Hemmed In: Responses to Africa's Economic Decline

Thomas M. Callaghy and John Ravenhill, editors

New York

Columbia University Press

1993

Bibliographic Data

3. Debt, Conditionality, and Reform: The International Relations of Economic Restructuring in Sub-Saharan Africa

D AVID F . G ORDON

My purpose is to explore the dynamics of debt, conditionality, and reform in Sub-Saharan Africa as they evolved in the 1980s. Although my perspective is regional, I nevertheless draw on specific examples to illustrate the analytical themes and empirical trends. While each of the topics under consideration has generated what is by now a fairly substantial literature, there have been surprisingly few attempts to examine them together. In particular, the discussion of African debt has generally been divorced from the issues of conditionality and reform. But, debt issues are not usefully addressed in isolation. The debt crisis is, at the same time, both a source of reform and a constraint to reform's having its desired impact of restoring economic health. I will argue that the debt crisis grew out of the very same policies that generated Africa's broader economic crisis, which is the rationale for both reform and the use of conditionality. Moreover, the external agencies that have been most active in promoting reform through conditionality-the international financial institutions (IFIs)-are, at the same time, the most important creditors in the debt crisis and have been given the task of providing further financial resources to the continent. While the IMF and World Bank almost always attempt to separate their roles as creditors and financiers from their role as promoters of reform, I will argue that the evolution of IFI conditionality in Africa has been linked to the evolving creditor and financier positions of the Bank and the Fund.

Debt

It is now generally accepted that Africa cannot grow out of its debt: despite the continually rising volume of total debt and debt service, the actual level of debt repayment is declining, and rapidly. While debt repayment levels had reached $12 billion in 1985, they had fallen to $9.5 billion in 1987 and have continued to diminish. ₁

While African debt has been the focus of special international attention since 1987, the crisis is unlikely to be speedily resolved. There has already been a great deal of de facto debt relief and many of the debt reduction plans under consideration will do little more than give this de facto relief de jure recognition. In addition, a very large, and increasing, proportion of African "problem" debt (that is, debt in countries with the least capacity to repay) is with the IFIs, who are, for reasons we will explore, loath to reschedule let alone reduce the debt. But if resolution is unlikely, improved management is both possible and necessary if economic reform efforts in Africa are to be more successful.

Africa's debt crisis was originally the result of the larger and more fundamental economic malaise-in a real sense the product of a generation of misguided economic policies that were themselves part and parcel of "crony statism." According to Callaghy, Africa became dominated by "a crony statism consisting of three interrelated characteristics: clientelist networks used to build support through the extraction and distribution of rents, the expansion of the size of the state, including the creation of an extensive parastatal sector, and the purchase of primarily urban support via state welfare services and subsidies. . . . Most African states were transformed into highly personalistic, authoritarian, but weak, administrative states-lame Leviathans-in which crony statism and a subordinated crony capitalism prevailed." ₂

But, if Africa's debt crisis originated in "crony statism," in the past five years, the extent of the debt crisis has transformed it into an important cause of continuing economic distress. The debt crisis reinforces the economic crisis in several interrelated ways. First, it absorbs a substantial proportion of crucially needed scarce foreign exchange in debt service payments. Second, it makes budgetary management more difficult by absorbing a large proportion of expenditures, especially in the context of needed exchange-rate devaluations. Finally, the imperative for reform is weakened when debt service acts as a high marginal tax rate on reform; that is, when the benefits of reform are accruing to external actors in the form of debt service payments.

Underlying the evolving issue of African debt has been the changing political relations between debtors and creditors. African governments tend to believe that, freed from the burden of debt, they would be able to achieve restored economic growth and development. Until recently, creditors in the IFIs believed that debt relief, by itself, would have limited effects in Africa. Similarly, the IFIs tended to be skeptical about the commitment of many African governments to fundamentally change the policies that created the debt crisis in the first place. For most of the 1980s, this sharp diversion in perceptions generated a conflictual atmosphere between African governments and the IFIs.

But, in the late 1980s, skepticism (often in the form of staff recommendations) took a back seat to other factors in determining IFI actions in Africa. Both the Fund and the Bank have been increasingly willing to undertake or maintain externally supported adjustment programs in the face of such skepticism because of their interest in a country's repaying existing obligations, political pressures from their shareholders, i.e. the leading Western governments, to continue to provide much-needed foreign exchange to financially strapped African governments, and their own desire to maintain the existing structural adjustment regime in which the IFIs continue to act as "financiers of last resort" to Africa in return for being excluded from debt reduction schemes. For example, in the final years of the Kaunda regime in Zambia, the IMF, which has substantial financial exposure in the country, heavily pressured the concessional donors to co-finance Zambia's adjustment program in the face of considerable skepticism (in retrospect, well-justified) about its viability.

The Origins of the Debt Crisis

Similar to other developing areas, the origins of the African debt crisis lie in the aftermath of the first oil crisis of 1973. But the effect of the oil shocks of the 1970s was somewhat different in Africa than the rest of the developing world. ₃ The rapid increase in oil prices was part of a much broader commodity boom that affected many of Africa's other major commodity exports such as bauxite, cocoa, coffee, cotton, tea, groundnuts, and uranium.

The oil boom transformed the availability of credit in the international banking system while the commodity boom enhanced international perceptions of African credit-worthiness. Demand for financing was relatively stable in the industrialized countries, but the oil price hikes generated a demand for external borrowing in both the oil importing countries, to cover the additional cost of oil, and in the heavily populated oil exporting countries, which sought to use their enhanced credit-worthiness to promote industrialization. Banks lent to both sets of countries because the new additional liquidity had to be on-lended or be slowly eaten away by inflation, while the commodity boom had generated the optimistic expectation that future export revenues would rise commensurate with new obligations.

In Sub-Saharan Africa, the explosion in commercial bank lending never reached the poorest countries. But even in those countries there was a rapid run-up in debt originating from export credit agencies in Western countries (committed to boosting their own exports to compensate for the rising costs of oil imports) and the IFIs. Borrowing in the mid-1970s appeared to make sense given the low or negative real interest rates that were the outcome of high rates of inflation, rising commodity prices, and excess liquidity.

This process, commonly referred to as the recycling of petrodollars, was seen at the time as a sign of flexibility and maturity in the international financial system, precluding the need for drastic domestic adjustment in oil-importing developing countries. Between 1973 and 1977, Africa's total volume of debt increased from $9 billion to $27 billion. But was petrodollar recycling really effective?

Even as late as 1980, when debt to GNP ratios for oil-importing countries were significantly increasing, the IMF defended the system: "the question of whether a country should seek further credit should not be answered by reference to statistical measures . . . Higher foreign indebtedness is sound policy for both lender and borrower because the higher level of investment financed by foreign borrowing will eventually be reflected in additional net export capacity." ₄

But, in Africa, the resources were not being effectively utilized as productive investments. Rather than promoting development, the windfall resources that flowed to Africa created unrealistic expectations, facilitated budgetary mismanagement, and produced "white elephants"-hotels, steel mills, new capitals, and palaces. Strong commodity prices allowed governments to substantially increase public expenditures, which became difficult to compress when, in the late 1970s, commodity prices fell. The productivity of investment was declining. In retrospect, it was during the commodity boom years, during which time African growth data looked good, that the seeds of later disaster were laid.

Given rising commodity prices and a rapid rise in foreign financial flows, much higher rates of economic growth should have been achieved during this period. Much of the growth was simply an expansion of government services; most of the rest was the Keynesian effect of short-term stimulation. The basis for sustained productive expansion was not being laid. Most important, foreign borrowing was not financing additional export capacity. This is an important difference between Africa's debt build-up and that of Latin America. In the period before the oil crisis, Latin America's export performance had been particularly poor, in volume terms increasing at only one percent per year. But there is little doubt that petrodollar recycling in Latin America, despite leakage through capital flight, was financing the expansion of exports. Between 1977 and 1984, Latin American export volumes rose by more than 5% annually. ₅ At the same time, African export volumes were stagnant.

When the commodity boom ended in the late 1970s, African countries continued to build up their volume of debt. Borrowing went to finance the second increase in oil costs (in 1979) and to avoid politically painful and socially disruptive cutbacks in public expenditure. The burden of domestic adjustment to declining terms of trade fell most heavily on investment as countries increased price subsidies and supports. The total volume of debt between 1978 and 1982-the commodity bust period-increased from $27 billion to $72 billion. Meanwhile, both the sources of financing and the recipients evolved. Commercial borrowing shifted almost solely to the oil-exporting countries; in the poorest countries, the IFIs, especially the IMF, became the major creditors. ₆ The IFI's financial stake in Africa increased dramatically in the 1980s. In 1982, IMF credits made available to Sub-Saharan Africa already totaled $4 billion. By 1988, they were more than $7 billion. The multilateral banks' stake (of which the World Bank constitutes the bulk) increased from $14 billion in 1982 to more than $30 billion in 1989. ₇

During the commodity bust period, the accumulation of new debt financed a substantial run-up in the current account deficit in Africa. The current account deficit is the excess of import costs and invisible payments over the value of exported goods and services. It is, by definition, financed by capital inflows and by drawing-down foreign currency reserves. It is normal and appropriate for developing countries to run modest current account deficits that are then financed from abroad. For instance, in the decade prior to the 1973-74 oil shock Africa ran a current account deficit of slightly over 20% of exports and Latin America ran one of slightly under 20% of exports.

