A Response to ActionAid International and Other Organizations By Thomas C. Dawson, Director, External Relations Department, IMF

September 30, 2004

By Thomas C. Dawson
External Relations Department, IMF
September 30, 2004

This letter and attachment were sent in response to a paper on the fight against HIV/AIDS prepared by ActionAid International USA and other organizations

Your organizations have recently released a report on the effort to address the terrible spread of HIV/AIDS in the developing world, "Blocking Progress: How the Fight Against HIV/AIDS is Being Undermined by the World Bank and International Monetary Fund." The report contains certain important findings and recommendations. Nonetheless, we find it regrettable that it also contains allegations, errors and misconceptions that do not reflect our consultations over time.

The effort to combat HIV/AIDS requires an ongoing debate over issues related to economic, public health and foreign aid policies. The IMF is always ready to engage in discussions that have the potential to enhance our understanding of these issues-and to improve our operational approach to them. Sadly, a report that could have sought common ground and made a constructive contribution to the ongoing exchanges has instead become a vehicle largely for recriminations and accusations.

Consequently, we feel obliged to issue our response publicly in order to clarify and correct many of the issues you raise. We still look forward to discussing this issue-and others-with your organizations in the future.

The policy brief "Blocking Progress: How the Fight Against HIV/AIDS is Being Undermined by the World Bank and IMF"1 concludes that the Fund is standing in the way of increased spending to combat HIV/AIDS. The report concludes that low inflation targets under the IMF-supported programs and, thereby, restrictions on overall public spending in low-income countries are making it difficult for countries to hire more doctors, nurses, and health workers, as well as to buy the medicines required to fight the HIV/AIDS pandemic effectively.

This note clarifies the issue of the IMF role in HIV/AIDS assistance. In this context, it sets the record straight with respect to the Fund's stance in the specific country examples discussed in the policy brief.

In sum, the policy brief is partly correct but fundamentally wrong in how it assesses the role of the Fund in the fight against HIV/AIDS. It is right in that it acknowledges some of the basic tradeoffs between macroeconomic stability and the fight against HIV/AIDS and because it concludes with some sensible recommendations, which we share. It is wrong in its core analysis of what the Fund does.

On reading the brief, it would be understandable to ask: how could the Fund behave in such a way?

· How could the Fund care so little about AIDS that it would prevent spending that would save countless lives?

· How could the Fund push such low inflation targets when (1) the cost is enormous in terms of foregone spending on HIV/AIDS and (2) the benefits are so low?

· How could the Fund not allow countries to spend concessional foreign loans or even grants in support of the fight against HIV/AIDS?

The answer is that all of those claims are wrong.

The IMF stance on the fight against HIV/AIDS

The IMF has endorsed the call by the United Nations Secretary-General for a global campaign in the fight against HIV/AIDS, and is collaborating with the UN community to expand country-level AIDS prevention and treatment programs. In December 2003, the IMF Managing Director noted that "the IMF must do everything possible within the context of its mandate to assist the agencies that are in the front lines of the fight against HIV/AIDS."2 Accordingly, the IMF is working to ensure that countries allocate adequate budgetary resources to priority expenditures, including for the fight against HIV/AIDS.

Central to this effort are the Poverty Reduction Strategy Papers (PRSPs) that are prepared by low-income country governments in collaboration with civil society and development partners. Measures to fight HIV/AIDs feature prominently in many of the PRSPs. These country-owned PRSPs provide the operational basis for concessional lending by the Fund and Bank and for debt relief under the Heavily Indebted Poor Countries (HIPC) Initiative. Since the inception of the PRSP approach in 1999, the share of poverty-reducing expenditures in 23 low-income African countries that have reached their decision points under the HIPC Initiative has increased on average by about 2 percent of GDP.3 Moreover, the IMF, in cooperation with the World Bank, is also helping low-income countries to improve their public expenditure management systems to ensure that funds, including those for all health and HIV/AIDS programs, are used efficiently and transparently.

Inflation Targets: What Do IMF-Supported Programs Really Advocate and Why?

