Preventing Financial Crises: The Case for Independent IMF Surveillance

Introduction

These are particularly uncertain global economic times. And my argument today-as we prepare the IMFC Deputies meeting in London at the end of this month in advance of the Spring IMFC Meeting of Finance Ministers-is that it is now urgent that we move forward in both implementing agreed reforms and shaping the next steps on both crisis prevention and crisis resolution.

The Institute of International Economics and its fellows have played a central role in this debate about reform. At the Federal Reserve and US Treasury, Ted Truman developed and delivered vital reforms; Fred Bergsten, Morris Goldstein, Michael Mussa, and John Williamson, to name just a few of the Institute's distinguished fellows, helped to set that reform agenda and continue to shape the debate.

Let me, on behalf of the UK, pay tribute to the role that the IMF has played over the past six years at the forefront of reforms to the international financial system. Michel Camdessus and Stanley Fischer shaped many of the reforms in the period after the Asian crisis of 1998. And I want, in particular, to pay tribute to the leadership of Horst Kohler, and to recognize the great contribution that IMF First Deputy Managing Director Anne Krueger has made with her work leading the debate on crisis resolution and through her strong and compelling arguments for a Sovereign Debt Restructuring Mechanism (SDRM). We must and will make progress on crisis resolution at the Spring meetings-and I will return to that subject later.

But we must also make progress on crisis prevention. And this is my principal topic today. I want to look again at the lessons we learned from the 1997-98 global crisis and the effectiveness of the reforms we introduced. I will argue that the lessons we learned from the Mexican, Asian, and Russian crises were the right lessons-and that these lessons have been reinforced by recent crises such as in Argentina and Turkey, the recent IMF surveillance review and the first report from the new Independent Evaluation Office.

Our work so far has stressed the fundamental importance of strong, credible, and transparent macroeconomic frameworks to provide the foundation for stability and growth; the need for internationally agreed codes and standards for sound policymaking that can be implemented by all countries; the rewards from greater transparency at the national and international level; and the need for a stronger focus on key vulnerabilities and risks, particularly in the financial sector.

But there is more to do. We must implement fully and deeply the ideas we had back in the late 1990s. But learning the lessons from the past-including the more recent past--demands that we go further and accept the need for a fundamental strengthening of IMF surveillance-making it more independent, authoritative, transparent, and accountable.

The international community has recognized that further reform of IMF surveillance is necessary. Last September's IMFC communiqué called for "ongoing work to ensure that surveillance in program countries reassesses economic developments and strategy from a fresh perspective." The challenge-as the UK Chancellor of the Exchequer Gordon Brown said last September-is to make the IMF as credible and independent from political influence in its surveillance of economies as an independent central bank should be in the operation of domestic monetary policy.

So today I want to set out how, in implementing this IMFC remit, the world community of nations can implement new reforms to build a new system of international economic governance and surveillance which can strengthen our ability to prevent crises and advance our shared objectives of stability, development, and prosperity.

Enhanced International Cooperation: IMF Surveillance At Its Center

Before I begin, let me focus on why crisis prevention is so important. The main case for reform does not rest on the need to protect taxpayers in the developed world. (Of course, protecting the interests of taxpayers is a vital objective for any Treasury. But lending to crisis countries is invariably repaid with interest.) The case for crisis prevention rests primarily on the need to promote social justice in the countries worst affected by crises and instability-to ensure that economic dislocation is minimized at times of instability, that private sector investors contribute responsibly to this outcome, and that governments are helped to manage their economies in a way that promotes the interests of all their citizens.

Some argue that it is, in fact, the IMF which is the obstacle to social justice, that its solutions impose the costs unfairly. I do not accept this view. We all recognize that, at times, mistakes have been made. But I believe that a reformed IMF represents one of the clearest expressions of a new internationalism-social justice at the global level. For every country, rich or poor, macroeconomic stability is not an option but an essential precondition of economic success, which is why it is important that the IMF promotes macroeconomic stability, sound and transparent institutions, and the building blocks of long-term sustainable growth as the foundation for development and prosperity. As the new Brazilian Finance Minister Antonio Palocci noted, explaining his government's commitment to credibility and responsibility in the conduct of economic policy, "Nations are not built in a hurry. They need solid foundations, laid brick by brick, of stability, clear rules, and sound institutions."

Social justice has been at the heart of our approach to debt relief and development aid for the poorest countries of the world. It is at the heart of the progress made on Highly Indebted Poor Countries and most recently Chancellor Gordon Brown's proposal for an International Finance Facility. A Facility designed to achieve both the additional finance-raising the amount the developed world spends on aid from the current $50 billion a year to $100 billion a year-and the value for money necessary to meet the internationally agreed Millennium Development Goals by 2015 so that every child has schooling; infant and maternal mortality is reduced and poverty halved. The UK Government hopes to work closely with other governments, business and nongovernmental organizations to develop this proposal further.