The trouble comes when the current account deficit increases to the point that it can be maintained only by unsupportable additional volumes of debt or by squeezing import levels. That is precisely what happened during the second phase of the emergence of Africa's debt crisis. A comparison here between Africa and Latin America is again instructive. In the four years after the end of the commodity boom, Africa's current account deficit as percent of exports more than doubled to 45%. At the same time, the percent deficit in Latin America (and this is just before the onset of the Latin debt crisis) was only 32%. More important, Latin America was able to rapidly reduce its current account deficit by a combination of export expansion and import contraction to only 4% in 1984. In Africa, the current account deficit as percent of exports also dropped, but much more slowly (to 32% in 1984), and the adjustment was totally on the import side. ₈

This analysis strongly suggests that what distinguishes the emergence of the debt crisis in Africa from that in Latin America is Africa's poor export performance. Was this a terms-of-trade problem, an access to market problem, or an international competitiveness problem? The short answer is a bit of the first, virtually none of the second, and a lot of the third. As regards terms of trade difficulties, economic rigidity and lack of responsiveness, especially in low-income countries, has created a particular vulnerability in Africa to the sharp vicissitudes of the international economy. In the 1980s this was exacerbated by low commodity prices and the vicious cycle of low import volume constraining export growth. A large part of Africa's export problem is due to the composition of its exports. Primary products are twice as important to Africa than to other developing regions and international demand for primary product exports has been much more sluggish than demand for manufactures. Many of these features are themselves the result of the policy regimes that have been followed in Africa.

But the main trade problem for Sub-Saharan Africa has been the loss of market share to other developing regions. During the 1970s, Africa's share of non-fuel exports of developing countries fell by more than half, from 19% to 9%. The World Bank has estimated that if Africa had maintained its 1970 share of non-oil primary commodity exports, its export earnings today would be $9 to $10 billion higher, a level that would cut its debt service ratio in half. ₉ Jonathan Frimpong-Ansah, one of Africa's leading economists, in a study of trade issues in Africa, concluded, "the evidence showed that poor producer incentives were the principal factor in the decline of export production and that over-valued exchange rates and excess producer taxation were the principal adverse components. The general conclusion regarding the falls in export volume in Sub-Saharan African countries is that they are largely due to internal policy factors." ₁₀

Africa's loss of export market share was largely the result of a particular characteristic of "crony statism"-its import orientation. The origins of import orientation lie in both politics and in development strategies of the 1960s and 1970s. As mentioned by Callaghy, "crony statism" was based upon maintaining urban support. One mechanism to do so was to promote the availability of relatively inexpensive imported goods. Overvalued exchange rates do precisely this. At the level of development strategy, overvalued exchange rates were rationalized because they allowed relatively inexpensive purchase of capital and intermediate goods to promote import-substituting industrialization. Thus, the demand for imports was fostered by government policy. At the same time, since imports were only lightly taxed and the non-trade taxation capacity of African states was limited, taxation of exports became the predominant way to generate revenues. This reinforced the effect of exchange rate overvaluation in diminishing the incentives for export production.

Import orientation at the same time generated powerful opportunities for rent-seeking investments and corruption. A key feature of "crony statism" is state control of access to foreign exchange through trade licensing. As African currencies became overvalued, black markets for foreign exchange were generated. In many countries, the differential between the official and black market rate became very high. In Nigeria, for example, by the mid-1980s there was a roughly five-to-one ratio between the official and black market exchange rates for the naira. In that context, anyone having access to foreign exchange at the official rate, through a trade license, could immediately return to the black market and trade it for five times his original naira holding. The scope for personal gain by importers was enormous, creating powerful incentives for investment into this largely nonproductive sector. There was also tremendous scope for corruption among those responsible for the issuance of the licenses. The alliance of politicians, bureaucrats, and traders, not surprisingly, became a key component of the coalition behind "crony statism" in Nigeria and other countries. ₁₁

The point of this exercise is to demonstrate that not only is there a direct relationship between African politics and the general economic crisis, but also that the "political origins" thesis applies equally to Africa's debt crisis. Africa's debt crisis, while immediately precipitated by the second oil shock, is the long-term result of its poor export performance; that poor export performance is a function of import orientation, which, in turn, is an important characteristic of "crony statism." Africa's economic crisis was not initially the result of debt; rather the debt crisis was the inevitable outcome of Africa's economic crisis interacting with key trends in the international economy.

This is not to suggest that the debt crisis is solely the "fault" of African governments. As discussed, high levels of borrowing made some sense, especially before 1978. Furthermore, creditors share the blame for the rapid rise in the volume of debt due to their aggressive marketing of both loans and export credits and their participation in "white elephant" schemes. Fluctuations in interest rates and the sharp decrease in international liquidity, which turned the high volume of debt into a crisis of unsustainable debt service levels were, of course, completely beyond the control of African governments. In the period of the commodity bust, it was almost inevitable that African governments would seek to avoid difficult domestic adjustment through continued external borrowing justified by the hope that commodity price levels would again turn upward. But the fact remains that had export performance been better, all of these factors would have been much more manageable.

Once the burden of debt became overwhelming, by the early 1980s, both international market forces (weakening terms of trade and the depreciation of the dollar) and the very efforts undertaken to manage the debt crisis (IFI programs and debt reschedulings) served to continually increase the volume of debt. By 1987, African debt totaled roughly $130 billion, with very little of the new debt incurred in the previous five years going to productive investment that could break the vicious cycle syndrome described earlier. ₁₂ Unfortunately, at the very time when African governments began to become sensitive to the "policy" origins of their difficulties, the evolution of the debt crisis moved increasingly beyond their control.

Structure of Debt, Debt Service, and Debt Reform Initiatives

The overall volume of African debt pales in comparison to that of Latin America or to the so-called heavily indebted countries (HICs) that have been the object of most international debt schemes. However, for Africa, the burden of debt is greater than that in Latin America and the HICs. Debt and debt-service levels have risen more rapidly for Sub-Saharan Africa since 1982 than for any other set of countries. Only 12 Sub-Saharan African countries have serviced their debts regularly since 1982. The World Bank now characterizes 31 African countries as debt-distressed. Moreover, the social impact of debt and recession in Africa, especially low-income Africa, has been substantially more severe (although it has probably been politically less destabilizing) than in Latin America.

Africa's debt, unlike that of Latin America and the HICs, never posed a threat to the international financial system. The volume is small and the proportion owed to commercial banks is much smaller (35% of African debt) than that of the HICs (70%). ₁₃ But the large debt overhang does threaten the viability of economic reform, both by making the benefits of reform (enhanced external balances) appear to accrue substantially to foreigners and by limiting the likelihood that reforms will have their desired impact of stimulating growth. The modest size of African debt (and its public character) and the threat it poses to economic reform appear to offer incentives for creative management. This is especially so given that, as discussed above, the overall level of debt repayment has been declining for several years and there is almost universal agreement that even current levels of debt service will not be sustainable.

But the modesty of the debt volume and Africa's insignificance to the overall international system limit the impetus for more concerted debt relief. For most of the 1980s, the main instrument for managing Africa's debt crisis has been reschedulings through the Paris Club (for public debt) and the London Club (for commercial debt). By 1990, thirty Sub-Saharan African countries had negotiated some 120 reschedulings for debt totaling close to $30 billion. ₁₄ This figure substantially overstates the actual debt relief, since for many countries it includes successive rescheduling of already rescheduled debt. Generally, an IMF program was a precondition for participation in formal debt rescheduling. The lack of realism in this method of debt management is reflected in the declining proportion of debt service that was being repaid even after rescheduling. Nonetheless, broader measures will not be easy to generate.

Negotiating more effective debt relief will be analytically complicated, time-consuming, and inevitably raise difficult issues of burden-sharing. Burden-sharing problems among the commercial banks held up implementation of the Brady Plan of modest debt reduction for the HICs. Put quite simply, each individual bank has an incentive to hold back from offering debt reduction in the hope that the reduction offered by other banks will restore the debtors' capacity to fully repay its remaining obligations. For African debt, the burden-sharing problems are quite different and occur among the three different types of creditors-bilateral donors, IFIs, and commercial lenders.

In order to better understand the burden-sharing problem and the limited possibilities for a radical reduction of Africa's debt service burden, let us examine the debt service and debt repayment experiences of the three types of creditors and how each has responded to calls for more radical options. ₁₅ Total debt service payments in Africa rose from $7 billion in 1981 to $12.2 billion in 1985. Despite the increase in the amount of debt service, the proportion of debt service owed (after rescheduling) that was actually paid decreased from over 80% to under 75%. After 1985, both actual debt service and proportion of obligated debt service paid sharply decreased. In 1989, out of a totaled scheduled debt service of $18 billion, only slightly more than $7 billion (about 40%) was actually repaid.

In 1990, bilateral debt constituted 39% of total debt and accounted for 33% of scheduled  debt service. For low-income countries these figures are even higher. But, at least since 1985, bilateral debt repayment has comprised less than 15% of actual debt service paid. In 1989, even after rescheduling, only 20% of the amount due on bilateral debt was actually repaid. Bilateral debt is the largest single component of African debt and is most amenable to debt reduction. For several years, the IMF and the World Bank have recognized that the conventional approach to African debt was insufficient and have urged bilateral donors to provide debt reduction or cancellation. This is in fact occurring. A number of bilateral donors, especially the Scandinavian countries, began in the mid-1980s to convert concessional loans to grants. At the Toronto Summit of the industrialized nations, in 1988, the United States and West Germany, holdout nations, agreed to the principle of relief for non-aid debt in Africa. While the bilateral debt relief programs were slow to get off the ground, by 1990 the pace of implementation had increased substantially, with the U.S., France and West Germany all putting very substantial debt reduction programs in place. Unfortunately, bilateral debt relief, while having made the important conceptual breakthrough of legitimizing the principle of debt reduction, has not had a large cash-flow impact, since it focuses on that portion of African debt which already had the lowest ratio of repayment.

