The IMF Transparency Code
by Barry Potter
February 23, 1999
Cape Town

Mr. POTTER: Good morning Ladies and Gentlemen. First of all, thank you very much for the invitation to come and talk this morning. It is a very welcome opportunity for us. I am one of a fairly small team which has worked on this IMF Code of Fiscal Transparency. What I would like to try and do this morning is as follows. First I would like to tell you about the background and the rationale for the Code, that is, how it came into being. Secondly I would like to talk about the structure of the code itself, what it is that we are trying to do with the individual elements. And then thirdly I'll try to explain why the Fund's role in transparency is to some extent limited and that in a very deliberate way we have tried to set this initiative up as something that other organisations and other people could seize upon and build upon in its different dimensions.

The background work started in about 1996. For those who have been involved in the budget field for some time you'll know that transparency was one of those concepts that we all paid lip service to for many, many years. It is a bit like motherhood, you sort of know it when you see it but if you actually sit down and try to define it, it is actually quite difficult. I think that it was only in 1996 that we began to have some idea of what it was that we were talking about in a more precise way.

That year the Interim Committee, which is simply a body that helps govern the IMF, first came out with the idea that it should be promoting good governance in all its aspects, improving the efficiency of the public sector. Then one started to see more references to transparency in Fund statements. And back in 1996, we were particularly talking about enhancing the transparency of fiscal policy by persevering with efforts to reduce off-budget transactions and quasi-fiscal activities.

I would urge you not to be frightened by the IMF's use of buzz words and code words for things that you would probably recognise in other terms. When we say off-budget transaction we are referring to all those transactions which people who work in the budget field have known about for many years, whether it is expense accounts, below the line transactions, or extra-budgetary operations. As for quasi-fiscal activities, if you don't mind I'll leave that one till later.

Why is fiscal transparency important? We would put emphasis on three factors. The first is the role of markets as it says here on the overhead,

• Markets play an important role in exercising discipline over policy makers in improving the quality of policy decisions.

In fiscal policy perhaps nothing matters quite so much these days as what the financial markets think you are doing and how well you are doing it, and to add to the financial markets I think you increasingly have to take into account the fact that the donors like to know what it is that a country is doing and how well it is doing it.

The second aspect of that - I think no-one, certainly in South Africa here, but right across large areas of the world, is any longer in any doubt that, however misguided the financial markets may be, the one thing they are not is slow. So if you appear to be making the wrong decisions they pretty rapidly get hold of you and they can transmit quite quickly, right across from one side of the world to the other. Since the whole problem of Asia and South America and the Russian economy and the other economies in former Soviet Union will come up again and again, let me not go into it too much now. But it is clearly an interesting aspect of what has happened in the last few years, that once the markets start to have doubts they react, and they react quickly.

The last point on the role of markets is that at the end of the day markets are driven by expectations. So it is not just important to be doing well, it is important to create an expectation that you will do well in the future.

Next is a bit of first year economics, but it is one that we all tend to forget. Economic agents make better decisions if their expectations are based on full information about the government's economic policy objectives and strategies. That is a long roundabout way of saying that whether you are in the Ministry of Finance, or you are in an individual line ministry, the more information you have about what it is you are trying to do, what resources you will have available and precisely when you will have them, the better the decision you can take about implementing your specific policy.

If the role of the market was one factor driving transparency, basically its companion was that transparency can itself improve economic policies and performance. And thirdly, and perhaps the proper corollary, a lack of accountability has contributed to economic policy failures and financial crises.

The problems that arose in Asia were not - and I want to make this quite clear - were not rooted in fiscal problems. In fact many of the countries did not have a fiscal problem, at least not as conventionally defined, an issue we will come back to. What we have seen is that the doubts that were created in markets in Asia had a very dire, sudden, sharp effect on those economies and on their currencies and those effects moved very quickly from one country to the next.

From 1996 the IMF got more interested in transparency. By 1998 we were coming up with a code of good practice on fiscal transparency, a declaration on principles. As many of you have worked for a ministry of finance or are otherwise involved in bureaucratic matters, you will recognise that something which might simply have been called the Fiscal Code ends up with a title of ten words, which captures the spirit of the sort of compromise that has to be struck when you are trying to derive a set of principles that are supposed to be universal across all countries. That is an issue that we will come back to.

I am going to do just three things in the short time that is allotted. I hope I have already set out the broad rationale for the Code. I am now going to talk about the structure and then I am going to talk about the implementation.

In terms of rationale for the code I think the basic point is policy transparency is quite fundamental in the way that an economy is run. What do we mean by that? If, as a government, you don't know where you are going how can you expect the people of the country to know where you are going? If, as a government, you hide things from yourself or as officials we hide things from ministers, how will the truth emerge? I have sometimes made the joke that I spent the first 25 years of my professional career assisting ministers in hiding what was going on and the last five years trying to unveil what was actually happening. But it is certainly the case, for very many years we spent a lot of time when I worked in Government, in the words of one of our former permanent secretaries, being economical with the truth. That is to say we did not give as much information as we actually possessed. It is worth just thinking about this for one second.

There is a case, not a terribly respectable one, but there is a case for not being transparent, because you pay some prices or accept some costs for being transparent. If you start saying, the economy is in a mess, we know the exchange rate is going to go down, we know that the inflation rate is then going to go up, and if you start telling people that, they will behave accordingly. For many years in the 60s and 70s and in the early 80s in Britain, we were often seeking to present a case in which the inflation rate was not as bad as we thought it ultimately might turn out to be. So there is a case for not being as completely open as you might be.

But I think the balance has very much turned the other way. I think the problem now is to attune the various agents in the economy to knowing what the inflation rate and exchange rate are likely to be, so we aim to prevent the markets from fearing it will be even worse than you actually expect.

• Fiscal management is essential for macro economic stability and growth.

I think you only have to look - and I would say the best cases are undoubtedly, I am afraid, at the moment in the former Soviet Union, where too long a refusal to recognise what was happening on the fiscal accounts, has been extremely expensive. Let me explain what we mean by good fiscal management. It has three fundamental aspects. First of all sound government finances. I think everybody understands exactly why that matters. For the IMF, with our concerns about monetary growth and exchange rates, it is the fact that a large deficit that is improperly financed adds to money supply, and adds to inflation.

