A Response to ActionAid InternationalBy Calvin McDonald
Advisor, African Department, IMF
May 17, 2007
Senior Education Research & Policy Analyst
It was good to have the opportunity to meet with you here in Washington recently and discuss with you and your colleagues Action Aid's recent report on the IMF, wage bill caps, and education, Confronting the Contradictions. The discussion was very concrete and helped us understand better each other's point of view. We would like to share with you some additional comments on wage bill caps and on the paper in particular. Your report describes the appalling state of the educational systems in three African countries and as such is a call to action to all of us working to improve these systems. As you mention, all three countries have reached a reasonable degree of stability with strong positive growth and should now "be turning their attention to achieving their broader goals, including the MDGs." The IMF's role is to ensure that this happens; we believe that by helping these countries achieve a "reasonable degree of stability" we offer them the best chance to grow faster and thus be able to put more students in better-equipped classroom with more qualified teachers.
But, as your report says, "getting more teachers in schools is more complicated than just finding the resources to hire them. There are several other hurdles in the education sector that must also be addressed." The report proceeds to effectively describe some of the daunting challenges faced by ministries and policymakers in these countries. Many participants in the development process-including the IMF-must to do a better job in addressing these challenges. We believe that our responsibility lies primarily in helping these governments achieve faster growth; increase their capacity to absorb aid; raise revenues; and encourage pro-poor spending. All of this must be achieved while trying to preserve the hard-fought gains provided by macroeconomic stability. To do so, scrutinizing public sector wage bills in many of these countries is simply unavoidable. These bills have too often been a source of macroeconomic imbalances because of unplanned, excess spending, and poor expenditure control. The lack of a central database in Sierra Leone that you describe is just one example. So governments use ceilings when the wage policy is not coordinated with sectoral priorities and the availability of money to pay for them.
We have described the many reasons for including wage bill ceilings in a paper that you often quote, Aid Scaling Up: Do Wage Bill Ceilings Stand in the Way?. We concluded that, based on country case studies for 2003-05, wage bill ceilings have not restricted the use of available donor funds, and offered a number of suggestions for increasing the flexibility of wage bill conditionality in PRGF-supported programs to respond to higher aid flows. But the paper also described how ceilings are not the best way to solve the underlying problems of a poor country's total wage bill. This is one of the reasons why, as we mentioned in our meeting (and as detailed in this recent article in the IMF Survey) the Fund is committed to use ceilings more selectively and transparently.
In fact, the number has already been declining. As of April 2007, wage bill conditionality was in place in 9 PRGF arrangements out of a total of 29, with quantitative performance criteria, the primary indicators for assessing program outcomes, used in four cases: Central African Republic, Chad, Malawi and Mali. Indicative targets are included in the PRGFs of Benin, Burundi, Niger, Sierra Leone and Zambia.
Going ahead, we are focusing more on working with governments and development partners to address the underlying issue of civil service reform. If and when the ceilings are part of an IMF-supported program-and here we would differ from your portrayal of a blanket practice of imposing them-there must be a solid macroeconomic justification. The Fund is firmly committed to explain such a decision and the decision-making process that led to it.
We should never forget that these economic considerations must be put in the context of the tremendous and urgent needs in crucial social sectors. To achieve the MDGs, there has to be a scaling up in services, often meaning not only better but also more spending to pay wages and training of nurses and teachers. The Fund has shown the flexibility to address these complex issues. You describe how, in your three country cases, the IMF has taken the needs of the education sector into consideration: "In Mozambique, the wage bill ceiling was expanded by 1% of GDP specifically so the Government could hire more teachers. The Government of Sierra Leone has been granted annual "top-ups" for hiring more teachers. In Malawi, the education sector was exempted from the recruitment freeze and an automatic adjuster was added to the ceiling to allow the payment of salary top-ups for health sector workers and hiring more health sector workers if additional donor financing became available through the SWAP framework."
You go on to say that "these were steps in the right direction and enabled Governments to make some progress in hiring more teachers. However, our research also showed that the Fund needs to move further and faster to allow Governments to reclaim control over hiring decisions."
This is the central issue. These countries, and many others in Africa, must reclaim control over their budgets. They must count on reliable sources of income, and on realistic spending plans that are based on sound sector plans. So we agree with you, and we must ensure that we do our part, and that we do a better job. But we would like to add: this is a responsibility of the IMF and the international community. The example of the health sector in Malawi is illuminating. As you point out, two factors changed the situation: "First, when the Ministry of Health successfully made the case for expanding the wage bill ceiling to hire more health workers, the IMF proposed an adjuster in the wage bill, allowing the salary top ups provided by DFID to be accommodated within the wage bill ceiling...The Ministry of Health developed a credible, long term plan for financing health and therefore there was great confidence in the capacity of the Ministry to deliver this plan. Accordingly ... the plan received substantial backing from donors, who significantly increased resources to the sector."
