Chapter 1 of this volume argued that a long-horizon supply commitment sells an option nobody books. Chapter 2 found the same instrument in a museum statute. This chapter reaches the third instance, and it is the one with by far the most published data attached — which is exactly why the data is worth separating from the argument it is usually delivered inside.
Two disciplines apply throughout. Every figure carries the page it came from, so a reader can check it against the book rather than against this page. And the distinction Moyo herself insists on is kept: her target is systematic government-to-government and multilateral aid, not emergency humanitarian relief and not charitable giving. A critique of her that scores against emergency relief is not a critique of her.
| Figure | Value | p. |
|---|---|---|
| Aid transferred from rich countries to Africa since the 1940s | ~US$1 trillion | 35 |
| … expressed per person alive on earth today | ~US$1,000 | 35 |
| Official aid as a share of public expenditure, continent | 10% | 65 |
| Official aid as a share of GDP, average African country | ≥ 13% | 65 |
| Development assistance as a share of Africa's GDP | ~15% | 36 |
The Marshall Plan is the argument aid proponents reach for first, and Moyo's answer is arithmetic rather than rhetoric: the two programmes differ by an order of magnitude in intensity and by a category in duration.
| Figure | Value | p. |
|---|---|---|
| Marshall Plan transfers, 1948–1952 | > US$13bn | 35 |
| … in today's terms | ~US$100bn | 35 |
| Peak flows as share of GDP, larger recipients (France, Germany) | 2.5% | 36 |
| Ceiling for any recipient across the five-year life | < 3% of GDP | 36 |
| Africa's assistance relative to the Marshall Plan at its height | > 4× | 36 |
| Marshall Plan duration | 5 years, then stopped | 36 |
| Africa's aid duration | ≥ 50 years, no terminal date | 36 |
The Marshall Plan had a term. Recipients knew the date the money stopped, and planned against it. Aid to Africa has no terminal date, which removes the incentive to build an alternative and makes the flow a permanent line of income rather than a bridge. That is an argument about term structure, not about generosity — and term structure is what Chapter 1 of this volume is about.
| Figure | Value | p. |
|---|---|---|
| World Bank study: aid used for purposes other than intended | up to 85% | 39 |
| World Bank loan funds implicated in corruption (US Senate hearing, May 2004) | ~US$100bn | 52 |
| … including other multilateral development banks | ~US$200bn | 52 |
| World Bank lending to developing countries since 1946 | US$525bn | 52 |
| … estimated misused | ≥ 25% (US$130bn) | 52 |
| Uganda 1990s: share of each US$1 of education spending reaching the target school | 20 cents | 53 |
| Capital departing Africa annually (corruption analysts' estimate) | ≥ US$10bn | 57 |
| … as a share of Africa's 2003 foreign aid receipts | ~half | 57 |
| Mobutu's estimated looting of Zaire (Transparency International) | ~US$5bn | 48 |
| IMF disbursements to Mobutu in the decade after the 1978 Blumenthal memo | US$700m | 53 |
The 1978 sequence is the one worth pausing on, because it is not about theft but about the lender. An IMF appointee to Zaire's central bank resigned inside a year reporting that creditors had no prospect of repayment. The Fund then extended Zaire the largest loan it had ever made to an African country. Data (p.53)
| Figure | Value | p. |
|---|---|---|
| 1992 World Bank OED: release of aid tranches | ~100% | 55 |
| … when country compliance with conditions was | < 50% | 55 |
| 1997 World Bank study: adjustment lending 1980–1996 to countries with poor compliance records | 72% | 55 |
| People employed across the aid apparatus (World Bank, IMF, UN agencies, NGOs, charities, government agencies) | ~500,000 | 54 |
Moyo's reading of that last row is the sharpest institutional claim in the two chapters: for most development organisations, success is measured by the size of the lending portfolio rather than by what the money did, and undisbursed funds threaten next year's budget. The pressure runs toward lending, and it does not switch off when the recipient is known to be unaccountable. (pp. 54–55)
| Figure | Value | p. |
|---|---|---|
| Poverty in Africa, 1970 | 11% | 47 |
| Poverty in Africa, 1998 — the period of peak aid flows | 66% | 47 |
| … people | ~600 million | 47 |
| Growth in the most aid-dependent countries, past thirty years | −0.2% p.a. | 46 |
| Zambia, counterfactual per capita GDP had aid since 1960 become investment and investment become growth (Easterly) | ~US$20,000 | 47 |
| Zambia, actual per capita GDP, early 1990s | < US$500 | 47 |
| Africans killed in civil wars, past five decades | ~40 million | 60 |
| Major armed conflicts in Africa during the 1990s / elsewhere in the world | 17 / 10 | 59 |
Correlation between rising aid and rising poverty is in the rows above. Causation is not, and Moyo's chapters argue for it rather than demonstrate it. The 1970–1998 poverty series in particular depends on measurement choices that have been disputed since publication, and the Zambia counterfactual is explicitly a counterfactual — it assumes every aid dollar became investment and every investment became growth, which no economy has ever achieved with any source of capital. Cite it as the upper bound it is.
