The four topics have been arranged in the right order of priority. Domestic savings are the most important source of development finance in most countries; by comparison, foreign inflows (whether private or official) are marginal and debt relief is least important. But Africa is different, and the seminar did not really explain why.

Starting with domestic investment, industrialized countries grew by investing the surpluses earned from agriculture and from trade. Can African countries make the shift from subsistence agriculture fast enough to generate the surpluses required, as China has successfully done? Since so much African trade is in the informal sector, it must be a mistake to penalize the “indigenous entrepreneur” as some speakers proposed; he must be encouraged to invest his profits in his own country. Broadening the tax base can come later. Why is Africa’s personal saving ratio so low compared with similar countries? Does the extended family, so beneficial in social terms, inhibit accumulation of savings? Governments cannot do much to help at this early phase; but they can at least encourage the spread of savings banks that have worked so well in Japan. As the formal sector expands, they can also provide a legal framework for private pension and insurance funds. But this requires new regulatory skills: a spectacular financial collapse could set the whole process back badly.

Foreign investments are needed to supplement local savings, but Africa has lagged behind in attracting direct investment. Indeed many foreign companies are now disinvesting. In most cases, portfolio investment is a long way in the future. I endorse Mr. Jegathesan’s “Ten Commandments,” but add that they have to be translated into legally enforceable contracts. The “Legal principles for foreign investment” approved by the Development Committee in 1993 is a useful model.

Africa is much more heavily reliant on foreign aid than any other region, and receives more aid per capita than many poorer countries. Yet the results have been disappointing. Without a demonstrated payoff, political support for aid will diminish further. More effective use of aid would help. But bilateral flows, and perhaps even IDA, are likely to shrink further in the next few years; African governments must realize this. Aid has recently been described as “a transitory phenomenon of the second half of the twentieth century” Maybe it is time to start thinking of “decremented budgeting,” and systematically building declining levels of aid into country strategy documents, at least as a low-case variant.

Africa has often argued that the burden of international debt has inhibited development. But thanks to repeated Paris Club and bank rescheduling, and a shift of aid into grant form, Africa has paid significantly less debt service, as a percentage of exports, than any other region. The problem is the overhang of the existing debt stock. For most of Africa, commercial bank debt is not a serious problem now. The new “Naples Terms” go a very long way toward removing the governmental part of the burden, if applied flexibly.

Debt to the international institutions is more intractable; and each component has to be looked at separately. The British proposal on IMF debt would successfully soften the terms of ESAF lending. But the principal still has to be repaid, or (in practice) refinanced by new lending from other sources. Where are the funds to come from?

Similarly, the “Fifth Dimension” has successfully reduced the interest cost of old IBRD loans to countries that are now “IDA-only” cases. Yet, the principal of these loans also has to be repaid. At present, it is implicitly assumed that new money, from IDA and from bilateral sources, will replace IBRD loans. Is it safe to rely on a continued flow of such soft money in the future? And, IDA credits also have to be repaid. Although the existing maturity profile does not look impossibly steep, these too will need to be replaced by new money, much of it from IDA itself. Would it be simpler to stretch out the existing credits over a further period of years?

Clearly, new lending by regional banks should increasingly be on “soft” terms to avoid a further buildup of debt; this will cause problems for the African Development Bank in the absence of funds for its soft window. The Bank also faces problems with its existing loan book, which the shareholders will have to face before the total African debt problem can be regarded as solved.