The example shows that when there is a clear commitment by the authorities and by the donors, the IMF not only is not the problem: it is part of the solution. As the situation in many African countries changes so does the Fund.
Your paper also addressed other key questions, such as the Fund's policy advice on aid absorption. The recent IEO report on aid to SSA found that for countries with low reserves (below 2.5 months of imports), incremental aid is on average used for strengthening the reserve position, and that when inflation exceeds 5-7 percent, programmed aid-based spending had on average been limited. The IMF does not have a specific policy threshold on aid absorption and does not apply rigid rules or one-size-fits-all: no across-the-board rule for using aid could do justice to the many aspects involved, and be appropriate for all countries. The Fund's analysis of the scope for using aid takes into account many factors, not just inflation and reserves: important considerations are aid volatility, the incidence of shocks, debt sustainability concerns, export competitiveness, the domestic debt burden and microeconomic capacity constraints to higher spending to ensure the aid can be effectively spent.
We are currently drafting two papers that look at the implications of scaling up on Fund program design and on fiscal policies. We hope that these papers will answer many of your questions and we look forward to discussing them with you at the appropriate time.
The wage bill ceiling is not set at 7.2 percent of GDP as indicated in the report. The wage bill ceiling is set at the time at which a budget framework is agreed with the authorities and reviewed at each program review, usually every six months. It accommodates the government's plans for new hires.
The Fund-supported programs in Malawi have emphasized social spending (health and education) since the country reached the HIPC Initiative decision point in December 2000. Two specific HIPC Initiative completion point triggers highlight this (HIPC completion was reached in August 2006): (1) share of health expenditures of at least 13 percent of discretionary recurrent budget. The share averaged around 12 percent in 2001/02-2004/05, but increased to 18 percent in 2005/06; (2) share of education spending in the discretionary recurrent budget of at least 23 percent. The share averaged 29 percent between 2001/02 and 2004/05, but dropped to 21 percent in 2005/06 as a result of the emergency food operations.
The ActionAid report missed an opportunity to discuss the real constraints facing the education sector in Malawi. The introduction of free primary education in 1994 has put a strain on the education system, with the result that the quality of education fell in the 1990s. The government has been unable to increase yearly enrollment of 6,000 students for teacher training on a sustained basis as was required by the HIPC completion point. Yearly enrollment increased from 3,000 in 2000 to 5,490 in 2005, but this was because two cohorts were enrolled in one year under a crash program. The double-cohort teacher training crash program has been discontinued because it led to a reduction in quality.
A new approach has now been adopted. Since September 2005, the Initial Primary Teacher Education program was introduced. The number of teachers deployed in primary schools after every two years will now be 6,000 (twice that under the previous two-year program). With donor support, teacher training colleges are also being expanded, and enrollment is expected to increase significantly over the next two years.
The paper argues that the IMF "continues to constrain teacher recruitment in 2007." This is incorrect. The 2007 budget reflects the hiring of about 10,000 teachers, in line with the sector strategy agreed between the authorities and donors, including the World Bank. The IMF does not decide on the level of teacher recruitment as it is outside the expertise of the Fund.
The section on the impact of IMF policies on education tries to evaluate the impact of wage bill ceilings in Mozambique in 2007. However, there is no wage bill ceiling, so this section is misleading.
The IMF mission regularly meets the Minister of Education during Article IV and program review missions, particularly budget missions. In fact, during recent Article IV discussions, the Minister of Education noted that human capacity constraints (both personnel and training capacity) and lack of long-term financial commitments by donors remained major impediments to expanding teacher recruitment.
There is no cap on the number of teachers in the Fund-supported program, contrary to the suggestion in the report that the number of teachers was capped in 2007 to 33,122. In discussing the ceiling on the total wage bill, the Ministry of Finance indicates the number of needed new teachers-as agreed with the Ministry of Education-and the wage bill is adjusted to accommodate these hirings. Thus, in 2006, the program projected a 17 percent increase in the wage bill in part to accommodate the hiring of an additional 3,000 teachers. It is expected to increase by another 14 percent in 2007, reflecting the cost of recruiting 2,000 additional teachers. The PRGF program includes an indicative target (ceiling) on the wage bill that has been set after discussions with the Sierra Leonean authorities and is based on their own estimates. It has been our understanding that the Ministry of Education was the main source of these estimates.
The paper incorrectly states that following IMF advice, the government has curtailed expenditure, bringing inflation rates down from 12.1 percent in 2005 to 5 percent in 2008. In reality, total expenditure will increase by about 2.7 percent of GDP to 27.3 percent of GDP by 2008 to allow a significant increase in development (capital) expenditure (from 5.9 percent of GDP to 9.6 percent) while protecting pro-poor expenditure.
Finally, the report incorrectly refers to the 2007 Article IV consultations; it should read "the 2006 Article IV consultations".