Transcript of an IMF Center Book Forum
Handbook on Fiscal Federalism

International Monetary Fund
Washington, D.C.
Wednesday, November 15, 2006

Moderated by Teresa Ter-Minassian, IMF

Panel Presentations:
Ehtisham Ahmad, IMF
Danny Leipzigers, World Bank
Vito Tanzi, Inter-American Development Bank

Transcript Contents

Introduction by Teresa Ter-Minassian
(4:00 p.m.)

MS. TER-MINASSIAN: I would like to begin by thanking all of you for coming to this most interesting economic forum to discuss or to present the book on fiscal federalism, which Ehtisham Ahmad together with Georgio Brosio from the University of Turin has co-edited.

Intergovernmental fiscal relations are of course a topic of growing importance around the word, and not only from the point of view of their importance for additional objectives of economic policy, good governments, location of resources, distributional objectives, but also from the point of - Fund standpoint because of their important macroeconomic implications.

And we in the Fiscal Affairs Department have been devoting considerable amount of intellectual resources to studying various aspects of this whole area of the public finances. This work began in the early nineties with Vito's support and at that point culminated in the preparation of The Book on Fiscal Federalism in Theory and Practice, which I edited and the FAD published in 1997, which is frequently referred to as "The Red Book." And I still hear periodically references to it.

It was a bit pioneering in the FAD because it focused especially attention on the macroeconomic dimension of fiscal federalism. And it included a lot of case studies, a survey of the state of - the Intergovernmental of fiscal relations around the world at the time.

It's now a very dated book, but fortunately a number of other important efforts have been made by the Fiscal Affairs Department staff, including a conference, which was conducted in 2000 at the behest actually of the Board. They expressed an interest in such an event and which was then reflected in a book co-edited by Vito and Ehtisham.

And this new book which Ehtisham has co-edited, it's a further step in the analysis of intergovernmental fiscal relations issues. And it is one which perhaps goes further than the previous ones in terms of the focus on political economy, aspects, and also on some of the newer issues that have come to the floor like, you know, distribution of natural resources across levels of government horizontally among members of the same level of government, the relation between intergovernmental fiscal relation and environmental issues, issues of governments and in particular corruption, and it also puts more emphasis on the role of good institutions in determining good fiscal outcomes in the countries that - well, different levels of government have important different functions.

So, I think this is definitely a sort of major new contribution to a growing area. It reflects of course a lot of theoretical contribution, but it also includes some chapters which are more of an applied nature. And some of the fiscal affairs experience with technical assistance in various member countries in the design and the reform of intergovernmental fiscal relation is also reflected in some of these chapters.

So, of course I am quite excited about it and I think this is a very good opportunity this afternoon for you all to hear from first of all, the main author and editor of the book about the main messages that come out from the various staples in it, and then to hear the comments of, on the one hand, Danny Leipziger, the Vice President for PREM and the World Bank. The World Bank has a long tradition of advising countries on various aspects of their intergovernmental fiscal relations, and of course, from my good friend and mentor, Vito Tanzi, who also has written quite a lot on this subject.

So with this brief introduction, I would like to turn to Ehtisham to start. You have 15 minutes or so for your presentation. Thank you.


Presentation by Ehtisham Ahmad (View presentation)

MR. AHMAD: Thank you, Teresa. When I came to the Fiscal Affairs Department in 1990, I was surprised that there was no work on fiscal federalism. And my distinguish colleagues, Ved Gumby and Milka Cofenegra said to me oh, you mustn't touch it with a barch (?) pole; it's too political.

So then I was very pleasantly surprised that within a year or so of my arrival in the FAD, there was Teresa's project to look at fiscal federalism issues, which culminated in the book in the late nineties and as Vito himself has written later in the volume which we published in 2002. And it focused on corruption.

That indeed is political, so the politics cannot really be divorced from discussions of the technical issues. And indeed, that's very much the starting point for this book. And I'm delighted that I have both my mentors here today, Vito and Teresa, both in the subject area, as well as colleagues in the Fiscal Affairs Department.

We were invited to prepare this volume by Edward Elgar and EE has a series of international handbooks. The idea of these international handbooks is to bring together the latest developments in the literature in different fields. And there has been quite a lot of writing in the intergovernmental area in the last 15 years.

So the areas really to - first is to look at the developments in the literature. And our own and at least my perspective was also to see the interaction with the policies set. What implications does the literature have for policy? And of course, this is a volume for researchers and policy makers, but I would just add that some of the chapters we have in this volume are more than just reviews. Some of these are actually seminal contributions in particular areas.

And I will not necessarily highlight the seminal contributions but just mention that some of these really take the literature forward in this volume. So that's actually, I guess, a labor of love for most of the people who have contributed to this volume, because we did it on a shoestring and there was no (off mike) for the authors or for the editors. So it is really a labor of love in this particular case.

So, where do we start? We see that the practicing of intergovernmental relations, particularly in the last 15 years actually begins to set the stage for the theory. The old theory actually does not explain many of the new constructs that we see around the world, so really there is an interaction from practice to theory.

And the second question is how does additional developments in the theory reflect practice and particularly policy making. And here particularly we focus as Teresa said on governments and institutions on effectiveness of the processes, particularly effectiveness of public services, equity efficiency, and political constraints in both revenue assignments and expenditure assignments and also the same thing on the design of transfer.

And of course this is the FAD and we very much believe that having a sustainable macroeconomic system is essential to carry the process forward in a systematic and efficient manner.

So let's turn first to the recent developments. Of course as you know, the early writings in the area, particularly Musgrave(?) and so on, were based on a conception of functions of government, which then lead to both the revenue assignments and expenditure assignments.

However, as we see, these are based on the assumption that governments are benevolent. And if you drop the assumption of benevolent governments, then you begin to ask the question as to what works and what doesn't. How do you achieve good governments? How do you achieve accountability?

And once you begin to ask these questions and then you begin to relate it to what's actually been happening around the world, the legal, political, and administrative issues become absolutely essential. And similarly, one begins to look at the importance of information, importance of reporting, importance of standardized reporting, importance of incentives and sanctions, because all of these come together in determining the incentives for good governments.