The policy brief by the NGOs claims that the IMF undermines the fight against HIV/AIDS by insisting that keeping inflation low is more important than public spending to fight HIV/AIDS. However, this claim is wrong. The IMF strongly supports macroeconomic stability because it is a necessary condition for economic growth and poverty reduction, without which lasting improvements can not be made in public health conditions.

However, there is no evidence that attempts to systematically target high inflation rates above a few percentage points will work: they will not create more growth or more room to spend on HIV/AIDS. When the higher inflation rate becomes embedded in expectations, it serves only to create uncertainty and complicate macroeconomic management.

It does not allow higher growth; on the contrary, growth is generally lower, and the paper leaves out much evidence that contradicts its arguments:

· The paper claims that IMF-supported programs target excessively low levels of inflation-"generally, below 10 percent, and often as low as 3-5 percent"-and notes that research, including by Michael Bruno and Bill Easterly dating back to the mid-1990s, has shown that higher rates of inflation ("up to 20 percent") have no discernable impact on growth.

· The Bruno and Easterly finding has inspired substantial subsequent research, including at the IMF. This work confirms that higher inflation robustly leads to lower growth. Further, it shows that this negative relationship starts at fairly low levels of inflation-somewhere between 2 to 11 percent.4

· Higher inflation targets are also a bad way to create room for more public expenditures, however worthwhile these may be. The amount of expenditure that the government can make depends on available financing. Higher inflation can generate resources for the government through the so-called inflation tax, but this tax is inefficient, regressive, and unreliable. There are much better alternatives in all countries.

· The IMF does not target extremely low rates of inflation. A recent evaluation by the Independent Evaluation Office (IEO) of the IMF concluded that IMF programs for low income countries did not show an excessive deflationary bias.5 The IEO found a clear tendency in program design against tolerating double-digit inflation. But in the majority of cases where initial inflation was between zero and five percent, inflation was projected to go up. In light of the available evidence, countries that follow PRGF-supported programs should reap the benefits in the form of higher growth.

Fiscal Policy Design in Fund-Supported Programs

Fiscal frameworks in PRGF-supported programs are designed to help countries mobilize and accommodate large sizable inflows of foreign aid, provided that macroeconomic stability and financial sustainability are maintained. Thus, the claim that the IMF's low inflation targets constrain the room for government spending on AIDS programs, even if donor funding is available, is wide of the mark. Its argument includes a series of misunderstandings regarding the IMF's operations:

· IMF programs in no way limit the scope for government spending out of domestic tax revenues, contrary to what the paper argues.

· Contrary to the paper's claim (page 20), IMF programs for virtually all low income countries treat foreign grants as a part of government revenue. This means that the receipt and spending of grant money does not raise the government deficit and, as a result, is not subject to program limits on the deficit or its financing.

· The vast majority of Fund-supported programs also include no limit on the spending of concessional project loans.

In sum, in the context of IMF-supported programs, available resources from domestic revenue and foreign grants can be used for government spending, including on the fight against HIV/AIDS.

Accommodating High Aid Flows in IMF Programs

The claim that IMF conditionality deters countries from using available foreign aid is also grossly misleading. A key objective of most IMF-supported programs is to ensure that the conditions necessary for absorbing foreign aid are in place, again, with a view to promoting growth and poverty reduction.

The IMF programs for low-income countries (i.e., PRGF-supported programs) clearly envisage that the fiscal and financing targets in programs will accommodate higher poverty-reducing spending in cases where additional foreign aid flows on concessional terms are available and absorptive capacity exists.6 While the impact of high aid flows on competitiveness and fiscal and debt sustainability should be considered, in most low-income countries these considerations should not pose an insurmountable problem in utilizing additional grant aid. Even in cases where governments believe that they need to cap aggregate spending on macroeconomic stability and sustainability grounds, additional spending for poverty reduction including for health programs could be accommodated by cutting back spending in other areas.

In many cases, a key factor limiting aid inflows in support of the MDGs has been a government's failure to follow through on policy commitments on which this aid has depended. Moving forward, it will also be essential that for this large-scale assistance, disbursements closely match commitments and that there is some assurance of a long-term flow of resources from the donors.


