Transcript of a press briefing on IMF Conditionality by Masood Ahmed, Deputy Director, IMF Policy Development and Review Department
March 21, 2001
Masood Ahmed, Deputy Director
IMF Policy Development and Review Department
Wednesday, March 21, 2001 9:30 a.m.
MS WHITE: Good Morning. My name is Kathleen White, with the IMF press office. We appreciate your braving the rain and wind to be here. This is the briefing on IMF conditionality. The four papers plus the Interim Guidance Note have been distributed. I hope you have copies, as well as of the Public Information Notice, which is now available. That PIN summarizes the views of the IMF Executive Board when it discussed the papers some days ago. The papers and the briefing are all embargoed for 12:30, Washington time, today.
Let me introduce now our briefer, Masood Ahmed, who is the Deputy Director of the IMF Policy Development and Review Department. On his right is Tim Lane, who is one of the co-authors of the papers, and the Chief of the Policy Review Division within the IMF Policy Development and Review Department.
I believe Masood has some comments before he takes your questions.
MR. AHMED: Thank you very much, Kathleen.
What I thought would be useful to do, in case you haven't had a chance to go through the PIN yet, and certainly haven't had a chance to go through the four papers, which are pretty thick, is just spend five minutes walking you through the process of where we are and what the main conclusions are and then open it up and take any questions that you have.
And as Kathleen said, the Board discussion that we had two weeks ago today was an important milestone in the process the Managing Director had launched shortly after joining the Fund, taking over the Fund, to try and move to a more streamlined and focused approach to the conditionality in Fund-supported programs. Those of you who recall his speech at the annual meetings last year would remember this was also an important element of his overall vision of a more focused approach to the Fund.
What these four papers do is try and take stock of how conditionality in the Fund programs has evolved over the last decade and what lessons we can now draw from that in terms of moving forward and making it more effective.
The first paper is an overview, a brief overview, which tries to pull all the different elements together; second is a paper that looks at the policy framework in more detail; third is the one that looks at structural conditionality; and the fourth looks at trade policy, one dimension of structural conditionality. All of these papers then were the basis for the discussion that took place two weeks ago.
Just a word on the history. Conditionality has always been associated with Fund financing. It is, in effect, the link between financing from the Fund and implementation of policies by countries that make use of Fund financing. But over the last two decades, and particularly over the last decade, there has been quite an expansion in the scope of Fund conditionality, particularly in the structural areas, and the structural areas in this case cover things like tax administration, government expenditure systems, financial sector, trade liberalization and the like.
What the papers do is look at the reasons for the expansion of this conditionality, and basically conclude that at the time there were three reasons I think which got us there. The first was emphasis on growth as a policy objective. In the '70s and early '80s, there was a lot of concern that Fund programs focused on stabilization and not enough on growth, and so the emphasis in growth led to a focus on structural areas.
Second, the Fund started getting involved with countries for whom structural reforms were a major source of the distortions that they faced. Notably, low-income countries, through the SAF and then the ESAF, subsequently became the PRGF, but also in transition economies where, as you remember, the main objective was, in many cases, structural transformation.
And the combination of these two factors led to a substantial increase in the number and type of structural conditions that are associated with Fund programs. The papers that you have detail this and give you a sense of both the scope of the increase and the magnitude of the increase in different ways. In the overview paper, there are two charts which actually summarize this. So sometime, when you get to the papers, it will be a nice way to look at what's been happening, in terms of the history.
When this was discussed at the Board, Directors recognized that there were these valid reasons for the expansion in the scope of conditionality, but there was also a discussion, quite a lot of discussion about the problems and concerns that have arisen as a result of this expansion in conditionality.
And the kinds of concerns that you have probably been aware of in other contexts, also, but that were discussed at the Board were that the scope of conditionality, having expanded into areas that are not the core areas of the Fund, being that sometimes we are spread too thin in the coverage of our program, that the programs may be trying to support too many changes at one time, more than the capacity of countries to try and carry them out, and that the conditionality may be too intrusive, and therefore undermine a sense of the ownership of those programs by the countries concerned.
