Mittwoch, 21. April 2004

Heads in the sand: The IMF discusses debt sustainability

On Feb. 2nd IDA and IMF staff submitted to their respective boards a paper entitled „Debt Sustainability in Low-Income Countries – Proposal for an Operational Framework and Policy Implications“ [1]. This paper has been discussed at the IMF board on Feb. 23rd and the results of this discussion has been published in Public Information Notice No. 04/34 [2] on the IMF website on April 5th.

These publications are the latest ones in a debate on debt sustainability, which the IFIs have been waging as a response to the critique of their understanding of debt sustainability, as they have been applying it in the context of the HIPC Initiative. Through 2002 and 2003 a few staff papers have already explored alternatives to the rigid application of an across-the-board debt sustainability threshold, which is a key element of HIPC. The 2004 paper, however, is the first in this series to include some quantitative benchmarks, which indicate the direction, which the IFIs obviously intend to take.

This paper summarizes some concerns of debt relief movements regarding this new concept:

1. The context: wrong projections in the past

The biggest merit of the paper consists in admitting that in the past sustainability calculations by the IFIs have been marked by a systematic over-optimism regarding future income and growth rates: Growth projections have on average been 5 percentage points ahead of reality [3]. These considerable errors have not only led to excessive borrowing on the part of the debtor countries as the paper rightfully deplores. Most of all it served to artificially reduce the need for debt relief under Paris Club and HIPC schemes, which otherwise would have come at the expense of (among others) the IMF and the World bank, i.e. the same institutions, which have actually made the projections.

Against this background “more disciplined” projection and analysis are essential, if the institutions do not want to loose the lucrative monopoly they have been enjoying regarding the assessment of countries’ debt sustainability.

2. What is it good for?

Anybody who expected a broad new concept, which might serve to overcome the limitations of HIPC and open the doors for a more realistic debt management scheme is being disappointed by the paper in two regards:

• The concept is intended to serve “low income countries outside the HIPC Initiative”. Thus it supposes that the HIPC business will go on as usual. Nor will a new approach be applied to middle income countries, which find themselves in severe debt distress situations like Argentina, Ecuador or Brazil. No convincing reason is given for either those limitations. The PIN states:
  
  

  The re-emergence of excessive debt levels would be a serious problem in low-income countries, although given these countries' reliance on official flows, the concept of debt sustainability is somewhat different there than in middle-income countries. High debt levels can be problematic as they may require debt restructuring and forgiveness which is disruptive and costly and the burden of a debt overhang may undermine urgent progress on policy reform. High debt levels also force lenders to allocate scarce concessional resources keep high debtors afloat, often at the expense of other deserving countries.

  
  Obviously there is no convincing argument in this, why MIC debt unsustainability should be different in character from that of Non-HIPC-LICs and theirs from that of HIPCs. Most likely, the IFIs are again looking out for some guinea-pig which would be as far off as possible from those cases which really cause them headaches.
• Secondly, the concept is not meant - as one might assume - to improve the practice of debt relief by multilateral and bilateral creditors, but rather “to inform future lending and borrowing decisions”. There is, of course, nothing wrong with improving borrowing and lending processes; however, isolating a debt sustainability concept from the question of how much debt needs to be cancelled under existing schemes, leads to the somewhat ghostly debate, which will be commented in paragraph (4).

3. Outcomes

For the first time, the IFIs dare to come up with actual figures, indicating, where debt sustainability thresholds should be located in the future. This is the more laudable, as it exposes the institutions to a broad range of critiques, including this one, from a civil society perspective. It is therefore not surprising that all the indications are coming with a huge set of reservations and a clear warning that nothing must be considered as definitive yet.

The staff paper reveals – though as clear-cut only in the appendix – that 40% of GDP and 180% of export earnings will be considered as “baseline” thresholds for a sustainable debt in NPV terms, along with a debt service ratio of 20% [4]. This comes in a more elaborated form and with the additional element that countries with good policy performance and governance records can sustain substantially higher debt levels elsewhere in the paper [5].

Two things are remarkable about these figures: First, they tend to be higher than actual HIPC thresholds. The “forward-looking” character of the framework, however spares IFI staf from explaining, why Non-HIPCs should be in a position to bear higher debt rations than HIPCs. Secondly, with the differentiated approach according to policy and governance performance the IFIs commit themselves to a logic where good performers are to receive more loans and run deeper into debt. This, of course, is understandable on the positive side, but raises substantial questions on the negative side: Who is actually going to give more concessional financing to cover financing gaps to the poor performers? This leads to the more fundamental questions to the IFI’s approach formulated in the next paragraph.

A key question, of course, is how the above mentioned thresholds have been identified, i.e. which methodology has been applied. The figures are the result of an ex-post analysis of historical debt distress episodes. The baseline figures have been identified as the median of a set of historical debt distress cases minus a safety cushion. Debt distress episodes have been identified by the authors as cases where

countries resorted to exceptional financing in any of three forms (i) significant arrears on external debt, (ii) Paris Club rescheduling, (iii) non-concessional balance of payment support from the IMF. [6]

This methodology, however, is flawed in the sense that it does not refer to debt distress as such, but already to remedies, which governments may have taken against it. It is not unusual that in their earlies stages governments undertake other measures to deal with an already emerging debt distress situation. Governments may, f.i. cut down “soft” public expenses, mostly in the social sectors, raise taxes or deplete official reserves. Going back to such – admittedly harder to identify – measures would most likely have brought the baseline indicators further down- rather below existing HIPC indicators than above them [7].