The paper's misrepresentations are disappointing, particularly when its four "issues for consideration" are matters on which the authors of the paper and the Fund substantively agree. In particular:

(i) Grant aid should be allowed to finance higher deficit spending: We have not been able to trace an IMF-supported program that does not treat grant aid in such a manner.

(ii) Need to make future grant aid more sustainable and predictable: This is an objective we strongly support.

The limits that IMF programs impose on the use of nonconcessional borrowing, both from domestic and foreign sources, and, in some cases, on the overall fiscal deficit, are often related to risks to debt sustainability. The paper, however, erroneously attributes all fiscal conditionality in IMF-supported programs to inflation objectives. In order to support growth and stability, it is critical to avoid a recurrence of the debt crises, which resulted in a major setback to economic development in many low income countries during the 1980s and into the 1990s. However, given that the objective is to avoid incurring costly debt obligations, the answer to such constraints should be a shift in aid inflows from loans to grants. Provided that donors are willing to accommodate such a shift, there would be no need to forego the projects at hand, including on HIV/AIDS.

(iii) Money is needed for public health systems as well as fighting HIV/AIDS. Again, we agree. Higher investment in human capital by low-income countries is critical for long-term economic development. The paper itself recognizes that issues related to building effective public health systems are "beyond the scope of the issue of IMF low-inflation targeting." If the idea is for Fund-supported programs to accommodate higher inflation targets, while this might permit higher nominal expenditures. it does not mean that expenditures in real terms will be any higher. If the concern is that IMF-supported programs constrain wage bills, with the exception of few programs where the wage level is either excessively high relative to other low-income countries or a rapid increase risks causing serious macroeconomic destabilization, Fund-supported programs do not have wage ceilings.

(iv) Other ways to Increase Spending to fight HIV/AIDS. The IMF provides extensive financial and technical support to help low-income countries improve their tax systems, so that they are transparent, efficient and equitable.

Notes on Country Cases

· No global funding for HIV/Aids projects has been rejected by Uganda because of overall budget limits. The NGO paper is wrong to say that a $52 million grant was "nearly" rejected by Uganda because of IMF budgetary constraints.

· Disbursements under the Global Fund have been constrained by administrative capacity, more than macroeconomic considerations. Of Uganda's approved funding of $201 million (out of a requested $370 million), only $18.6 million has been disbursed to date. There were long delays in sending a Global Fund team to verify Uganda's procurement practices, and Uganda has had four Global Fund portfolio managers in one year. Administrative capacity and complex bureaucratic procedures have also been an issue in Uganda: for example, the Country Coordinating Mechanism (CCM) was established separately from the Uganda Aids Commission. Donors with direct responsibility for Aids/HIV issues in Uganda (UN AIDS, UNICEF) are working to reduce these obstacles.

· The IMF has played an active role, with other donors, in defending health spending. The government's draft 2004/05 budget proposed a cut in the health spending from Ush 386 billion in 2003/04 to Ush 330 billion in 2004/05 (down from 2.9 percent of GDP to 2.3 percent of GDP). The IMF worked with the donor community to restore Ush 49 billion to the health budget (over 0.3 percent of GDP). Of this, roughly half was government financed and half donor financed.

· The IMF has supported increased salaries for health care workers. The 2004/05 budget includes an additional Ush 33 billion (0.2 percent of GDP) for higher health care salaries.

· Spending on poverty-alleviation programs, including health care, is protected in Uganda by assignment to a special fund (Poverty Alleviation Fund) that is exempt from within-year spending cuts and reallocations.

· Uganda has achieved a great deal on HIV/AIDS. HIV prevalence has been reduced from 18% in 1992 to 6.2% in 2002. But more needs to be done on treatment. About 25,000 people are currently receiving anti-retroviral treatment out of a needed 120,000.

· The economic programs that the IMF has supported in Zambia under the PRGF have not included limits on government hiring in health or education, as alleged by the NGO report. Teachers, doctors and nurses were excluded from the freeze on civil service hiring in the government's Memorandum of Economic and Financial Policies dated November 2002. (available at www.imf.org). The program supported by the PRGF arrangement approved in June 2004 does not include a freeze on hiring in the public sector.