But just as social justice has driven our approach to debt relief and development aid, we should be reminded of the social and human cost of a failure to address the challenges of financial crises. The way forward is not to cut cooperation across the world but to strengthen that cooperation, modernizing our international rules and reforming the institutions of economic cooperation to meet the new challenges-and in doing so create a global economic system that recognizes the rights and responsibilities of all the parties involved.

This commitment to prosperity and social justice was, from the beginning, at the core of the Fund's mandate. As Gordon Brown said in his 1998 Harvard lecture, speaking of the need to rediscover the high ideals which drove the founding fathers of the Bretton Woods institutions:

The leaders of 1944 believed that global action on a new and wider stage could advance a new and worldwide public purpose of high ideals rooted in social justice: to achieve prosperity for all by each cooperating with every other with new international rules of the game that involved a commitment to high levels of growth and employment. In short, the job of every economy was to create jobs for all.

The founders of Bretton Woods resolved that the failed policies of laissez faire which resulted in vast inequities and recurring depression from the 1870s to the 1930s would not be repeated. Untrammeled, unregulated market forces had brought great instability and even greater injustice. In the post war era governments had to work collectively if they were to achieve either justice or stability.

The Challenge to IMF Surveillance and the Making of a New Consensus

Effective IMF surveillance of both global economic risks and national economic policies has been fundamental to that vision from the very start of global economic cooperation. The founders of Bretton Woods recognized that effective surveillance over national policies was critical to the international stability and cooperation on which prosperity and growth depended. The original Articles of Agreement established the importance of exchange of information, and gave the Fund authority to hold consultations with countries that had not lifted exchange restrictions. With the eventual collapse of the fixed exchange rate system and the expansion of capital markets, IMF surveillance or oversight over members' policies became the centerpiece of international economic cooperation
International economic surveillance was rightly seen an essential tool for strengthening crisis prevention and promoting global stability and growth. It offered a unique means to assess national and global economic prospects, to detect vulnerabilities at an early stage, to provide policy advice, and more broadly to promote international economic cooperation. And, following the Second Amendment of the IMF Articles of Agreement in the 1970s, the IMF was uniquely placed to undertake surveillance with its responsibility under the new Article IV to oversee the international monetary system and the economies of what is today a membership of 184 countries.

The ultimate test of whether surveillance is effective is whether it leads to better policies, both in the country concerned and in the international system more widely, so that risks and vulnerabilities are minimized and growth is more sustainable globally.

An effective surveillance mechanism needs to be authoritative, comprehensive, focused, influential and accountable. This is essential in order to ensure that surveillance detects problems at an early stage, has an appropriate impact on members' policies, and so helps to strengthen crisis prevention and promote stability and sustainable growth:

• Authoritative: It must command international respect, be a source of international expertise and best practice, and provide sound and credible policy advice;
• Comprehensive: It must cover all the issues that are relevant to economic stability;
• Focused: It needs to be well focused, highlighting the most important risks and vulnerabilities, to ensure problems are detected at an early stage;
• Influential: Its recommendations need to be reflected in the policies of member countries if problems are to be addressed at an early stage, crises averted, and uncooperative practices avoided;
• Accountable: To retain the necessary influence and legitimacy, it must provide a sound basis for its findings and recommendations, be prepared to justify and defend them, and respond to criticisms and suggestions for change.

This need for effective surveillance is even more important in a world of more global capital markets both for countries concerned and for the international system-with faster rewards for success, weaknesses punished harder and faster than before. Globalization does bring new opportunities for all economies. Greater trade and investment flows lead to lower prices and more choice, larger markets and economies of scale and faster adoption of new technology. More competition between firms and exposure to world best practice, free movement of capital generates a more efficient allocation of resources. Thus globalization can play a major role in enhancing growth and living standards.

However there are risks associated with globalization that need to be properly managed by policy makers. Globalization brings new opportunities but also can open new inequalities that need to be anticipated and addressed in developed and developing counties. And recent financial instability demonstrates the risks associated with volatile and liquid global capital markets. Opening up fully to globalization too early and too quickly, without the necessary preparation, is dangerous. Countries must put in place appropriate policies and institutional frameworks first.

So strong macroeconomic frameworks are an essential prerequisite for stability, economic growth and prosperity in a globalized world, a point that we set out in detail in the UK paper Macroeconomic Frameworks in the New Global Economy that we submitted to the November meeting of the G20. And IMF surveillance of these macroeconomic frameworks and policies is even more important today than it was when the system was set up 50 years ago.

