of their loans (both bilateral and multilateral) to the poorest and most heavily indebted countries would never be repaid. Even in the face of figures showing an obviously unsustainable debt burden and mounting poverty and devastation, the G-7 continued to reaffirm a debt strategy that eschewed debt reduction and instead looked to indebted developing countries to reschedule their debt obligations while pursuing stringent adjustment measures (G-7/G-8 Research Archive at www.toronto.edu/g-7).
The only glimmer of a prospect for debt reduction was made in the Paris Club forum for government creditors. In 1988 these official creditors agreed to "Toronto terms" (followed up by successor "London terms," "Naples terms," "Lyon terms," and "Cologne terms") laying out a menu of options through which creditors could modestly reduce the debt service obligations of their poorest borrowers subject to stringent conditionality. The lack of greater action in respect of the world's poorest, most indebted countries stood in marked contrast to the more decisive actions taken in respect of the middle-income, transition, and emerging countries whose situations more directly impacted on the economies of powerful industrialized countries (Evans 1999, Serieux 2001, and see chapter two).
Even the very modest reduction in debt service achieved by the Paris Club through a lowering of the interest rate on rescheduled debt was vociferously opposed by several creditor governments, among whom a consensus had to be reached in the Paris Club. Three arguments dominated negotiations. These would recur throughout all debates on debt in the 1980s and 1990s.
A first argument against debt relief was that it was wrong to let countries off paying. The sanctity of contracts had to be upheld, and not to do so would invite other (e.g., middle-income debtors) not to repay. This was a principle particularly emphasized by Germany in negotiations.
A second argument against debt relief was that it would undermine IMF conditionality by heralding incentives for failure.
A third argument against a new approach to debt was that creditors could not afford the cost of debt reduction, particularly in the straightened circumstances of the fiscally contractionary 1980s.
Against the arguments for not changing the status quo were the simple facts that many debtors were not repaying their debts, nor meeting their IMF condi-
i66 THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA 167
tion into the realm of multilateral debt. The United Kingdom, the Netherlands, and the Nordic states in particular began arguing for a reduction in the debt owed by the poorest, most heavily indebted countries to the IMF and World Bank. Opposing their stance (predictably given previous rounds of discussion) were the United States and Germany.
The official positions of the Fund and Bank were extraordinarily conservative at this point. The IMF remained resolutely opposed to relief in respect of debts owed to it. In the World Bank, although some staff set up a taskforce that gave a realistic appraisal of the urgent need for a radical new approach—and indeed outlined one—they were blocked by senior management who opposed any change in the status quo (World Bank 1995).
Inching toward a New Strategy
When the Heavily Indebted Poor Countries (HIPC) Initiative was launched in 1996, for the first time major creditor countries agreed that debt owed to the multilateral institutions by the poorest countries would need to be reduced. Nonetheless, the initiative was a poor and unworkable compromise reached among creditor countries. To be eligible for relief, a highly indebted poor country had first to undertake three years of structural adjustment (the technical conditions of which were drafted by the Fund and Bank staff) and exhaust all traditional debt relief, at which point the country could be considered for relief, which would become available only after three more years of adjustment.
Like all its predecessor debt strategies, HIPC required heavily indebted countries successfully to undertake deep economic restructuring and long-term improvements in performance even as they continued to be hobbled by a crippling burden of debt. To quote the IMF, countries had to tackle "the whole range of factors currently limiting their growth performance, including poor infrastructure, the lack of effective policy making institutions, and governance problems" (IMF 1998b). As in the case of Senegal discussed above, to ask governments to do this with no resources and in the context of hostile politics catalyzed by stabilization, economic contraction, and increasing poverty was to ask the impossible.
Although unworkable, the HIPC initiative highlighted the three elements that were required for a change in the debt strategy. First, there had to be new ideas about how to reduce debt. Second, there had to be resources available to do it. Third, there had to be a revision of conditionality to fit the new strategy.
New ideas were provided by technical work done by economists within the Fund and Bank and other development agencies. In the words of a senior British official engaged in the negotiations at the time: "Many of the individuals in the institutions had come to the conclusion that debt reduction was needed. This was not the policy of some key shareholders and therefore not of the IMF and World Bank management, bur the staff played important roles behind the scenes in giving support to the UK and other initiatives" (Evans 1999, 274). The new technical work opened up the possibility of a policy change. Once the technical basis
for debt reduction was established, the Fund and Bank needed to resolve two further issues.
The new debt strategy needed to be financed. One proposal was to use a revaluation of a portion of the IMF's gold stocks. However, this was opposed by the United States and also by Germany whose opposition led protesters to lay mock gold bars outside the German embassy in London. The United States was reticent in spite of the U.S. Treasury secretary's repeated declarations of support for HIPC (U.S. Treasury 1998a, 1998b). It would not be until 2000 that the U.S. administration finally made its first ever request to the Congress to agree to a contribution toward multilateral debt relief.