That is not the only aspect of good government finance. Second, the debt burden has become a huge liability for many countries, which is basically diverting resources that they might otherwise have for current consumption, including consumption on health and education. Thirdly, particularly, but not just for the developed countries, there are real costs associated with the high taxation that is necessary to cover existing deficits and to meet the debt burden that was associated with past deficits. You need efficient fiscal policies to achieve balance in government finance and you need fiscal transparency to achieve efficient fiscal policies.

If you wanted a single, one-line, alright four-line, definition of what fiscal transparency is, it is this. It is being open to the public about the structure and functions of government, your intentions on fiscal policy, the accounts and the projections. I would describe this as - if you want to put it a different and even shorter way — as telling the full story. Not being economical with the truth, but putting out as much as you genuinely can. We believe that fiscal transparency leads to a better informed public debate. It makes governments more accountable for the design and implementation of their fiscal policy. So the idea of the Code was to provide some kind of benchmark. The word score card I referred to earlier, but let's used benchmark for the moment, for assessing fiscal transparency, with the idea that it should promote greater fiscal transparency.

Keeping one eye on the time, let's move on to the structure of the Code. All you have to remember in this presentation are the four basic principles. I am going to go through each of these and explain within them something like 20 or so specific practices of policies which relate to each.

Let's start with the first.

• The roles and responsibilities of and within government should be clear.

Again you would think that is not something that is objectionable, but in many countries there are some very, very woolly boundaries between what constitutes the government and what constitutes private sector. Those boundaries are ill-advised. That is, any greyness in those boundaries is ill-advised. They should be as clearly defined as possible.

The second of our four principles.

• Comprehensive reliable information on fiscal activities should be make available to the public.

Almost every word in that little phrase is important.

Comprehensive is important, ie you shouldn't be leaving anything out. Reliable is important. Anybody can come up with information. The question is, is it soundly based? Note fiscal activities - not the fiscal deficit, not central government, not general government - anything that is in its nature fiscal, a point we'll come back to. And not, should be made available to the press or made available to parliament, but to the public.

Let's go on to the third. This is about budget processes.

• Budget preparation, execution and reporting should be open.

All of those phrases are important. One of the biggest problems that we come across in doing work in the IMF and in the budget, is that there are still a large number of countries which invest considerable administrative resources in preparing a budget and the budget is dead in the water before the first month is over, because it is based upon unrealistic revenue projections. It is based upon processes which simply do not work and from day 1 it is the actual budget execution which determines the budget priorities no matter what kind of effort has gone into budget preparation in advance. You need to understand and be open about the limitations of your own budget preparation systems.

The first three principles you would think should tell you the whole story. We know what the government is doing, we know is has got to produce information, we know it has got to be honest about what it is doing. The last principle, however, could be summed up in the classic phrase, "Never trust the government". You need to have independent assurances of the integrity of the fiscal information. And that is a principle which is not as cynical as it may have sounded. Even the most - in fact all the most advanced countries - have in place systems of check and balances to make sure that the government cannot get away with being economical with the truth for very long.

Let me go back to quasi-fiscal activities which can be a little hard to follow. The point about the last two boxes on integrity, you notice that the square box, independent assurances of integrity, is placed next to them. That is to try to get the point that the top half, as it were, of the slide is contained within government. It is the bit down below, Public availability of information that is outside of government. The importance is that the public availability of information and the independent assurances of integrity should rely on people other than the government to deliver what is required.

I have covered the four basic principles. Let me now go into the individual practices included in the Code. Again the experienced bureaucrat will recognise the sort of wording that comes in here. The good practices included in the code represent a standard that most countries should seek to meet. We played around with words like minimum standard, basic standard, basic practice, good practice, standard practice. This was the most that I think one could reasonably use, given the universality of this initiative. That puts restrictions on how detailed you are. The point is that, in what I am about to talk about, we have set what we believe to be basic minimum standards.

I am going to go through the practices under each of the four principles. In each case there is one slide setting out what is required, so let's go on.

The first one is the simplest.

• The boundary between the government sector and the rest of the economy.

Governments interfere with the private sector directly in three ways - through laws and regulation, through taxation, through transfers. What this basic point says is ideally we want to be able to measure each of these three. Regulation is by far the most difficult of these to measure but it is an important dimension not to be forgotten.

Next standard:

• The allocation of responsibility between central and sub-national government, between the executive, legislative and judicial branches of government should be clear.

I will just make three points on that.

First there is a wonderful phrase which people in the United States use and that is the reference to the "unfunded mandate". What this really means is that it is all very well to be clear about responsibilities but that does not mean that you can pass them between central and local government and simply say, it is up to you, you sort it out, without matching the funding to what they have to do.

Secondly, a point we often make is that the judicial branch in particular has to be independent of the executive. If it is not, then much of what you are trying to achieve in fiscal transparency cannot properly be done.

Third. You need some kind of mechanism for managing your budgetary and extra-budgetary activities. The Fund isn't saying - let me make this quite clear - that extra-budgetary funds are bad or should be banned, that there should be no health funds, or pension funds or social security funds or road funds. What it is saying is that if you have such institutions then they should be recognised as being a part of fiscal activity. They should be consolidated when you are trying to measure general government. They should have the same budgetary practices and roles as a central government.

Next standard:

• The relationship between the general government, the central bank, public financial institutions and non-financial public enterprises should be clear.

Again I am going to duck this one because we need to pick it up later. The essential point is that if you think of fiscal activity as being regulation, taxation and the spending of public money, you have also to consider that it isn't just general governments and local governments who do such things. A central bank that is directed to give credit to an enterprise at something below the market rate is indulging in a fiscal activity. A public enterprise that is required by the government on its own books, to run a health clinic or to run some kind of education class is undertaking a fiscal activity. And trying to get at and measure those is what quasi-fiscal activities are about. A big problem the Fund has right now in this area is trying to identify exactly how you include in the fiscal accounts the cost of restructuring banks in countries where those sectors have gone into some kind of sharp decline, or as in one or two cases, pretty well total collapse. The rules on this are not clear.

The last two points are back to motherhood statements. The sort of thing you couldn't possibly object to.