The IFIs comprise a little over one-quarter of Africa's total debt volume, but represent an ever-increasing percentage of actual debt service paid. In 1987, one-third of debt service paid went to multilateral creditors; by 1990, the share had risen to over one-half. Even though the World Bank and the IMF have steadfastly refused to formally reschedule their obligations, let alone reduce them, arrears to the two institutions are building up. In 1989, the IMF and the World Bank received only 85% of payments due to them. For a large number of low-income African countries, IFI debt constitutes more than half of their total obligations.

The Fund and the Bank have not accepted any obligation to share in the burden of debt relief, arguing that internationally mandated guidelines in their charters, their revolving character (in the case of the IMF), and their special role as a source of liquidity for Africa (in the case of the Bank) must exempt them from any overt debt relief. Beneath these public explanations are the two institutions' fears that debt relief in Africa might be the thin end of the wedge that would legitimize similar actions for Latin America, Eastern Europe and the other HICs.

The IFIs have, however, generated mechanisms to increase their flows to Africa and to increase the concessionality of such lending. As discussed, the IMF became heavily involved in lending to Africa in the early 1980s. Because of repayment schedules, by 1986 there was a substantial negative flow (about SDR 1 billion) from the region to the IMF. To balance this, the Fund created the Structural Adjustment Facility and Enhanced Structural Adjustment Facility to recycle repayments from governments to the IMF back to Africa on highly concessional terms. By the end of 1990, SDR 2.5 billion had been committed under SAF and ESAF, but disbursement has been slow, and through 1991 the negative financial flows from Africa to the IMF continued. ₁₆ In response to the desperate financing needs of Zambia, the Fund also created the Rights Accumulation Program, which, while not an explicit form of debt relief, allows the Fund to engage in operations in countries in which they are heavily exposed.

The World Bank, at about the same time, inaugurated the Special Program of Assistance, which, among other things, seeks to devise approaches to rescheduling nonconcessional debt without accelerating the growth of total debt volume. In 1988, the Bank began to grant interest payment relief to 13 low-income African states on loans received years earlier. The Bank has also committed an ever-expanding volume of its concessional IDA loans to low-income Africa. Net transfers from the World Bank to Africa reached $2 billion in 1990.

Nonetheless, if a more concerted approach to debt reduction in Africa is to be developed, it will inevitably have to involve some form of action on IFI debt. Several options are possible for this, but none appear feasible or likely. One option would be to replace existing loans with ones offering more concessional terms. Another would be the creation of a new international facility to buy out the debt owed to the IFIs, replacing it with new loans on more concessional terms. A third would be for bilateral donors to directly pay some of the debt obligations of African states to the IFIs. Each of these options would demand substantial new resources from the bilateral donors during a time of severe budget limitations, new commitments to Eastern Europe, and the apparent decision by the donors that bilateral debt relief is the appropriate tack for them to take on the African debt problem.

Commercial creditors account for some 35% of the total debt volume. Not surprisingly, this is concentrated among oil-exporting and other middle-income countries. Nigeria alone constitutes one-third of all long-term commercial debt in Africa. Private creditors absorbed less than 50% of Africa's debt service in 1989, down from 68% in 1983-85. Commercial credits can be divided into long-term and short-term debt. Short-term debt is basically trade finance for which, despite arrearages, debt reduction is not feasible. For long-term debt, middle-income and oil-exporting countries have kept up fairly high levels of repayment, while low-income countries are building up substantial arrearages. The arrearages on commercial debt that Africa has generated have further damaged African trade performance by necessitating cash payments, externally confirmed letters of credit, or price premiums for goods sent on credit because of delays in payment. Those countries keeping up payment of their long-term commercial debt have every reason to keep doing so, since it allows them continuing access to international capital markets. For low-income countries, the levels of long-term commercial debt are so modest as to not engage the attention of commercial banks in formal debt reduction schemes, though there has been some swapping of these assets. The World Bank has set up a debt reduction facility to expedite the use of swaps to reduce commercial bank debt. Nevertheless, African debt to commercial creditors does not appear to offer much scope for reduction.

While in the late 1980s there was a plethora of proposals for additional debt reduction in Africa, my analysis helps explain why there has been, in fact, so little scope for radical new initiatives. Such initiatives would have demanded the new commitment of donor resources and shared agreement among all of the key creditors. Neither has been or is likely to be forthcoming. Except for bilateral debt reduction, the financial resources involved would have to be up-front payments to create the facilities to finance the reductions. Thus, Africa's debt crisis is unlikely to be fundamentally resolved in the near future. The most likely scenario is a continuation of reduced actual debt service prompting increased official debt forgiveness. While widespread bilateral debt relief is on the agenda, it cannot by itself address the continuing squeeze that the debt crisis places on cash flow.

Ironically, at a time when responses to Latin American debt have moved away from the "liquidity" approach focusing on new financing, a liquidity approach to Africa's financial problems may be the only politically viable route. This approach will necessitate a continued rise in the levels of bilateral assistance and enhanced funding from the IFIs. For the IMF, disbursements of ESAF funds can be sped up to reverse the net flows back to the IMF. For the World Bank, IDA funding can be enhanced and directed ever more strongly to Africa. For this approach to play a positive role, the increased transfers need to be highly concessional so as not to add to the debt service burden and they must be directed to reviving Africa's productive capacity, especially in the export and import-substituting sectors.

But the liquidity approach is not without its drawbacks. It does little to lower the debt overhang that countries face and might, in fact, increase the debt service obligations that countries face in the medium term. In addition, while more politically manageable at the international level, the liquidity approach will reinforce the domestic linkage between policy reform and debt, thus maintaining the politically unsustainable impression that economic reform is being undertaken to placate external creditors rather than to promote national development interests. This leads us directly to the issues of the viability of conditionality and the outlook for economic reform in Africa.

Conditionality and Reform

In Africa, the international financial institutions and the donor community have been at the forefront of efforts to promote economic restructuring in order to restore sustainable economic growth. Controversy about the role of the IFIs has focused on their use of conditionality to leverage policy change. The use of conditionality expanded in scope after the first few years of the 1980s, following the failure of the IMF's efforts to reverse Africa's economic decline through its traditional instrument of the short-term "standby" arrangement. In the aftermath of this failure, the World Bank, as well as some of the bilateral donors (especially USAID), took the lead in what in fact has been a much more ambitious attempt to restructure African economies through conditional programs.

The predominance of conditionality in Africa in the 1980s generated the widespread misperception that external actors have been the major, if not the only, source of economic policy change. While the donor community has taken the leading role in setting the agenda within which African governments have responded to the continent's economic crisis, other factors have also motivated economic reform. The very fact of economic decline has weakened the political payoffs of "crony statism" in many nations, thus leading to some agreement on the need for reform if not full consensus on its content. Agreement about the need for reform has also been promoted by the general reorienting of global strategies for growth-a new belief in the efficiency of markets, a more open stance toward the international economy, a larger role for entrepreneurship-that has affected other areas of the developing world as well as Eastern Europe. Glasnost and perestroika in the former Soviet Union, the remarkable changes transpiring in Eastern Europe, and the rise of the "four tigers" of East Asia have energized a growing number of African intellectuals today, much as the Cuban revolution and the Chinese Cultural Revolution did a generation ago. Finally, Africa's debt burden, illustrating the need to generate a more viable policy framework, itself serves as an incentive to reform. A young generation of economic technocrats imbued with the belief in economic reform is emerging all over the continent. So while the role of conditionality in promoting reform has been substantial, it has hardly been the only factor at work.

By the late 1980s, within Africa there was a broad consensus about the necessity for economic reform. But, while the need for economic reform in Africa has been almost universally proclaimed (even by the architects and beneficiaries of "crony statism"!), it remains a bitterly contested terrain of public policy. The economic and technical problems of reform are far more complicated than many thought. Almost invariably, the specifics of reform programs lie firmly in the intellectually sticky realm of "second-best" solutions. A 1990 World Bank Working Paper by Harvard economist Dani Rodrik argues that the economic theory behind liberalization efforts in developing countries is embarrassingly weak. ₁₇ Similarly, Toronto economist G. K. Helleiner, a long-time observer of Africa, has argued that there is little common meaning attached to "structural adjustment." ₁₈ Equally problematic are the social and political ramifications of reform, touching as they often do basic elements of the social fiber of African societies. These ambiguities about the nature and impact of economic reform lie at the root of the debates concerning conditionality.