So, for these reasons, there was an agreement that we now need to move towards a more streamlined and focused approach to the application of conditionality in Fund programs going forward. And the aim of this streamlining is not to weaken the implementation of Fund programs, rather it is to ensure that countries have the maximum possible scope to make their own policy choices, but ensuring, at the same time, that the Fund's financing continues to be provided if those policies, which are essential to the objectives of the program, continue to be implemented.
So it's a focusing, sharpening of the prioritization, but it's not about weakening program implementation.
There are three elements to the streamlining which Directors have endorsed. The first one is to focus on the policies that the government would undertake, which of those would be covered by formal conditionality. Historically, we have tried to make sure that conditionalities applied to those policies which are relevant to the objectives of the program, but the test of relevance has been applied with quite a lot of latitude.
And the interim guidance note, which Kathleen referred to, and which is now in effect as the basis under which programs are being developed, raises the threshold of importance, so that measures which are critical to the objectives of the program would continue to have conditionality associated with them, but there would be a much more parsimonious, if you like, approach to applied conditionality to other policy measures, which may be relevant, may still be very useful, but which are not critical to the objectives of the program.
Now there's obviously a subjective element in trying to figure out what is critical--where the boundaries of critical stop and important or relevant start, but the presumption in the program design would shift from one of comprehensiveness to one of parsimony or selectivity. And, if you like, the burden of proof would be much stronger on trying to include additional structural measures as formal conditions in the design of programs.
The Executive Board agreed that, in some sense, this is going to be a bit of learning by doing to get the balance exactly right as to where we are all comfortable at drawing the line between critical and relevant. And as a result, in the upcoming programs, there will be a specific focus on trying to clarify why certain conditions were included and why certain were not included, and this would be a focus, if you like, a real-time assessment of whether we've got the balance right in the programs in the next few weeks and months.
The second way in which we are going to streamline our conditionality is by relying more on other institutions in areas which are outside the core areas of the Fund--notably, in the case of the World Bank, where we've always had a partnership, but where we now want to rely much more systematically on the Bank to support governments in implementing structural and social reforms in areas outside the core expertise of the Fund, and this includes things like public enterprise restructuring, privatization, labor market reforms, sector reforms in energy and agriculture, the sorts of things that have comprised a significant part of the total conditionality in Fund programs, particularly Fund programs which have had a longer focus in, as I said, in transition countries and in PRGF countries.
Many of you will recall that in September the Managing Director and the President of the World Bank issued a joint statement with an objective of enhanced cooperation, better partnership between the Fund and the Bank.
And some of you will also have been following the work in low-income countries, wherein, the context of the poverty reduction strategy papers, the PRGF facility in the Fund and the Bank's operations are now much more closely aligned. And that we see as a particular fruitful avenue for better division of program content and conditionality in the structural area between the Bank and the Fund.
And the third way in which we would like to proceed with streamlining conditionality is to reduce the number of formal conditions associated with the steps leading up to a structural reform, rather than to the end point of that reform itself. Just to give you an example, in the case of Mauritania, there were 19 specific structural benchmarks or prior actions associated with the introduction of the value-added tax. And going forward, we see that there is some scope in reducing the degree of monitoring associated with these individual steps leading towards the implementation of a particularly important measure, such as a value-added tax.
Obviously, the countries often see a great deal of value in having these steps laid out in detail, but we believe, and the Board feels, that this can be better done in the context of technical support or technical assistance work, rather than necessarily linked to program conditionality explicitly.
Finally, let me say that, in addition to these three ways of streamlining, there is a fourth related element which the Board also endorsed, and that was to try and clarify the boundaries of Fund conditionality in a little more precise way because, for a couple of reasons, they have become blurred in the way in which they are perceived, certainly outside.
