Information about Sub-Saharan Africa África subsahariana

Matthew Martin

I. Patel
Published Date:
December 1992
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

African central bankers and IMF officials have shared at this symposium a wish to improve the preparation, negotiation, and monitoring of adjustment and external finance. There are two key ways to achieve this: by reducing transaction costs and by improving information. The first means rationalizing the staff time—and funds—both groups spend in interminable rounds of multiple negotiations. The second means improving the timeliness focus, and accuracy of economic data and analysis behind adjustment programs. The two obviously interact: rationalizing negotiation would free time and money for analysis and monitoring; and better information would save time in negotiations.

Reducing Transaction Costs

To put together a typical package of adjustment and finance, most African governments have to negotiate with the following groups: (1) the IMF and the World Bank, for adjustment loans; (2) the Paris Club, for (usually only Organization for Economic Cooperation and Development (OECD)) bilateral debt; (3) the London Club, for rescheduling (usually only long-term) debt to banks—-and more recently, for reducing the debt; (4) consultative groups, for (largely OECD) aid disbursement; (5) each non-OECD creditor and donor government separately, for debt relief and aid plans; (6) many other commercial creditors, for short-term debt, uninsured suppliers credits, and import and bridging facilities not in the London Club; and (7) OECD government export credit agencies, for new loans. Each requires extensive preparation, negotiation, and monitoring.1


For adjustment agreements, the government must provide accurate and up-to-date data and analysis on recent and future economic policy; a letter of intent for the IMF; a letter of development policy for the Bank; and (for low-income countries) a policy framework paper (PFP). The Paris Club demands a memorandum on economic policy, a request for relief terms based on recent data, and a speech on these issues. The London Club asks the same as the Paris Club, but data and possible terms are different, and the economic policy memo has usually to be rewritten, owing to poor coordination (of which more shortly). Consultative groups require a report on economic policy, assessing aid needs, usually two or three other reports (on public investment or social sector programs, individual policy issues)2 and several speeches. Other creditors need data, requests and speeches, and sometimes formal documents. Each separate meeting can take three or four days to prepare.

The workloads of the IMF and the World Bank are equally heavy. They involve IMF staff in Article IV talks, on reports of recent economic developments in the country, briefing papers, draft letters of intent, and (usually annual) requests for loans under a stand-by arrangement, the structural adjustment facility (SAF), or the enhanced structural adjustment facility (ESAF).3 World Bank country staff write initiating memoranda, draft letters of development policy, and president’s reports (one or two a year) for adjustment loans; public investment programs, country strategy papers, and country economic memoranda. Both institutions have to agree on a joint policy framework paper for low-income African countries.4

They also have to attend meetings of the Paris Club, London Club, and consultative groups (and often those of other commercial creditors). For each of these, the IMF and World Bank prepare a separate assessment of the government’s debt relief or aid needs and a speech. Their data specialists often have to help the Paris and London Clubs reconcile debt data with those of the government. The Bank also writes a report on the economic program for consultative groups and usually several accompanying documents to match those of the government.


Adjustment negotiations can take three years on an initial program, but more usually last from 9 to 15 months. IMF missions spend 10 to 15 weeks a year in the country, and government staff are in Washington for up to 10 weeks. World Bank structural adjustment loans require more missions and longer stays. If the initial program is implemented as planned, monitoring of implementation and negotiation of the next program run in tandem: each quarter sees one or more IMF review missions, several Bank visits, and several government trips, with the frequency mounting as a new program approaches. Countries with implementation problems may have from 6 to 12 months between programs, with three or four missions each way. Paris Club formal multilateral talks are mercifully short—only one or two days, though prior lobbying by African governments and the IMF and the World Bank may occupy several days. African governments then discuss as many as 19 bilateral follow-up agreements with creditor governments. The Paris Club sets deadlines for concluding these talks, but not one of these was met in 1981-90. Most agreements take from 6 to 24 months, due largely to data reconciliation and administrative problems on both sides (including difficulty in scheduling talks between other debtor negotiations).

London Club rescheduling talks took as much as 30 days over a period of 12 to 18 months, depending on the degree of disagreement between the banks and the government.5 However, banks and African governments are experiencing a long and frustrating delay in the International Development Association (IDA) buy-back facility, due to data and legal problems, and to creditor bank reluctance to reduce debt: negotiations have taken 45 days or more, and 24 months to conclude.