And really, if you look at the work of the bank and the FAD in the last few years, the focus on good governments and accountability very much is supported by the developments (off mike) -- good governments, accountability.

Coming to the institutional arrangements in policy challenges around the world, I would say that this is a universal phenomenon. You have, of course, federal countries, large federal countries in the rich parts of the world and the advanced countries. You also have countries like India and Brazil, where federalism is alive and thriving. There are other countries which are federal in nature but where there are tremendous strains and stretches, like my own country in Pakistan or in Nigeria. There are issues of conflict and some other parts of the world where federalism actually has not worked.

But we also look at unitary states where there is a degree of autonomy and a movement towards more devolved responsibilities. Countries like China and Indonesia are actually unitary states even though they look like federal states. All countries in Latin America or Africa, which are unitary states and yet there's a process of devolution. Spain and Italy, unitary states moving towards more federal constructs. The whole process in the European Union, where you not only have national governments and local governments, but a supranational government. How does the supranational government begin to affect the policy making circumstance?

Then of course you have the whole range of issues dealing with post-conflict countries. Will the centralization hinder or solve problems in post-conflict countries? These are issues which are of relevance. Iraq, Sudan, you know, the Balkans, all of these issues arise. Do you decentralize? Will it make life more or less difficult?

Similarly, you have issues in developed countries. In the OACD we have a network of the bank, FAD, and OACD, which is beginning to look at issues in Germany, in Austria, in Australia, and France looking at the process. These are not issues cast in stone. This is a continuous process in the advanced countries and in developing countries. So in a sense, what we have is a full range of countries, enormous range of issues, and a process which is a continuing one.

As I said earlier, the political economy considerations are fundamental. You have interactions between levels of government, between the national government and sub-national government. You may have interaction at the horizontal level given the mobility. And again, institutions and accountability are important.

And of course, Vito has written a lot about corruption. You can have corruption at the center. Of course, the scale is extremely large. Similarly, you can have corruption and capture at the local level lobbies. And capture at the local level is not unheard of and is a constraint in the way policies are designed and implemented. And of course, you know, long time ago in the Federalist Papers, James Madison warned about the issue of factions and being a constraint on federalism.

So the issues that the volume addresses of traditional issues, intergovernmental interactions and competition. It's a question of sequencing. How do you sequence it? And of course, going into the traditional areas like expenditure institutions and policies, revenue assignments, financing issues, transfer design borrowing, and macroeconomic stability. And then the special issues as Teresa said, the development contacts, the issue of natural resources, the environment, corruption, keeping countries together, and political institutions. These would be all issues that are addressed in the book.

So briefly, to go over some of the key themes, constitutions and legal framework are the starting point to discuss and examine the interplay between levels of government. And this is something that the (off mike) of this area, Albert Breton writes about beautifully in the volume.

And quite often in this interaction you can get asymmetric responses. You get asymmetric responses in countries like Indonesia, where you may have special arrangements for natural resource producing regions or even in federal states, you may have more powers or more functions in particular states than in others. And this is something which is nicely written by Roger D. Congleton. Issues of horizontal competition, again and in fact mobility nicely summarized by B.F. Selmon(?). On the whole area of political economy is an excellent paper by Ben Lockwood. I would recommend that very strongly.

And of course all of this results in spacial interactions. And there are a number of papers in this area on the causes and affects of tax competition, Revelli and Wonsome(?). And then the contractual dimension, this is again a new area in intergovernmental finance. Ikenburger(?), DuFray(?), and also Bernschpun(?) talk about these issues.

So, in all, what you find in the new literature is that the responsibilities and assignments come about as a result of bargaining between different levels of government and contract rather than any set principles for each government.

So if you are a policy maker, what are the key issues for assignments and outcomes? And of course, the question is who does what. And here again, you can redraw both on the normative theories of fiscal federalism as well as the positive theories of fiscal federalism in looking at expenditure assignments. These are much more complicated now that you bring in political economy issues. There is a very nice paper by Bernard Dafflon on this and the size of government.

Bernschpun looks at the scope for contractual federalism, which is now becoming extremely important, not only in Europe but also the mechanism for generating good governments in developing countries. The issue really is can you match these, can you monitor these new contracts. Can you actually implement them? Do you have the legal framework to do so?

And of course, there are issues with asymmetric operations. Can they work? Should they be designed? Congleton again does a very nice job for that in this volume. Govinda Rao discusses decentralization for countries that were formerly centrally planned or have vestiges of central planning, India in particular and also the traditional economies.

And there's an extremely important paper by colleagues from the bank based on their work for the World Development Report, Stuti Khemani, I recognize here. Looking at service delivery, what works, what doesn't, who benefits? Now these are very important questions for both the bank and the FAD, which should govern the design of policy making and the implementation. And these are critical not just for the bank and the FAD. They're critical for the countries. They're critical for other bilateral (off mike). And again, this is an area where I think a lot more work needs to be done, and we do have some followup issues here.

Coming to the issue of again sequencing of measures, there are a number of issues about whether you devolve -- you start the devolution process with the resources or not. If you start with resources, there's no guarantee that these additional resources may be used effectively. Again, this paper by Shantayanam Devarajan and

Stuti Khemani, and Shekhar Shah again points out. There's also an excellent paper by Bernard Barton on the issue of capture, which again prevents achieving some of the objectives that one may well otherwise have thought.

And of course, if you start with resources before you start with responsibilities, you can build up macroeconomic imbalances. And to us, that is of course a fundamental problem.

But coming back again to the sequencing and policy making, for good governments you need good public financial management. And of course, good principles of public financial management are things that the bank and the FAD are actually working on day to day in virtually all parts of the world. So, we should really build on the tools that we have at hand and we point to the importance of intertemporal(?) decision making.

Now there is an issue about whether or not one moves to performance budgeting. The performance budgeting issue comes about through the contracts. So the contract federalism literature, again Bernschpun points to the importance of this, looking at outcomes and outputs, rather than outputs. Essentially, it's a desirable direction to go, provided you have the preconditions to make it work. And the preconditions for performance budgeting, particularly at the sub-national level are actually quite severe.

It all boils down to whether or not you have consistent information, reporting, and tracking. This is really - these are really the critical public financial management building blocks for good governments and for enhanced competition, which is the basis for the new political economy literature.