A final necessary element for a new debt strategy was to rewrite conditionality. The existing approach was not working. However, its failure provoked two different responses among economists within the Bank and Fund (as well as among critics outside the organizations). Innovators argued that the institutions should reconceive conditionality to ensure greater "ownership" by borrowing countries. Traditionalists argued that the institutions simply had to be tougher in applying existing conditionality.
When UK chancellor Gordon Brown proposed a review of HIPC in September 1998, the traditionalists feared that this would lead to wrong criticism or dilution of the institutions' prescriptions. On this view, the continuing lack of growth in Africa was not due to any problem with the content of conditionality. Rather it was due to the failure of governments to restructure and provide incentives to the private sector. In the words of the World Bank in 1994, "Even among the strongest adjusters, no country has gone the full distance in restructuring its economy" (World Bank 1994, 1). The crisis in Africa was "predominantly a consequence of the failure of domestic policy and of the institutions the state helped to develop and sustain" (Sahn 1994, 366). The solution lay in tough love and the more stringent application of conditionality—as enshrined in the 1996 HIPC.
The weakness in the traditionalist approach was that it neatly split sound economic prescription (the work of the Fund and Bank) from practical implementation and sustainability (the duty of the borrowing state). It sidestepped the fact that regardless of who was to blame, Fund and Bank conditionality was simply not working in Africa. By contrast, innovators, particularly within the Bank, began to open up and consider what this failure suggested about both the content and the process of defining conditionality (World Bank 1996).
The New Strategy—A Revolutionized Washington Consensus?
In 1999 a new, enhanced HIPC was launched that would potentially affect some thirty-four African debtors.4 It was heralded as "deeper, broader and faster" than
4 It included the following highly indebted poorest countries in Africa: Angola, Benin, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Comoros, Congo, Cote d'lvoire, Democratic Re-
164 THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
165
tions. Furthermore, most creditors were being forced to extend new credit to them anyway. Two British participants in negotiations among the G-7 recall having laid out these arguments to other participants (Evans 1999, Lawson 1992.)
In practice even once the new Paris Club rescheduling framework was in place, it did little to address the seriousness of the debt crisis in the poorest, most heavily indebted countries. Under the Toronto terms creditors could reduce debt service by about a third (this proportion increased under subsequent terms) by choosing from a menu consisting of partial reduction, rescheduling, or rescheduling in a way that would reduce debt. The debt figures from eligible countries reveal that creditor actions under this agreement did little to moderate these countries' increasing debt burdens (see figures 6.1 and 6.2 above).
As countries faced mounting difficulties in meeting their obligations to the IMF and World Bank (which if they failed to meet, would cut off funding), bilateral creditors deferred repayments to themselves and essentially provided loans to countries so that they could repay the IMF and the World Bank. Development assistance budgets thus rapidly turned into funds being directed to the IMF and World Bank. Zambia offers a good example of what was happening. Between 1991 and 1993 Zambia made a net transfer to the IMF of $335 million in an effort to pay past debts—a sum which Oxfam points out was equivalent to total government spending on health and education (Oxfam 1996). Zambia was making its repayments mainly from foreign aid. With a 14 percent current account deficit, Zambia had zero debt servicing capacity. The $335 million earmarked for debt repayments in 1993 was more than half of the $550 million or so pledged to that country in development assistance over the same period.
The position of the IMF and World Bank was becoming less and less tenable in the 1990s. Loans conditional on stabilization and structural adjustment had not catalyzed new flows of finance nor growth nor better debt sustainability in heavily indebted countries. Throughout the 1980s new loans and conditionality had ensured that most poor debtor countries did not fall into arrears (as the IMF terms it) or nonaccrual status (as the World Bank describes a country more than 180 days in arrears on its payments). However, as the debts of the poorest countries mounted and were gradually becoming dominated by their debt to the Fund and Bank, the unsustainability of this debt ultimately risked eroding the institutions' own financial credibility.
As "preferred creditors," the IMF and World Bank enjoyed being first in line for repayment throughout the 1980s. Debtors had to repay them in full or face being cut off from all other debt financing, including trade credits. Until 1984 the IMF had only ever taken action on three cases of nonrepayment (Cuba, Egypt, and Cambodia), all of which were due to powerful political circumstances (Boughton 2001, chap. 16). The World Bank had never declared a member country in "nonaccrual status" until 1984.
in April 1984 the IMF faced three borrowers overdue by more than six months in their repayments and a further eight overdue by at least six weeks (Boughton 2001, 757-846). In that same year, the World Bank placed Nicaragua in "nonaccrual status," and by 1989, nine countries with loans comprising 4 percent of the Bank's portfolio were in nonacrrnal statue (Vdr.nr or -.1 ,„-0 _..»<_t/-.
zie 2002). Simply put, the institutions were beginning to face a debt crisis of their own.
In the IMF by 1990, eleven countries were in protracted arrears (payments in arrears for six months or more) to the tune of nearly 14 percent of outstanding Fund credits (Boughton 2001, 764). In total, at the end of December 1998, some forty heavily indebted poor countries had outstanding and disbursed debts of US$39,247 billion to the World Bank group (mainly IDA), of which $746 million were in arrears, and US$8,192 billion to the IMF of which US$1,660 billion were in arrears (IDA/IMF 1999). Clearly both institutions now urgently needed to reduce their nonperforming loans and thereby any risk to their own financial credibility.