Consultative group meetings are also short (two or three days) but World Bank and government staff face lengthy prior bilateral lobbying of donors (sometimes even pre-meetings to test likely pledges) and subsequent talks to confirm that pledges will be committed and commitments will be disbursed. Other talks can take several years of intermittent meetings, depending on the number of creditors, the insistence of other creditor governments on preferential terms, and the divergent interests of commercial creditors.


In theory, monitoring should occupy several staff full-time in tracking Paris Club bilateral agreements, debt-service payments, aid commitments and disbursements, balance of payments changes, trade finance and export credit availability, and import costs. In practice, it often has lowest priority, owing to the simultaneous demands of preparation and negotiation. The exception is adjustment program implementation (as described above). As one senior African policymaker said to me in 1987: “In order to meet IMF targets, and prepare and negotiate new programs and finance, our team needs to be in five places at once and have 36 hours in a day. Where do we get the time to monitor finance?"

Poor Synchronization and Coordination

Each group of creditors makes decisions separately from other groups and discusses rescheduling and new money separately;6 in addition, many creditors will only talk bilaterally. This raises major hurdles for synchronization and coordination.

IMF and Paris Club talks usually mesh fairly well, with the Club meeting just before or after the IMF formally approves a loan. London Club talks are often badly out of sequence: banks delay signing when the IMF suspends disbursements, until a new program is agreed; sign agreements just before programs break down (and have to abandon them); or do not sign before programs break down (and therefore receive no payment). The same applies to agreements with other commercial creditors—though occasionally they occur just before an IMF-supported program, and are made part of its arrears reduction target. Many consultative groups meet just before IMF loans are formally approved, to maximize donor pledges of aid, due to faith in the country’s commitment to adjustment. But others are held later, and therefore are delayed or undermined by program implementation problems. Many are held when programs have been (or are about to be) suspended, and most donors respond by minimizing new aid pledges.

Paris Club, London Club, and consultative group meetings have no predetermined order, but their actual order can be a key influence on whether the adjustment program is adequately financed. If the Paris Club or consultative group meets after formal IMF Executive Board approval of the program, the program starts with an unfilled “financing gap,” not calculated to inspire Board or creditor confidence. If the Paris Club comes before the consultative group, Club refusal of softer terms can put an extra financing burden on aid. But the reverse order is worse: aid pledges at consultative groups then have to be made assuming a certain level of debt relief and if the Paris Club is less generous than expected, programs may be underfunded before they begin.7

Poor synchronization and protracted talks can have severe negative effects on packages. Given the frequent sudden changes in the balance of payments experienced by many African countries, results of the different talks cannot fit together into an integrated “financing gap” calculation unless they end almost simultaneously. So the IMF and Bank have to recalculate gaps for each meeting. This can undermine creditor confidence in the calculation, reduce debt relief or aid, and leave a gap to be closed by import or reserves cuts or arrears on debt service.

Transaction costs in adjustment and financing negotiations are far too high, due largely to lack of synchronization and coordination. They place an inordinate burden on the staff time of all sides, and encourage duplication of data compilation, analysis of the economy, and document writing. They also encourage a short-term outlook where African governments and the IMF and the Bank cannot plan beyond the current annual program, and creditors look for short-term repayment. Thence the “short-leash” approach of reschedulings with one- to two-year consolidation periods, aid pledges for only one or two years, and annual SAF or ESAF programs, which further increases the workload for all sides.

A recent IMF paper described the resulting overwork and frustration of many African government officials: “In many African countries, policymaking has deteriorated to a state of constant crisis management…,” with beleaguered government officials scrambling from day to day to meet debt-service obligations (Greene and Khan (1990, p. 13)). Many IMF and Bank staff interviewed also criticized current procedure. As one said in Washington last November: “We are here to help African governments design adjustment policies which will enable long-term development, not to rush from meeting to meeting with little time for in-depth analysis.” In other words, the cumulative burden of constant talks leaves everyone less time to focus on the key issues: identifying and monitoring economic policies that are feasible for African governments and external finance gaps which creditors can fill.