The revenue assignments again, this is where the new theory perhaps differs a little bit from traditional theory. Ambrosanio and Bordignon point to the fact that if you have taxes on mobile bases, it may help in getting accountability at the sub-national level. But that's quite different from the standard prescription, which we have actually been propounding in different parts of the world.

So the question is this new recommendation valid and if so, under what conditions. And I think again here, the preponderance of the new theories would agree that administrative requirements, the requirements of a tax administration are what will dominate eventually. So if you cannot administer a VAT at the sub-national level, you shouldn't do it even though you might wish to do so on theoretical grounds.

Similarly, on the issue of revenue sharing, do you share individual taxes or do you share pools of resources? And here again, the political economy issues may differ from standard prescriptions. And what you see around the world in Australia and China is attaching - is the sharing of individual taxes, which is driven by political economy considerations.

And here again, you know, there may be differences between the old theory and the new theory, which have implications for policy making. But throughout this, you want own source revenues, local own source revenues. And this is particularly important for accountability. And again, Ambrosanio and Bordignon, an excellent paper really points to the importance of own source revenue in generating hard budget constraints and getting intertemporal fiscal discipline.

So again, I think this is one area where the old theory and the new theory come together. The importance of own source revenue at each level of government is critical in getting good and accountable outcome.

On financing, there's an excellent paper by Bordway(?), who looks at the equity and efficiencies considerations of transfer design. I strongly recommended. And then a paper which I have done with Bob Searle, which looks at the management of transfer systems.

And here again, there are some lessons. And these are perhaps not affected by the new theory or the old theory. These I would suggest are universal lessons, like you don't put in gas selling transfers because these remove the incentive for levels of government to produce anything efficiently. And therefore, if you are trying to set up transfer systems which are neutral, you do need to look at institutions. You do need to look at institutions.

You also need to be careful about special purpose transfers. There are a lot of emphasis on special purpose transfers by international agencies, by bilaterals. The problem with special purpose transfers is that you cannot always monitor the outcome. You cannot really make sure that these are used for the purposes they're designed to be used for.

And of course, they override local preferences. So if you have a system of special purpose transfers to finance local government, in a sense you are pulling the legs from under the decentralization person.

And on borrowing, of course our recommendation in the FAD has been that you have to put it in overall framework and that there are important issues in looking at the overall macroeconomic risks coming from sub-national borrowing. An important thing is to have sub-national hard budget constraints. And the seminal work done by Teresa and John Craig in her volume, which we refer to in our paper with - my paper with Maria Albino-War. So again, in this, the important standardized information is again critical.

Now, a number of emerging issues are also addressed in the volume. We have of course the issue of how do you manage natural resources, do you share natural resources. This again is something which is quite difficult. It's a central issue in Iraq. It's a problematic issue in Indonesia and in Nigeria. What are the implications? What are the policy questions? Again, Georgio Brosio and again in the transfer paper with Bob Searle would look at some of the questions arising.

The issue of decentralization and corruption, I refer to Vito's paper in the 2000 volume. Again, do look at that. It's a very nice survey by Anwar Shah and the World Bank, which looks at it in terms of what the literature says. A very balanced paper, it's well work looking at it.

Effects of decentralization and poverty reduction, Bernard Barton's paper is again addressing this issue and similarly, the bank paper by Stuti, Anwar Shah and others. The question of environmental challenges, excellent paper by Silvana Dalmazzone, again, of Turin University, which looks at certain aspects of decentralization and certain aspects of environment concerns and models how one might actually begin to look at it in different contexts. Again, something which is interesting, it's a very, very important area for research given the emphasis by my other mentor and friend, Nick Stern recently in the Environment Report.

Decentralization and Maintaining National Unity is an excellent paper by Richard Burdon(?), Bob Ebel(?) from the World Bank. Again, very readable and very relevant in the discussion for instance in Sudan or in Nigeria or in New York. Some of the issues are set out quite nicely.

And finally, the issue of political institutions, we didn't - it's largely a paper on political economy written by economists. But there is a paper by a political economist, Mr. -- Professor Gallagher from Melbourne, which again is quite nicely put and looks at the political evolution of different types of countries, federal and unitary.

Where do we go from here? I shouldn't presume to speak entirely for the FAD, but we have a number of research projects underway. We are working with the United Nations Development Program and a number of research institutes in Europe, taking further this issue of effectiveness of service delivery. And we're looking using household service data at what's been happening in Eastern Europe. Has access to local services improved a lot and under what circumstances? And again, this is an ongoing research building on some work we did - started. Sungi Cooper(?) had started looking at questions for Africa and we have a series of paper on Uganda and Ghana and Ethiopia.

We begin to ask the question what has actually happened with decentralization in the last 10 or 15 years. What has happened to access? What has happened to budgets? So putting together an eclectic use of information from different sources, we're beginning to piece together some of the constraints and challenges which we face going forward.

And that said, there's work ongoing with the UN. There's work which we hope to be doing on a case by case basis with the bank. And then there is a conference to follow up on this focusing essentially on the political economy and service delivery. Because I think that's really where the whole process is going. It's what works and what doesn't. How do you judge whether or not you're doing the right thing? The answer is really have people's lives improved or not and what are the constraints.

So those are the sort of questions I think going forward one would want to begin to do more collaborative research and then hopefully, begin to derive more general policy perfection, begin to evaluate again the advice we're giving - getting on assignments on management of the process.

So I think I'll stop here.

MS. TER-MINASSIAN: Thank you very much for this nice and clear and comprehensive overview of the main contributions of this book.

I would like now to call on Danny Leipziger. Danny, are you ready?

MR. LEIPZIGER: Sure.

MS. TER-MINASSIAN: Thank you.

Danny Leipzigers's remarks

MR. LEIPZIGER: Well, good afternoon. Thank you for asking me to show up. I think there are three reasons to do so: one, to show that Bank Fund collaboration always works; second, I never say no to any of Teresa's invitations; and third, Ehtisham actually worked at the bank once upon a time. So, there are plenty of good reasons to be here.