The arrears crisis initially brought out different responses in the Fund and Bank. At first the IMF's response to arrears—at the behest of its powerful shareholders led by the United States—was to try to penalize countries in arrears through both financial and nonfinancial means in order to deter countries from not repaying. In 1985 the board raised the interest rate charged on outstanding obligations (the "rate of charge"), thus passing the full cost of arrears onto all borrowing countries making repayments. This was soon altered so as to pass the extra cost directly onto those countries in arrears through "special charges." However, in 1986 a new burden-sharing arrangement was agreed in response to the argument that the membership as a whole had approved arrangements that had subsequently gone wrong and therefore the whole membership should bear the cost. This paved the way for the Fund to work toward a more cooperative strategy to help the arrears countries to return to a more sustainable course (Boughton 2001, 812).
The World Bank at first muddled through in negotiations with countries in arrears. It was able to use its concessional lending arm—the IDA—to disburse new credits to severely indebted, low-income countries so as to ensure that they kept up with their IBRD loan repayments. In several cases economic decline made non-repaying borrowers eligible for such credits. In the short-term IDA became "a means to bail out the Bank" (Kapur et al. 1997, 1067). In 1991 the Bank announced a new "carrot and stick" arrears policy. Countries would be encouraged to keep repaying through a waiver of a part of interest charges on a year-by-year basis. However, the payment deadlines would be tighter as would the penalties attached to these deadlines (Kapur et al. 1997, 1064).
Throughout the 1990s the IMF and the World Bank fought to maintain their preferred creditor status vis-á-vis private and official creditors. They also fought a little with one another. The Bank worried that countries would use their loans to repay the Fund, and the Fund worried that the reverse would happen. These concerns led to detailed negotiations and agreements between the two institutions regarding their respective roles in supporting and receiving payments from borrowers in arrears (Kapur et al. 1997, 1071).
In summary, by the mid 1990s it was clear that the financial credibility of the IMF and the World Bank could be threatened by members' failure to repay. At the same time, a small number of powerful member countries, urged on by an
168 THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA 169
the existing H1PC. It was faster in the sense that most countries would get relief at an earlier point in the process. It was deeper because the amount of assistance would be determined by their actual debt position at that time (the so-called "decision point") whereas previously, debt relief was based on a debtor's projected position at a later time (the "completion point"). It was broader because condi-tionality would now be defined by a country's poverty reduction strategy anchored by a "broad-based participatory process." And where countries needed more time to develop such a strategy, the initial relief could be based on an interim strategy setting out their commitment to and plans for developing such a strategy (IMF 1999c, International Development Association 1999 and 2.002).
Change had to take place. By 1999 it had become clear that the 1996 HIPC was failing. By 1999, the debt of HIPC-eligible countries had quadrupled (from about $59 billion in 1980 to about $205 billion in 1999). On average, countries now faced debt burdens more than four times larger than their export earnings, and equivalent to more than their entire GDP (Birdsall and Williamson 2002). These facts mobilized debt-relief politics in industrialized countries.
Throughout the 1990s many NGOs had been monitoring the work of the IMF and World Bank and calling for more action on debt. However, in the late 1990s they became better organized and visible, and began to mobilize serious levels of public support on the issue. At the G-8 Summit in Birmingham in May 1998 an astonishing seventy thousand Jubilee 2000 supporters formed a Human Chain around Birmingham City Center urging the meeting of world leaders to forgive the debts of the world's poorest countries. As national and international media covered the event, even its organizers were amazed by the number of people, churches, charities, and civic organizations who had come out to demonstrate on the issues of debt and poverty.
Capitalizing on their success, by the end of 1998 a high-profile NGO campaign under the umbrella organization of Jubilee 2000 dominated the international media debate about debt. Although they addressed themselves to the IMF and World Bank, the real impact of their campaign was on voters within powerful creditor countries. "When a plea for debt relief becomes the common cause of a coalition that embraces both the Pope and the pop world, creditors should take notice," wrote the Financial Times in their leader of 17 February 1999. "The case for appropriate and radical action," the newspaper continued, "is compelling. Debt servicing imposes an impossible burden, particularly in Africa. Mozambique spends more on repaying debt than it spends on health: this is a country where one in five children die before the age of five. In Tanzania, payments consume more than the entire primary school budget" (Financial Times 17 February 1999, 21).