Next standard:

• The legal and administrative basis for the budget and taxes should be clear.

This is fundamentally about avoiding too much discretionary action, too much exercise of judgement, too many rent seeking opportunities, which is the Fund's classic word for corruption.

Similarly, the last standard about good ethical standards of public servants includes the same sort of considerations.

Let me go on to the next principle, which if you remember was about the public availability of information. This is about fiscal information that should be provided and making public commitment to the timely publication of such information.

The basic minimum standard that is being suggested is that governments should account for all central government including extra-budgetary activities. That is, if you have a road fund, a health fund, an unemployment fund, a social security fund, you should be stating what is going on in that fund and consolidating the account in general government.

How many people, I wonder, in the audience realise that the much heralded success in the United States of closing the Federal deficit has other interesting dimensions. For example, without the social security fund the Federal Government would be still in deficit. Second, including state governments (which are mostly in surplus), the general government position would have looked better in the past. So the question of a deficit or otherwise depends on exactly what measure you are looking at and the important point is that while you should look principally at central government, you should always look at consolidated general government also.

Next standard.

• You should, as a basic requirement in presenting your budget, show what has happened in the two preceding years and show what you expect to happen in the two next years.

It is still the case that in a very large number of countries people in year T - let's say now, 1999 - are preparing a budget for the year 2000, using the provincial outturn data that they had for the budget of 1998. As I remarked earlier on, in too many countries the 1999 budget would be dead in the water within a month because it is based upon unrealistic projections. But by continuing to base each new budget on the budget of this year, which often tends to be the case, you get into a situation in which you perpetuate a myth about the resources that are actually available.

This is the most difficult point of them all and I apologise in advance before it comes up. This is economists' stuff, but it is important and let's quickly go through what each of these means.

Next standard:

• Contingent liabilities, tax expenditures and quasi-fiscal activities.

Governments are perennially caught out by having to meet some expenditure they did not anticipate. It matters not whether that is caused by war, or by storm damage, or by somebody triggering a guarantee that the government had given many years ago that it would repay, or whether that was an explicit guarantee that the government had signed, or some kind of implicit guarantee. A good example would be where a local government has got itself into trouble, the central government has no formal obligation to repay that debt but it nonetheless feels an obligation to do so. I think that was basically an issue in Brazil in the last few weeks.

Another contingent liability that people often forget is the judicial review. Governments draw up a law, they put the law into action. Somebody says, that is not how I think this law should be interpreted. The case goes to court, the plaintiff wins and suddenly there is a large liability on the government. All the Code is suggesting is that where you have these contingent liabilities, make a note of them, try to understand them, try to work out what the likely effect is going to be in any particular year, and try to allow for that in your budget. Let me say now, this is not a perfect art. You will get it wrong many times. What you can say is it will be even worse if you do not allow for it and you do not have a record of all the liabilities that are out there.

Tax expenditure is again slightly economist jargon, but it is the concept that, if you think in terms of government activities (regulation, tax and expenditure), sometimes the relief of tax is basically the same as having given a subsidy to a company. So any kind of investment incentive or special taxation regime has a fiscal cost and what the Code is doing is encouraging people to recognise that.

Now that said, let me quickly move on and say that most OECD countries can't produce a list of their tax expenditures, let alone anyone else because you have to define what the base is and so on. But it is an area where I think there will be increasing attention: what is being achieved through the tax system through incentives and special regimes.

Quasi-fiscal activities we have talked about. What are the central banks or development banks or non-financial public enterprises doing that are basically forms of fiscal activity?

Next standard.

• Governments need to be clear about their debt and their financial assets.

Again you would think this would be crystal clear, but it is remarkable how many countries still get caught out by not having a clear record of what debt they have and when it is due to be repaid.

The last standard: advanced release dates is actually picked up better later. Let's move on.

The third principle: open budget documentation, budget estimates, budget implementation and monitoring and fiscal reporting. All of these should be fairly straightforward.

Next standard:

• The budget should specify fiscal policy objectives and include an assessment of sustainability.

This covers a whole range, from the sort of fiscal policy rules that the advanced countries like New Zealand and Australia, (and to some extent the UK, have now with their golden rule), right down to just the need to estimate what you think is going to happen in the next couple years. The Fund, and the World Bank does the same, always puts emphasis on having a medium-term expenditure framework and a medium-term fiscal framework. Know where it is that existing policies are going to take you. Next standard: comprehensive micro-economic framework. That is just part of the same thing.

The next:

• assessment of the fiscal risks.

I can't speak for many other countries but I know that when I was working in the UK many years ago we used to produce the annual budget statement with a certainty which seemed almost religious and we would state things in terms of what we were expecting to happen with not the tiniest grain or shade of doubt. It seemed that we were setting up for the next year not what might happen, what could happen, but what would happen because the government had said that it would happen. In difficult questions about the exchange rate and interest rate, we would reply confidently that these were based upon our best assessments, after lots of professional work. Of course three months later something would have happened and it would be out of date, but it didn't matter. For the two to three weeks in which the pressure was on from the press you sat there looking as omniscient as you possibly could.

What this is about is to acknowledge that we live in an uncertain world. It is not to say that people should start saying, well the exchange rate will probably be 10 : 1 but it might be 12 : 1 or it might be 8 : 1. Or the inflation rate is targeted at 10% but it might be 20% or it might be zero. That is not what this is about. It is about acknowledging that there are risks on exchange rate and inflation. But also acknowledging things like you don't know what contingent liabilities are going to arise. You may not know what the demand for a new policy instrument is actually going to turn out to be. There may be some genuine uncertainties and it is simply saying that governments ought to be perhaps a little less arrogant and a little more forthcoming about the uncertainty of the world we all live in.

Next standard:

• Following recognized accounting standards.

This is an important if tedious point. Accountants, whether in the public sector or the private sector, seem to me to have developed a creativity which may seem admirable, but which is also questionable. Anyone who doubts that has only to look carefully at the way that various European Union countries qualified to meet the Maastricht criteria. You can produce different numbers within the constraints that international classifications can put upon you. Budget classifications tend to follow a fairly simple standardisation which is partly derived from something the IMF put out called GFS. But there are increasing moves towards international accounting standards, particularly the work done by a group called IFAC. This is not some attempt by the IMF to impose accrual-based accounting on everyone. On the contrary. While I think everybody recognises that is the direction in which government accounting is going, nobody is suggesting you can get there quickly. It is saying that where there are international standards everybody should follow them. I think if more are applied, the scope for creative accounting that we have seen in recent years will gradually diminish.