Yet, while developments in the global economy and global capital markets have emphasized the need for effective surveillance, so they have also called into question the effectiveness of the Fund's surveillance and advice. In particular, the succession of primarily capital account crises since the mid 1990s raised concerns about the authority, comprehensiveness, focus, influence and accountability of the Fund's surveillance.

First, on its comprehensiveness and focus, a common feature of all the crises since late the 1990s is that they started from weaknesses in their domestic economic policy framework in the fiscal or financial systems often combined with unsustainable fixed exchange rate pegs upon which macroeconomic credibility was pinned. In each country-Thailand, Korea, Indonesia, Malaysia, Brazil, and Russia-large and rapid capital outflows exposed inappropriate macro-frameworks and fragile banking sectors, with massive costs for budgets and economies and ultimately their people. But in each case Fund surveillance had not ex ante stressed the weakness of domestic or financial sector frameworks or questioned the sense of linking credibility so clearly to a fixed peg.

Second, the social impact of crises on countries raised questions about the IMF's approach, its quality of advice and accountability-it was criticized for advocating rapid capital account liberalization and paying insufficient attention to the foundations for stability and growth; for applying a one-size-fits all model that was insensitive to countries' individual circumstances and needs; for advocating policies that served primarily the interests of creditors; and for being insufficiently open to outside views and advice.

And third, the frequency of crises raised questions about the influence and effectiveness of the IMF's approach. Traditionally, it had sought to exert influence through peer pressure "behind closed doors". IMF documents and discussions remained confidential, despite the fact that increasingly the rest of the international community emphasized the role that greater transparency could play in encouraging governments and markets to address risks at an early stage, without compromising candor.

By the Annual Meetings of 1998, there was a clear recognition of the enormity of the challenge we faced to restore stability and confidence in the international financial system and a need for reform. A new consensus began to take shape at those meetings, subsequently expressed in the seminal G-7 statement of October 1998 a few weeks later.

Ministers resolved to put in place a new framework for global stability, including a framework of internationally agreed codes and standards for monetary, fiscal, and financial transparency; greater transparency at the IMF itself; stronger cooperation on financial sector issues, including through the Financial Stability Forum; a new framework for crisis prevention and crisis resolution based on a partnership between public and private sectors; and a stronger focus at the IMF and World Bank on the social costs of crises.

There has been considerable progress since then. There is evidence that the risk of contagion from financial crises is less than might have been the case only a few years ago, and that investors are beginning to discriminate more between countries. The dispersion of spreads on emerging market debt has increased and the high degree of cross-country correlation of financial market returns seen at the time of the Asian crisis and Russia's default has significantly moderated.

Emerging markets are also adopting strategies to insure themselves against contagion, by, for example, building up reserves, or shifting to more flexible exchange rate arrangements. While bond spreads have risen in some countries in Latin America in response to recent difficulties, the effect has been much less marked in those countries, which are seen to have the strongest policy framework; and there has been little impact on spreads in emerging markets in Asia.

But more needs to be done to educate the private sector and ensure adequate information is available to enable efficient investment decisions to be taken. More recent events in Argentina and Turkey over the past three years, and the instability and risks that many countries still confront, means there can be no complacency.

So let me set out progress we have made and the next steps we need to consider on codes and standards, greater transparency, vulnerabilities and also the institutional changes we need to consider in order to strengthen IMF surveillance.

New Rules of the Game: Codes and Standards

First, the new system of internationally agreed codes and standards. Relevant to all countries-rich and poor-covering fiscal policy, monetary policy, banking supervision and other key aspects of public policy, they emphasize the importance of clear and sound long-term policy objectives; a precommitment to long-term stability through institutional arrangements and procedural rules; and maximum openness and transparency, providing a foundation for stability in the global economy.

As IMFC Chair Gordon Brown has said, "These new rules of the game are not incidental to the financial architecture for the new global economy: they are the financial architecture for the new global economy. This is the way in which we can deliver global financial stability in a way consistent with national sovereignty."

The codes require an effective and authoritative mechanism to monitor their implementation. To achieve this, the IMF, working with the World Bank, has adopted a program of Reports on the Observance of Codes and Standards (ROSCs). These reports summarize the extent to which countries observe the codes and are becoming an increasingly important part of the IMF surveillance process. In just 4 years, over 300 ROSCs have been completed for more than 80 countries. And of these, over 70 per cent have been published. This is a significant achievement.