Reducing the Costs

In the 1980s, governments, the IMF, and the World Bank often helped each other to reduce the workload. The institutions sometimes drafted letters of intent or development policy for governments to adapt, or advised them on preparing for meetings. Governments provided the institutions with data and information on bilateral contacts with donors and creditors.

As a first step, such cooperation could be formalized in the joint preparation of a single document and data for each negotiating meeting of the Paris or London Clubs and consultative groups. Though adding some time to reconcile different positions, this would save time in separate document-drafting and data compilation. It would make data and analysis authoritative and put them largely above further negotiation, freeing manpower to design and monitor adjustment.

A single negotiating meeting on finance would be preferable. Precedents are available from Mauritania in 1985, where Arab donors pledged aid and debt relief at the consultative group; and from several round table meetings where donors pledged aid and an overall amount of debt relief (which was later confirmed in bilateral talks). It could include all commercial and government creditors that now negotiate bilaterally and all multilateral creditors. It could have separate caucuses for different groups—or agencies of creditor governments—if needed to retain confidentiality.8

The World Bank’s recent effort to discuss debt informally at consultative group meetings and to include non-Paris Club creditor governments could be a precursor for a single meeting in Paris, with administrative support from the IMF, the World Bank, ex-Paris Club Secretariat, key banks, and other data experts. If the average meeting lasted from 7 to 10 days and agreed on outline terms with creditor governments, it would save 10 to 20 days of annual time.

Finally, joint monitoring of finance would give it much more priority: a joint team could track balance of payments developments, export credit and aid, new loan flows and terms, aid disbursement, and debt-service arrears. This would build on the current work of donor-government, joint monitoring committees on aid disbursement.

Improving Information

Quality of Existing Information

Economic data and policy analysis on African economies improved dramatically in the 1980s, owing to strenuous efforts by governments and the IMF and the World Bank. However, most countries still face major faults and omissions in both areas.9


A recent Bank paper described the unreliability, absence, and lack of timeliness in much of the data on Africa, owing to the new demands of multiple users and budgetary, resource, training and staff constraints (Chander 1990). Oxford’s External Finance for Africa (EFA) project has gone into detail on balance of payments and external finance data and found that nobody can be expected to know how much external finance an African country needs.

Trade data are often inaccurate when compared with trade partner statistics,10 or initial statistics have to be revised repeatedly by factors of 25 percent for two or three years. Many countries compile disaggregated data only on one or two nontraditional exports. In many cases we do not know the percentages of world market prices that countries actually receive for their export products. African countries routinely pay 30 percent or 40 percent more for their imports than other developing countries, but in many countries no actual import price series are available, and we do not know enough about import composition or sources to identify causes of overcharging. The same is true of service payments, where detailed figures are often estimated or delayed. Loan and grant disbursement schedules are often unknown. Debt service is underestimated, due partly to creditor reluctance to divulge data until rescheduling talks. Efforts and omissions are therefore large and frequently revised, and some balance of payments items are derived as residuals.

Onto this are built problems with projections. It becomes impossible to project potential nontraditional exports in detail, leading to projections of aggregate percentage rises. Aggregate import data allow only broad elasticity-based analysis of import needs and import use efficiency. Import prices are assumed to fall with liberalization and rationalization; trade-related services are assumed to correlate with trade levels; and external finance is projected based on subjective assumptions about negotiations. In all of our project countries, actual data have not lived up to these assumptions: projections have been systematically optimistic. The actual shortfalls have necessitated import and reserves shortfalls or debt service arrears, undermining the feasibility of (and donor confidence in) the adjustment program.

Economic Analysis

Tony Killick’s forthcoming book, Adjustment Policies in Low-Income Countries, reinforces the case that on adjustment policies for the exchange rate and the agricultural, industrial, or financial sectors, there is still considerable uncertainty on details of implementation methodology, scale and timing for individual countries. As Elliot Berg (1991) has recently put it: “Many of the intellectual underpinnings of l980s-style adjustment lending are contested—such fundamentals as the feasibility of export-led growth, the efficacy and beneficence of deregulated markets, and the desirability of market-determined interest rates or exchange rates.”