Also, there's some familiar faces in the audience and obviously, there were some contributions by bank staff to this book. I think it's a very valuable useful book. The only concern I have maybe is the price tag. And since Ehtisham said that the authors get nothing, there's something in the value chain that doesn't quite add up.

But that notwithstanding, I think it's a very useful guide to practitioners and policy makers. And it focuses on a number of issues that are important to the Bank.

And it presents, I think, the state of our knowledge on institutional arrangements and competition, new issues about contracting, general decentralization concerns that we have.

Having said that, I think there are some existing challenges to fiscal federalism. I think Ehtisham has referred to some of the political economy concerns. Obviously, in the Bank, the countries that we work on, we're confronted with weak local accountability, some issues of corruption, problems of local service delivery. And this all makes the job of the World Bank tougher. And I think it's useful to have this kind of handbook available.

I want to refer to a couple of the chapters in particular. I think the Bartom(?) chapter lays out some very cogent reasons why the traditional assumptions that surround fiscal federalism may not hold for developing countries. In other words, the arguments of subsidiarity usually assume that citizens have a certain voice, that there's a flow of information, that responsibilities are given at the same time as resources are given.

And all these assumptions don't necessarily prevail and therefore, there are additional challenges put upon people working in this area who realize that the local level is responsible for a lot of service delivery, which is necessary for some of the objectives that the World Bank has, but there are some impediments in that we need to work more on capacity monitoring and reform at the local level. So, I think there's a lot in that chapter.

Also, I would be remiss if I didn't refer to the chapter done by the WB Art Team(?) 2004, which does review our understanding of local service delivery and why resources alone don't always result in services being delivered.

And then there's a chapter by Ehtisham and Bob Serle, which I think also points out some of the concerns that one might have about the lack of data and the ability to track where the money goes and what the results are.

Going forward, I think Ehtisham has mentioned that there are some issues that perhaps the handbook doesn't have yet the definitive answers to. And so I thought I'd at least mention a few of these. I think one of them is that the political economy and local administrative capacity weaknesses do imped local governments from generating their own source revenues and makes them pretty dependent on the transfers. I think that's a basic issue.

I think the question of yardstick competition and whether or not that is really effective in promoting broad decentralization reforms may leave certain weak or poor governments behind. And the question is what do you do with those.

Third, I think there's the issue of unfunded mandates. It's always a concern with international institutions as well as local levels. And I think it's a quintessential problem because a lot of the problem is passed through to the local levels. So think about infrastructure for example. Much of the infrastructure investment is undertaken at local levels, but the local level budget may be dominated by the wage bill. So the national problem is translated down to the local level where it's not going to be dealt with.

I think there are issues of how to monitor and evaluate how well the centralization has gone. I think there's ample room for debate, but I think the main thing is how do you evaluate. What tools do you use?

Fifth, I think there's an issue related to local accountability of elite capture and corruption. I think the traditional view that voice and information at the local level is somehow a sufficient antidote is not always the case. And certainly, this whole area is of keen interest to the bank, and we have a new governance and anti-corruption strategy that will deal, we hope, with some of these issues.

Finally, Ehtisham mentioned the Stern Report on Global Climate Change, and I thought I would just raise this as sort of a provocative point, which is that if you think that urbanization is pretty much a irreversible trend and climate change problems are irreversible trends, there's a bit of a collision here because much of the pollution and much of the congestion and much of the problems that will cause this continuation of a climate change problem for the next 20 years is happening at decentralized local levels.

Even were the international community to be sufficiently enlightened to deal with the Stern Review kinds of issues, it's not exactly clear how they would do that, you know. Whom do you tax? And how do you put the money to the proper use? So there's an incongruity I would say between the nature of the problem, which is a global public good kind of problem and where it's actually occurring, which is a very localized problem. So, just to add that to the list of things that I think are worthy of further work.

So, to conclude, I would say that this is a very useful book. It demonstrates that people from the bank and the Fund can work together very well. And to my mind, I also want to congratulate Ehtisham, not only for putting this book out but also for showing that you don't have to be Italian to be a good fiscal economist. Thank you.

(Drop words)

MS. TER-MINASSIAN: As a chair, I should know that you have to have your speakers on.

Vito Tanzi's remarks

MR. TANZI: Thank you very much, Teresa. Maybe I should start by mentioning that around 1990 or so, Teresa talked to me about the possibility of doing some work in fiscal federalist and I said, you know, you're out of your mind. Nobody is interested in the topic, you know. Why do you want to waste your time? And fortunately, Teresa was insistent on this and you know, we started doing some work and then eventually, she came out with this very nice book, which has had quite a bit of influence. And I'm glad that she insisted and that I gave in at the time.

Well, I want to have - to structure my comment - by the way, I didn't have all that much time and since I got the book, I've had to travel continuously, so I had very little time to organize my thoughts. I read very carefully the introductory chapter. That is really a very, very nice chapter. And I skimmed through the rest of the book, so the comments will be based on this very preliminary and kind of superficial review of the book.

But I want to make - to organize my comment in three parts: you know, one, to make just a general comment - few comments on the book; second, to identify a few general points that I think are of some importance; and then, if I have time to go on some comments on specific chapters.

Well, first of all, the book is really very handsome and very important. In fact, I regret that I'm not in it, you know. And I cannot complain because I was asked but at the time I was so busy that, you know. And now I definitely regret it. It's very important.

It will for sure prove useful to scholars and hopefully to policy makers. Maybe scholars are richer than policy makers because they will be able to afford buying it. I think that every millionaire in the world should buy it, you know. And you have to be a millionaire to afford the price. I will not mention the price, otherwise Ehtisham will be embarrassed by it, you know.

But well, mainly most of the topics are included in a fair and balanced way. But perhaps if I could make just one minor marginal negative comment to the book, there is a chapter that deals with macroeconomic aspects of the - of the fiscal federalism, but I think that's fiscal policy. Fiscal policy could have received a little bit more attention, you know, especially in a world where local government may become responsible for mineral resources. And the price of this resource can fluctuate, so there are issues of this kind that could have attracted a lot of attention and they didn't.

But in any case, this is really very minor problem. And I want to complement the editors, who were also the authors of several of the chapters.

