Comments on CGD Report on IMF Programs and Health Spending by Abdoulaye Bio-Tchané, Director, African Department, IMF

October 5, 2007

By Abdoulaye Bio-Tchané, Director, African Department, IMF
At the Center for Global Development
Washington, DC
September 7, 2007

I would like to begin by thanking Nancy Birdsall for inviting me to participate in this presentation and to comment on the continuing important work of the CGD.

I would also like to congratulate David Goldsbrough and the other members of the Working Group. Like the CGD, we feel strongly that a more focused and effective engagement in low-income countries is critical to understanding policy options. That is why it has so large a role in our medium-term strategy. We are a learning organization, and we appreciate criticism from any corner—as you know we've been learning a lot lately!—but perhaps especially from former colleagues who understand our objectives, point of view, and even the constraints we face.

Our goal is to help our African members achieve their development objectives. Better health is clearly one of them. Focused as we must be on a core set of macroeconomic issues, we do not venture into sectoral decision-making. But as we advise member countries on areas like fiscal sustainability in the presence of aid flows, we are doing what we can to make space for them to exercise their development choices.

We have been engaged on these issues for many years. Indeed, the report draws on recent Fund work, both by our Independent Evaluation Office and by regular Fund staff.

I can say comfortably, we agree with all six of the report's major recommendations for the IMF, as well as the lessons it draws for other stakeholders.

On the first and second recommendations: We agree that it is a central part of our job to help countries explore policy options. As our recent Board guidance has made clear, we need to develop alternative aid and spending scenarios, at the request of the authorities. This is something we are doing literally every day. Just recently we have prepared scaling-up scenarios for Ethiopia, Madagascar, Malawi, Mali, Mozambique, Rwanda, Sierra Leone, and Zambia. Admittedly, we struggle here a bit because the costing estimates—for which the responsibility extends beyond not only our mandate but also our expertise—are often not available. One thing is clear: the momentum for scaling up has slowed as new aid commitments, envisaged in the Gleneagles Summit, have not yet materialized.

On the third recommendation, we believe we can and should do more to promote fuller information about our expectations for aid levels. Our Board has made it clear that our baseline scenarios should generally represent our best estimate of the amount of aid that can be expected. We have also repeatedly urged donors to give more aid. In the next Regional Economic Outlook, we will compare the results of our country-level aid forecasts for the next four years with the Gleneagles commitments. The report is not yet public, but I think I can safely say that, so far, there is little sign that the Gleneagles commitments will be met.

Fourth, we agree, and our Board has recently affirmed, that wage bill ceilings are to be used only in exceptional circumstances. Because we have learned that wage-bill ceilings, though sometimes useful, they are not always the right tool, only 10 percent of current PRGF-funded programs now have them as quantitative performance criteria.

• The wage-bill ceilings in the Malawi program have been instrumental in helping stabilize the economy. This in turn has helped promote growth and reduce interest on domestic debt from 25 percent to 10 percent of total government expenditures. That kind of improvement has made room for large increases in propoor spending. As macroeconomic stability becomes entrenched and management of the wage bill and the civil service improves, it is time to start thinking about removing these ceilings.

Fifth, the report presents a convincing case for expenditure smoothing. Recent IMF Board guidance has emphasized that, when aid falls short, essential spending should maintained. But consider a situation like Zambia's. In 2003 Zambia had such low reserves that there was virtually nothing to draw down to replace the aid shortfall. Another side of the same problem is that when reserves are not adequate and spending is ramped up with every surge in aid, when aid recedes the cuts that become necessary may do more harm than good.

As to the final recommendation, we want to be as transparent and proactive as possible in discussing the rationale for our policy and the assumptions underlying programs. An upcoming Board paper, one in a series dealing with these issues, will discuss our role in the PRSP process.

So we do share the main conclusions of the report. But we also think we have played a more constructive role in helping the authorities consider a range of policy choices that the report is willing to give us credit for. The report sometimes confuses the important point that the Fund can do better, with which we agree, with the argument that we have unduly constrained country policy choices, with which we disagree.