The EFA project has examined policies to promote exports, rationalize imports, and mobilize external finance. It has found that successful export policies stress overcoming structural (payment, processing, marketing and transport) problems, export promotion and market development, foreign exchange retention, and import-duty drawback, as well as devaluation and producer price incentives. Results are often distorted by switches from parallel to official market exports, or by weather, and have usually been disappointing because they paid too little attention to price incentives or to the motivations of farmers (particularly smallholder) and manufacturers (particularly small-scale).

Import rationalization has been equally elusive. Market allocation of foreign exchange and imports has had to be gradual. This was partly for political reasons, but also reflected the illiquidity of private sector importers; their inability (or unwillingness due to competing sources of foreign exchange) to cope with the complex specification, tying, procurement and letter-of-credit procedures necessary to use market-allocated funds and their lack of access to imports due to domestic distribution problems. Amounts and forms of external finance often remain incompatible with adjustment priorities and needs. Import support aid to governments is delayed by problems with procurement, documentation, tying, counterpart funds, donor or recipient implementation capacity, and poor coordination leading to duplication, which delay disbursement. The Bank is still having difficulty convincing donors to provide aid directed to the private sector through liberalized import systems. The Program of Action to Mitigate the Social Costs of Adjustment (PAMSCAD) in Ghana, and similar social programs elsewhere in Africa, face slow project design and implementation. There is no mechanism to provide trade credit on local currency terms most importers can afford or foreign currency terms central banks can afford.

Improving the Information

These findings argue for more work on data compilation, analysis, and projection. This can be helped by expanding practical, closely focused technical advice and training, institution-building for specific units, and incentives for data compilers and analysts. The newly established African Capacity Building Foundation (ACBF) could play a lead role here. Equally, we need to rationalize the demands of multiple users and the work programs of data compilers.11 The aim should be to establish one authoritative data base (for example on balance of payments, debt, and finance), using all available expertise of the IMF, Bank, OECD, Commonwealth Secretariat, debtor governments, and other independent advisors. The stress should be on key data needed for adjustment and development apart from those on standard targets, closer identification of interactions between investment, import composition, and growth.

One key mechanism can vastly reduce external finance and balance of payments shortfalls: the deliberately pessimistic balance of payments scenario and the contingency mechanism in Zambia’s program of 1989—90, which allowed excess copper earnings to offset finance shortfalls, should certainly be repeated. In addition, rescheduled debt and new loans could be denominated in SDRs to reduce interest and exchange rate fluctuations. However, other nations will require cheap, total, and timely contingency and compensatory finance; and the compensatory and contingency financing facility fulfills only the last of these characteristics.

Many other participants at this symposium have suggested more humility in recommending policies, and more caution in implementing them: similar safeguard mechanisms to offset unexpected results or exogenous events could easily be devised for other standard adjustment program targets.

Programs need still more stress on overcoming structural export supply problems and on gradualism in import liberalization. They also need to focus on the precise motivations of exporters and importers, the effects of the parallel market, how to optimize import composition without distortionary policy, how to diversify import sources and end-users, how to reduce import prices, and how to increase service earnings and reduce service payments.

The 1988–91 successes in untying aid in the World Bank’s Special Program of Assistance, for example, establishing prototype standard procurement, documentary, and counterpart methods in Mozambique, show how much can be achieved relatively quickly. But more work is needed to raise donor implementation capacity and channel donor funds to priority projects and liberalized import systems. This implies greater consultation between recipients and donors in deciding priority projects and routes to import liberalization: lower transaction costs in existing negotiations could provide the time. Importers also require extensive training in using liberalized import systems. We need to find ways to get aid rapidly to the poor, to the social sectors and to environmental protection and institutional reform. Finally, we need to replace commercial or ECA trade finance loans, which most low-income countries cannot afford, with concessional trade credit.

The Wider and Longer-Term Picture

For most individual African countries, these problems are daunting and the solutions are urgent. They cumulate for the continent and the world. In terms of transaction costs, many IMF, World Bank, and donor/creditor staff cover several countries and face constant meetings. They also have to attend overall “tour d’horizon” sessions of the Paris Club, which discuss recent developments in individual countries’ packages—for these they have to prepare assessments of ten or more African countries, based on constant monitoring; and methodology sessions that decide possible changes in terms and operating procedures—for these they have to decide their positions and whether they wish to lobby for changes. Bank staff have to assess recent developments for each country to prepare meetings for the Special Program of Assistance. Staff of both institutions have to assess developments in country and regional lending in order to make internal decisions on their lending program (or on arrears clearance).