Creditor governments had rejected multilateral debt relief for both ideological and financial reasons. Ideologically, opposition focused on the adverse con-
public of the Congo, Ethiopia, Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Liberia, Madagascar, Malawi, Mali, Mauritania, Mozambique, Niger, Rwanda, Säo Tome and Principe, Senegal, Sierra Leone, Somalia, Sudan, Tanzania, Togo, Uganda, and Zambia.
sequences of weakening contractual obligations undertaken by borrowers and on the impossibility of ensuring that relief would be well used. This was memorably expressed by Senator Phil Gramm:
It is a pretty hard sell to talk about forgiving billions of dollars of debt to countries that borrowed money from us and, in too many cases, simply squandered or stole it, and now they do not want to repay it. They riot, they protest, they demand, but those things do not work in College Station, Texas. In College Station, Texas, when you borrow money from the bank or finance company or from your brother-in-law, you are expected to pay it back (Gramm to U.S. Senate 18 October 2000).
In the upper echelons of the IMF and World Bank, this argument had its attractions but it was being rapidly superseded by the simple fact that large amounts of debt on their books were now recognized as unrepayable—putting at risk the financial solidity of the institutions. And among the wider public within industrialized countries the argument for "no relief" was rapidly losing sway.
In industrialized countries, opponents to debt relief soon found that they were losing the argument to a groundswell of public opinion. In late 1998 the new center-left government in Germany turned around that country's traditional opposition to debt reduction, and let it be known that debt relief would be showcased at the G-8 Summit to be held in Cologne in 1999 (Elliott 1998). Tellingly, no fewer than five of the eight countries attending the Summit produced debt relief proposals (Chote 1999).
In the United States, as Senator Biden would later declare to the Senate, the campaign to reduce debt drew together right-wing Christians such as Reverend Pat Robertson and left-wing legislators such as Maxine Waters (Address in the U.S. Senate, 12 Oct 2000). In a more complaining tone in his closing remarks to the 106th Senate, Senator Phil Gramm declared, "I had a group of holy people come to my office the other day to lobby for this debt forgiveness. I do not think since Constantine the Great called his ecumenical council in Nicaea has there been a larger gathering of holy people in one place than the people who came to see me about supporting debt forgiveness" (U.S. Senate, 18 October 2000). The impact of public pressure such as those faced by Senator Gramm helped to unblock a new approach.
Public pressure on the debt strategy focused on two features that would shape a new approach. The first was poverty reduction. The original HIPC had mainly left poverty alleviation to other agencies and processes, requiring only that the international financial institutions monitor progress toward the OECD 1996 Development Assistance Committee goals of poverty reduction and social development. The old approach reflected the view that poverty alleviation was best "supported by the international community through various instruments, including lending, policy dialogue, and social expenditure reviews" rather than through explicit IMF and World Bank programs (IMF 1998b). The enhanced debt relief initiative changed this.
In 1999, the IMF joined the World Bank in voicing a new focus on poverty,
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA 171
recognizing "increasing evidence that entrenched poverty and severe inequality in economic opportunities and asset endowments can themselves be impediments to growth" (IMF 1999c). In large part the new focus on poverty was a direct response to the concerns of people and governments within industrialized countries. Two decades of indebtedness in Africa had exacerbated poverty within the poorest countries. Any new debt relief initiative would have to demonstrate that it was attempting to remedy this.
The second new element of the enhanced strategy was its explicit commitment to let countries and their peoples "take the lead." Bank and Fund conditionality was to be based on strategies developed locally with the active participation of civil society and NGOs as well as donors and international institutions. The key concepts driving the new process would be "participation" and "ownership." "Participation" captured a new "on-the-ground" approach to working with local communities and nongovernmental organization. This mirrored what was happening in Washington, D.C., as more NGOs became involved in input, advocacy, and the monitoring of results of HIPC.
"Ownership" captured a rethinking being undertaken within the Bank and Fund as to how each might improve the commitment of governments to reform and thereby the effectiveness of conditionality. It dovetailed with a public anxiety about the institutions imposing harsh terms on governments. The new emphasis on ownership permitted the institutions to respond both to critics of their harshness and critics of their ineffectiveness.
The role of NGOs in influencing the debt strategy was significant. From 1998 onward large well-organized NGOs (mostly from industrialized countries) successfully carved out a place for themselves in the official review of HIPC. Their campaign for debt reduction targeted both the IMF and the World Bank, accusing the institutions of failing to listen or heed the views of people within the most indebted countries. Many NGOs presented themselves as proxies for otherwise marginalized people, at least in negotiations at the international level. In this role, they played a very active part in the review of HIPC. To cite a joint report of the Fund and Bank:
From the very beginning the HIPC process has benefited from consultation with civil society in all parts of the world. . . . Recently, a number of organizations have produced detailed and insightful analyses on the HIPC Initiative and debt relief more broadly. We want to build on this existing consultative process as we carry out this year's comprehensive review. (IMF 1999c)
The Fund and Bank proceeded to formalize the input of NGOs in the 1999 review, creating a broad-based consultative exercise managed at the headquarters in Washington. This shaped the rhetoric of the enhanced debt strategy. It also further established a pattern of engagement with NGOs—whether at headquarters in Washington or in the field—in formulating poverty reduction strategies.