Next standard:

• Stating objectives of government programmes.

What this is saying as a principle is that where a government introduces a new programme it should say so and it should say what it is trying to achieve. Of course there are all sorts of techniques out there. Where do we start? We have output based budgeting, resource oriented budgeting, performance budgeting. We have accrual based budgeting and so on. The Fund has no difficulty with those. We are not seeking standards in aspects of participation or output orientation or anything like that. What the Fund is saying is there is a basic minimum standard. Be clear, be honest. When you have introduced a new policy, state what it is trying to do. As another dimension of that, please distinguish in your budget statements between ongoing policies and new policies.

Next Standard:

• Fiscal indicators.

It is too easy just to rely on the cash fiscal deficit as telling you where you are. To know where you are you often have to look at various different fiscal indicators and they will partly depend upon the circumstances that you are in. Without getting into any great detail, the Fund also considers measures like the primary deficit, which excludes the expenses on debt interest, things like a cyclically adjusted indicator, and increasingly something we have taken to calling the augmented balance, which is our latest attempt to deal with the costs of bank restructuring and the burdens that they can impose upon the fiscal sector - it is simply saying that the days when you could put up the revenue expenditure, calculate the cash deficit and say that's it, that is the fiscal gap, have gone. You have to be rather more sophisticated.

Next standard:

• Clear specification of execution procedures.

This is simply about having budget manuals, accounting manuals, financial regulations in place.

Next standard:

• Fiscal reporting.

I am going to cover that in the next slide, so let me not waste my scarce time doing that just now.

The last principle, independent assurances of integrity, is about institutional mechanisms to ensure the quality of fiscal information.

Next standard:

• There should be a national audit body.

First it has got to be independent of the executive and that means not appointed by the executive or controlled by it. Secondly it has got to report to the legislature on its findings, not the executive. Thirdly, just saying that you have met the financial or proprietary conditions on the budget is no longer enough. There has to be much more work in auditing, about saying whether a programme is meeting the objectives that you set out for it. Is it good value for money? Is it being efficiently run? Is it effective?

Next standard:

• Openness of macro-economic forecasts to independent scrutiny.

Let me give you the range here. What the Fund has in mind is not necessarily the kind of complex arrangement that, for example, I know the UK had when we used to produce the projections. There was another group of economists working in the private sector, who could take the treasury model and play around with the variables and come up with a different outcome. Recently the National Audit Office was asked to conduct a separate audit of the treasury forecast and so on. That is one extreme. At the other end it is as simple and open as going to the leading university and saying, please tell us what you think is going to happen to the economy in the next year and we, the government, will publish a statement: these are your forecasts and these are ours. It is simply trying to constrain the government's ability to tell a story.

Next standard:

• Independence of the National Statistics Office.

I would put this point very simply, given that I have only got five minutes. There should be no-one who can influence when or whether any set of statistics on the economy, that are regularly published, are published. No-one should be able to say, delay that for a week or, don't publish it this week or, we have to change the way this particular indicator is calculated, therefore we are not publishing it for six months. Those kind of decisions should be matters for professional judgement only. No political influence.

The last standard is:

• statistical standards

These are voluntary codes, the Special Data Dissemination System and General Data Dissemination Standard set by the Fund, but they are about the timeliness, the reliability and the periodicity of information. In other words, how often are reliable and comprehensive data published.

Let me go on now to the last section of this presentation. Implementation of the Code. It is voluntary, that is the first point. Second, we know that is has to be general because it has a lot of countries to cover. Three, we know that technical assistance is required in many cases. I'll come back to that. Four, let me try and define our limited rule.

Let's start with voluntary implementation. Let me make this absolutely clear: this is a Fund initiative that is voluntary. It is not compulsory. There is no score sheet. It is meant to assist the Fund in two of its three activities. The Fund does three things: surveillance work where it looks at each economy; programme work where we provide financial assistance for economies with Balance of Payment problems; and technical assistance. Whether you subscribe to the Code or not, does not affect your eligibility for Fund technical assistance or for Fund programme assistance. So it is voluntary.

To help people we have a manual which you can get hold of on the Internet. We have also put out a questionnaire which we encourage people to fill in. Let me emphasise that we encourage people to fill it in. It is not the IMF coming along like a Time and Motion man and simply ticking off and saying you scored X on this or Y on that. It is for countries to look at and compare themselves with the standard and say, well do we meet this or do we not? We hope that many will do that and then come out with some kind of evaluation report on how they have been doing and if they do, the intention is to put that on the Transparency Web Site http://www.IMF.org/fiscal.

Let me make another point before I leave that. There is something that you need to think about. Having put this initiative in place and having got these questionnaires in place, it is like a free good, like a public good on the Internet. There is nothing to stop other people working on this, deriving their own score cards, including the people that lend money to countries. It is an interesting aspect.

The focus is on a minimum standard that should be attainable by all countries. We always recognised that, for example, Anglophone budget execution systems and Francophone systems - while not totally different - have a lot of different dimensions. We were trying to cover every single country, from the greatest in terms of their progress made to those that still have a long way to go. So how do you, as an individual country, decide where you stand? We would encourage people to look at basically how do you relate to another country that looks to have a broadly similar background. And thirdly, in light of that, technical assistance can play a role.

We recognise technical assistance is required. We can do some of it, but not that much, because frankly it is not a big area of IMP work. The division that I am responsible for does most of this work. There are precisely ten of us, so it is not exactly something where the IMF can go around giving huge amounts of technical assistance. It is partly for that reason that we have been encouraging other international organisations or national organisations to work with us and the OECD, for example, is now developing a code, a sort of higher standard code. We have encouraged people to pick up on all sorts of different dimensions. Some people have an interest, for example, in issues like quasi-fiscal activities. Other people have picked up on contingent liabilities, participatory budgeting, different systems of budgeting. We always recognise that would be the case. We encourage it to happen. We are not against it.