· The case of Zambia illustrates that a key factor limiting aid flows has been the government's failure to follow through on policy commitments on which this aid has depended. Moving forward, it will also be essential that disbursements closely match commitments and that there is some assurance of a long-term flow of resources from the donors. In Zambia, the disbursement of donor assistance has been a small share of that committed. Donors are currently taking steps to coordinate better with the authorities to improve the reliability of aid delivery.

· The government wage bill in Zambia, which rose sharply from 5.3 percent of GDP in 2000 to 8 percent of GDP in 2002, has increasingly been a source of pressure on the fiscal deficit and a key issue in policy discussions with the IMF. This increase in the wage bill has been primarily due to large wage increases rather than new hiring. In April 2003, the government granted wage increases and introduced a new housing allowance. These increases were not part of the budget approved by parliament. If these amounts had been paid in full, total remuneration of government employees would have exceeded the budget allocation of 8 percent of GDP in the 2003 budget by 2.5 percent of GDP.

· The IMF staff and the government of Zambia agreed that payment of this large wage bill and related allowances in 2003 would have jeopardized the economic revival that is critical to reducing poverty. Zambia's economy grew by an average of 4 percent per year from 2000 to 2002. Per capita income thus rose after more than two decades of economic decline and rising poverty. But this recovery remains fragile, with inflation still running at around 20 percent. Payment of the high wage bill would have endangered the recovery and compressed priority spending. Specifically:

  • Financing the wage bill through monetary expansion would have brought even higher inflation, which would have had negative effects on growth and reduced the real incomes of the poor, who are least able to protect themselves from inflation.
  • Attempting to meet the increased wage bill through budget cuts elsewhere would not have left sufficient funds for priority spending to support the government's poverty reduction strategy.
  • Financing the increase in wage bill through additional government borrowing would have resulted in further increases in interest rates, thereby increasing the burden of domestic interest payments, and leaving still less room for priority spending.

· Against this background, understandings were reached on a program for 2003 that was monitored by IMF staff as a precursor to a new PRGF arrangement. The government reopened negotiations with public service unions on housing allowances, in an effort to contain growth in civil service compensation in 2003 and set the basis for returning the wage bill to a more manageable level in 2004.

· As a result of the large and unplanned increases in wages and related payments in 2003, some ministries found that their budget allocations did not allow for recruitment of new staff during the year. However, the government could have made use of the savings from the removal of some 4,000 employees from the payroll to hire more staff in priority sectors. But in the event, new hirings were not focused on health and education.

· Structural factors have also impeded the timely recruitment of staff. For example, there are currently some 9,000 teachers waiting to join the teaching service. However, there are also over 7,000 teachers ready to leave the service but not able to do so because the government could not afford to pay their terminal benefits. They have therefore remained in the payroll. This problem is not unique to Ministry of Education. The IMF, together with Zambia's development partners, is working to find ways to resolve this structural problem that has impeded hiring in some key sectors.

1 Issued by a group of NGOs (ActionAid International USA, Global AIDS Alliance, Student Global AIDS Campaign, and RESULTS Education Fund) in September 2004.

2 IMF Press Release N0. 03/208.

3 IMF, Heavily Indebted Poor Countries (HIPC) Initiative - Status of Implementation (August 20, 2004).

4 Khan, M. and A.S. Senhadji, Threshold Effects in the Relationship Between Inflation and Growth, IMF Staff Papers, Vol. 48 (2001), and Ghosh, A. and S. Philips, Warning: Inflation May be Harmful to Your Growth, IMF Staff Papers, Vol. 45 (1998).

5 IEO Evaluation Report on PRSPs and the PRGF, Box 4-7 (July, 2004)

6 IMF, Aligning the Poverty Reduction and Growth Facility (PRGF) and the Poverty Reduction Strategy Paper (PRSP) Approach - Issues and Options (March, 2003).


Public Affairs    Media Relations
E-mail: publicaffairs@imf.org E-mail: media@imf.org
Fax: 202-623-6278 Phone: 202-623-7100