It is not easy to define conditionality precisely. In general, it refers to the agreements between donors and recipients that exchange financial transfers (either grants or loans) by the donors for policy changes by the recipients. But the specific relationship between the resource transfers and the policy changes is a source of disagreement among observers of conditionality. Donors themselves tend to portray this relationship as reinforcement, i.e. that the resource transfer provides an added incentive for the recipient government to implement policy changes to which it is already committed. Critics of the IFIs have viewed the relationship as imposition, i.e. that the donors utilize financial transfers to enforce inappropriate policy changes on otherwise unwilling governments. Academic analysts have tended to conceptualize the relationship as one of purchase, i.e. that the donors "buy" reforms that governments, for one reason or another, would otherwise hesitate to make. ₁₉

Conditionality in Africa presents the following paradox: while in the course of the 1980s there was an ever-expanding set of conditions placed upon donor resource flows to Africa, evidence suggests that there has been increasingly less donor influence over policy outcomes. In other words, we have witnessed ever more conditions, but less and less effective conditionality. While the IMF has sometimes been able to generate policy reform on narrow stabilization measures, including devaluation, the use of the financial leverage of conditionality on the part of the Fund, World Bank, and the bilateral donors has proven too blunt an instrument and a wasting asset for promoting broader economic reform. While it has been effective in placing the issue of policy reform on the agenda of many African governments and in initiating some reform efforts, it has played a far less positive role in sustaining reform.

It is useful to make a distinction between the financial leverage of "pure" conditionality and the broader instruments of influence that, in practice, have accompanied IFI and other donor policy-based loans and grants in Africa. Financial leverage, while the most visible element within conditionality, is part of a much broader pattern of donor influence on economic reform in Africa. Where conditional programs have been effective, the effective wielding of intellectual and political influence has also played an important role. ₂₀

The IFIs have been the main conduit of the diffusion of the ideas of economic liberalism in Africa. This has taken place through a number of mechanisms, some linked to conditionality, others independent of it. Directly linked to conditionality are the formal "policy dialogues" that the IFIs engage in with all recipient governments both over individual programs and concerning the broader overall economic policy context. Extensive training programs that the IFIs conduct are attended by both middle- and senior-level government technocrats. In addition, many senior African policymakers have actually worked in the IFIs. Finally, the IFIs and the other donors involved in policy reform have provided extensive technical assistance support, especially to core economic ministries such as central banks, ministries of finance, and ministries of planning. Taken together, these sum to a tremendous intellectual impact on the way economic policy is perceived, especially by technocrats.

This influence has been particularly important given the breakdown of Africa's traditional public-sector led and inward-looking development strategy. In the context of Africa's economic decline, government technocrats have been forced to consider alternative models and strategies. The financial dependence of African governments on the IFIs gave these agencies unusual access to economic policymakers. Over time, the range of interactions described in the previous paragraph helped to substantially change the terms of the policy debate in a large number of African states. Consider Nigeria and Tanzania as examples. A decade ago, the economic discourse in both countries were conducted in categories (Marxist-socialist for Tanzania; nationalist for Nigeria) far different than those (neoclassical economics) that dominate in the IFIs. Today, the terms of the economic debates, both domestically and between those governments and the IFIs, has evolved dramatically in the direction of the categories propounded by the IFIs. These changes would not have come about without the ongoing interactions with the IFIs.

Another, and more sensitive, form of influence by the IFIs and donors is political influence. According to their charters, the IFIs are supposed to be strictly apolitical. They have never really been apolitical, but the rise of conditionality in Africa forced them inexorably into an even more active political stance. Increasingly, as the 1980s wore on, the donors most directly engaged in policy reform coordinated closely with individuals at the highest technical levels of the bureaucracy. In many policy reform efforts, a key political role was played by "credible technocrats," individuals having the ear of both senior government officials and respected by the donor agencies as well. Such technocrats have been crucial to reform efforts in Tanzania, Madagascar, Kenya, and Nigeria among other countries. In several countries, the most prominent being Nigeria, IFI representatives have played an important role in coordinating and enhancing the political influence of these technocrats. In other countries, such as Malawi, while the IFI political role was less direct, a central aim of policy reforms proposed was to enhance the political influence of senior technocrats.

Thus, while financial influence is the most tangible form of IFI and donor influence, intellectual and political influence have also been important components of donor efforts to promote economic restructuring through conditionality. While in theory conditionality is generally analyzed as the exertion of financial leverage, in practice it has involved varying mixes of financial, intellectual, and political influence. In fact, I will argue that, for a number of reasons, the influence that derives to donors from the financial leverage of "pure" conditionality is quite limited and is highly unlikely to form the basis for successful efforts to promote economic reform. As the 1980s drew to a close, the IFIs appeared increasingly aware of the limits of financial leverage, arguing that country "ownership" was necessary for successful reform. ₂₁

What donors involved in conditionality have been less willing to concede are the inherent problems that the process entails. In general, the entire conditionality "game," whereby donors attempt to "buy" as much reform as they can while recipient governments attempt to get as much money from the donors as they can with as little reform as possible, draws government attention away from the serious need for economic restructuring by creating a context in which the benefits of reform became identified as increased donor resources rather than improved economic performance. Decisions concerning economic reform too often become responses to external pressures and attempts to maximize external resource flows rather than efforts to grapple with imperative domestic problems. ₂₂

Furthermore, the financial leverage of conditionality becomes vulnerable to manipulation as donor interests expand beyond the single goal of promoting economic reform. Once would-be promoters of reform develop large stakes as creditors in a situation in which repayment is problematic, their reform goals are in danger of becoming subordinated to their creditor interests. Similarly, since the IFIs and the donor agencies have assumed the task of providing a financial cushion for Africa in the context of an unsustainable debt situation, the conditionality that is attached to their financial transfers over time has lost much of its credibility. The evidence suggests that external finance can either be an instrument for cushioning the debt burden or a lever for promoting economic restructuring; it is very difficult to do both simultaneously.

In rest of this section, I will explore the origins of conditionality in Africa, dissect the mechanics of conditionality, and offer some explanations for why its influence has been more apparent than real. I will also examine the validity of the critiques of conditionality made from the left, i.e. that it threatens the sovereignty of African states, and from the right, i.e. that it actually impedes adjustment by providing African governments the leeway to continue "business as usual." Finally, I will explore some of the reasons why conditionality in general has been a frustrating experience for both the donors and the recipients.

The Rise of World Bank and Bilateral Conditionality

The use of conditionality in Africa by the World Bank and some of the bilateral donors has been driven by a range of factors. The World Bank's 1981 Berg Report paid special attention to policy problems as a source of economic distress. Critical among these problems were poor public sector management, a bias against agriculture, and trade and exchange rate biases against exports. ₂₃ The Berg Report's focus on the policy roots of Africa's economic crisis was consistent with several large cross-national studies that appeared at about the same time, all of which argued that a country's international trade regime was an important source of differential growth in the Third World. Countries with "open" trade regimes consistently outperformed those with more "closed" regimes, both in times of international economic stability and in times of international shocks. Africa was seen as the prime example of a region where efforts to promote import-substituting industrialization led to adverse economic outcomes. ₂₄ Within the World Bank, a series of project assessment reports concluded that the relatively poor outcome of Bank projects, particularly in the agricultural sector, had been due to inappropriate overall economic policies that thwarted even the best designed projects.

The Berg Report argued that growth and development in Africa could be reignited only through a process of "structural adjustment" centered on realigning overvalued exchange rates, improving price incentives (especially in agriculture), limiting the role and improving the performance of the public sector, and energizing entrepreneurship in the private sector. The implication of all of this for the World Bank was the need to design lending instruments that could improve the quality of projects by giving the Bank a larger influence over the general policy environments in African countries.

Many of the bilateral donors, for their part, were concerned that traditional foreign assistance activities (focusing on projects) were not working. A series of empirical studies suggested that foreign assistance, in theory intended to supplement investment, was slipping over to support increased consumption. ₂₅ This, combined with a more generalized "aid fatigue," led several of the bilateral donors also to be interested in new forms of assistance.

Coincidentally, the onset of the economic crisis led African governments to also seek different forms of funding from the World Bank and the bilateral donors. Countries sought resources that were flexible, not tied to particular projects, were fast-dispersing, and could be used to maintain import capacity. They sought non-project assistance. Thus, at the same time as recipients sought more flexible resources, donors sought influence over policy. The basis was laid for a new instrument: the non-project, policy-based loan or grant.

The World Bank's use of conditionality in Africa was also influenced by critics of the IMF "demand management" approach to Africa's external payments crisis. Critics argued that Fund programs were inappropriate to the problems of African countries for two reasons: first, their time scale was unrealistically short; second, their approach was too narrowly focused on the financial sector rather than the "real economy," the supply-side issues of enhancing responsiveness and growth. ₂₆ The failure of IMF programs in the early 1980s lent weight to the critics. While these views never gained full support at the IMF, (despite the stretching of IMF programs to include "supply-side" issues), they did find a sympathetic ear among policy-makers at the World Bank. Rather than propose an alternative to IMF programs, the Bank developed an instrument to supplement the demand-restraint and external balance-oriented IMF programs with "supply-side" measures.

Conditionality-based non-project instruments were the outcome of the confluence of donor assessments of the sources of Africa's economic malaise, recipient demands for more flexible non-project funding, and the influence of alternative stabilization strategies. The most prominent of these instruments was the World Bank's Structural Adjustment Loan (SAL). The typical SAL sought a range of policy changes: improvement in producer prices, especially in agriculture; the reduction or elimination of consumer subsidies; the liberalization of international trade through quota and tariff reduction; the liberalization of domestic trade through reducing licensing and price controls; the reorganization and streamlining of government agencies, including reducing the size of the public service; the restructuring of education and health services and the introduction of cost-recovery schemes; the restructuring and sometimes privatization of state-owned enterprises; and the development of multi-year investment plans.

