In a recent IPC One pager, Terry McKinley asks why Africa is being constrained in spending aid (IPC OP No. 34, 2007). He implicates вЂњrestrictive, inflation-focused monetary policiesвЂќ sponsored by the IMF as the primary agent blocking fiscal expansion. I agree that the IMF has been overly conservative in formulating fiscal strategy in many programmes with low income countries. However, the core of the problem is not inflation targets. The main issue is the implicit assumptions the IMF makesвЂ”often without much supporting evidenceвЂ”about how the real economy will respond to changes in fiscal deficits and public spending.
Our understanding of the links between macroeconomic policies and ultimate objectives such as growth and progress towards the
MDGs is quite limitedвЂ”much more so than the IMF, or its critics, often imply. In light of this uncertainty, humility is in order and choices on the utilization of additional aid should, to the maximum extent possible, be left to national political processes to decide. For example, following debt relief, the range of fiscal paths (for deficits, spending etc) that are feasibleвЂ”in the sense of not risking renewed debt distress or macroeconomic instabilityвЂ”is greater than many IMF programmes would suggest.
A recent report by the Independent Evaluation Office (IEO) of the IMF, вЂњThe IMF and Aid to Sub-Saharan AfricaвЂќ has triggered controversy about the IMF approach. The report suggests that IMF programmes target only 27 cents of each dollar of additional aid to be channelled to higher public spending. But results vary widely from country to country. The study, which used regressions to examine the relationship between expected increases in aid and programmed uses of that aid, allows for two important insights: i) results explain only a small part of the variation across programmes; and ii) they do indicate that the IMF programmes depend critically on a countryвЂ™s starting conditions. (Both results suggest the IMF is not pursuing a вЂњone size fits allвЂќ approach).
If external reserves are low (less than 2 Р… months of imports), virtually all additional aid is programmed to be saved in higher reserves. If reserves are higher, but domestic macro-conditions fail a high test of stabilityвЂ”which the IEO proxied by inflation of 5 percent or lowerвЂ” the vast bulk of extra aid (85 cents on the dollar) is channelled to reducing domestic debt. Only when reserves are high and domestic macro-conditions are highly вЂњstableвЂќ, most additional aid is programmed for higher fiscal spending.
These results suggest a very conservative policy stance. The IMF is right to take account of the level of reserves and domestic macro conditions when considering how additional aid should be used, but the degree to which these factors influence aid allocation seems excessive. Using part of any initial increases in aid to rebuild reserves is appropriate, but the share allocated to reserves should depend on how long the higher aid is expected to last. IMF programmes seem to assume that all aid increases will be temporary.
Unfortunately, the IEO results have been interpreted as implying that the level of inflation targeted in IMF programmes is the main issue. The IEO used the initial inflation rate, rather than a measure of domestic public debt, as the sole indicator of domestic macro stability because the internal database used by the IMF on performance under its programmes does not have good data on domestic debt. This lack is certainly shocking given the emphasis the IMF has placed on reducing such debt levels and makes it harder to tell what is really driving the fiscal design. However, both recent CGD work and earlier IEO evaluations suggest that the design of IMF fiscal programmes is heavily influenced by several implicit assumptions about how the economy will respond to fiscal expansion or tightening. In practice the country-specific empirical evidence is often quite limited:
Addressing these three aspects of IMF programme design is, in my opinion, much more important than the concern about excessively low inflation targets. That is not to say that the IMF approach to inflation and monetary policy has always been right. It might well have given too much emphasis to achieving very low inflation in Africa. But reversing the recent gains on inflation is not going to yield higher growth, and might detract attention from other, more fundamental problems with the way IMF fiscal programmes are formulated.
Programmes often assume that lower fiscal deficits, especially with domestic financing, will lead to higher private investment (through lower interest rates etc). In practice, the private sector response depends on many other policies and country-specific factors. IMF programmes have systematically overestimated the size and speed of such responses.
The longer-term supply-side effects of higher public spending are, with some commendable recent exceptions, largely ignored in many macroeconomic frameworks. The main IMF focus is on the shorter term; programmes frequently lack concrete medium-term expenditure plans whose longer-term effects can be analysed. Yet, assessing the appropriate medium term fiscal path requires some judgment on the likely impact of higher public spending.
Many programmes combine a conservative approach to initial projections of aid (to avoid programmes being underfinanced)
with conditionality that calls for higher-than-expected aid to be saved and temporary aid shortfalls to be matched by spending
cuts. Such an asymmetric approach reflects implicit assumptions about the duration of aid increases; it also tends to downplay the
costs of temporary disruptions to spending, which can be high (e.g. in the health sector). Only recently have programmes begun
to give greater emphasis to expenditure smoothing.
David Goldsbrough is chair of a CGD Working Group on IMF Programs and Health Spending.