The two ways in which they have become blurred is, first of all, the Letters of Intent that are part of the programs--that forward programs from governments to the Fund are used, in a number of cases, by government to lay out a broad, their overall program, broad program design, the scope of which exceeds the focus of the Fund's conditionality and the reviews that the Fund would do that are linked to its financing.
Obviously, governments do this because they find it useful sometimes to signal their overall program credibility to markets, sometimes because it's an important instrument for laying out a comprehensive vision that governments have of their agenda. But what we are proposing to do now is, in cases where the scope of the Letters of Intent is broader than the coverage of the Fund program, to very explicitly set out within the Letter of Intent what the subset is that will be the focus of the Fund's program and of the reviews that the Fund would carry out so that there would be more clarity of the boundaries.
And, similarly, the other area in which there is some degree of blurring is to clarify the role of things like structural benchmarks and prior actions, where it's not always clear to people whether these are or not strictly related to the continuation of financing from the Fund.
That's, insofar as the background and the main thrust of the changes that we put into place. The process of going forward is that the discussion that we had two weeks ago, as I said, is an important milestone, and now we are going to continue on two tracks.
First, we are going to continue with the application of this more streamlined approach to upcoming programs. And as I mentioned, the Board will be reviewing these programs in sort of real-time, with a view to giving us guidance on how far we've got the balance right, in terms of the threshold for criticality that we're using to assign specific conditionality to policy measures.
In parallel, we're also going to put out the papers that lay out our analysis and our approach on the Web, with a view to getting more public discussion and feedback on the papers and on these issues. As you know, this is a topic on which there is a great deal of interest, and which there is a great deal of interest outside amongst academics, amongst civil society organizations. We think it's very important to get the benefit of their work and their reactions to this.
We are also planning a major seminar on this topic here in Washington in May, which will bring together people from different perspectives, looking at the issue of conditionality, and we will obviously have an opportunity to get guidance from ministers on this topic at the International Monetary and Financial Committee meetings which are going to take place at the end of April.
So the conclusions then of the seminar, guidance from Ministers, the feedback we get on the electronic, comments that we get through the website, we will convey back to our Executive Board as part of their next discussion of this topic in June. They will also have, during this time, as I said, the benefit of the feedback from the review of programs which are coming to them for consideration.
After that review in June, the process will continue then. First, there are a number of issues on which additional work is needed, which are set out in the PIN, which we will be doing over the course of the summer.
The ultimate or the eventual objective, if you like, of this exercise is to review and revise the guidelines on conditionality, the formal guidelines on conditionality, in the Fund, with a view to aligning them more with the realities of today and to give due emphasis to the need for streamlining and country ownership. And we hope that this work will be brought back to the Board in the second half of this year. So that's roughly giving you a little bit of a sense of the time table going forward.
So that, I think, gives you background information on this, and now I will be happy to take any questions that you have.
QUESTION: Just a couple of quick questions. While the Board has got more discussions to do, these four suggestions are being taken on board [inaudible]. Could you just give us a couple of examples for which would be, in the typical case, a critical benchmark and which things are never intended--intent would not be critical.
And then, also, do you know of any instance where, just anecdotally, where a country was punished by financial markets for being perceived to miss a benchmark which, in actual fact, was in the Letter of Intent, but was not a critical benchmark in the IMF program?
MR. AHMED: Let me give you an example to illustrate what we mean by critical and what we mean by something which is not critical. I'll give you two examples to contrast that.
If you take a case like dealing with the problems in the cotton sector in Mali, where the cotton sector accounts for about a third of the population, in terms of employment, half of the exports, and where there has recently been, because of problems in the organization of the state monopoly that is responsible for the purchase of cotton from farmers and then their export, a major problem in terms of an implicit tax on farmers that account for about 4-percent of GDP, and where, as a result, farmers have cut back on production, and cotton production last year is down, I believe, by about a third, dealing with that is critical because not dealing with it has a major impact on the macroeconomic viability and growth prospects for Mali. So we can't work out a meaningful growth and macroeconomic viability scenario for Mali without dealing it.