In return for agreement to a set of mutually developed policy reforms by the recipient country, SALs offered non-project "free" foreign exchange: that is, a direct infusion into the central bank, whose local currency counterpart directly augmented the government's budgetary revenues. Conditionality in SALs concerned implementation of the policy changes themselves, rather than meeting aggregate targets (such as reducing the rate of inflation) as in IMF programs. The World Bank later initiated Sector Adjustment Loans (SECALs), a similar instrument targeted at sectoral (agriculture, industry, education, export) rather than macro-level adjustment. At the same time, it encouraged bilateral donors to co-finance Bank programs by pledging their own resources, which are added to the total financial package. A number of major donors have done this, while the U.S., U.K., and France have, in addition, undertaken their own conditionality-based policy programs.

Conditionality-based lending is designed to enhance growth in three ways: by improving the policy environment, by directly increasing the availability of foreign exchange, and by catalyzing other foreign exchange flows, both private and public, concessional and nonconcessional. These increased foreign exchange flows ease the import constraint which, in turn, is supposed to facilitate a quickened response to the reforms undertaken. Conditionality was seen by the IFIs as particularly important in Sub-Saharan Africa, since the drying up of commercial flows signaled that international capital markets had essentially declared the region noncreditworthy. Conditionality was intended to decrease the likelihood that external finance would play the role that so much of it had in the 1970s: allowing countries to escape the imperatives of adjustment, providing "rents" for privileged groups, and ending up as capital flight. Unfortunately, it does not appear that the architects of conditionality seriously explored the question of how reform would work domestically, how donor influence would have its desired effect, and whether the leverage offered by conditionality would be practical. It is to these issues we now turn.

The Analytics of Pure Conditionality

To better understand how conditionality works, let us begin by presenting a simplified economic model of pure conditionality. ₂₇ In the model, economic growth is constrained by a lack of efficient investment, which in turn is a result of government policy. The indirect, but primary, goal of conditionality is to catalyze the restoration of external investment flows from private sources. Foreign investors refuse to lend because they fear that their loans will not be repaid. They believe this because they recognize (correctly) that governments have a strong preference for present consumption (including, for the purpose of simplicity, rent-seeking investment) over future consumption. Thus, financial inflows will inevitably go to consumption rather than investment. The direct goal of conditionality is to bind the recipient government to alter its consumption/investment preference function toward investment. Thus, conditionality functions not only to enhance the viability of the loans to which it is attached, but also to improve the investment climate for all creditors, thus catalyzing restored external private flows. Similarly, conditionality promotes a context in which assumptions about foreign aid as additional to domestic savings become realistic, thus promoting increased foreign assistance.

This economic model is consistent with a range of real-world perceptions of conditionality. It provides a basis for the widely held belief that conditionality involves ceding sovereignty to the IFIs. It also provides the theoretical underpinning for the widespread notion that an IMF agreement is akin to the Good Housekeeping Seal of Approval for government policy, increasing the attractiveness of a country to foreign investors. It can explain IFI optimism about the potential for policy reform in Africa and IFI frustration that recipient countries generally didn't believe that conditionality was in their interests. The model is consistent with IFI and donor beliefs that conditionality enhances the creditworthiness of countries.

But, as the 1980s evolved, conditional agreements became separated from increases in private external flows, especially in Africa. This suggests that conditionality has become less credible in the eyes of international capital markets, and that there might be flaws in the economic model of conditionality presented above. In fact, the model sinks or swims on the assumption that conditionality is, in practice, binding: that is, that governments will implement conditional agreements. How valid is that assumption?

While African states have had a powerful incentive to enter into conditionality-based agreements-their desperate need for the foreign exchange that accompanies such agreements-they have much weaker incentives to implement the conditions that they have agreed to; nor do they always have the technical or political capacity to do so. ₂₈ The weakness of these incentives derives from several different sources: characteristics of the international system, incentives in the IFIs and donor agencies, and politics in African countries.

The conditionality process is weakened by the fact that governments often believe that noncompliance will go unpunished. In the "anarchic" international system, the legal constraints upon sovereign nation-states are minimal. In theory, the IMF and the World Bank should be able to ensure compliance with programs by threatening to withhold future funding if the conditionality attached to existing programs is not implemented. But, are the IMF and the World Bank really the tough financial "cops" they are often made out to be?

Here is where the multiple roles of the IFIs-promoters of reform, major creditors, and "financiers of last resort"-come into conflict. The international pressures on the IFIs to continue to supply liquidity to African states, largely growing out of humanitarian concerns about African poverty and that Africa not fall further behind the rest of the world, undermine the IFI's ability to sanction noncompliance with conditional agreements. The situation is worse for bilateral donors such as USAID who often have more obvious political stakes that preclude sanctioning. For instance, the U.S. strategic commitment to President Mobutu of Zaire limited its ability to exert leverage in favor of economic reform. Moreover, given that debt repayments may be put at risk if programs are canceled, the IFIs themselves had a growing disincentive to enforce conditionality as their financial exposure in Africa increased throughout the 1980s.

This bias against sanctioning noncompliance is reinforced both by the difficulty in monitoring compliance and by the bureaucratic incentives within donor agencies. SAL conditionality is often very difficult to monitor, with the possibility that reforms enacted might be countermanded by other policy initiatives outside of the scope of programs. The resources provided for monitoring and evaluation of conditional programs are minuscule compared to the task involved. It is no exaggeration to say that the Bank often does not really know if governments are complying with its conditions. Partially for this reason, Fund staff are often openly disdainful of the World Bank's claim to be effective in applying conditionality.

Staff members of the IFIs have seen that their path to a successful career is through participating in the design and implementation of successful policy reform programs. They thus have a strong incentive to portray the conditional lending activities in which they have been involved in the best light possible. Similarly, at an institutional level, given the controversy attached to conditionality, the IFIs have a broad institutional interest in enunciating the positive. This is particularly true for countries that have good relations with the IFIs, repay their debts, and have been described as successful adjusters, such as Kenya (until 1990) or Ghana. In such contexts, recipient governments have gained a good deal of flexibility in how (or whether) they implement conditionality-based programs.

Evidence from recent studies supports this interpretation. Several comparative studies agree with Tony Killick's finding that the IMF has experienced considerable difficulty in ensuring that its programs are implemented. ₂₉ A major reason for this is, as noted by Bienen and Gersovitz, that the penalties for partial compliance are not great. ₃₀ If anything, the World Bank and the bilateral donors are even more flexible. Gates, in a study of Bank and USAID conditionality, found minimal risks for noncompliance. ₃₁ IMF, World Bank, and bilateral programs are continually renegotiated.

Of course, sometimes there are sanctions for nonimplementation; the IFIs are hardly "paper tigers." When a government publicly repudiates a program, as Zambia did in 1987, IFI funding does get cut off. Also, governments do have to show some real efforts in order to qualify even for policy-based programs. IMF programs are sometimes discontinued, and World Bank SALs generally are undertaken only in the context of an IMF program. But discontinuance, in and of itself, has not heavily damaged a country's ability to reapproach the IMF and the World Bank and renegotiate a new program. The World Bank's main form of sanction in adjustment lending is to delay dispersal of funds, not a particularly powerful lever of influence. The point of this discussion is that recipient governments are aware of the very limited sanctions for nonimplementation and are thus less likely to implement the agreed-upon conditions. The stop-go nature of the implementation of conditional agreements presents a picture quite different from the theoretical model of conditionality, in which it is assumed to be binding. It is thus not surprising that conditionality has lost credence with global financial markets and has not had the "catalytic" role of generating additional private investment it was designed to have.

Within the IFIs, there has been a growing recognition that the limited likelihood of sanctioning noncompliance substantially weakens the financial leverage of conditionality. This is expressed empirically in the growing emphasis in donor discussions of economic reform on ascertaining recipient government "commitment" before initiating policy-based programs. ₃₂ If conditionality was more binding, this would not have arisen as a concern. The World Bank and USAID have begun to engage political scientists as staff members and consultants to work on policy-based programs. But judging commitment is very difficult indeed, largely because of the incentives generated by the very lack of sanctioning. "Commitment" is likely to be susceptible to the "game" aspects of conditionality, and to be very difficult to predict with much confidence.

The limited likelihood of sanctioning creates a context where elites have an interest in expressing a commitment to reform even where one does not exist. Generally, the policies attached to conditional financing challenge the interests of key components of the coalition behind "crony statism." For example, devaluation hurts anyone who has had privileged access to undervalued foreign exchange. This gives rise to what Miles Kahler has labeled the "orthodox paradox" of conditionality: how can external actors convince governments to change policies that are economically damaging but politically rational? While state policy and political coalitions produce many of the distortions that conditionality seeks to change, external actors must nonetheless work through the instruments of the state. ₃₃ A growing literature has argued that economic restructuring programs have not been more effective because the dominant elites have a vested interest in the status quo and thus don't want them to work. ₃₄ In the 1980s, in Africa, the cases of aborted adjustment in Liberia, Zaire, and Somalia are examples in which entrenched elites gained access to considerable conditionality-based external resources without moving beyond the initial stages of implementation of reform packages.