I can't emphasise enough that having set matters in motion, the Fund is not seeking, not even trying, to control how far it goes or who runs with it. We have always recognised that the Code is going to be more effective if other international organisations support it formally and in practice.

Next slide. What we will be doing is trying to promote the concepts of fiscal transparency in connection with both surveillance work and the technical assistance. I say again it is not any part of our intention that this should be linked to the programme work. The whole idea is a voluntary initiative, setting some standards for countries to look at, to understand where they are, to try and spot their own shortcomings, to try to take it forward, working on this.

I will just finish by saying that I have made a copy of all these slides and there is a short paper that we also have, which I hope will be circulated to everybody. The material is available on the Net. Anyone who wants to look at that questionnaire can do so. You can draw down the manual, which is a big document and goes into many of the points I have made, in much more detail, if you wish to. The web site address is actually included in the material that is being circulated, so it should be possible to get it from that. Thank you.

Mr. KRAFCHIK: Thank you Barry. I am going to ask Richard Humphreys from the Centre for Policy Studies to respond briefly and then we'll open the session for questions.

Mr. RICHARD HUMPHREYS - CENTRE FOR POLICY STUDIES

Before making a few introductory remarks about the IMF’s code, I would like to thank the organisers for the invitation to participate in this very timely conference. More relevantly, I would like to thank the Ford Foundation for the extremely important role they have played in funding this, and similar conferences, in South African and in other parts of the world. As in other countries undergoing a transition away from authoritarian rule, the Ford Foundation has played a very crucial role in South Africa in strengthening many policy research and advocacy organisations. Understanding budget issues is a key element of their research work.

I would like to make two general points about the IMF code.

The first is that as an undergraduate in political science I was taught that politics, following Harold Lasswell, is about who gets what, when and how. Budgeting is clearly the essence of politics, since it involves choices by the government, even though it might be portrayed as a very technical exercise. Thus budgeting is about politics and politics is about budgeting.

The second is that I do happen to think that the IMF deserves credit for having drawn up this code. When reading it, I was struck by the realisation that its contents are applicable not only to states in the developing, under-developed or Third World spheres, but it is in fact equally applicable or of relevance to advanced economies and the budgets that these states craft annually. The code notes that many of these advanced economies would have some difficulty in meeting the requirements of the code. I make this point purely to try to move debate away from what might be a rather narrow perception that the IMF’s prescriptions, so to speak, are aimed at one sector of the global economy. As to more specific comments about the code:

First, I would like to raise the issue of the voluntary nature of the code and whether it is in fact connected to programme or technical assistance. Mr. Potter noted that it was not; but I would simply raise, for purposes of debate, whether in fact it should be more than just a voluntary code. In short, is there not some merit in going beyond its voluntary status to linking it more directly to programme or technical assistance? Many persons in the audience might think that it should remain a voluntary code, but some others might question some of the consequences of leaving it as a voluntary code. Perhaps we could debate this issue in greater depth during discussion.

Second, how does the growth of sub-national governments in many parts of the world impact on the IMF’s links to simply national governments? One consequence of the growth of devolution preferences during the last decade has been to vest greater budgeting powers in sub-national governments; how does the code take this into account? The recent Brazilian crisis, which triggered renewed jitters amongst emerging markets, was caused, it should be noted, by a state government defaulting on a loan to the central government. While this might be an extreme example, it does still point to the ability of sub-national governments to impact on broader budgeting processes. In South Africa itself, three of our nine new provinces have experienced severe financial crises during the last two years which affects the integrity of their budgets.

Third, the universality of the IMF’s code was noted during Barry Potter’s presentation. But, while this very universality is understood, does it not undercut the impact of the code in very specific situations? I make this remark because I am always struck when participating in these international, comparative, cross-country gatherings how very different the political, institutional and economic situations are from country to country. One has to understand these situations, and the limits that follow from them, before one can start to generalise comments across the globe, so to speak.

Finally, I would agree with Mr. Potter that civil society organisations have a very important role to play in facilitating debate in different countries around the contents of the code. Donor organisations can also play a key role in this respect, if they agree with the broad trend of the code’s contents.

Mr. KRAFCHIK: Thank you Richard. It is now open for questions from the floor to Barry Potter or to Richard.

QUESTION: One point is about contingent liabilities, these liabilities should not only include the public sector. I think this is an important issue. You mentioned Brazil and sometimes you have to consolidate also local governments which I think is very important. But the other contingent liability that you should, actually we all should be aware of, is some liabilities in the private sector in those key sectors which we will never allow to be or to go bankrupt like the banking system. You mentioned some banks but you never mentioned commercial banks. In Mexico these contingent liabilities and in many other countries it has happened somehow in Japan and in some other Asian countries, the commercial bank is also a contingent liability. It is very hard to analyse what is going to be the burden for the public sector I know, and I understand that. But I think it should be worth mentioning that there are some other aspects of the private sectors that will represent a liability. On the other hand also, we should be very careful also with the liabilities in the privatisation process. Like for example in some other countries when you privatise roads or where you are privatising electricity or something like that, if the projections to get a future income, a flow of income, are not well specified, then some countries are bailing them out. So this is a contingent liability that I think should be taken into account.

There are a couple of very quick points I would like to make. We should have accounting standards internationally but also it is very important that we have standards, similar accounting standards among levels of governments. In many countries it is very hard to compare, for example, investments across various governments. So it is very important to stress that also among levels, different levels of governments, we should also have standards and across some national governments also. At least I know in Latin America in general, most national governments have their own accounting standards and they are different, they differ from each other and that makes it very hard then to compare different states or different municipalities.

Finally, I would like to stress that it is very important for this Code to have not only fiscal indicators but performance indicators. I think this is a challenge in which the IMF should play a very important role as well as the NGOs. Thank you.