For the Bank and Fund as institutions (i.e., for their management and staff), the new participatory approach had direct political benefits. Increasing openness
to NGOs both softened critiques of the organizations and enhanced the leverage of management over the creditor governments in which most of the powerful, well organized, and mobilized NGOs were based. Robert Wade proposes to us that the new alliance "may be understood, in part, as an attempt to build a broad constituency of support precisely so that the Bank is not completely beholden to the U.S. government and a narrow range of US 'gotcha' NGOs" (Wade 2001). However, as Wade goes on to note, much of the new alliance and broadened Bank agenda was "largely rhetorical and aimed at satisfying external Part I [i.e. creditor] entities rather than intended to have any effect on the goods and services delivered to the borrowers in return for loans" (Wade 2001).
A second effect of the new alliance with NGOs is that it boosted the resources available for the institutions to use in implementing the new debt strategy, as the World Bank heralds on its website:
In 1999, the Jubilee 2000 global coalition and hundreds of other interested NGOs participated with the Bank in a six-month review of the HIPC Initiative. The contributions to the HIPC review from civil society directly resulted in the doubling of debt relief pledged by international creditors, accelerated implementation, and the linking of relief to poverty reduction strategies, (www.worldbank.org)
At the global level the new participation of NGOs opened up the debate about debt relief and mobilized political support in industrialized countries, forcing governments and the international financial institutions to engage with a wider audience and a wider range of interlocutors.
At the national level in borrowing countries, the participation by NGOs provoked a new debate about who participates and why in IMF and World Bank consultations and public outreach exercises. Critics argue that a very selective process of engagement has emerged that privileges some groups over others and too often bypasses the broader "civil society" (Scholte 2001). A second criticism is that too often the new participation excludes or marginalizes existing political institutions such as political parties and parliaments (Eggers, Florini, and Woods 2004). The growing antagonism between Southern and Northern NGOs, and indeed the intense suspicion on the part of Southern governments of Northern NGOs, exacerbated these problems.
The other key element of the enhanced HIPC was "ownership," a concept that emerged from several in-house studies and external evaluations undertaken in both the IMF and the World Bank that detail the degree to which traditional conditionality was not working. "Improving ownership" was seen as a way to ensure greater national commitment for policies and increasing public accountability through policy debate and better monitoring of expenditure and outcomes.
The clearest expression of the new ownership approach lies in the 1997-98 World Bank's Comprehensive Development Framework (CDF), although there had previously been significant informal discussion about ownership, both within the Bank and even more so within the DAC and donor community. For
172
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
173
example, the "radical" initiative on aid relationships in Tanzania, which centered on ownership, and in which the Bank was directly involved, preceded the CDF. Within the IMF, rather a latecomer, the ownership issue received a significant impetus from the external ESAF review. Drawing on earlier research into the Bank's relations with its borrowers (World Bank 1989^ World Bank 1996b), the CDF aimed to "put the country in the driver's seat, both 'owning' and directing the policy agenda, with the Bank and the country's other partners each defining their support in their respective business plans." The key was to find "mechanisms to bring people together and build consensus." The Bank's role would be to support the process, which would forge stronger partnerships allowing for strategic selectivity, a reduction of wasteful competition, and an emphasis on the achievement of concrete results (World Bank 2001b).
The rhetoric of ownership is powerful. Much more difficult has been practical clarification of how to operationalize and muster strong staff support for the new approach. In its early renditions, increased ownership was frequently understood by staff to mean that they should better explain conditionality and its rationale to local groups (Piciotto and Weaving 1994). In respect of its poorest borrowers, the Bank's own findings highlight serious difficulties in attempting to alter its policies and its modus operandi (World Bank 2002c). In spite of these problems, the most recent study of the Bank's evaluation department provides empirical evidence of why greater ownership will lead to more effective development assistance (Operations Evaluation Department 2003).
For the IMF the new participatory approach to negotiating conditions for debt relief posed a yet more substantial challenge. The institution voiced its desire "to be ready to assess new approaches and to recognize and support a healthy process of experimentation and innovation. Fund staff will be open to considering alternative adjustment paths, taking into account their impact on the poor" (IMF 1999c). However, this would always be difficult for an organization used to monitoring concrete specific actions through intensive internal review processes. The new framework called for a more fluid approach and one that required the institution to balance several competing "key features." The Fund would have to narrow its approach to ensure more selective structural conditionality and more emphasis on measures to improve public resource management and accountability. At the same time, however, the framework calls on the Fund to broaden its approach so as to embed poverty reduction in overall strategies, ensure budgets are pro-poor, and undertake social impact analysis. The IMF, with its emphasis on fast crisis response, is bound to have a great deal more trouble with participatory processes (Boughton and Mourmouras 2002).
At a more general level, ownership poses a larger challenge to both the IMF and the Bank, requiring them to undertake a degree of self-denial—to facilitate specific outcomes but at the same time to abjure from imposing conditionality. To sharpen up their expertise but to hold it back in preference for the new broader, country-based, participatory approach to designing policies. Fund staff speaking ro assessors of the new initiative voiced fears that in the end this tradeoff is one of "ownership" versus "quality" (Adam and Bevan 2001, 4). In practice, however, the result has been to change little.