But the influence of entrenched interests is not the only political factor that limits the likelihood of effective implementation of structural adjustment programs. Even those leaders who are committed to reform face severe political difficulties. Recall the "time consistency" problem-the preference for consumption over investment-that is the rationale for conditionality. The root of African governments' preference for immediate consumption over investment is the political fragility of African states. The imperative of regime maintenance often clashes with conditionality's effort to shift government expenditure (and societal incentives) from consumption to production. The classic empirical example of this is risks attached to removal of consumer subsidies on basic goods, such as bread or flour. African politicians, like all others, have difficulty looking beyond the short term. Parallel to the "time consistency" problem is the "coalition" problem, the fact that while the benefits of adjustment tend to be marginal but broad-based, the costs are sharp and focused. The theory of collective action suggests that, in such circumstances, "losers" will politically mobilize against the reform to a much greater degree than "winners" will mobilize in its favor. Unsurprisingly, it has therefore been extremely difficult in Africa to gain political momentum for economic restructuring. ₃₅

Even those elites who want economic restructuring to work-and their numbers are increasing in the face of the collapse of alternative Marxist models and the success of the East Asian NICs-do not always have the ability to ensure implementation nor do they wish to pay the inevitable cost of implementing such programs. ₃₆ In several instances, conditionality-based agreements entered into in good faith by governments were not implemented because of a combination of limited technical skills and bureaucratic blockage. This is an especially difficult problem in complex institutional reform efforts such as privatization and budget and tax reform. Implementing structural adjustment is an inherently difficult task; it involves changing standard operating procedures, challenging vested interests in the bureaucracy, and establishing new relations with the private sector and nongovernmental organizations.

Finally, ruling elites do not necessarily accept the technical analysis upon which the policy conditionality is based. They may not believe that their undertaking the policy changes indicated will achieve the intended outcomes. The record of IFI programs in Africa during the economic crisis suggests that this analytical skepticism is justified. Many governments have completed one stabilization program only to have to return to the IMF in a year or two to undertake a new program, having failed to achieve balance-of-payments stability that is the goal of Fund programs. If anything, the technical analysis behind World Bank and bilateral donor "structural adjustment" programs is weaker still; and the Bank can only point to Ghana as the single example of sustained, effective "structural adjustment" in Africa.

This entire analysis suggests that the incentives for fully implementing conditionality-based agreements are not particularly strong, and that the ability of IFIs and other donors to "buy" economic reform through the instrument of pure conditionality is very limited. At the same time, it challenges the notion that donor conditionality has diminished the sovereignty of African states by allowing the IFIs to assume control over economic policy. While donor financial leverage, in the context of Africa's economic deterioration, could put the issue of economic restructuring on the agenda of African governments, it was capable, by itself, only of generating initial stabilization measures.

Nonetheless, in a limited number of African countries, economic reform activities have been quite substantial. How can we explain the extent of reform initiatives in these countries in recent years despite the narrow political base of support for such activities? The primary explanation for this is domestic. When substantial economic reform has taken place it has been because political leaders have seen the consequences of failing to undertake reform, i.e. the continuation of a downward economic spiral, as being more risky than those of undertaking it. But, in almost every case where this process has been at work, there has also been a dimension of external influence. To understand this, we need to return to the distinction among different forms of donor influence. While the role of financial leverage as an instrument of donor influence is indeed limited, when combined with intellectual leverage, it has played a significant role. A more satisfying model of conditionality in Africa would focus on the interplay between the three levels of influence: financial, intellectual, and political.

A rough "ideal type" of how conditionality in Africa, when successful, bilateral donor "structural adjustment" programs is weak has worked in practice is as follows: policy dialogue, both formal and informal, and joint technical analyses form the basis for establishing consensus and commitment on the part of senior technocrats for particular economic reform endeavors. These technocrats then persuade their political mentors of both the costs of the status quo and the need for such changes; their position reinforced, either directly or indirectly, by their relations with the donors and the knowledge of availability of donor resources. While the ultimate weight of donor influence is financial, it is wielded through the intermediate steps of intellectual influences, especially on key technocrats, and through generating "transnational coalitions" with those technocrats that enhance their political weight in policy decision-making.

Conclusion: The Impact of Conditionality on Economic Restructuring

What, then, has been the impact of conditionality on economic restructuring programs in Africa? The impact can usefully be divided into themes involving economic outcomes and those involving political process. There is a large and growing literature that tries to calculate the impact of policy-based operations on economic trends. ₃₇ This is a methodologically challenging issue. Conditionality agreements involve both policy changes and resource transfers. Disentangling the impact of these two elements is very difficult, involving the use of sophisticated models of individual economies. Ideally, one would wish to answer three questions: first, what has been the impact of conditionality on a country's policy environment; second, what has been the impact of these policy changes on economic outcomes; and third, what has been the impact of the financial resources transferred. On none of these issues is there a broad consensus.

One approach to these questions is to ask what would have happened in the absence of conditionality. This is a tough question to answer because it necessitates the construction of several counterfactual scenarios. The first is an international counterfactual. What would have happened in Africa had the international financial resources that went along with the conditionality agreements not been forthcoming? A simple approach to this is to subtract the value of conditionality-based external resources from total external resources and then calculate the further import contraction this would generate and its impact on levels of GDP. But this assumes that conditionality-based flows were completely additional to other external flows, which is almost certainly not the case in aggregate terms though it may be a good way to think about each specific loan or grant. This method also assumes that debt service payment is not higher under conditionality-based programs. That has certainly not been the case. In numerous countries that have, for one reason or another, ended conditionality-based programs with the IFIs, among the first changes is a lowering of debt service payment.

A number of World Bank reports on adjustment in Africa make the strong case that those countries engaged in sustained, externally supported adjustment programs have performed better than those without them. But, the Bank has been criticized for not analyzing the independent effect of higher resource flows to "adjusting" countries. Better performance might be a result of enhanced resource flows rather than policy changes. The Bank itself accepts that improved external flows are important to the success of policy reform efforts. But, a recent econometric exercise undertaken by British researchers concluded that there was not strong evidence that the finance accompanying policy-based agreements caused improved growth performance in countries implementing such agreements. ₃₈

The second counterfactual, a domestic policy one, is even more challenging. This explores what the policy actions of a government would have been in the absence of conditionality. To what degree does conditionality drive policy? This is a question that can be posed both for policy formulation and policy implementation. In principle, one can conceive of a situation in which conditionality agreements embody nothing beyond government's own policy preferences. Officially, the IFIs contend that conditionality is merely an instrument of discipline in the implementation of government-generated policy goals. This appears to be contradicted by the active role that IMF and World Bank officials have personally played in drawing up the economic restructuring programs in numerous African countries. The degree of variance between a government's own preferences and what is embodied in conditional agreements is probably quite high between countries.

The third counterfactual derives directly from the second; that is, what would have been economic outcomes in the absence of the policy-changes that were engendered by the conditionality-based agreement?

The difficulty of constructing realistic counterfactuals makes the systematic assessment of the impact of conditionality on economic outcomes problematic. What is more amenable to systematic assessment is the extent and nature of economic restructuring that has actually taken place. There is now a large literature of case studies on economic restructuring in Africa, as well as a number of comparative and aggregate assessments, some from the World Bank and others from academic researchers.

In examining the 1980s economic restructuring efforts of ten representative African countries, David Sahn and his colleagues at Cornell University found that, "Measured against the failed policies that predated the reforms, considerable progress has been made . . . Both in terms of policy change and performance outcomes." ₃₉ This finding is consistent with that of the World Bank's recent Third Report on Adjustment Lending, which concludes that policy reform and adjustment support has restored growth in actively adjusting African countries to the "moderate levels" of the 1970s. ₄₀ Both the Cornell and the World Bank study agree that two elements of structural adjustment that appear to show considerable, and sustained, success in a number of African countries are exchange rate reform and pricing and marketing reform for food crops.

Outside of the CFA zone, virtually all countries that have undertaken donor-supported adjustment programs have succeeded in lowering the spread between official and parallel exchange rates, and, indeed, in depreciating the real effective exchange rate. This is significant since many critics of adjustment had argued that it would not be possible to depreciate the real exchange rate in African circumstances. More recently, there has been a growing trend for countries to go beyond periodic devaluations to establish market-oriented exchange rate regimes, either through legalizing foreign currency trading or through some form of foreign exchange auction.

The importance, both economic and political, of exchange rate reform should not be minimized. The exchange rate is the single most important price in any economy, and over-valued exchange rates distort all other prices. Experience in a range of developing countries has emphasized the importance of exchange rate depreciation to the success of associated trade reform efforts, especially tariff reduction. Politically, exchange rate depreciation and the ending of official monopolies over the allocation of foreign exchange restrict what was one of the most important sources of rent-seeking behavior. The ultimate sustainability of exchange rate reform is yet to be secured, however, given that, in virtually all cases, the existing exchange rates are maintained by very high levels of foreign aid, while export responses have generally been less than expected.

Given the centrality of agriculture in most African countries, the incidence of success in food pricing and marketing reform is also very significant. The impetus to reform of food pricing and marketing policy was the practical breakdown of the state-dominated marketing channels and the shift of both production and trade into the informal sector. Thus, in many countries, the political benefits of the old system had already largely worn away. One of the most interesting findings of recent studies is that the cost of raising incentives for farmers has generally not been at the expense of consumers. Part of the reason is that, in many countries, few consumers in practice had access to officially priced food. Part of the reason is that lower marketing margins generated by increased efficiency throughout the marketing chain has allowed both producers and consumers to gain. It is not surprising, then, that food policy reforms have tended to be sustained.