And these codes are relevant for developed and developing countries alike. Last week, the Board discussed the UK's Financial Sector Assessment Program. Under the FSAP, we were assessed against a number of the codes and standards modules, including Banking Supervision, Securities Market Regulation and Insurance Regulation. We were also the first country to be assessed against the new code on anti-money laundering and combating terrorist financing. Our experience of the FSAP process was a very positive one. It highlighted good practice in several areas. But it also identified areas where more work could be done by our authorities, for example in our insurance sector. With the conclusion of the FSAP, the UK has completed 8 of the 12 codes and standards modules, and we have asked to be assessed against the remaining 4-the first country to commit to undertaking all 12.

But there are weaknesses that still need to be addressed. These include:

• First, the uneven take-up of ROSCs-over half of IMF members still have to complete a ROSC module and there are strong regional variations in the take-up. We need to find ways to give other countries greater incentives and capacity to adopt these international best practices; increasing the resources available for technical assistance will be critical to achieve this;
• Second, we need to ensure that ROSCs provide advice that is acted upon: one way to achieve this would be to include in all Article IV reports a summary of which ROSCs a country has undertaken, the extent to which the ROSC criteria were met, and the main follow-up actions; the findings of ROSCS should also be reviewed regularly;
• Third, the need to go further in strengthening the coverage of areas outside the IMF's core expertise such as the non-financial corporate sector, corporate governance and accounting and auditing. This emphasizes the importance of greater access to outside expertise and cooperation with the World Bank, the FSF, WTO, ILO and other organizations. The IMF has demonstrated that it can provide the organizing framework for the overall surveillance of codes and standards, including those in areas outside its core expertise; and
• Fourth, ROSCs should be published and we need to encourage the private sector to make greater use of ROSC assessments in their pricing of financial instruments. S&P and Fitch now explicitly use them in their rating assessments, but this is a practice that we want to see become more widespread in order to tackle contagion by giving investors the information to differentiate between good and bad policy regimes.

The codes will only be as effective as the quality of the assessment, so it is important that the effectiveness of ROSCs is regularly monitored and assessed as part of a broader assessment of the effectiveness of Fund surveillance.

This is a long-term strategy: the codes and standards challenge governments at all levels of development. We need to build up our capacity to apply international standards at the local level, to provide technical and financial support to countries undertaking difficult reforms, and to help countries use the framework to overcome weaknesses and demonstrate their strengths to the markets.

And we also need to do more to strengthen the incentives for good practice and to tackle contagion. That is why the UK supported the US proposal in 1998 for a Contingent Credit Lines (CCL)-a potentially important part of the Fund's crisis prevention armory, acting as a precautionary line of defense to be made readily available against balance of payments difficulties arising from contagion. The key insight underpinning it was the need to incentivize good policy. It confirms the principle of using IMF financing proactively to recognize good performance and help safeguard a strong policy framework in the face of turbulence in international capital markets. However, there has been no take-up-due either to issues of design or concerns about market signaling. These issues are being examined in the IMF's ongoing CCL review. To take forward this review, the Fund has consulted widely with officials from a range of countries (including potential CCL applicants) and private sector participants: we support this approach and look forward to discussing the conclusions of the review. In this ongoing debate, we must ensure that we do not lose sight of the underlying objective-the need to incentivize good domestic policy.

Greater Transparency

Second, enhancing transparency is critical to our work on strengthening crisis prevention, helping to ensure that vulnerabilities are addressed at an early stage and strengthening the accountability of the international financial institutions. Greater transparency would not only benefit countries, but strengthen public confidence in the IMF's policies and operations-making clear the basis for the Fund's findings, opening the Fund up to outside critical review, and enhancing its accountability, reputation, and credibility.

And the IMF and its members have made real progress in increasing transparency, with increasing numbers of Article IVs now published. As recently as 1998, Article IV reports could not be published. Between 1999 and last year, over 200 were published for well over 100 members. Statements by the Chairman of the Executive Board and Press Releases are issued routinely following Board discussions of programs, and last year the Board agreed to a presumption that all policy papers will be published.

The UK believes that there would be strong benefits in going further to enhance the transparency of the Fund. Having experimented with the voluntary publication of Article IVs, we believe there is now a strong case for moving to a presumption in favor of publication. We believe there is an even stronger case for IMF missions' concluding statements to be published immediately after surveillance missions, as in the UK and many other countries. And we believe there is a good case for moving to a presumption in favor of the publication of ROSCs, to increase their effectiveness further.

We need to consider carefully the argument that there is a potential tension between greater transparency and the Fund's role as a provider of candid and frank advice. It is encouraging that the Fund's recent review of transparency found no pattern suggesting that publication had led to significant disparities in the coverage of sensitive issues between published and unpublished reports. We believe that candor and transparency can be, and should be, taken forward together. Nevertheless it will be important in the future to continue to assess the impact of transparency on the candor of the Fund's advice, as part of the broader assessment of the effectiveness of Fund surveillance.