The Impact of the New Strategy
In 2003 the Jubilee Research group described the progress of HIPC as "glacial," referring to the fact that only eight countries (rather than a projected twenty-one) had reached completion point and therefore benefited from stock-of-debt reduction (Jubilee 2003). According to the IMF and the World Bank, two obstacles have rendered progress very slow. First, there has been a lack of adequate funding for the initiative. Creditor governments have failed to convert support for the initiative into "firm commitments" and to provide adequate "topping up" funding to increase debt relief, especially since so many HIPC countries are suffering from the global economic downturn and a fall in commodity prices (International Development Association 2002). Furthermore, many countries need further grants, particularly the most vulnerable countries, some of whom are too far in arrears and too conflict-affected even to qualify for interim assistance.
Critics argue that the Fund and Bank could themselves put more of their resources into debt relief—net income from the Bank and gold sales from the Fund (Jubilee 2003). The Bank and Fund respond that this would weaken their capacity to provide financial support to low-income countries, including the HIPCs (International Development Association 2002). It would also cast some of the costs of debt relief onto their borrowers.
The lack of committed funding to debt relief means—to cite a report by the Bank's operations evaluations department—that the Fund and Bank cannot hope to improve both debt sustainability and poverty alleviation in the most heavily indebted countries. Debt sustainability requires redesigning the ways resources are delivered to countries—something the institutions can do. Poverty reduction requires increasing the resources delivered to poor countries—for which the Fund and Bank must rely on other aid flows (Operations Evaluation Department 2003, 57). On this logic, the enhanced debt strategy is a less radical revision than it seems. Constrained by resources, the Bank and Fund must still rely heavily on indebted countries "adjusting" so as better to be able to service their debts and invest in poverty alleviation.
A second reason for the slow progress on debt relief lies in the conditionality attached to HIPC. The new conditionality requires countries to produce a poverty reduction strategy in a consultative and participatory way. Where this is difficult, countries can use an interim procedure to access relief. However, the old conditionalities must also still be met. In the absence of much greater funding, adjustment is still the mainstay of the debt strategy. And this continues to be difficult and contentious.
To qualify for debt relief, countries are required to meet macroeconomic targets in respect of inflation, fiscal balance, and their external position, as well as to implement structural and sectoral reforms (IMF zooie). Critics accuse the IMF of applying these targets to crisis-ridden countries with far too great a vigor. They argue that the Fund is forcing countries off-track for debt relief by applying overly conservative fiscal targets and focusing too much on privatization in the conditionality attached to HIPC. Some fourteen of the nineteen program countries in the HIPC process have fallen "off track" at least once (Jubilee 2003, 3).
174
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
175
To some degree the institutions' own review tells a similar rather pessimistic story. Summarizing the experience of HIPC-eligible countries, the staff of the IMF and World Bank detail slippages in fiscal policy and delays in structural reforms and privatization measures to explain why several countries, such as Ghana, Malawi, the Central African Republic, Senegal, Rwanda, and Tanzania, have been slow in qualifying (IMF/World Bank 2002).
A fundamental question here is whether the conditionality—for these countries at this time—is right. Or, more pointedly, what explains the priorities reflected in the ongoing conditionality? Does it reflect the needs and economic conditions of each country? Or is it being overly shaped by the resource constraints of the institutions?
The enhanced HIPC promised a new process for fashioning conditionality_
one that focused on participation and ownership. Yet, the evidence suggests that in practice little has changed. Take two countries discussed earlier in this chapter—Senegal and Zambia. Each has run into difficulties in its path toward eligibility for debt relief. In Senegal the IMF and World Bank write that external debt stock indicators have worsened significantly as a result of lower export projections than anticipated and that progress in economic reforms has been slow (International Development Association 2002, 65). Particular sticking points in Senegal's performance have been in respect of privatizing the peanut industry and privatizing and deregulating electricity.
About Zambia the institutions are more sanguine: "All structural performance criteria and benchmarks were met" they announced in 2002 (International Development Association 2002, 69), heralding the government's commitment to a speedy privatization of the Zambia National Commercial Bank (ZNCB) in a later press release (IMF 2002). However, low world prices for copper and the closure of mines in Zambia have meant lower export earnings in Zambia and therefore an ongoing unsustainable burden of debt.
Instantly recognizable in the cases of Senegal and Zambia are two factors that have marked Africa's last two decades: the devastating impact of external factors and in particular lower world prices and markets in commodities; and the unabated continuation of structural conditionalities whose urgency, sequencing, and efficacy in countries facing extreme economic (and often political) crisis are at the very least a matter of debate. The argument is not that in the abstract structural adjustment is wrong. It is that countries suffering extreme political and social stress do not enjoy the conditions necessary for all such reforms to be beneficial. These pre-conditions include the core infrastructure, political capital, and transparent, effective institutions, which are necessary in order to proceed with wholesale programs of privatization and liberalization.