4. Which policies?

The PIN summarizes the recommendations of the staff, which have been endorsed by the directors as follows:

Directors broadly endorsed the key elements of the debt sustainability framework: (i) the analysis and careful interpretation of actual and projected debt-burden indicators—including measures of the debt stock in present value terms and debt service—under a baseline scenario and in the face of plausible shocks; and (ii) an assessment of these indicators in relation to indicative thresholds taking into consideration the quality of the country's policies and institutions.

First of all, this interesting conclusion lets one ask, how actual and projected debt burden indicators have been interpreted in the past. And if this has not been done “carefully” what reason directors have to assume that this could be done by the same people who so far have done it without due diligence. Directors assume [8], that

Fund staff, in close consultation with World Bank staff, would begin cautiously applying the part of the framework that relates to the analysis of the long-term debt dynamics, using the proposed templates, as background for Article IV surveillance in low-income members.

Contrary to the idea that the frameworks new elements will introduce a new world of analytical sharpness and coherence, stress tests have already been part and parcel of Debt Sustainability analysis in Art. IV consultations before. Unfortunately, they did not serve to give an impression of sufficient dues diligence by avoiding f.i. reliance on overoptimistic growth expectations within HIPC completion point documents . Again: what is going to be different when the same exercise is being done by the same people?

Another “motherhood and applepie” suggestion is then added to this one, demanding that countries which might face debt distress situations should receive higher concessional financing, most desirably in the form of grants. However, the Fund immediately clarifies that it is not going to be a source of this grant financing [9]. Some directors rightfully suspected that “prospects for additional grants may not be favourable.” This suspicion, however, is not an answer to the question which the staff had already raised towards directors in its paper [10]: “What options do Directors see, if grants are not sufficient to ensure that a country can obtain the financing required to meet the MDGs while maintaining debt sustainability?”. No options have been suggested by directors so far. To the contrary: With their heads in the sand, staff and Board are approaching a situation, where f.i. an unsustainable post-completion point HIPC like Bolivia or Uganda will be unable to close substantial financing gaps via concessional financing as suggested in the paper. Bolivia in particular has largely taken recourse to non-concessional multilateral money over the past two years and is therefore firmly on its way into a new (this time largely multilateral) debt problem. In that situation “informing the borrowing and lending decisions” of the country and its creditors is of no great help, but rather adding insult to injury. The only real way forward would have been to re-connect a revised understanding of debt sustainability with a perspective for enhanced debt relief.

5. Some light

The overall direction of the process within the Washington institutions is highly problematic, because it gives the impression of a new approach, while it remains largely on the tracks of the old worn-out procedures and avoids concessions in terms of debt relief. However, there are also a few merits in the paper, which may turn out to be useful for a real reform debate, whenever this will start:

• The Paper admits the need to include domestic public debt and private external debt into the concept of debt sustainability. It is quite strange that this correct and timely understanding is then de facto kept out of the proposed framework, as no thresholds for these categories are indicated. The staff paper meritoriously provides alarming figures regarding the rising internal debt problem of some low income countries. However, it fails to understand that debt sustainability must be a coherent concept. There is nothing sustainable about the debt of a country, which complies with thresholds laid down in the staff paper, but spends half of its public revenue for domestic debt service.
• The paper also indicates the need for a reform of the NPV calculation methodology. The Niger case, as well as conceptual analysis by Debt Relief International [11] and others, have pointed to the problematic nature of NPV as a concept.
• Indexing debt service to GDP growth, export earnings of other critical indicators of economic capacities via a range of instruments is being considered. Though no conclusions are drawn, it is useful that this debate has been kicked of (again).
  

6. Papers, more papers

What can be expected after the spring meetings?

Directors asked staff to return with two papers that would form a basis for further discussions on these issues. One, to be prepared jointly with World Bank staff, will give further consideration to issues related to the sustainability framework itself. A second, to be prepared by Fund staff in consultation with their World Bank counterparts, will examine the operational issues for the Fund, especially in setting debt limits in Fund-supported programs; it is envisaged that a similar operational paper will be prepared in parallel at the World Bank. [12]

7. The real issue

Despite the methodological flaws and the protracted approach, when it comes to financing and policy implications, the fact that the debate has been triggered, is to be welcomed. Whether it will lead to useful results which will go beyond the HIPC mission creep, will in essence depend on one crucial question: Will Bank and Funds members recognize the contradiction between both institutions’ roles as lenders and as “independent experts” assessing countries’ debt situation. As long as either debt relief or new lending will be based on the expertise of institutions, which will themselves eventually incur losses from those decisions, no realistic understanding of debt sustainability can be expected, projections will remain wishful thinking [13], and the indebted countries will pay the price.

Jürgen Kaiser, erlassjahr.de, April 21st 2004

[1] Hitherto referred to as „Staff“
[2] Hitherto referred to as „PIN“
[3] Staff pt.19 fn 10
[4] Staff app. II; pt.2
[5] Staff table 2, p.21
[6] Staff, App.I pt. 2
[7] Not to mention truely non-distress or successfull relief cases like Germany in 1953, where debt to GDP was brought down elow 7% and the debt service ratio to below 3%.
[8] PIN p.3
[9] Staff p.35/36
[10] Staff p.52
[11] Aguilar, J.C. (DRI): Critical Assessment of Existing Debt Mechanisms; paper presented for the Debt Experts Meeting; Dakar Oct. 17th 2003
[12] PIN p.4
[13] Compare for an illustration the alternative stress scenarios as applied in the staff paper with the „historical“ scenarios, in order to assess how far the IFIs even under this new approach are still away from the „reality check“. Staff annex II, p.68.