Now, let me deal with some general issues, you know, that I deal with about six or seven of these issues. First, the literature on fiscal federalism is always reflected in American bias. And this is recognized in more than one place in the book. It has been essentially influenced by American institutions and characteristics. By the way, I think that the same thing has happened to economics in general and not just to the literature on fiscal federalism.

Most people study economics in the U.S. and most people who write about economics are American or at least American adopted - adopted American like me or Teresa, or others, or Ehtisham. And we think that the world is just like that. Well, when you get out of Washington and you go to Nairobi and place like this, you realize that the world can be very different, so that the framework that you have developed with American institutions may not be terribly relevant.

I was amused in one example of this is the Tibu Hypothesis(?). I personally always thought that it was a big joke, but then I see the index of the book and there are 26 references to it, more than any other person, you know. Now, the idea that people pick up their luggage and move from one place to another every time that the taxes changes is just not, you know. You go to Italy and you know, for example, the big problem of Italy that people never move. It doesn't matter what happens, you know. And this happens in many other parts of the world.

The other thing, the United States came out of independent political - the United States was formed out of various independent political territories or jurisdictions that agreed to unite freely. A federal government was superimposed on them, so it's not a question of decentralization, was question of centralization of process, was the other way around.

For a long time, most public spending was done by the states. You know, if you look at the statistics of the U.S., until the twenties or thirties, much of the spending was at the state level. The Federal Government was responsible for foreign policy, so they paid the salaries for the ambassadors and they spent some money for defense. Besides of these two, there was very little else at the federal level.

The United States adopted a constitution that was difficult to change, so this is another interesting characteristic. It's very, very difficult to change the American Constitution. It's also created a powerful Supreme Court that had the independence and the power to interpret the Constitution in order to fix the rules and the fiscal arrangements were maintained.

The Federal Government grew out of - the Federal Government grew out of a new activity such as pension and public health. Let's not forget that until the thirties, the Federal Government did not have expenditure for pension, did not have any expenditure for health, which are really the two expenditures which have grown enormously in more recent years and have made the Federal Government appear more important than the state. But that's really a recent development.

Still, frictions continue between state and Federal Government in spite of the Constitution and the Supreme Court. This friction at one point lead to a Civil War and they have become more frequent with the passing of time. But existing institution help to deal with them, so they - you know, the institutional arrangement, the Constitution, and the Supreme Court at least provide a solution which may not be acceptable for everybody, but it's a solution to problem.

But the second point I wanted to make is that most countries under growing fiscal decentralization do not have the American historical background and institutions. Fiscal arrangement between different levels of government tend to be essentially contracts. You know, you have a contract. Like most contracts, they can not contemplate on future developments. You know, when you enter into a contract, you've made some - you call somebody to remodel your house, you enter, you make some agreement, but then a strange thing happen along the way. You cannot anticipate what will happen.

And so, in a dynamic world, in world that is changing all the time, the only certainty is that the future will be different from the present. And the only certainty is that the contract we have agreed with will become - will come under friction, under pressure with passing of time.

Arrangements made today, say on spending and taxing are not likely to be adequate tomorrow. That's a clear political process for a set of clear political rules is necessary to make the needed changes peacefully and efficiently. If this process or these rules are missing, chaos will be the inevitable outcome. Who will settle the disputes? And we've seen a example of this in many countries from Indonesia to Nigeria to other places occasionally.

Well, the two point I want to make is that the movement towards fiscal decentralization came at the same time as the movement towards privatization. You know, if you look at the timing, I mean, we started worrying about decentralization in 1990 when we started worrying about privatization.

In a paper I wrote several years ago, I argued that privatization is often an alternative to fiscal decentralization. The question is who can better deliver services to the citizen, the decentralized fiscal entities or the private sector. This is a relevant question, and the process of fiscal centralization that tend to characterize the afterward period was largely connected with - I'm sorry - the process of fiscal centralization, not decentralization, fiscal centralization, that tend to characterize afterward period was largely connecting with the growing role of the state in the economy.

If we had not had a changing role of the state in the economy, all this centralization would not have taken place. And today in the United States, the local government would still be spending much more than the central government. This process shifted the functions and responsibilities from the private sector to the government and specifically to the central government.

Privatization would do the reverse. Decentralization is an alternative clearly to privatization. So the question is which one of the two is the better route. And this probably is an aspect that also the book maybe could have - a region of the book could explore it at some point.

The choice between privatization and decentralization should have much to do with whether market failure or state failure at the centralized level is the greater problem. And it's difficult to answer. In developing countries, both don't work well. So which one works worse is an open question.

The fourth point and it's related to the both question, the choice between privatization and decentralization has little to do with equity. There is not evidence that decentralization has made the systems more equitable. So decentralization does not necessarily make a system more equitable, so the choice really would come on an efficiency ground.

The fifth point is that there is a lot of talk about tax assignments to sub-national government. And the book contains discussion of tax competition. But here again, there is an American bias. The truth is that for a variety of reasons, mostly (off mike), the taxable capacity of local governments is very limited. Thus fiscal decentralization almost always requires large transfers of revenue from national government to their sub-national governments. This brings out all sorts of problem which are discussed at length in various parts of the book.

The sixth point is that the transfers become inevitable when the national government gives mandates -- and this has been the point that Danny has made earlier - to local governments to finance given activities such as education and health. When the taxable capacity of sub-national governments is limited and it's different for different sub-national governments when you have poor and rich regions, the decision of how to restore vertical balances becomes important and difficult. Most formula have much to be desired. Political difficulties can clearly be enormous here.

The seventh point, it has been recognized that the push for decentralization is often a push for separation. And we have to recognize this. And very often when there is strong pressure on the part of an area for decentralization, say Indonesia, parts of Indonesia, you know, it's really that a certain group don't feel comfortable with the country they are with and they would really like to be different countries.

The cultural, ethnic, or religious diversity often pushes for this decentralization. When a diverse region has important natural resources and especially exportable resources, pressure for decentralization become very intense. Look at the war in Nigeria and in the past, what happened in Indonesia, et cetera. But again, does this pressure hide the wish for political divorce? They say they want to separate.

Additionally, push for centralization can be promoted by unhappiness with policies of the national government. Sometimes this is another source. When the government is doing a very bad job, well, in some region, people begin to think that we can do a better job.