The report, for instance, glosses over the uncertainties that invariably prevail at the onset of IMF programs, and their effects on outcomes.

• Take the case of Mozambique, where the unpredictability of aid was a major important concern. The 2004 IMF program targeted fiscal consolidation because we expected aid inflows to fall after the aid surge of 2000-2003. And in 2004-05 aid inflows were not only lower than in previous years, they were even below what was projected. (Can you believe the IMF may not have been conservative enough?) The fiscal consolidation required by the program limited recourse to monetary financing and helped reduce inflation to single digits. This has reduced pressures on domestic interest rates and led to a substantial increase in credit to the private sector. Once aid began to be scaled up in 2006, projections were revised upwards and illustrative scaling-up scenarios were drawn up.

• When programs began in Rwanda and Zambia, donors were obviously reluctant to step up their engagements. Once they began to change this minds, the staff did explore alternative options for higher spending.

Another point: initial conditions matter more than the report recognizes in determining appropriate policies.

• Rwanda's NPV of external debt was estimated at 326 percent of exports at the end of 2003, even after full delivery of the assistance committed at the Fund program decision point. When debt levels are clearly unsustainable, additional borrowing even on concessional terms has to be ruled out. Our programming for scaling up therefore had to await the prospect of further debt relief and additional grants, which depended as I mentioned on a change in donor attitudes.

Also, something else the report does not recognize: in the end the authorities make the decisions about how money is or is not spent, and they are not necessarily economically naive. They often prefer to avoid overly optimistic aid projections so they can later avoid the need for unpopular expenditure cuts. In general, the decision to depend on uncertain foreign donations to finance key government functions is not easy. The needs are great, but the international community should not dismiss the aspirations of those countries that wish to travel a somewhat different path. The fact that the authorities in Mozambique prefer to reduce aid dependency should not be dismissed, as the report does, as "political rhetoric."

In our experience the authorities are often concerned about real appreciation and potential adverse effects on growth. This has limited the use of aid in some cases. Turning again to Mozambique, Rwanda, and Zambia, as shown in the report, the authorities have accumulated many more reserves than the IMF program required. In these circumstances, they may find it difficult to spend all the additional aid without causing macroeconomic difficulties, because the same dollar cannot be used both to accumulate reserves and finance additional spending. We are working with the authorities to bring fiscal and monetary policy in line with their strategy on aid use.

In sum, our experience, working on scaling up in Fund programs, is that the Fund has generally been supportive of efforts to spend and absorb aid when initial conditions allow. As I suggested earlier, we do not arbitrate between competing sectoral needs. We can help the authorities make room for more social spending, but they decide where they will spend it. Nevertheless, let me highlight the finding in the report that increases in health spending were on average larger in low-income countries that have Fund-supported programs than in those that do not, even controlling for the possibility that having a Fund program attracts more aid. We interpret this finding as direct evidence that Fund programs have not constrained health spending in these countries.

Before concluding, let me mention one important point that the report might usefully have covered in more detail. In my view, administrative capacity constraints, rather than excessively tight macroeconomic policies, may be the main factor constraining rapid increases in spending on health. I have seen this in my own time in Benin as finance minister. With experienced managers in short supply in the health sector, doctors and nurses are often drawn into administrative duties for which they have not been properly trained. We have all seen the enormous difficulties that can emerge when spending exceeds administrative capacity. Think of the governance issues and waste in health spending in Niger and Uganda. Both domestic and external constituencies rightly demand better. That might imply a more measured pace of scaling up. Indeed a survey conducted by the CGD and the International Aids Economics network of international health professionals found that among the most important obstacles to spending available resources on health were poor national coordination (mentioned by 28 percent of respondents), shortcomings in the health care system (14 percent), and absorptive capacity (8 percent). Fund spending limits were mentioned by only 1 percent of respondents.

Let me conclude. We are now at a stage where many low income countries have some stability in their economies. As the report says, we are adapting our work to these new circumstances. This and other CGD reports are a factor in our continuing education. As we learn, we are, I believe, making significant progress and becoming ever more useful to our members.

Thank you.


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6220 Phone: 202-623-7100