In conversations with country experts in both the IMF and the World Bank, it is clear that many staff in both institutions know this. There has doubtless been a great deal more rhetoric about ownership than actual change in practices. However, there have been some significant changes by donors at the country level where an increased proportion of aid from some countries is taking the form of budget support or sectoral programs. For example, the UK Department for In-
ternational Development reports that some 15 percent of its bilateral aid program is being disbursed in budget support and other forms of program aid (DFID 2004a, 117-18 and 162-63). Furthermore, at least in respect to some countries, there is an improvement in donor coordination (Renzio 2004, OECD/DAC 2004). However, there are larger institutional imperatives, which prevent the core approach from changing.
What Is Driving the IMF and World Bank?
Three obvious tensions arise out of the way the Bank and Fund might adapt to better achieve their mission in Africa. Each takes us back to the institutions' sources of power and autonomy explored in the first two chapters of this book. In the first place, each institution has long relied on its specialist "expertise" as a rationale for conditionality and as a source of influence in persuading governments to undertake reform. In giving advice, experts in each institution face very powerful incentives not to deviate from standardized prescriptions. The more standard the template of conditionality they negotiate, the less any individual staff member will have to justify his or her actions. It is a risk-averse strategy for staff whose time is short and whose expertise is more theoretical than empirical.
Equally, for each institution a template makes life easier. It makes it easier to claim that all borrowers are being treated equally. Furthermore, the closer the institution's advice reflects a consensus among professional economists, the easier it will be for the institution to justify its prescriptions in terms of specialist expertise. All of this will be threatened if ownership and participation were genuinely to take hold in the modus operandi of each institution.
A genuine local ownership of policies, resulting from broad local participation in Africa, would likely produce more complex and diverse policy packages. The Bank has recently noted the tension "between the Bank's country focus and its implementation of more comprehensive and rigorous operational standards" (World Bank 2002b, vi). Taking steps away from their professionalized economic expertise takes each institution into uncharted territory—not just as economists but also as institutions with norms, practices, and structures, which have developed because they are useful to the institutions.
A second tension in the new mission in Africa takes us back to the financial structure of the Fund and Bank and the nature of lending they undertake. Institutional reform and poverty alleviation take considerably longer to achieve than the kinds of macroeconomic and microeconomic structural adjustment measures that have been promulgated in conditionality to date. The ESAF review discussed above demonstrated that the institutions tend to focus their energy and hard conditionality on short-term monitorable targets rather than broader long-term goals.
A new broader mission in Africa, which strengthens the processes of decisionmaking rather than just focusing on targets, will be much more difficult to implement, measure, and monitor. Institutional change, degrees of participation and
i76
THE GLOBALIZERS
MISSION UNACCOMPLISHED IN AFRICA
177
ownership, and poverty alleviation are all multifaceted and complex goals. A recent study of the "new approach" to IMF conditionality found that the new poverty reduction strategy papers "have tended to be rather general, weakly prioritised and of variable quality" (Killick 2002.).
The essential question is whether short- and medium-term conditional financing instruments can achieve the longer-term goals of the IMF and the World Bank. If they cannot, the institutions need to retool or to delegate to other institutions better placed to undertake the longer-term mission.
A third and final tension in the IMF and World Bank's mission in Africa is that between "borrower ownership" and "donor control." The aid community is now discussing longer-term and more concessional lending or grants that will not necessarily be channeled through the Fund or Bank—such as the Global Fund for AIDS/HIV Tuberculosis and Malaria and the new Millennium Challenge Account. The contradiction arising from these new funds is that they offer yet another donor-controlled modality of development assistance. Just as some economists are drawing the lesson from debt relief that aid works best where it is fungible (Birdsall and Williamson 2002), the new "global fund" model of governance proposes assistance that is less fungible and more highly directed and controlled by international donors. Indeed, policymakers within Africa have argued that this new approach further hollows out any possibility of genuine participation or ownership in the budgetary process within countries (Tumusiime-Mutebile 2002).
The Bank and Fund risk being caught somewhere between their new mission and a new model of financing development. Their new mission attempts to inject enhanced participation and ownership into their work and necessarily devolves (or will in the future devolve) responsibility and control to borrowers. Yet the new model of development funding injects greater donor control or at the very least greater donor scrutiny of "concrete, provable results."
The problem for the Fund and Bank is this. They are under increasing pressure to demonstrate results. This is captured in the ferocious critiques of the institutions mounted both by the U.S. Congress through a commission it appointed in 2000 (Meltzer Commission 2000) and by the remarks of the then-incoming U.S. secretary of treasury Paul O'Neill (Blustein 2001b). Yet their mission is being rewritten to set goals that are, by definition, more difficult to evaluate and to prove successful.