Well, the last point, the general point is that public goods are not strictly municipal, regional, or national. Some extend over different sub-national governments. The bad habits are not fitting precisely the jurisdiction's limitation that will be imposed in the countries. The (off mike) are always prevalent.

Also, the uses of some natural resources, such as water, can be diverted from one region to another. This is a big issue in China, for example, today. These issues are likely to become very important to require national policies.

Now, I don't know whether I have a few minutes for some comments on the specific chapters.

MS. TER-MINASSIAN: Yes.

MR. TANZI: Just a few random comments very quickly. Chapter one by Lockwood shows that with the decentralization, there is a better matching to preferences of citizen. If this is the case, why not go all the way and create a new country? It would be even better, you know. So this is a question that really deserves to be asked. I'm highly skeptical that the centralization increases accountability, but this is an issue on which I might be wrong and obviously, depends on the country. You should not generalize.

I think that James Madison, which is mentioned in the book, was right when he warned about the power of local lobbies. Every time I go to my home town in Italy, I realize complaints about the mayor and so forth, assigning use of lands to the friends and company. I never took seriously the Tibu argument as I said, but competition does lead to more experimentation and possibly to the emergence of best models and emulation.

This is a very important aspect of decentralization which deserves to be mentioned, the possibility to allow experimentation, say in the school system, in the health system. When you have a national system, if the national system is wrong than you have a big problem. But if you have a lot of systems that compete with each other, then a new idea always will come in.

In chapter two by Pierre Salmon, the tax competition across national jurisdiction may not be too important if local taxes have benefit taxes and much of the revenue are transfers from a national grant. If there is tax competition that uses revenue from some taxes, the problem will be tax degradation more than tax losses.

This, I think, is an important point that I want to make. There is too much discussion that if you have a tax competition, the local government will lose revenue and therefore, things will not be done. Well, most of the time, this is not what will happen. They will replace bad taxes for good taxes. So, they will still keep collecting what they were collecting before but in a very bad way. Sometimes they will replace taxes with regulation, with a bad regulation. So this is an aspect which has to be recognized.

Chapter three, I fully agree with Breton emphasis on institutions and rules. The problem is that many countries are not capable of producing good rules and institution. So, to tell a country to behave and do create a (off mike) may not be terribly helpful if they are not capable of doing it. But it's important to keep emphasizing the need for rules and so forth.

Chapter five by Roger Congleton concludes that the assignment of responsibilities derives not from principles but from bargaining and competition among governments. I would add that the role of history at the beginning. I think history has played a large role in many of these arrangements in Brazil, United States, et cetera.

Chapter six by Eichenberger and Frey on the functional overlapping and competing jurisdiction, this is a classic academic proposal. Let me put it this way. The only question that I have is why not - well, the first question is why not privatize the functions. You know, if you are talking about creating an institution, say the fire department, why not privatize the activity? You know, subsidized people cannot afford to pay the fee. I find the proposal a little bit too messy for the real world. And the need of flexibility, that does not exist. You know, the idea that you can continue to reorganize an institution to fit a new need, well, it's interesting but perhaps not very practical.

Chapter seven by Paul Bernd Spahn, a substitution of rules and interactive comments by contract, contract tend to be messy too and often lead to conflicts. They definitely require a strong arbiter. So this - it's an important paper as long as you recognize that somebody must have the power to be the arbiter.

Well, chapter eight by Pranab Bardhan is an interesting paper that shows the skepticism about the benefit of decentralization when, as is often the case, institutional legal framework is weak. In this case, the version of funds is likely as Ehtisham is very often mentioned to me that maybe money that was allocated to education ends up with building swimming pools, so this is a problem that we have to recognize.

Chapter 10 on the nexus between decentralization and improve service delivery acknowledges the existence of elite capture, which is the point about medicine earlier. Elite capture is a real problem for sub-national governments.

Chapter 11 by Bernard Dafflon on expenditure assignment, do we need more than two layers, central government and municipalities? I think this is an important question, you know. In come countries, in Italy for example, we have four layers, you know. At least one of the four, nobody seems to know exactly what they do or why you need them. The other one is last of skepticism. Clearly, you need municipalities and probably clearly, you need a central government. But do we need something in between? It's not quite clear why we do need that.

Can we assign expenditures in a dynamic context in which the role of the state keeps changing with time. This is something that worries me. We know that the role of the state because of technology, because of new things about what the state should do keep changing all the time. And so, can you assign on a permanent basis responsibilities that will with time no longer be adequate?

The Chapter 12 by Ambrosanio and Bordignon, revenue assignment, the need to decide the tax assignment with the transfer system, here, there are several possibilities. The assignment can involve own taxes. You give specific taxes to the local government and of course the question whether you can give enough taxes to cover the cost. You can share the tax basis. You can share individual tax revenue, not the basis, but the revenue itself, or you can share total tax revenue. Well, these obviously are very important alternatives.

There are also dynamic elements. Once you assign taxes, the elasticity of those taxes may be very different from the elasticity of the expenditure that you are supposed to cover with those taxes. You don't know what the elasticity will be over time. Maybe the elasticity of the expenditure much higher than the elasticity of the tax. So inevitably over time, you run into problems.

And there are of course economics of scaling tax administration. You know, it's really efficient to have regular taxes in every single region or should you do it at the national level? Is it efficient to have income taxes unless you have a base which is a national base and you can share at local level?

The chapter 13 by Wilson, main issue is tax competition. In tax competition, if competition effects incentive, I would have put more emphasis. That's really where the real competition comes. It's not in the tax system itself, but in the incentive. Every region try to exploit advantage from other regions. This has happened in Argentina. It's happened in the United States and elsewhere.

In chapter 14 by Boadway on intergovernmental transfers, need for transfer to equalize the taxable capacity, but of course, the rich provinces will oppose or limit the size of transfers. You know and then it becomes a question on where is the political power. In Italy, the discussion today about decentralization and fiscal federalism is that the people from Milan don't want to subsidize the people from Calabria so that, you know, when you need to have this large transfer and you make them explicit through decentralization, well, in a unified system, they're implicit. They're not evident. But when you make explicit, you may give ground to those who oppose it. So this is some point something that we need to recognize.