| Figure | Value | p. |
|---|---|---|
| Countries permanently graduated from IDA since 1960 | 22 | 37 |
| … of which African | 3 | 37 |
| Botswana aid as share of national income, 1960s | ~20% | 38 |
| Botswana average real per capita growth, 1968–2001 | 6.8% | 38 |
| Botswana aid as share of national income, 2000 | 1.6% | 38 |
| Botswana GDP per capita, 2002 / sub-Saharan average | US$8,170 / US$1,780 | 34 |
Botswana is the case both sides claim. Moyo's reading is that the growth arrived as the aid share collapsed from a fifth of national income to under two per cent, and that what did the work was trade openness, monetary stability and fiscal discipline. A proponent reads the same rows as aid successfully financing an early stage and then being outgrown. The rows do not decide between those readings, and a chapter that pretended otherwise would be doing what this volume exists to stop.
Strip the polemic and one mechanism is doing most of the work, on page 58. In a functioning economy the middle class pays taxes and receives accountability in return. Aid weakens the government's financial dependence on its own citizens, and a government that does not need its citizens' money owes them correspondingly less.
Chapter 1: an offtake contract sells the producer's option to transact with anyone else. Chapter 2: a museum statute removes the option to return. Chapter 3: aid substitutes a counterparty who is not the citizen, and extinguishes the citizen's option — the withholding of revenue — which is the only leverage most people ever hold over a state. In all three the transaction is visible and the extinguished option is on nobody's books.
Two corollaries follow, and both are measurable rather than rhetorical.
The encumbrance ratio has a fiscal twin. Aid at 10% of public expenditure and 13% of GDP (p. 65) is precisely the quantity Chapter 1 asks for, applied to the accountability relation rather than to a commodity: the share of a government's operating position that does not depend on its own population. Nobody reports it under that description.
The cost of holding it is real and payable. Uganda's central bank issued US$700m of aid-related sterilisation bonds in 2005; interest alone cost the Ugandan taxpayer US$110m a year (p. 65). That is a recipient paying to hold money it was given, and it is the clearest single figure in either chapter.
Stated at its strongest, because a volume that only publishes numbers it likes is a pamphlet.
The most-cited critique of aid to Africa rests on a table of figures assembled in 2009, and this chapter could not locate a published, updated version of that table. Not a rebuttal, not a vindication — the same rows, recomputed. Seventeen years of argument, and the arithmetic underneath it has not been refreshed in public. That is the same defect this volume found in commodity commitment and in custody instruments, in its third domain running: the position is argued continuously and the ledger is never reissued. Open
Chapter 2 admitted it was written from Newark without asking anyone. The same admission is owed here and is worse, because this chapter is about a policy whose subjects are living and reachable. Nothing on this page was checked with anyone who has received aid, administered it, or had a public budget built on it. The figures are transcribed accurately and the reading of them is untested against a single person with standing to correct it. That correction is invited, and it is worth more than the chapter.