In terms of information, putting together faulty data from individual countries makes it impossible to analyze sufficient needs for debt relief and new money or the need of other low-income, poverty-filled, or debt-burdened developing nations. As data improvements through the SPA have clarified the needs of low-income Africa, so the pressure has grown to find additional finance to minimize rationing among countries. In other words, there have been two problems with “adding up” individual countries’ needs: (1) country estimate + country estimate = continental guess; and (2) better estimate + better estimate = continental shortfall.

Coordination, synchronization, and better information in individual countries’ talks will provide the time to study the continental and global picture. They will also provide more security for all sides to plan for the long-term, encouraging debt relief and new finance, and adjustment programs for periods of five to ten years. Such planning would replace day-today crisis management and enable more analysis of how to raise the quantity and quality of domestic savings and investment; to enhance regional cooperation; to reduce dependence on import, aid and commodity exports; to improve human capital, basic needs, income distribution and prospects for the poor; to protect the environment; and to maximize political freedom. By improving negotiating procedure now, we can free resources to focus on these key issues for African development in the 1990s and beyond.


    BergElliot“Comments on the Design and Implementation of Conditionality in Adjustment Programmes”in Restructuring Economies in Distress: Policy Reform and the World Banked. by ThomasVinod and others (Oxford; New York: Oxford University Press1991).

    • Search Google Scholar
    • Export Citation

    ChanderRameshInformation Systems and Basic Statistics in Sub-Saharan Africa: A Review and Strategy for Improvement World Bank Discussion Papers No. 73 (Washington: World Bank1990).

    • Search Google Scholar
    • Export Citation

    GreeneJoshua andMohsinKhanThe African Debt Crisis AERC Special Paper No. 3 (Nairobi: African Economic Research ConsortiumFebruary1990).

    • Search Google Scholar
    • Export Citation

    JaycoxEdward and others“The Nature of the Debt Problem in Eastern and Southern Africa”in African Debt and Financinged. by CarolLancaster andJohnWilliamsonIIE Special Report No. 5 (Washington: Institute for International Economics1986).

    • Search Google Scholar
    • Export Citation

    KillickTonyThe Adaptive Economy: Adjustment Policies in Low-Income Countries (Washington: World Bank, forthcoming).

    MartinMatthewAfrican Debt Negotiations: No Winners (London: Macmillan Press and New York: St. Martin’s Press1991).

    MartinMatthew and othersExternal Finance for Africa Project: Overall Report (Oxford: International Development CenterJune1991).

    YeatsAlexanderOn the Accuracy of Economic Observations: Do Sub-Saharan Trade Statistics Mean Anything? PRE Working Paper WPS 307 (World BankNovember1989).

    • Search Google Scholar
    • Export Citation

    YeatsAlexanderDo African Countries Pay More for Imports? Yes World Bank PRE Working Paper WPS 265 (World BankSeptember1989).

1Must of this section is based on Martin (1991).
2For example, at a recent meeting, one country had to prepare reports (with the World Bank) on the economic program, food security, debt prospects, and foreign exchange management.
3For a few countries, it has also involved assessing export shortfalls and requesting loans under the compensatory and contingency financing facilities.
4One government with large London Club debt spent almost three months in constant negotiations (as did three staff members of the IMF and World Bank) in 1986.
5The one notable exception was at the consultative group meeting for Mauritania in March 1985, where Arab donors attended and made pledges on both aid and debt relief for 1984—85—though OECD donors did not follow their example.
6As happened for Zambia in 1984; see Jaycox and others (1986).
7The meeting could also include pledging an amount of export credits—as export credit agencies did for Nigeria in 1986.
8Most of this section draws on Martin and others (1991), which presents the findings of the External Finance for Africa Project, a joint study between five African governments and the International Development Centre, Oxford.
9See also Yeats (1989a).
10See also Yeats (1989b).
11Chander (1990) also suggests that data improvements, and how to organize and fund them, should be a main item on the agenda of consultative groups.

    Other Resources Citing This Publication