And of course, (off mike) distribution if the objective of distribution is decidable, how to achieve it. Well, I'm almost through. The paper by Ahmad and Searle on implementing grant system is a very important paper. The important point is that in a dynamic world, there would be need to change size and nature of grants. But when grants are prescribed by constitution, this may require constitutional changes. But if constitution are easy to change, their rules become arbitrary. So you have a problem there. You need a strong constructional framework, but at the same time, a strong framework prevents you from making changes that you should make.

To chapter 16, on that, okay, this is - obviously, the main point here is the question of more or less the suspicion that the government will step in when you engage in debt and you have a soft budget constraint, then you have problem.

Chapter 17 by Brosio on revenue from natural resources, the issue is of who gets the revenue, how to insulate the revenue from fluctuating prices. You know, if you transfer the control of this resource at the local government and you have the price of copper that goes from 40 cents to 3 dollars and a half or whatever, what do you do? At which level you have stabilization fund? Should you have them?

Chapter 18 by Silvana Dalmazzone is very, very important. I think it's a very nice chapter and which makes some point which will become more and more important with time.

Chapter 19 by Anwar Shah, we have engaged in this discussion with him on whether decentralization leads to more or less corruption. He says less; I say more. It's a fine chapter, very fair. So it presents an alternative.

And Chapter 20 by Bird and Ebel, the centralization is a way of keeping countries together. The question inevitably is why is this important in a world of which is globalized, where the need to have a strong defense is no longer there, is it really so essential that you try to keep the country with scotch tape together by giving more power to them. I'm not sure of that, you know.

Well, Thank you very much. I think I've talked too much.

MS. TER-MINASSIAN: Thank you very much, Vito. And this is because you only spent a little bit of time reading the book. I wonder what you would have done if you had spent a lot of time.

I would like now to sort of open the floor to the audience. You can ask questions. You can make comments. You can object vehemently to things that were said. Just please identify yourself and use the microphone because we are being audio taped and otherwise, it will not be possible to get the benefit for the audience - for the future audience that are going to look at this tape of your comments and questions.

Thank you.

Question and Answer Session

MR. MILLER: Hi. I'm Vic Miller. I've worked internationally, but I never worked with states and information flow with the federal government and the states in terms of fiscal federalism. If I could just add to one thing that Vito said about the U.S. Constitution, it is very strong, but it's very simple. It was created at a time when the Federal Government had almost no roles. So when you talk about governmental roles in the United States, you're really talking about state constitutions. And you have multiple layers of Constitution. You won't see education or health in our national constitution. The role of government in serving people is at the state level Constitution.

But I have a basic question which hasn't been addressed and which we face in the United States and probably elsewhere, and that is the role of fiscal federalism and trade. Because we are continually told at the state level that trade will require us to reduce the flexibility of state governments because in the WHO or something else will require the Federal Government or the Federal Government should for efficiency purposes reduce the differences among states so that it can better interact on a trade basis.

If you could address that, I'd appreciate that.

MR. GALLAGHER: Good afternoon. I'm Mark Gallagher. I've recently been appointed chief of USAIB's worldwide fiscal reform program. I very much appreciate this book coming out at this time, and I'm anxious to buy it despite the very high price, and I will be getting the 50 percent discount.

As I was listening to the presentations, I was waiting the whole time wondering why didn't we get to this piece that was written by Byrd and Ebel, which it turns out is because it was the last piece. And I think it's something that really needs to be looked into much more, especially since I'll be working very much with the U.S. Government.

Clearly, fiscal federalism and its bizarre format was a requirement of the Dayton Accords(?) to keep the Bosnia entities together. Clearly, fiscal federalism probably should have investigated more as a means for serving as the spokes of keeping Afghanistan as a viable country. And Ambassador Peter Gallbreth(?) has been speaking lately about how Iraq is already dividing into three sub-country organizations. And it's at a different level pay scale we could say that the decisions have been taken that these countries must remain as individual countries.

So, I'd like to see if there is some further distance we can go in seeing how federalism and it's - how it could be used as the spokes to keep together a wheel of a country that might otherwise fall apart. And I don't think it's always necessarily great. I mean clearly the creature that came out of the Dayton Accords is not a good way to - it's a peace accord but it turns out to be a constitution and it's almost unbearable.

If any of you would address that, I'd appreciate it.

MS. TER-MINASSIAN: Would anybody else like to comment?

MR. AHMAD: I think the value of having Vito on the panel cannot be overemphasized. And again, Thank you very much, Vito.

Just a few points that I'd like to make. One is that most of the points you have raised, most of your authors would agree with. And I think that Tibu Model was referred to so often not so much to recommend it but more as a critique. And indeed, the political economy literature that essentially drives most of the modern literature is based on the critique of the previous literature including Tibu. So, the fact that it's mentioned 23 times or 26 times doesn't necessarily mean that it's being recommended 26 times.

Also, in terms of the background of the authors, there are three Americans out of 29 authors. And the - I would just emphasize that there's a predominance of Italians particularly in discussing the issues of revenue assignments and market failure.

And on the issue of - again, as you said, there is a U.S. bias, I really wouldn't accuse Bourguignonne of having a U.S. bias or perhaps Rivelli, but not necessarily -

MR. TANZI: I was not referring to the book. I was referring to the literature.

MR. AHMAD: The literature, yes. Certainly on the literature, of course. In the literature, there is very much of a bias and new literature is actually coming out in Europe because you're seeing the new constructs in the European Union. And the old structures which people were talking about relating to the United States are no longer relevant to this particular context.

And that's where the new literature essentially has got its impetus. It's largely European. Of course, there's a lot of literature not coming out of developing countries and the two institutions here. I'm partially to blame for that.

As I said, I think many of the points you've raised, the contrast between the United States and Germany, I mean that was very much the central point of Bernschpun's paper. When you're looking at different constitutional constructs and you look at how approaches differ and how you may wish to take different approaches when you have constitutional arrangements, which in Europe and particularly Germany are different from those in the United States.

I won't go into, you know, point by point discussion here, but I think I'll come back to the point that was made by Mr. Gallagher that it's important actually to look at what countries want in the fund and bank, I guess, respond to requests from countries. And I guess there is a presumption that the countries want to stay together, so quite often our technical assistance is looking at ways which improve the functioning of multi-tiered governments, whether it's Bolivia or Peru or it's Bosnia or Iraq.