QUESTION: I think it would be particularly interesting to see how this Code would be received in a country like Nigeria. So we congratulate the IMF for their efforts in this area but I would just like to pick up on one point, which is the fact that you would not like to link it to your programmes and asks what happens in a country like Nigeria where corruption is endemic, where it is systemic, where IMF programmes have failed consistently as a result of a lack of fiscal transparency? Where even all the structural adjustment programmes and things like that that have been prescribed by the IMF and various donor organisations have failed and they know that it is precisely because of a lack of standards that they can hold anybody to. I am rather surprised that the IMF would do all this work and at the end just leave the momentum to dissipate by not insisting that it is linked to their programmes. I know that at the Joint Ministerial Conference that was held in Hong Kong - I think that was in September 1997 - the IMF and the World Bank came out with joint statements that, from then on, any country that didn't do something about the corruption that they were facing, would cease to get support from the IMF and the World Bank and if they could go that far as to make such a statement and then on a code, which they claim should be the minimum standard, they now find it non-enforceable, can we really call it a standard then? Perhaps it should be the minimum ideal, because since it is not - you can't enforce it, you can't police it, you can't do anything with it except hope, and really we have always had that hope because these countries that flout these minimum standards, is not from a lack of knowledge. In Nigeria there are professionals in most of these areas, only they decide to do something completely different with their training. So I think really that the IMF should revisit this issue and consider making it linked to their programmes. Thank you.

QUESTION: These are actually two questions here but one is virtually the same, just stated in a different way.

You said that this is standard which I believe you said most countries should seek to meet and in phrasing it that way I wonder if it doesn't - I understand the reasons, the need to be culturally sensitive and politically sensitive - but in phrasing it that way it seems to undercut the force of the Code. I mean if you look at the power of things like the Universal Declaration of Human Rights, it is precisely because that language is not there. I mean this is a Code which all countries are expected to try to meet and so I was curious about why you only say that "most countries should" and that would seem to imply that there are some cases where some countries shouldn't and how and when you would make that decision.

The second thing was when you were making the case for the Code at the very beginning you used primarily market rationales for justifying it and I can think of lots of reasons to justify a Code, the primary being that governments are spending taxpayers' money, either current taxpayers' money or to the extent that they take out loans, the tax revenue of future generations are being committed. That would seem to be the principal reason for economic decisions to be open and accountable, primarily to your own citizens. Because it seems to me that by using a market justification, it again does not weaken the case because it is not entirely clear to me that markets automatically reward open and transparent economic decisions and automatically punish those that aren't. I think you could take the cases of Indonesia, Thailand, South Korea or Brazil. You could probably have a pretty healthy debate about and depending on what time in history you are looking at, you may come to different conclusions about why the markets reacted the way they did to those countries. Clearly the markets treated Indonesia and Thailand very well for a long time and then all of a sudden they did not. It is not entirely clear to me whether that is simply a judgement about transparency versus other factors of economic performance.

Mr. POTTER REPLIES: I think probably the first thing I learned from those questions is that you should never cut your presentation short even when you are out of time.

There were two further slides, which cover two of the points that have been made. One specifically refers to the Hong Kong 1997 Declaration and the other one refers to the fact that we have a separate Code in preparation that is covering monetary and financial policies, which deals precisely with this point about the commercial banks and our concerns about them, what sort of transparency should be set for banks and when it is that a government should have some kind of responsibility for them. Let me go back on what seems to me to be the central point that is being made, which is this voluntary nature of the Code and the linked programme activity. Let me explain a bit about how Fund programmes operate.

Essentially the Fund lends money, as you all know, for balance of payments support and in turn it requires certain conditions to be met. In the past, those conditions were related to such things as the control of monetary aggregates or the control of the fiscal deficit. For some time now and predating anything that was done upon the Fiscal Transparency Code, the Fund has also required certain performance benchmarks or prior actions or performance criteria, which relate to fiscal policy. They may relate to introducing a specific tax measure or modifying a tax measure or they may relate to some aspect of the fiscal accounts.

Now, while there is no intention to say that unless you subscribe to the Code you will not get programme money, there is every intention to say that when we are looking at the situation in any country, if we are basically, as the banker in this situation, as the lender, unhappy with the way in which fiscal policy is being specified or is being conducted, including on issues of corruption, then we may well either refuse to lend at all, which we have done on a number of occasions, or require that certain changes be made which would pursue fiscal transparency. So there is a mechanism available to us, which is being used to promote fiscal transparency. But it is not a question of saying you have to subscribe to the whole Code in order to get the programme money. Let me give you some real examples. You'll forgive me if I don't mention the countries. We have said to at least one East European country, that we would be prepared to lend to them and to continue giving support provided they cut out, or cut down the number of extra-budgetary funds and start to consolidate their fiscal data in a reasonable way. That is essentially part of following the Fiscal Transparency Code.

In a number of African countries we have gone in and said, this is basically a quasi-fiscal activity which we want you in the long-term to close down, but in the short-term we want to see the accounts consolidated with general government. So there are all sorts of little individual aspects of the Code that can be pursued in programme work.

That said I go back again to the fundamental point that I made. This is a voluntary initiative which we hope many other organisations will pick up, including NGOs, including the regional banks and including the World Bank. We would see the means by which this Code is implemented or carried forward as not primarily coming through these sort of conditions on IMF programme work, but on the contrary being taken forward through technical assistance, partly by us, partly by other people. Many of the points that have come up are related to that.

The gentleman from Mexico raised the question of local governments, as did Richard. Again, we ourselves are somewhat constrained by our own mandate. The IMF has a mandate to work with national governments. It does not have a mandate to go and negotiate with local governments. We do talk to them, we have to. The Brazil example is one obvious case where it is necessary to look at that. But the World Bank, for example, is now taking the Code and looking to see to what extent the same principles should be applied to sub-national governments, local governments and state governments and how these should be modified in some way or another. And it is right that they should. It is right that somebody should have picked that up and had a look at it. Let me see if I have missed any other points I want to pick up.

Oh yes, performance indicators. Again that is not something that the Fund objects to. How could it? I mean we are writing about that all the time, different missions are going to different countries and saying performance indicators can be an important part of the way that the budget process works. But it is not something that you could have laid down as a universal standard at this stage. Many of the countries we work with in the former Soviet Union would not be able to produce any kind of performance indicators at this stage. If we had set that kind of requirement in place we would have gone beyond what is realistically achievable.

But I do hope that there will be more specific standards arising not so much from us, but from other organisations taking this, building on particular aspects and encouraging people to follow them.

