What relief actually freed up

The spending question no one answered cleanly

Debt cancellation under HIPC and MDRI was framed, from the start, as a mechanism for redirecting fiscal resources. The logic was direct: money no longer owed to creditors is money available for schools, clinics, and roads. That framing shaped the completion point conditions, which required each country to show a poverty reduction strategy before relief was locked in. What the framework never resolved was how to measure whether the redirection actually happened.

The difficulty is fungibility. When a government's debt-service bill falls, the saving enters a consolidated treasury account alongside every other revenue stream — taxes, aid, natural resource receipts. Nothing in standard public finance automatically tags it and routes it to health spending. The IMF and World Bank tracked HIPC and MDRI relief flows in their monitoring reports, but tracking relief disbursed is not the same as tracking relief spent on the poor. The two are separated by a budget process that produces numbers only as good as the underlying fiscal data.

Poverty reduction expenditure as a constructed category

To get closer to an answer, the joint IMF–World Bank HIPC framework introduced the concept of "poverty-reducing expenditure" — a budget line that countries were expected to report over time. The category sounds precise; in practice it was defined inconsistently across countries and across years. Health and education were usually included. Infrastructure was sometimes in, sometimes out. Defence was generally excluded but the boundary moved. When the category's definition shifted, trends in the series shifted with it, making comparison over time or across countries unreliable.

The underlying fiscal data also varied in quality. Countries at the decision point — the earlier stage gate before final relief — often had public expenditure management systems weak enough that what the budget said and what was actually spent were different figures. Arrears, off-budget expenditure, and donor funds channelled through project accounts rather than the treasury all complicated the picture. A country could show rising poverty-reducing expenditure on paper partly because its classification system had improved, not because more money was reaching services.

The additionality problem

Donors added their own complication. Many of them counted debt relief as part of their official development assistance, which meant that as relief was delivered, other aid flows could fall without donors breaching their ODA commitments. Whether this substitution occurred — and to what degree — was contested at the time and remains difficult to establish cleanly from the data. The OECD's Development Assistance Committee tracked aggregate ODA figures, but disaggregating debt relief from new-money flows required assumptions about what the counterfactual aid level would have been. Counterfactuals do not appear in national accounts.

On the growth side, some economists argued that reduced debt-service obligations freed up foreign exchange as well as fiscal space, easing a balance-of-payments constraint and supporting investment. The evidence from country studies is mixed. Where debt stocks were so large they had already been trading at deep discount on secondary markets, the actual cash-flow saving from cancellation was smaller than headline figures implied — the creditors had effectively written the debt off already, and the country's borrowing costs had priced in the distress regardless.

What the monitoring left out

The most honest description of the monitoring exercise is that it measured what governments reported spending in categories that governments themselves defined, against a relief figure that was calculated at the point of cancellation rather than traced through the fiscal system year by year. That is useful information — it is not nothing — but it is not a causal account of what debt relief freed up for whom.

Household survey data, had it been timely and consistent enough, might have provided a partial check: if poverty-reducing expenditure rose and service delivery improved as measured at the household level, the chain of evidence would be stronger. In most HIPC countries, the survey coverage was too sparse and too infrequent to close that gap.