Stronger Focus on Vulnerabilities

Third, it is vital to ensure that IMF surveillance focuses on the key risks and vulnerabilities, paying more attention to the sources of growth (including trade), exchange rate regimes, medium term debt sustainability and external financing needs.

The Fund's Biennial Surveillance Review in April 2002 and the IMFC meetings identified specific areas where surveillance could be strengthened. These included: more rigorous assessments of potential vulnerabilities, with particular attention to debt sustainability and the private sector's balance sheet exposure; more candid and comprehensive assessments of exchange arrangements and exchange rates; focusing on the global impact of the policies of systemically important countries, including trade; expanding financial sector surveillance to the entire membership; and stronger coverage of relevant structural and institutional issues. Article IV reports will also now assess the actions taken by authorities in response to past and existing Fund advice.

We welcome the Fund's work on enhancing its coverage of key risks and vulnerabilities, including making better use of market information, developing a new mechanism for monitoring the risks of capital market crises, improving and applying the framework for assessing debt sustainability, and the continuing expansion of the joint Bank-Fund Financial Sector Assessment Program (FSAP).

The Problem of Surveillance in Program Countries

In each of these areas-codes and standards, transparency, and identifying vulnerabilities-progress has been made, but there is further to go. And critical to achieving our goals in each of these areas in the need to go further in strengthening the institutional arrangements for IMF surveillance.

Recent work by the Fund itself and the Independent Evaluation Office suggest that- particularly for program countries-there is a case for further reform to bring greater independence and objectivity to the surveillance process-what First Deputy Managing Director Anne Krueger has called "the fresh pair of eyes issue."

To achieve greater objectivity, the Biennial Surveillance Review drew the conclusion that surveillance of countries with active Fund programs should bring a "fresh perspective" to the Fund's assessment of the country's policies and prospects.

This requires a stepping back from the program framework. There is wide acceptance that Article IV consultation discussions sometimes failed to do this, thereby limiting the potential effectiveness of surveillance.

The IMF has subsequently agreed a new set of surveillance guidelines, covering surveillance in program countries. This should entail (i) a comprehensive assessment of economic developments, not narrowly focused on program targets; (ii) a candid analysis of the short- and medium-term outlook, including a thorough discussion of risks and vulnerabilities; (iii) a stock-taking of the policy strategy and the effectiveness of the measures taken; and (iv) a candid account of the dialogue between staff and authorities on key policy issues and on the program strategy.

We welcome these changes-but they do not go far enough. We believe it is essential to work up further options for strengthening surveillance in program countries, by providing not just periodic Fund reassessments but regular and independent reassessments of economic developments, prospects, and policies and of the program framework.

The importance of achieving greater objectivity and independence was also highlighted by the Independent Evaluation Office's first report on prolonged use of IMF resources-countries which have had continuous Fund programs for many years. The report noted that its case studies suggested that "by and large, surveillance failed to play a major independent role in prolonged user cases." It recommended that steps should be taken to strengthen further surveillance in prolonged user cases, going beyond the revisions to surveillance guidelines, including:

• clarifying the role of surveillance in program cases;
• greater operational separation between Fund program and surveillance activities in prolonged use cases, although this raises delicate trade-offs; and
• seeking a second opinion-including from outside the IMF-on key policy issues that appear to be contributing to prolonged use.

The Independent Evaluation Office report also found that "there is evidence that internal incentives in the IMF encourage overpromising programs. This results from both the relatively short time frame of programs, forcing optimistic assumptions about the pace of adjustment and also from a desire to maximize the program's catalytic role. This led to a tendency to downplay risks. Even when, as was often the case, they were well identified during the internal review process, the assessment of risks was not candidly presented to the Executive Board."

The Case For More Fundamental Institutional Reform

These cases and reports highlight, in our view, a deeper structural problem in the current surveillance process. The current structure of the IMF treats program design as an extension of surveillance. Yet there is evidence that the lack of a clear distinction between lending and surveillance activities creates the wrong incentives and diminishes the effectiveness of surveillance.

Working to reach agreement on a program and to restore confidence, there may be incentives and pressures on the Fund to be overoptimistic in its surveillance of the risks. Once it has agreed the program and disbursed the loan, the Fund as the subsequent monitor of performance may face pressures and incentives that prevent it from stepping back to provide a candid assessment of the sustainability of the program framework. And the blurring of the surveillance and program mandates makes it difficult for the Fund to be held accountable over a period of time for the quality and effectiveness of either its surveillance or its program design.