A final comment needs to be made about the relationship between the IMF and the World Bank. The new debt relief initiative has brought the Fund and Bank into greater areas of potential conflict with one another. As had occurred before, the respective roles of the IMF and the World Bank had to be very carefully negotiated and elaborated (a proposal for them to work together on joint programs had failed to gain support). The documentation about HIPC subtly reveals the tension between the institutions. To quote the Fund's definition of their relationship:
I he staffs of the Fund and Bank will need to cooperate closely and seek to present the authorities with a coherent overall view, focusing on their traditional ar-
eas of expertise in line with past agreements between the two institutions. ... In order to fulfil their role in assisting in preparation of the macroeconomic strategy, the Fund staff will need to be able to interpret the work of the Bank and other institutions. However, consistent with the views of the Board, the Fund staff will not attempt to supplement or substitute for Bank work in poverty analysis or the development of social policies. (IMF 1999c)
In substance the World Bank has been assigned to take the lead in advising countries on the design, cost assessment, and monitoring of poverty reduction strategies; the design of sectoral strategies and structural reforms such as privatization and regulatory reform; the strengthening of institutions including public expenditure reviews; and the provision of social safety nets.
The IMF's role is to lead in its traditional areas of responsibility such as in promoting prudent macroeconomic policies, structural reforms in areas such as exchange rate and tax policy, and issues related to fiscal management, budget execution, fiscal transparency, and tax and customs administration. The division of labor leaves many areas of overlap, mostly notably in governance issues such as "establishing an environment conducive to private sector growth," trade liberalization, and financial sector development.
Beneath the polite language of collaboration, liaison between the IMF and the World Bank was found to be "seriously deficient" in the external evaluation of ESAF (IMF External Evaluation 1998, 34). The expert reviewers found that as the Fund broadened its agenda into areas of the World Bank's expertise, it was still not working closely with the Bank on the ground. While "expressions of goodwill" abound, no attempt had been made to undertake the "major institutional change" necessary if technical advice were to be improved. Indeed, they found no evidence of even the minimal requisite formalization of procedures for cross-institutional teamwork and decision rules (IMF External Evaluation 1998, 34).
The IMF and World Bank have found it extremely difficult to facilitate successful economic growth, development, and policy reform in line with their conditionality in Africa. This is puzzling from the outside because the institutions look very powerful vis-a-vis Africa. They have leverage due to their resources and knowledge. Their borrowers in Africa are among the least likely to have access to alternative sources of finance. Powerful shareholders are less likely than elsewhere to override the authority of the institutions to meet their own geopolitical goals. In sum, Africa is the one region in which we might expect the staff of the institutions to act relatively independently of the ideologies and preferences of the most powerful member states. Under these conditions, the technical expertise and research of the IMF and the World Bank might well be expected to come to the fore.
The experience of sub-Saharan Africa highlights weaknesses within the international financial institutions. Those weaknesses cannot all be attributed to political pressures from outside. In Africa countries seem to have been poorly served by the research and lending practices of the Fund and Bank. The most recent evaluations undertaken by the World Bank staff and the IMF's independent evalua-
178
THE GLOBALIZERS
tion office highlight the shortcomings of their respective missions to date (Independent Evaluation Office 2002, World Bank 2002c). Too often specific policy advice has been fashioned according to easy blueprints rather than hard research—ideological presumptions rather than tested theories. Certainly the institutions have had limited resources with which to fashion policies for poor, indebted countries. But even within those constraints, it would seem that they economized on staff time in designing programs for their most needy borrowers, they were very slow to seize and shape the issue of debt relief in respect to sums owed to themselves, and most poignantly of all even after two decades of engagement their main borrowers in sub-Saharan Africa seem no closer to the promise of economic growth, and are still highly indebted to the IMF and World Bank. This experience, set alongside that of emerging market economies such as Mexico, and transition countries such as Russia, discussed in earlier chapters, necessitates consideration of how each institution could be reformed.
Chapter 7
REFORMING THE IMF AND WORLD BANK
The IMF and the World Bank are extraordinary international institutions. Their money, remit, and expertise endow them with a power about which other international organizations can only dream. The fact that they are automatically funded and earn income from their lending and investments gives them a degree of independence unrivaled by other institutions. They exist to foster global monetary cooperation and financial stability, facilitate international trade, promote high employment and sustainable economic growth, reduce poverty, and improve the living standards of people in the developing world. They can pursue these aims through lending and conditionality as well as through research. They employ the largest number of applied economists of any institution in the world, aggregating an awe-inspiring bank of economic data and applied research.
Critics assail the IMF and World Bank from all sides. They accuse the institutions of being U.S. dominated. They charge them with peddling poor quality economic advice. Each institution is indicted for supporting or even promoting corrupt and oppressive regimes. This chapter reviews and builds on the evidence assembled in this book. It argues that the work of the Fund and Bank is affected by the preferences of their most powerful members, by their own bureaucratic motives, and by politics within countries with whom they work. They are not purely technical institutions. Economists are naive to assume that the Fund and Bank would work better if they were insulated from the hurly burly of politics. The agencies cannot escape the political decisions and debates they sit within. What would make them more effective is a governance structure that better mediated the competing interests they face. This chapter seeks to lay out that structure.
What Drives the Institutions?
Three forces drive and shape what the IMF and World Bank try to do. First, there are the interests of their most powerful member countries—very much led by the