While the successes of structural adjustment in Africa are important, the fact remains that, in many ways, the overall results of economic reform in Africa have been less than encouraging; and the sustainability of reform efforts remains tenuous, especially given the new environments of political liberalization. The World Bank concedes that the growth rates achieved for the actively adjusting countries of Sub-Saharan Africa (around 3.5% per annum) lag well-behind those for adjusting countries in other regions. Export and savings responses have been weak and, not counting donor-provided resources, investment levels are still exceedingly low. In particular, the private investment supply response that will be needed to sustain economic recovery and growth has not been forthcoming. A group of "adjustment stars" has yet to emerge in Sub-Saharan Africa, and adjustment efforts in the best performing countries remain dependent upon extraordinarily large donor transfers (for instance, donor resources constitute between 15 and 20 percent of Ghana's GDP).

A major explanation for the limited impact of adjustment is that existing programs have not been very successful in promoting reform in several crucial areas such as fiscal policy and the public service, the regulatory environment affecting private investment, and export crop pricing and marketing. David Sahn concludes that "In many instances, policy change has lagged behind rhetoric as implementation of reforms has often proved more perilous than planning them." ₄₁

Fiscal deficits have proved very difficult to attack, on both the revenue and the expenditure side. Even those countries that have had some success in reducing budget deficits have tended to do so in ways that negatively affect medium and longer-term development potential. In virtually no country has fiscal reform enabled the budget to become a real tool for effectively managing the development process. Budget cutbacks have tended to target investment rather than consumption, with maintenance being especially neglected. Civil service reform programs have tended to lag far behind schedule, with virtually no success in either cutting the public sector wage bill nor in providing senior technical staff with competitive wage packages.

Reform of the regulatory environment affecting private investment has also been lagging. In many African countries, the lack of complementary reform measures in the regulatory environment have limited the impact of trade reform and industrial sector reform initiatives. In many donor-supported reform programs, there has been detailed conditionality for exchange rate reform and for trade liberalization, while the conditionality for regulatory reform has been limited and fuzzy.

The experience of substantial reform in food crops contrasts with that of limited reform in export crops. The roots of this difference lie in export crops' continuing role as a major source of taxation revenue for African governments. While the incidence of export taxation has diminished in many countries, governments have generally seen this as all the more reason to maintain their ability to tax what they can. Thus, the reform imperative has been much weaker in the export crop sector than it has been in the food crop sector. As a result, the export response to adjustment has been limited. The lack of success in export crop policy reform thus has quite serious implications for the overall sustainability of the reform process.

Taken together, the limited extent of reform in the budget and overall public management, in the regulatory regime facing private investors, and in export crop pricing and marketing, present a picture of reform outcomes substantially less successful than that discussed earlier for exchange rate and food crop policy. The reform cup in Africa can be thought of as either half full or half empty. But, either way one looks at it, the cup has not been full enough to provide an enabling environment for private investment.

What has transpired under conditionality-driven structural adjustment programs in Africa is thus not fundamental economic restructuring but "partial reform." In most Sub-Saharan countries that have initiated reform, the dynamics of the reform process in the 1980s led to neither a collapse of adjustment efforts nor to fundamental transformation and dynamic economic growth, but rather to a suboptimal mix of partial measures. In a typical "partial reform" syndrome, a willingness to initiate adjustment measures is not supplemented by the basic institutional and attitudinal changes needed to carry through a transformation to market-oriented and private sector-led growth. Adjustment efforts have some success in eliminating the worst distortions and in restoring low-level economic growth, but do not really transform either policy-making or the overall economic environment. The "partial reform" outcome is consistent with what most African incumbent regimes desired from donor-supported reform programs. Robert Bates has argued that, from a recipient perspective, structural adjustment is a last-ditch effort to reform a regime from within, while minimizing political costs, in order to restore its viability in a new economic environment. ₄₂ In other words, for governments, the goal of adjustment was "partial reform."

Donor conditionality has played an important role in the "partial reform" syndrome. On the one hand, conditionality, broadly defined, has played a predominant role in initiating economic restructuring in Africa. Donor pressures have been successful in promoting reform in areas that are technically blunt, politically manageable, and institutionally non-complex. The exchange rate is a prime example of this. But donors have also had a tendency to sustain the "partial reform" syndrome, because of their strong incentives to maintain donor programs even in the absence of thorough and effective implementation, and the inability of conditionality to affect so many of the factors that determine the success of adjustment.

But the broader political impact of conditionality has been destabilizing to incumbent regimes and has played an important role in bringing about the demise of so-called "development dictatorships" in Africa. This impact has been felt in several different but often mutually reinforcing ways.

Conditionality-driven programs propelled the expansion of the nongovernmental sector. According to Naomi Chazan, "Many governments, cognizant of the role of the voluntary sector in the provision of essential services, relaxed some of the restrictions on organizational life in order to relieve them of direct responsibility for public welfare." ₄₃ This led to flowering of both local-level voluntary development organizations and intermediate organizations all over the continent, breaking the organizational monopoly that most African"development dictatorships" had imposed, and creating the beginnings of an organizational counterpoint to the state.

Conditionality also facilitated the expansion of the informal sector by weakening, in practice if not always expressly, the ability of the state to restrict informal activities. The informal sectors became, in many countries, the most dynamic area of the economy. The political impact of this, in turn, was the development of a significant resource base outside of the control of the state.

Donor programs, even when not fully implemented, reinforced the effect of the economic crisis in reducing the patronage available to rulers and the amount of rent-seeking in most African political systems. This threatened the control mechanisms in most African states. As van de Walle explains, "The essential problem for state leaders during the reform process is to maintain control of the clientelist networks on which they have based their power, even as they decrease the cost of those networks by ousting old clients or curtailing their access to rent-seeking." ₄₄

Paradoxically, in most African states, even as structural adjustment programs were initiated to limit patronage and to restore the fiscal balance, corruption increased and budget deficits worsened. Two processes appear to be at work. First, the state's internal discipline collapsed as the system of rewards and loyalty that previously held the system together frayed. This is particularly evident in the growing inability of many African states to raise taxation revenues, even in the context of donor-supported revenue-enhancement projects. Second, the threat of major policy changes gave a powerful incentive for those administering the systems to get as much as they could in the fear that the tap would soon run dry. As the political crisis hit in the late 1980s, both of these trends were exacerbated as it became "every person for himself," in the expectation of imminent political change.

At the same time as the state began to fray from within, structural adjustment programs became the target of many of the urban middle-class protest movements that later became more deeply politicized. Protests generally attacked cuts in housing allowances, school stipends, and price increases for basic goods. But while this has often been interpreted as deep-seated opposition to adjustment, the reality seems somewhat more complex. The logic of economic reform efforts suggested to the same urban middle-class the need for deeper institutional and political change, and the protests quickly shifted from being against adjustment to being against incumbent regimes. This two-sided impact of reform is seen in the ambiguous attitude of many of the new political movements toward economic reform efforts. For example, while in opposition, the Movement for Multi-party Democracy (MMD) in Zambia criticized some of the government's stabilization initiatives as being too harsh, while at the same time calling for fuller implementation of structural adjustment measures! Since taking power, MMD has already implemented reforms that go far deeper than any envisioned by Kaunda's former regime.

Thus, the overall political impact of donor-supported structural adjustment was to interact with broader international trends to destabilize existing African regimes. This fact raises problems with the conservative critique of conditionality. This perspective charges that conditionality-based external resource transfers enable governments to continue to avoid necessary reforms by providing them with "breathing space" that they otherwise could not have. ₄₅ Clearly, some elements of the analysis that I have presented, especially the limited sanctions for noncompliance and the counterproductive impact of large volumes of balance-of-payments financing are consistent with such an interpretation. ₄₆ But what about the larger picture? Does conditionality lubricate reform or does it offer an escape from it?

The cumulative political impact of donor conditionality-destabilization-appears to be quite different from its immediate political impact-breathing space. The evidence from Africa suggests that, except in cases where conditionality-based agreements were really motivated by donor security interests and thus had almost no impact, it is difficult to argue that, in the longer term, conditionality-based resource transfers have effectively protected incumbent regimes from facing the imperatives of adjustment.

Nonetheless, given the high expectations of the early and mid-1980s, the record for externally supported economic restructuring efforts in Africa has been disappointing. But the fact is that nowhere in the world has fundamental reform ever been substantially driven by external actors. In both Asian and Latin American countries, fundamental reform took place only when domestic leaders put in place programs that went far beyond anything suggested by the IMF and the World Bank. This has not yet happened anywhere in Africa.

While the donor community generally has been well-motivated in assuming ever greater responsibility for responding to Africa's economic crisis, it may have been mistaken to do so. First of all, as the analysis presented in this chapter has shown, its ability to address such large issues is inherently limited. More importantly, taking on such responsibilities has had a serious downside-encouraging a growing "dependency syndrome" in which Africans assume that their problems can and will be addressed by outsiders.

In the 1990s, with the end of the Cold War, the increasing demand for resources from Eastern Europe and the former Soviet Union, and the growing "aid fatigue" in Western capitals, the likelihood of the donor community continuing to expand its role in Africa is limited. This is not necessarily a bad thing. For the donors, and donor conditionality, have "hemmed in" Africa in a very peculiar way; not by imposing inappropriate strategies or policies, but by substituting external pressure and financial resources for domestic leadership and an indigenous process. In the 1990s, with luck, the processes of African democratization and relative external disengagement may provide a more conducive environment for a development breakthrough than did the 1980s pattern of massive external financial involvement and omnipresent donor conditionality. This is not to say that African economies can recover without external support, but that such support can, at most, play a supportive role in Africa's development efforts. The main initiatives for African development must come from Africans themselves.