Moreover, there is currently no formal regular mechanism for assessing whether the Fund is providing objective, rigorous, and consistent standards of surveillance across all member countries-program and nonprogram countries. This highlights an additional structural problem.

In principle, the Executive Board discussions of Article IV reports should provide this perspective through peer review. But again there is a blurred division of responsibilities. Executive Directors are responsible for both lending and surveillance decisions. In addition, while responsible for ensuring the effectiveness of the Fund's activities, Executive Directors have responsibilities to their authorities. The External Evaluation of IMF surveillance (1999) noted that Fund staff observed that Executive Directors tended to be defensive about the countries they represent, and that other Directors deferred to this, partly because they expected the same deference in return in due course. Peer pressure can, in short, become peer protection.

But even if such considerations never affect Board decisions, the case for change remains. There is a clear parallel with central bank independence. The case for an independent central bank is that delegation of responsibility for setting interest rates to an independent body can solve the time inconsistency problem whereby governments are tempted to exploit the short-run output-inflation trade-off for short-run gain. It is not that Finance Ministers are necessarily badly motivated. It is the suspicion that they might be that damages credibility. Similarly the case for moving to an independent surveillance regime is that it would remove the suspicion that surveillance might not be objective and candid.

Some might argue that this parallel does not work-that decisions need to be taken by the IMF Board to give proper legitimacy, while the Board must have the discretion to respond to changing economic events and circumstances without being boxed in by over-rigid rules or procedures.

But the modern case for central bank independence also recognizes that elected Finance Ministers have a proper role underpinning the democratic legitimacy of the central bank-hence in the British model of central bank independence the Chancellor of the Exchequer both sets the inflation target for the Bank of England to pursue and appoints the members of the Monetary Policy Committee.

And the modern case for central bank independence also recognizes the need to go beyond the old debate between rules versus discretion and recognize that a robust framework for policy that maintains stability must also be able to adapt appropriately to shocks.

In a world of fast moving capital, long-term stability requires an overall framework that constrains macroeconomic policy to achieve clear long-term and sustainable goals, but which gives discretion to respond flexibly to shocks. If policymakers have a sufficiently credible commitment to long-term stability, then they will be able to exercise discretion in response to shocks without damaging long-term expectations.

The new British model of central bank independence is neither based on complete discretion, nor fixed rules, but on "constrained discretion". It recognizes that the discretion necessary for effective economic policy-short-term flexibility to meet credible long-term goals--is possible only within an institutional framework that commands market credibility and public trust with the government constrained to deliver clearly defined long-term policy objectives and maximum openness and transparency.

So the new British model of central bank independence has five key features:

• a strategic division of responsibilities: with the elected government setting the wider economic strategy and the objectives for monetary policy, while monthly decisions are passed over to the central bank;
• a single symmetric inflation target: with no ambiguity or dual targeting;
• independent expert decisions: with monthly decisions to meet the government's inflation target taken by an independent Monetary Policy Committee;
• built-in flexibility: with the Open Letter system to allow the necessary flexibility so that policy can respond in the short-term to surprise economic events without jeopardizing long-term goals;
• maximum transparency and accountability: with monthly minutes published and individual vote attributed and with a strengthened role for Parliament.

Central to this framework is the separation between the government setting the objectives for monetary policy in terms of the inflation target whilst monthly decisions on interest rates are passed over to the central bank. This removes the suspicion that policy was being manipulated for short-term motives. As Deputy Governor Mervyn King said in his 1999 Belfast lecture, "the rationale for handing operational responsibility for setting interest rates to the MPC is that it is better qualified to make those decisions than elected politicians, whereas elected politicians have the democratic legitimacy to choose the target."

I believe we can and should apply this "constrained discretion" approach to global economic management and crisis prevention and resolution.

Just as in domestic economic policymaking, so in international economic policymaking this means breaking from the sterile debate of the 1990s between rules and discretion-the false choice between fixed and rigid rules on the one hand and a completely ad hoc approach on the other.

Crisis prevention or resolution cannot be based on rigid policy. There must be discretion for the Fund Board to act as circumstances dictate.

But credibility, predictability and tackling moral hazard demand that this discretion needs to be constrained by much clearer procedural rules. And that means that, in addition to reforms to crisis resolution currently being discussed and the new rules recently agreed on access to financing from the Fund, learning the lessons from the past, including the more recent past, demands a fundamental strengthening of IMF surveillance to make it more independent, authoritative, transparent and accountable.

As in domestic monetary policymaking so at the international level, IMF surveillance must be-and be seen to be-credible and objective and that means it must operate within an institutional framework that provides a credible guarantee of candor.