QUESTION: I think the presentation was technically superb and if you could tell us what software package you use it will help.

I just think this is welcome from previous IMF documentation and language. You know it was suffused with medical jargon around prescriptions and taking your medicine and even on some occasions shock treatment. I must confess that when I heard about this Transparency Code I thought, oh IMF is going to take a good look at themselves, but I have been disappointed, it is about governments and not about themselves. So sorry Mr. Humphreys, I do have an axe to grind. I mean ESAP, Economic Structural Adjustment Programmes - in Africa we call them Extreme Suffering for African People, ESAP. But just coming to your presentation I don't think some of the questions were answered. The way you look at the market for example, as if all countries are captives of this global market, they are not given any kind of agency in a particular way. Also you make statements like markets play an important role in exercising discipline, but I am afraid the past while has shown us that the international market plays an important role in exercising global anarchy. I mean if you look at the role of speculators etcetera, I can go on on that. That is the first question.

The second question is, it is not so clear when you talk about the greyness or the boundaries between private and public, that these need to be defined and I don't think the question from Mexico about privatisation was really answered and I am interested in that.

But the third area is you know there is this thing about technical arguments and that there is no politics involved. There isn't any history involved in the presentation I have seen. Social justice issues are left out of the equation. In South Africa we followed - and I am very interested to listen to the next input about South Africa and how we could apply this model. But I mean we followed the three-year rolling budget model as proposed. We have looked at a comprehensive macro-economic strategy and we are working towards a number of areas around transparency but we still make decisions like spending R30 Billion on armaments when you know there is so much poverty in this country. So I think decisions like that, how do you decide whether to build a casino because it might make accounting or budgetary or economic sense or build creches which are required? I can go on but, just finally, when you talk debt as well, when you mentioned that, shouldn't we look at the context, shouldn't we look at it historically? How did this debt arise? In the South African context we heard Warren talking about the Odious Debt. Those kind of questions are left unanswered. Thank you very much.

QUESTION: You used the word "motherhood" a lot, which I thought was appropriate and true. Much of the basic approach you laid out was very useful and to those who may not be economists probably quite insightful. Indeed I can think of many examples of governments in East Africa, which have actually taken on in large measure some, if not many, of the prescriptions being set out in the guidelines. For example, many of them have now introduced or beefed up national audit offices. They have actually integrated the work of those offices with the legislative oversight through their public accounts committees. My colleague maybe wants to say more about that. There is a better track record on publication of financial statistics etcetera. In other words it is arguable in many cases, often in countries which have IMF lending programmes in place, that countries have already gone some significant way down towards these guidelines. However, I would suggest it is probably arguable how effective these have been. In other words it is a classic problem. Much of the architecture of what you are proposing may have been put in place because it may well make finance ministries look a lot better in dealing with IMF missions, but to what extent can you monitor compliance and the effectiveness of the sorts of measures that have been put in place.

The second point is that I found it quite revealing and in some ways encouraging that you saw these guidelines as not being entirely voluntary. These guidelines can be translated into conditions and built into conditions for lending programmes and that would seem to be quite obvious and reasonable. Though at the same time it is not obvious to me how the conditions that you establish are internally scrutinised or monitored. In other words do you make public, for example, the monitoring of those guidelines and their implementation?

The third related point concerns that of the media and whether in your technical assistance you hope to support the media. Thank you.

QUESTION: I was going to ask if your use of the word fiscal meant that you were making a distinction between fiscal policy and monetary policy and you have answered that by suggesting that there is a code to be expected that will cover monetary issues. But maybe you could elaborate on that just a bit and let us know if we should expect a call for transparency of government actions related to issues like money supply and exchange rates?

Mr. POTTER: Let me first of all pick up a couple of points that were made earlier, which were repeated.

I should have made clear at the beginning that while obviously this question of markets and the influence that they have is important, the fundamental reason for transparency is of course internal accountability between governments and their electors, between governments and their legislature, between officials and their ministers that they work for. That is why you need to be transparent, because if you don't then it is very difficult to hold people accountable. Just simply tracking down who did what can be very, very difficult. Within the last ten years perhaps - no more than that - it was possible for a permanent secretary in the United Kingdom to refuse to undertake a particular measure at the request of his minister because he believed it to be illegal and to allow the decision to go to the courts and for him to be found correct. That is a good classic example of proper accountability being put into place.

Secondly this whole question of whether the market has a perverse or a beneficial effect. That they tend to overshoot, to overreact is beyond dispute. Nobody I think could deny that in the light of what we have seen or the ways in which some very sharp exchange rate movements have been partly reversed. As presumably many of you will know it is true that in all three of the large Asian cases the exchange rate fell and then has come back by roughly 50%, of that fall within the last two years.

The issues that were raised by the gentleman from South Africa will always be with us. The question of priorities, whether it should be military expenditure versus schools versus casinos, that is not something for which the IMF has a responsibility. I can't make that clear enough. That is a matter for a government to decide. Governments draw up their priorities, not the IMF, though we are opposed to unproductive expenditures. Ultimately, governments decide what they are going to spend their money on. What the Code says is, fine, if you must buy tanks and not build schools, be transparent about it. State what it is that you are doing. Make it clear. That is the point about transparency. Transparency is not about trying to impose some particular set of judgements about the priority that should be followed. It is about saying that when you have taken such decisions, for whatever reason, be open about it, say what it is that you have done. That is what we see as being at the core of transparency.

How do we monitor compliance? Let me distinguish between the technical assistance work and the programme work. Where we have given technical assistance, where people have filled in the questionnaire, the compliance with the Code will be fully observable to the public. We are encouraging people to fill in these questionnaires and to put the questionnaire back on the web site. So in that sense we will be able to observe what is or is not being done according to the country authorities.

Do we really know whether it is being done? I think that depends. At certain levels, at the more aggregate levels we do, because there are the annual surveillance trips. In many of the countries with programmes there are resident IMF representatives. They can go, they can check out whether something is actually being delivered or not.