On the other hand, if you take an example like sugar refining in Mozambique, where the Fund has had a view for many years that this is a distortion and that it's not pro-poor either because the government resources that are used to subsidize sugar refining in Mozambique could be better deployed for inoculation or primary health aimed at the poor. Nevertheless, the scope of this distortion is not so large that if you don't deal with it you cannot still continue with a high- growth and macroeconomically viable medium-term scenario.
So, in that case, it's important. We continue to think it's important. We, in our work, will continue to highlight the costs of that, both in terms of government revenues and their alternative uses, but we do not see that making that a formal condition of Fund financing is justified because it's not critical to the objectives of the program.
I don't have any illustration of the second question you asked of a country that has been punished by the markets for missing a benchmark that was in the program, but not critical. If I think of one, I'll come back to it.
QUESTION: One thing the papers bear out is that the IMF believes it was almost forced into broadening conditionality because some of the aspects essential to program objectives were covered by the World Bank, but the lack of conditionality at the World Bank in their programs meant that the IMF had to take up the slack, as it were, and ask or set benchmarks in that area.
Is there any work being undertaken with the World Bank to improve their conditionality or at least to achieve a clearer understanding, whereby the IMF doesn't have to do the work of two institutions in that regard?
MR. AHMED: There is now work underway with the World Bank to try and get a more coordinated approach, as I mentioned. Particularly in low-income countries, that work is quite advanced. As you know, the Bank has recently put forward a new instrument, which is the poverty reduction support credit, which is intended to complement the PRGF of the Fund in low-income countries in a way that will allow the Bank to systematically cover structural and social issues outside of the Fund's core areas.
In the middle-income countries, that work is proceeding, but in a less-structured way, as there is no similar framework. However, as some of you who follow the Bank will know that they have recently been putting to their Board some proposals on the Bank's role in middle-income countries, the objective of which would be to also provide a basis for more enhanced collaboration with the Fund in the same way.
I should also say, however, that the objective is not simply to transfer a certain block of conditionality from the Fund to the Bank. The objective is also to ask the question, what was first question here, which is regardless of which agency is involved in supporting and monitoring conditionality, you've got to first ask the question is this particular policy measure really critical to the objective or not, and that's part of a broader focus of ownership.
QUESTION: Could you tell us what some of the new conditionality might mean like in a specific case like Indonesia that is already an IMF adoptee and what that might mean if they need to come back to the Fund for future funding?
MR. AHMED: I'm not sure I can usefully give you the specifics of what it would mean for any particular country, but the way in which we envisage this process moving forward is that, as new programs come to the Board, we will be asking or applying, if you like, this more strict test of relevance and criticality to looking at the specific conditions that would be included in programs.
The second issue which applies to a lot of countries, including the one you referred to, Indonesia, is this concept of the Letter of Intent, what it means and what part of it is covered by the specific conditionality of the Fund. As you know, a number of Letters of Intent have also included very detailed policy matrices which sometimes go into scores of measures that the governments intend to carry out, but only some of which have a direct bearing on the scope of the program supported by the Fund.
And, consequently, going forward, what you will also see in Letters of Intent is the scope of those policy matrices much more aligned with the measures that are the focus of the Fund's work, rather than sort of broad detailed policy matrices of the kind that you've seen in the past.
QUESTION: How will all of this actually affect the volume of lendings to be made in incoming months and years? Will there be any change in your priorities, in terms of country programs of regional outlook?
And my second question is, this was actually discussed earlier this month at the Board, will there be follow-up discussions to it, and can you give us a time line on how it will be discussed?
MR. AHMED: I don't see any link between this work and volumes of Fund support going forward, in that the objective of this exercise is not to make access to Fund resources easier or more difficult. The purpose of this exercise is to focus the conditionality on those measures which are critical to the objectives of the program, and therefore to improve the ownership and implementation of programs.