The case for making the IMF as independent from political influence in its surveillance of economies as an independent central bank is in the operation of monetary policy is that credibility comes from demonstrating that surveillance is-as a matter of institutional design and procedural rules-impartial rather than influenced by political considerations be they from shareholders, directly from or through the Board or from internal pressures.

I stress that the case for greater independence is not an attempt to fetter the discretion of the Board to make proper program lending decisions more than is sensible to properly guide expectations and tackle moral hazard. The objective is instead to make sure that Board decisions are made-and seen to be made-on the basis of open and candid surveillance about sustainability and risks, and to protect both the Board and IMF management from being dragged into decisions, which-on the basis of objective evidence-they would not want to take or publicly justify.

So we believe there is a strong case for institutional reform. Our objectives must be:

• First to ensure that surveillance is, and is seen to be, independent of decisions about program lending and the use of IMF resources. We should consider the full range of options.
• Second, to establish the IMF as independent from political influence in its surveillance of economies as an independent central bank is in the operation of monetary policy.

Enhanced transparency, both of procedures and substance, is essential to boosting credibility. So there is a case for publishing staff reports at the same time they go to the Board.

But transparency must be matched with clearer procedures and a clear separation between surveillance and program lending. So we need to re-examine and clarify the objectives, procedures and responsibilities of the Executive Board and IMF Management.

How this can be achieved should be a matter for debate and not something to be dictated. Indeed we need a wide-ranging discussion of the best ways to achieve these objectives. The IMF Board paper last year canvassed opinion on various options to achieve greater independence of surveillance in program countries, from clearer guidance by the Board on the role of surveillance to an "institutional firewall" within the IMF-separating program and surveillance activities, including consultation missions headed by a mission chief from a department different from the area department. On the broader question of how to deepen accountability, one way forward would be for the IMF Board or even the IMFC to have a formal responsibility to set an annual surveillance remit, with the IMF staff and management reporting back regularly on their performance against that remit.

For the long-term we also need to consider more radical institutional reforms to achieve the two objectives I have set out. And, while this is not the time for detailed proposals, we can be clear about our long term goal-a surveillance framework would deliver the independence, accountability, authority, and legitimacy on which effective surveillance and international economic cooperation depends.

Lessons from Argentina

I want to end by showing how these arguments apply in a real world example-the recent case of the collapse of the Argentinean currency board over the course of 2001. The scale of the economic and financial crisis that Argentina has experienced over the past few years and the impact of the crisis on the poorest and most vulnerable in Argentina, has again highlighted the very real human impact of crises, and why prevention is better than cure.

The causes of the Argentine crisis have been well discussed, here at the institute and in the recent IMF Article IV. Loose fiscal policy, combined with a rigid, fixed exchange rate peg and a lack of flexibility elsewhere in the economy produced a series of ultimately irresistible pressures. A clear demonstration of the need for credible macroeconomic frameworks.

The experience of Argentina shows how, in a world of mobile capital and open markets, unsustainable policy regimes quickly become unstuck. But the experience of Argentina also clearly demonstrates the case for a further strengthening of IMF surveillance.

Exiting from the convertibility regime was always going to be disruptive and entail a high cost. Unsurprisingly, successive Argentine governments were unwilling to consider moving to alternative exchange rate regimes, given the centrality of convertibility to the policy framework. But we need to examine the effectiveness of Fund surveillance and ask whether, in hindsight, it should have taken a firmer position on policies critical to sustainability.

Indeed, the IEO's work program for the coming year has already recognized the crisis in Argentina "has raised a number of questions about the effectiveness of the IMF's crisis prevention efforts and the quality and impact of its policy advice" during a decade in which "Argentina was continuously engaged in IMF-supported programs for most of the period since the adoption of the Convertibility Law in 1991 and was often presented as a success story."

It will be important that the IEO focuses on the quality, timeliness and impact of the IMF's policy advice on the currency board and 'exit' strategies, fiscal policy and debt sustainability and on the structural policies that were needed to maintain the currency board arrangement.

Had surveillance reports more effectively highlighted Argentina's problems at an earlier stage the decisions that the Argentine authorities would have faced would not have been any easier. But by forcing the Argentine government and the IMF to face up to them earlier, it might have prevented a crisis on the scale that subsequently emerged.

Nor, with the public availability of more transparent and credible independent surveillance, would there have been such a one-sided public debate throughout the spring and summer of 2001.

This again highlights the point that, while the Board must retain the discretion to make its own program decisions, the availability of transparency and objective surveillance can strengthen the hand of both IMF management and the Board in the face of wider pressure.