And indeed, we are engaged with the bank and treasury and others in looking at options for Iraq. And we are doing, you know, sort of beginning the investigations. We've been looking at some of the options for a number of years. We're going back, looking at not just the intergovernmental issues, but in particular how the management of natural resources impinges on the options for decentralized operations. Decentralized operations, as you correctly pointed out, are a natural fact. They have happened.

The question is how do you keep the country together because there is a presumption that the country is being kept together. And then how do you make it work more efficiently? It's not easy. But there are issues involved in - there may be some technical solutions, which might help the political dialogue. And that's quite often really where we begin to do work at the margin.

And I would add that the work done by Amil Sunley(?) and Bob Ebel in the context of Sudan, the technical work essentially helped in getting the agreement between the north and the south. By putting some of the technical solutions on the table, you are then able to see that oh well, these things can actually work, that if there is an equitable sharing mechanism, the Kurds might actually say well, it's okay then to have central management of the petroleum resources.

It's a question of trust and it's a question of being able to demonstrate that some of the technical options are actually workable. And that's really what our day to day concerns are. And it was change of countries.

And you know, you go to Bolivia, you go to Peru, the question is there's been a lot of emphasis on decentralization, but if there's been no impact for the poor in the vast majority of municipalities, then you have a problem. And you have a problem which has a political impact.

And therefore, looking at mechanisms which actually do ensure that you get better service delivery and they may not necessarily be decentralized options, so essentially, you're looking at mechanisms. You're looking at assignments. You're looking at the assignment of responsibilities. You're looking at options for revenues. And revenue assignments at the margin do not necessarily require a local government to administer themselves.

As Vito himself mentioned, you can have a piggybacking on income tax which is administered centrally. Similarly, there may be other options up with (off mike) and so on. A number of options have been tried out in different parts of the world. What will work in each individual case depends on the case. And really, one has to start from the beginning, look at the constitutions, look at the institutional structure, look at the administrative capabilities, and then come up with something that is reasonable and workable.

So our approach in our work going forward is to look at what works and what doesn't. And that comes back to Danny's point in work that the bank has initiated, which we think is actually rather important. What works, how does it work, do you have alternatives? So those are the practical issues that we - I think that this is the work that we've been doing in the Fiscal Affairs Department and under Vito's supervision and now under Teresa's supervision is to look at sensible practical approaches to making the quality of life better, meeting the goals of particular countries, ensuring that the process doesn't result in wastage and corruption but actually gets down to the nitty gritty of improving standards of living.

So I think I'll stop there in terms of the response to you. And we'll have a long discussion bilaterally on specific points, but I think you'll find that most of the individual authors would by and large agree with what you've said with the exception of not many are Americans and many Italians.

MS. TER-MINASSIAN: Okay. Thank you, Ehtisham.

Danny?

MR. LEIPZIGER: I'll make three quick points, one on the trade issue. I mean obviously there's a price to pay for states to conform to international trade agreements, but there's also a benefit. I mean if you're a state, you're not going to be able to ensure quality standards or make sure that the electric plug that you just bought from Canada or Mexico is not going to blow up. So, there are pros and cons.

On the spoke wheel thing, I mean this is a politics issue. But it just seems to me from the fiscal federalism point of view, fiscal federalism, no matter how well designed, is not going to solve a political problem that is not solvable by economic means. So I think the issue here is to make sure that for situations that are viable, you don't create forces that will tear something apart that would normally be held intact. But obviously, if the politics were so overriding, I don't think any fiscal federalism intervention is going to solve some of the problems of the countries that you mentioned.

And just thirdly, on one of Vito's points, I think one could probably put together some reasonably good reasons why - maybe not in Italy, but in other countries - there is a rationale for something in between the municipality the nation. I mean certainly in Brazil, for example, you've got municipalities that are so small that they're just not viable. And there are a lot of scale issues as I'm sure you know.

The elite capture point that you made might be even more valid in a small area than with some checks and balances by a province or a state, whatever you want to call it. And then you have the issue, at least from the World Bank point of view, that the poor can move. So if it comes in an issue of moving the poor from one municipality to another - I mean, but there are many reasons why one could argue that there's something in between. I'm not saying it needs to be a fiscal federal structure, but administratively, there are probably some valid arguments for something in between those two levels of government.

But anyway, that's all I have to say.

MR. TANZI: Maybe just to make the point that I'm very happy to see this book. You know, it will become a reference book for many people. And I hope that it will have a lot of influence. My concern is that maybe I'm concerned that I developed while I was for a while in the Italian Government was the fact that many of the decisions that are made are not made by expert, are made by politicians who very often don't buy this book, not because of the price, but because of their own particular interests.

So, I hope that the book will have an impact on the people that are involved in making decision and will - I'm - somehow I continue to be a firm believer in the market for ideas. I think that the better ideas over time prevail over the bad ones. And I hope that this book will help the better ideas to prevail over the bad ones. Thank you.

Closing remarks by Teresa Ter-Minassian

MS. TER-MINASSIAN: Well, I won't try to sum up such a rich and interesting discussion. I agree with Vito that this book will have - or it has the potential to have a very good impact on moving forward the reflection and the policy advice - at least the policy advice that international institutions or bilateral technical assistance provides - gives to countries.

Maybe it will not have a direct immediate impact on a number of politicians that, as you said, an agenda which is not necessarily dictated by the best public interest, but forming - helping to form a public opinion and including through media, through academic circles and increasingly there is fortunately a sort of two way street between academia and policy making I think holds some promises that the reflections that have been included in this book will have a positive impact.

I think today's forum an initial step in that direction. And I want to thank you again, all of you, for participating. And well, I would not necessarily urge you to buy the book given what I've heard, but maybe we will - at least some libraries will require a sufficient number of copies to make it worth while for the publisher.

* * * * *



IMF EXTERNAL RELATIONS DEPARTMENT

Public Affairs    Media Relations
E-mail: publicaffairs@imf.org E-mail: media@imf.org
Fax: 202-623-6220 Phone: 202-623-7100