Now, to the extent that ownership improves implementation, implementation makes financing more available to countries that have already embarked on programs, you might see some effect, but I don't see that there's any direct link envisaged between the two.
Insofar as your second question of what is the process going forward on the follow-up, as I mentioned, we will now be gathering comments from the electronic dissemination that we're doing, from the conference that we have planned in May, from the IMFC we get guidance from Ministers, and we will bring this back together, along with the review, if you like, the experience of the review of programs that will be coming forward in the next few weeks to our Board. This will all come back to our Board in June for the next discussion.
And then, thereafter, the eventual objective is to move towards a review and revision of the guidelines of conditionality, which I am sure would also require at least one further Board discussion, and possibly more than one, but this will probably take place during the course of the second half of the year.
QUESTION: You mentioned the reasons why conditionality expanded or began to expand about a decade ago. Are you now saying that that expansion is a mistake? That seems to be the subtext of what you are saying.
And looking ahead, is there now a risk that, by cutting down on conditionality, that the growth won't come about, the very reason that this expanded, won't come about in the future?
MR. AHMED: I think whenever you look back, the question in one's mind obviously is how do you draw the lessons from the past, and was it a mistake at the time? Is there a risk of overshooting now?
I think that as the papers set out, the reasons for getting involved in structure of conditionality were primarily linked to the importance of structural reforms for growth and the fact that--I think it was mentioned by somebody in an earlier question--that, in many cases, Fund programs became the primary instrument for carrying that structural focus, structural conditionality.
There are now, I believe, two lessons that we have drawn from this, which will change the way in which we've applied it.
First, in a simple way, we've got a better framework for dealing with the Bank and with other partners. So, while structural conditions still remain important for growth, they don't all have to be covered in the context of Fund-supported programs. There are other people that we can rely on.
Second, and equally importantly, because a measure is important for growth and government intends to undertake it, it doesn't necessarily follow that applying formal conditionality linking continuation of financing to each and every one of those measures is necessarily the most effective way of supporting governments in that endeavor. And I think one of the lessons of the work that's been done on conditionality and ownership, which is also summarized in the policy paper that is part of the package that is being released today, is that, in a number of ways, if program conditionality is excessively detailed or intrusive, it may actually go to the point where, rather than supporting ownership, it may act as a hindrance to it and may slow it down. So it's trying to get the balance right on how conditionality should be related to structural programs, not a comment about whether structural measures are important for growth.
Now will we get the balance right? Will we go too far in the other direction? That's precisely the focus I think of our case-by-case work over the next few months, and that's very much going to be the focus of what the Board will be looking at when they review these cases in the sort of real-time assessment that I mentioned earlier.
QUESTION: Another fact brought out in the paper seems to be the fact that conditionality, no matter whether it's streamlined or extensive, is only as effective as the political will of the government that is actually going to implement it.
What sort of, you know, when you actually come to putting a paper forward that actually becomes a blueprint for conditionality going forward, is that going to be brought out as a main point? Because, while it's one thing to say, you know, we are streamlining conditionality, it's always an evolutionary approach to meet the conditions of the time.
But one condition that is sort of manifest in all of this is the fact that some governments are very good, some governments are very weak. Whether a program works or not may not depend on whether you have a list of 100 conditions or 10, it's how effective that government actually is.
When you look at the blueprint, would that be a sort of founding principle to this, recognizing the fact that a condition is only as effective as the government implementing it?
MR. AHMED: There is a lot of evidence that conditionality is less important for outcomes than the ownership that governments have and the commitments they have to implementing their own programs and, in some ways, there is no substitute for government ownership.
There was a considerable amount of discussion of that in the Board, and indeed the papers that have been put out include a lot of questions and issues that that raises.
One direction that that walks one into is to be more selective in ensuring that there is a minimum degree of government ownership of the programs as part of the basis for going forward with financing from the Fund and from other agencies. And I think many agencies are now putting a great deal of emphasis on ensuring that such ownership is--demonstrated ownership, if you like--is a priori for moving forward.