Argentina is not the only recent example of a crisis exposing weaknesses in ex ante surveillance. The recent case of Turkey, for example, also demonstrates some of the dangers that can arise when surveillance fails to pick up key risks. As we all know, Turkey's December 1999 disinflation program ended in crisis when its fixed exchange rate regime collapsed in February 2001, with huge impacts on the balance sheets of a range of banks, which had taken unhedged foreign currency positions. Pre-program surveillance did actually identify some risks in the financial sector. But the link with the proposed exchange rate regime was not adequately made-despite the recent experience of bank losses under a fixed exchange rate regime in the Asia crisis, and despite a World Bank team being in the country to do work on reform of the financial sector.

I am not suggesting that in an alternative world with independent surveillance, these crises would necessarily have been prevented or that difficult decisions for both domestic governments and the Board would have been any easier. But these decisions would have been taken on the basis of a clearer and more transparent understanding of the underlying position-and possibly at least limiting the scale and severity of the crises.

Conclusion

The recent experience of Argentina and Turkey are an important reminder that, however much we try to strengthen our procedures and surveillance, crises cannot be prevented from happening. They are a fact of life in a turbulent world. This is why better crisis resolution mechanisms are important.

Any of us who have lived through the series of emerging market crises over the past decade can recognize that radical improvement in the institutional architecture for crisis resolution is necessary to move beyond the current unsatisfactory and "ad hoc" approach to crisis resolution in order to reduce moral hazard and bring greater certainty to creditors and debtors. Reform can benefit creditors-by allowing the retention of greater value than is currently the case-and debtors-by speeding up the process and reducing the needed level of domestic adjustment during the crisis.

There is probably no single method of achieving our goals on crisis resolution. That is why in September last year, Ministers at the IMFC called for pursuit of a twin track approach to crisis resolution. One track involved encouraging the development and inclusion of Collective Action Clauses in international sovereign bonds to strengthen creditor co-ordination. Significant progress has been made in devising model clauses, with Mexico last week incorporating CACs into a bond issue for the first time which we hope will prove a model for other emerging market issuers. The other track was the development of a 'concrete proposal' for a Sovereign Debt Restructuring Mechanism by the IMF to ease orderly debt restructuring. We strongly welcome the progress the Fund has made towards the development of such a proposal and are keenly looking forward to seeing a concrete proposal in advance of the IMFC meeting next month. As the Ministers themselves noted last year, these approaches are truly complementary.

The crisis resolution debate is challenging, raising important and difficult questions as well as strong feelings, and will be a central issue at the IMFC meeting in April. No new procedures will be operational quickly: the legal processes to implement an SDRM will take time; CACs will only bring their full benefits as the debt stock turns over fully. And a code could address a broader range of issues than SDRM, including crisis prevention. In this context, there is a need to clarify the rights and responsibilities of debtors and creditors, not only in times of crisis. That is why private sector and official engagement on a code of good conduct for debtors and creditors is to be encouraged.

And-of course-we must recognize that being clearer ourselves about what the international community can-and cannot-do in a crisis will help to shape the incentives that will in turn make crises less likely. We can do much more to make them less likely, to create the right incentives for the countries themselves and the international financial institutions to take action before a problem becomes a crisis.

That is why we see merit in the case for discussion on crisis resolution to be accompanied-as the private sector make clear in their draft code of conduct-with more effective steps to strengthen crisis prevention. Because-as this draft code also recognizes-the need to send clear signals to avoid moral hazard is only part of the steps we have collectively identified to prevent crises happening in the first place.

So-going with the grain of the reforms of recent years-there is a case for further institutional change to strengthen crisis prevention. We need to ensure surveillance is dispassionate and independent, and not colored by time pressures or program responsibilities. One approach is to provide a fresh pair of eyes for surveillance in key countries with active programs. But, in the longer-term, it might be worth reviewing more fundamentally the institutional architecture of the surveillance and program functions of the Fund and Bank.

People often say that the key institutional change we need to consider is a merger of the IMF and World Bank-because of the growing role of the IMF in development issues and increasing overlap. I am not going to comment on this today other than to say that it has been clear in recent years that the IMF and World Bank have complementary but separate roles in the development agenda and are working increasingly well together.

My argument is that the key institutional issue is not a merger but a greater separation between the surveillance and lending functions-an argument relevant to the Fund and Bank alike.

Experience has shown that it is not possible to prevent all financial crises, but prevention is both better and less costly than picking up the pieces afterwards. Making its surveillance more independent would be a significant step the Fund and its members could take to improve the chances of detecting problems early, and encouraging countries to take the actions needed to prevent financial crisis.

The Fund has complimented the UK for its bold decision in making the Bank of England independent in 1997. Now it is time for the Fund and its members to show that they are not afraid to consider institutional change for the Fund itself to help it play its full role in delivering the stability upon which global prosperity increasingly depends.