When it comes to the monitoring of the conditions of programmes, are they secret? The answer to that is yes, they are. They are not made available to the media. That has been part of the way in which the organisation has operated. It gives the money, it requires certain conditions. Those conditions are private between the IMF and the government. There is nothing to stop the government from revealing what they are, nothing at all and many governments do that and do that to good effect. It has been clear for many years that a number of governments, faced with an IMF programme, get into a situation where the IMF requires them to do something. It is often easier for the government to say, we don't really want to do this, but the IMF insists, so we have to do it in order to meet the conditions. I think the IMF has always recognised that that is part of the way in which these games are played. But the decision on whether to open up the conditionality in terms of expressing what is actually happening is a decision for the government. You might logically come back to me and I wouldn't blame you for one moment for saying, but surely if you are in favour of transparency, you as an organisation should be in favour of making those conditions public? I suspect that is exactly the way that matters will develop. But at this point, that is not the situation.

Let me turn lastly to the monetary code. Without revealing too much, I can say that my colleagues who work for the monetary side of the Fund have had far more difficulty in drawing up the code than we had on the fiscal side.

I think on the monetary side nobody is really too sure exactly how to go forward. There is a much greater acceptance of the Fund's mandate in fiscal matters, ironically enough. I am sure you have all heard this old joke but people who work in the IMF know that it doesn't really stand for International Monetary Fund, it stands for Its Mostly Fiscal, because at the end of the day most of the problems that you deal with are fiscal in nature.

On the monetary side there is the separate question of the role of people like the Bank of International Settlements, and whether they should have some role in setting financial regulations. We are having difficulty not so much in dealing with the government aspect of the Code. The question of how government deficits are financed and being transparent about that and the ways in which a fiscal deficit is being financed, that is not a problem. But on the other side of the house, for example, in looking at the commercial banks and what can be done there, that is more difficult to work out.

QUESTION: Good morning, I want to thank you for coming a long way to share your perspective with this group. I suppose it is a good sign of how much you want to encourage other groups to take up what you have done and make something of it. I also want to say, before I say anything else, that I appreciate that you are working for the International Monetary Fund and have certain constraints that emanate from the role that the Fund plays worldwide.

Having said that I want to give you the feedback that I agreed with Brad Smith when your original presentation seemed to be focused mostly on the importance of transparency for government discipline and the role of lenders. And then you responded to him by saying that we wanted transparency in order to attain democratic accountability in some way and you seemed to concede that there were additional reasons that transparency was important. The question I have is whether you would elaborate on the relationship between transparency and accountability? If I can make a distinction which is over-simplification of reality, one can have all the information in the world but if there aren't mechanisms in place to put government decisions to the inspection of the people that are ultimately going to be affected and provide opportunities for debate for political competition for hearings in a timely manner, then you will fall short. I am sure you have debated additional mechanisms that might be in place in order to allow for the kind of internal national discipline that you seem to agree was important. Perhaps you could comment on that.

QUESTION: I just have one question. Could you name five countries which meet all the rules that are in the Code, or which are the closest? You mentioned that in the UK it was difficult or different a couple of years ago and you also mentioned these are the minimum rules. Thank you.

QUESTION: First of all I must say thank you for this presentation. I am liberal by my opinions, but last year in Russia when my liberal friends were in power we made some progress. But now I can see they are real liberals and maybe one of the reasons that they are faring not so good is also Russian. That was the idea of liberalism of 16 or 17 years and I think now we have a good example of liberalism of 19 years or maybe second century. For this reason I support your approach and I think that, for example our centre, would be very glad to organise a translation of this Code to Russian, with some commentary from Russian experience. I think also that a network of participation of this programme for budget control and maybe a good network for some time of feedback for your projects and this kind of collaboration would be very useful. Thank you very much.

Mr. POTTER: Let me cheat by answering the last two questions, which are slightly easier to handle first.

On Russia, we are actually trying to get the Code translated into several languages, including Russian and I am hoping that that is going to be available quite soon. If there is any difficulty in trying to get it done into Russian then if you write or you simply E-Mail the number then we can arrange to have it done. That is something that needs to be done. We have long recognised it.

Perhaps an unexpected answer to the question from the gentleman from Poland. Can I give five examples of countries that meet the standards. No. I should have explained this. What we started to do was to ask for volunteers to fill it in. Two countries were so confident of their degree of transparency they could barely resist sending something in the next day. We are in the process of writing back to them saying that impressive though it is, we don't feel it quite gets all the way, which will not be a very welcome message. Part of the problem is that I think I said earlier that some of these issues are really quite complex on a technical basis. One is likely to founder on the basis that it does not really have an estimate of its tax expenditures. There are also some questions about whether others, for example, meet some of the data reporting standards that are required.

Even though it is general, the basic minimum standards are still quite ambitious, not so much in terms of their detailed requirements but in terms of their coverage. That is making it difficult for countries to meet it. It was always intended to stretch countries. There would have been no point in setting a Code which 85% of the membership could have met overnight. The idea was deliberately to stretch and to push people as far as we reasonably could go.

The last question, or rather the first question from Mr. Lipsky. This is always a difficult question for us to answer. Irrespective of what you think of the Fund and what it does, it is constrained by its Articles of Agreement and by what it does as an organisation. Its role, at the end of the day, is to offer balance of payments support. Of course over the last 40 years it has become very clear that that requires you to get involved in many, many aspects, particularly in the fiscal aspect and in the monetary aspect and also the trade aspect.

When we were looking at fiscal transparency we very much approached it in terms of our own mandate and in terms of our own Articles of Agreement and that is why you will see the emphasis that I put initially on the market aspect of it because that was something that was very prominent in our minds. After all, we like it no more than you like it when the financial markets go pushing huge sums of money from one part of the world to another and destroying exchange regimes in the process. That benefits no-one except possibly them and even then you have to question it. I think that you have to see the Code as having originated partly from a concern that it was, in part, a lack of reliable information in financial markets that was leading to this overshooting. We have also recognised, as I think I did when I put it now, the fundamental economic case in terms of just good information and the way in which that allows you to operate your budget processes. Of course we have always been aware of the link back to political accountability. But frankly, it is a subject where the Fund is less prominent. We are very conscious, however, and have deliberately, I would suggest, set it up that other people and other organisations such as Idasa, can pick up the ideas in transparency, link them to the questions of political choice and work from there.

Mr. KRAFCHIK: I would like to thank Barry Potter and Richard Humphreys very much.