Having said that, I should also say that it's not always easy to establish the extent and depth of ownership. Ownership by government is not a monolithic, uni-dimensional thing. Because even within countries who owns the program and how broadly that needs to be owned and is owned is something that will vary from country to country and from type of program to type of program.
And, similarly, I think there are going to be cases where the Fund will continue to be faced with difficult choices where ownership of programs may be less than complete, in some ways. We have to have parts of the program where there is broad ownership and others where it takes more time to reach that ownership, but the costs of delay may be very high.
So I think those are the kinds of trade-offs that people are conscious of and need to work through, but the basic premise that ownership is essential for successful implementation, particularly of the kinds of structural reforms, institutional changes that are the basis for long-term growth, I think is one that underpins a lot of the analysis in these papers.
QUESTION: Can I ask you to tell us a little bit about what prompted the review now and to what extent was it related to some of the criticism the Fund faced after the '97 Asia crisis, the Meltzer Report and the criticisms of some of the people now showing up in the Bush administration?
MR. AHMED: The Fund looks at different elements of its work over time. We have periodic reviews of surveillance. We had a month ago discussion at the Board of the Fund's experience in governance. So it's not a unique or unusual event for the Fund to take stock of a particular dimension of the effectiveness of its programs.
In this particular instance, the added impetus for this work, as I mentioned, is the strong push by the Managing Director to take a look at the effectiveness of conditionality under Fund programs and to see if there are ways in which one could streamline it, and improve, and therefore contribute to improving also ownership of those programs.
And much of this work was initiated early summer last year, and it's been in the making since that time, brought to closure in this phase for the Board now.
QUESTION: I wonder if you could address the example of Turkey and its recent augmentation of its loan from December until a month ago. I mean, it seems like this process of streamlining had already been suggested by Mr. Kohler, and yet in that program, I would say, and I think most people would acknowledge, very specific structural benchmarks along the way.
Looking back on that, would you see that as an example of excessive conditionality and how would that change in the future?
MR. AHMED: I'm not going to address the specific example of Turkey, but I will use that question to address the point that you're raising, which is I think that there are two parts to what you say, which we need to go forward with. One is, in each country case, even after a streamlined approach, the coverage of structural conditions will vary quite a lot.
Let me give you one example, as a general balance. If you look at the PRGF programs that have been approved by the Board, since the Interim Guidance Note was issued last September, on average, the number of structural conditions in them is significantly lower--I believe it's about a third lower--than the PRGF programs prior to that date.
But, equally importantly, even in the set of programs that were approved since September last year, there's a couple where the number of structural conditions, particularly structural conditions relating to public expenditure management, public resource management, are significantly higher than the average. There's still quite a lot of variation across countries. And I think that's exactly the kind of balance one has to get right, depending on each country case, and which is what the Board will be looking at going forward.
The second point that you raise is an important one, also, which is how do we deal with programs that are underway, ongoing? And there was a discussion in the Board as to whether or not we should now be looking at sort of every program quickly to try and apply this approach or whether this is an approach that gets phased in, as new programs come in.
And the consensus of direction from the Board was that this is something we need to now phase in, as new programs come in, and there's no intention now, as a result of a particular Board discussion last week, to suddenly take stock of all programs and look at the particular content or conditionalities, a gradual process of phasing them. Of course, this is something which has been phased in, as I said, since last September, but the process is going to continue to be a gradual one.
QUESTION: I'm just looking for a simpler example of critical versus noncritical, and Turkey is an obvious case. Obviously, in Turkey's case, the banking sector reform would be critical. What would not be critical for Turkey as a condition?
MR. AHMED: I don't have enough detail on the Turkey program to use it as an example for this.
MS. WHITE: Any last question? I think we've covered a lot of ground.
Let me remind you, again, of the embargo for the briefing, and the papers, and the PIN until 12:30, Washington time, today.
MR. AHMED: Thank you all very much. And if you do need any further information, please feel free to come back to us.