---

Note 1: Percy S. Mistry, African Debt: The Case For Relief For Sub-Saharan Africa (Oxford: Oxford International Associates, 1988), p. 4. Back.

Note 2: Thomas Callaghy, "Lost Between State and Market," in J. Nelson, ed., Economic Crisis and Policy Choice: The Politics of Adjustment in the Third World (Princeton: Princeton University Press, 1990), p. 258. Back.

Note 3: For an excellent survey of debt-related issues in the Third World see Jeffrey Sachs, ed., Developing Country Debt and Economic Performance (Chicago: University of Chicago Press, 1989). Back.

Note 4: International Monetary Fund, World Economic Outlook 1980 (Washington: International Monetary Fund, 1980) p. 76. Back.

Note 5: Rupert Pennant-Rea, The African Burden (New York: Twentieth Century Fund, 1986), pp. 13-16. Back.

Note 6: For an excellent discussion of African debt issues by two senior World Bank staff members, see Charles Humphreys and John Underwood, "The External Debt Difficulties of Low-Income Africa," in I. Husain and I. Diwan, eds., Dealing With the Debt Crisis (Washington: The World Bank, 1989). Back.

Note 7: Mistry, African Debt, p. 16. Back.

Note 8: Pennant-Rea, The African Burden, p. 14. Back.

Note 9: Humphreys and Underwood, "The External Debt Difficulties." Back.

Note 10: Jonathan Frimpong-Ansah, "Sub-Saharan Africa and the International Trade System," unpublished manuscript, 1988, p. 44. Back.

Note 11: See Callaghy: "Lost Between State and Market," for a discussion of the politics of adjustment in Nigeria. Back.

Note 12: Joshua Greene, "The External Debt Problem of Sub-Saharan Africa," IMF Staff Papers, Washington, D.C., December 1989. Back.

Note 13: Mistry, African Debt, p. 8. Back.

Note 14: World Bank, World Debt Tables 1990-1991 (Washington: World Bank, 1991). Back.

Note 15: The data used in the following section are primarily drawn from Percy Mistry's excellent study that is cited earlier and supplemented by recent IMF and World Bank figures. Some of the calculations are my own. Back.

Note 16: World Bank, World Debt Tables, p. 92. Back.

Note 17: This working paper was later published. See Dani Rodrik, "How Should Structural Adjustment Programs Be Designed?," World Development (July 1990). Back.

Note 18: G. K. Helleiner: "Structural Adjustment and Long-Term Development in Sub-Saharan Africa," unpublished paper, 1989. Back.

Note 19: For a recent restatement of the "reinforcement" position by senior World Bank staff see Vittorio Corbo and Stanley Fischer, "Adjustment Programs and Bank Support: Rationale and Main Results," unpublished manuscript, August, 1990. For an example of the "imposition" perspective, see Robert Browne, "Conditionality: A New Form of Colonialism" in Africa Report, September 1984. For the "purchase" argument, see Paul Mosley, Conditionality as Bargaining Process: Structural Adjustment Lending 1980-1986 (Princeton: International Finance Section, Princeton University, 1987). Back.

Note 20: For a similar argument see, M. Kahler, "International Financial Institutions and the Politics of Adjustment," in J. Nelson, ed., Fragile Coalitions: The Politics of Economic Adjustment (New Brunswick: Transaction Books, 1989). Back.

Note 21: See the World Bank's Long-Term Perspective Study on Africa, Sub-Saharan Africa: From Crisis to Sustainable Growth (Washington: World Bank, 1989). Back.

Note 22: For examples of this syndrome, see the country case material presented in P. Mosley, J. Harrigan, J. Toye, Aid and Power: The World Bank and Policy-Based Lending in the 1980s (London: Routledge, 1990). Back.

Note 23: World Bank, Accelerated Development in Sub-Saharan Africa (Washington: World Bank, 1981). Back.

Note 24: Ibid. Also see Michael Roemer, "Economic Development in Africa: Performance Since Independence and a Strategy For the Future," Daedalus (Spring 1982). Back.

Note 25: For the original economic theory of development assistance see H. Chenery and A. Strout, "Foreign Assistance and Economic Development," American Economic Review 56, 4 (1966). For a review of the empirical studies on the impact of aid see Paul Mosley, "Aid, Savings and Growth Revisited," Oxford Bulletin of Economics and Statistics, 42 (1980). For a recent overview see A. Krueger, C. Michalopoulos, V. Ruttan, Aid and Development (Baltimore: Johns Hopkins University Press, 1989). Back.

Note 26: See Tony Killick, et. al., The Quest for Economic Stabilization: The IMF and the Third World (London: Heinemann Books, 1984). Back.

Note 27: This model is heavily influenced by the recent work of Jeffrey Sachs, published both by the National Bureau of Economic Research and the World Bank. See, for example, "Conditionality and the Debt Crisis: Some Thoughts for the World Bank," unpublished manuscript, 1986; and Efficient Debt Reduction (Washington: World Bank, 1989). Back.

Note 28: For a theoretical discussion of the broader issue of which this is a subset see Robert Putnam, "Diplomacy and Domestic Politics: The Logic of Two-level Games," International Organization (Summer 1988). Back.

Note 29: Killick, The Quest for Economic Stabilization, pp 251-255. Comparative studies include K. Remmer, "The Politics of Stabilization: IMF Standby Programs in Latin America, 1954-1984," Comparative Politics (1986); S. Haggard, "The Politics of Adjustment: Lessons From the IMF's Extended Fund Facility," in M. Kahler, ed., The Politics of International Debt (Ithaca: Cornell University Press, 1986); and J. Zulu and S. Nsouli, Adjustment Programs in Africa: The Recent Experience (Washington: Occasional Paper No. 34, IMF, 1985). Back.

Note 30: Henry Bienen and Mark Gersovitz, "Economic Stabilization, Conditionality and Political Stability," International Organization 39, 4 (1985). Back.

Note 31: Scott Gates, "Micro Incentives and Macro Constraints on Development Assistance Conditionality," unpublished Ph.D. dissertation, University of Michigan, 1989. Back.

Note 32: Vittorio Corbo and Steven Webb, "Adjustment Lending and the Restoration of Sustainable Growth" (Washington: World Bank, 1990). Back.

Note 33: Kahler, "International Financial Institutions and the Politics of Adjustment." Back.

Note 34: Callaghy, "Lost Between State and Market"; and Gates, Micro Incentives and Macro Constraints. For a discussion or the Zambia case see Kenneth Good, "Debt and the One-party State in Zambia," Journal of Modern African Studies 27, 2 (1989). Back.

Note 35: See Center for Strategic and International Studies, The Politics of Economic Reform in Sub-Saharan Africa (Washington: CSIS, 1992). Back.

Note 36: Within the World Bank there is an increasing focus on the problems of implementing policy reform programs. For discussions of the issues involved see John Nellis, Public Enterprise Reform in Adjustment Lending (Washington: World Bank, August, 1989); and Barbara Nunberg, Public Sector Pay and Employment Reform (Washington: World Bank, October, 1988). Back.

Note 37: Ibid., Mosley, Harrigan and Toye. Also, Mohsin Khan, "The Macroeconomic Effects of Fund-Supported Adjustment Programs, IMF Staff Papers, 1990. For Latin America, see M. Pastor, "The Effects of IMF Programs in the Third World: Debate and Evidence from Latin America," World Development (February 1987). For Africa, see B. Ndulu, "Growth and Adjustment in Sub-Saharan Africa," unpublished manuscript, 1990. Back.

Note 38: Jane Harrigan and Paul Mosley, World Bank Policy-Based Lending 1980-1987: An Evaluation (Manchester: Institute for Development Policy and Management Paper, 1989). Back.

Note 39: David E. Sahn, "Economic Crisis and Reform in Africa: Lessons Learned and Implications for Policy," in D. Sahn, Adjusting to Policy Failure in African Economies (Washington: Cornell University Food and Nutrition Policy Program, 1992). Back.

Note 40: World Bank, Third Report on Adjustment Lending (Washington: World Bank, 1992). Back.

Note 41: Sahn, "Economic Crisis and Reform in Africa." Back.

Note 42: Robert Bates, "The Reality of Structural Adjustment: A Skeptical View," in Simon Commander, Structural Adjustment and Agriculture (Portsmouth: Heinemann Educational Books, 1989). Back.

Note 43: Naomi Chazan, "Africa's Democratic Challenge," World Policy Journal (1992). Back.

Note 44: Nicolas van de Walle, "Rent Seeking and Democracy in Africa," unpublished MS, 1992. Back.

Note 45: James Bovard, The Continuing Failure of Foreign Aid (Washington: Cato Institute, 1986). Doug Bandow, "What's Still Wrong With the World Bank," Orbis (Winter, 1989). Back.

Note 46: An earlier essay that agrees with some elements of the conservative critique while rejecting its ultimate validity is E. Berg and A. Batchelder, "Structural Adjustment Lending: A Critical View" (Washington: World Bank, Country Policy Department, January 1985). Back.