Speech

De Tocqueville's "Dangerous Moment": The Importance Of Getting Reforms Right, Remarks by Anne O. Krueger, First Deputy Managing Director, IMF

September 10, 2004

Introduction

Thank you for that kind introduction, David. I am delighted to be here, especially when I see quite a number of familiar faces in the audience.

I want to start by explaining that reference to De Tocqueville. I hope both you and Alexis de Tocqueville himself will indulge my having taken his phrase a little out of context. De Tocqueville, as many of you probably know, was actually arguing that embarking on a reform process could prove a dangerous moment for bad governments—the implication being that this could prove their undoing.

My theme this evening is somewhat different. I still want to argue that when governments introduce reform is a dangerous moment. But my focus is on the difficulties that all governments inevitably encounter in judging what reforms to introduce; when; and how fast. Opportunities for reform are infrequent and if critical efforts go wrong, reforms get discredited. Once that happens it can be difficult to get another chance to introduce reform. And successive reforms failures make each subsequent effort that much more difficult—and more costly.

So I want to examine what we have learned about the way economic policy reform can most successfully be implemented. I want to look at what we mean by economic policy reform: I would argue that successful economic reform is much more wide-ranging than we realized until relatively recently. And I want to suggest some ways, based on past experience, in which governments and policymakers can best cope with the problems of uncertainty that invariably accompany the reform process.

The impetus for reform

Past policy failures provide the motive for reform. In the postwar era that saw the creation of the multilateral economic framework—including the IMF—many of what we now call the emerging market countries pursued dirigiste policies. Import substitution was actively encouraged. State owned and managed enterprises were given a central role in economic activity. Government spending was high, and high levels of taxation and regulation discouraged entrepreneurial activity. Some of these policies were pursued in many industrial countries too.

The fundamental weaknesses of these policies gradually became apparent. Economic growth slowed or stagnated; or a crisis erupted. Living standards rose only slowly if at all. Poverty increased. Private sector enterprises had little or no incentive to compete in overseas markets and sheltered domestic industries—often monopolies—could remain high-cost and turn out low-quality products. Economies that pursued such policies experienced slowing economic growth or stagnation over a long period, often punctuated by balance of payments crises and "foreign exchange shortages". This often provided the political impetus for reform.

In some cases, policy failures were so acute that they precipitated major economic or financial crises. This was the case in India in 1966 and 1991; in Turkey in 1980; in Mexico in 1982. In these and other cases, the old policy stance was not consistent with continuing economic growth.

Defining reform

As the need for change became widely accepted, reform became part of every politician's vocabulary. No political program seems complete without a plan for economic reform. Economic success has long been recognized as making a crucial contribution to the political success of democratic governments. The well-being of ordinary citizens, measured by rising real wages, full employment and price stability, is seen as an important factor in determining electoral winners and losers. The number of democratic states has increased markedly in the past twenty or thirty years, of course. And even in those states with authoritarian governments, there seems to be a greater desire for economic growth, and the rising living standards and falling poverty that such growth brings.

So the need for reform has become almost a mantra for many politicians and policymakers. To be successful, or to remain so, economies need to be constantly adapting. Economic policy needs to enable economies to become flexible so as to adapt to change in consumer tastes and demand as incomes rise; in technology; and in the external environment.

But what do we—or more precisely, those policymakers—mean by reform? Not every small change can deliver the sort of welfare improvements that policymakers say they want. Not every change in the direction of economic policy will have a beneficial impact on economic performance. I would define reform as a measure, or set of measures, that are expected to result in a significantly improved framework for economic activity because of the way these measures alter behavioral incentives and thus improve the alignment between private behavior and economic welfare. These changes can be direct-tariff reductions or the elimination of quantitative restrictions on imports. They can be institutional: such as the introduction of a proper commercial code, or the establishment of an effective financial regulatory framework. Good, well-planned reforms will involve both types of change.

And reforms deliver the best results when they are complementary. A series of changes is likely to bring far greater returns than a one-off shift in policy direction. Trade liberalization and a move to a floating exchange rate regime will together have more impact than either one carried out in isolation. And the two can bring greater benefits when accompanied by reforms that increase labor market flexibility. The whole really is greater than the sum of its parts.

But reforms haven't always brought the improvements anticipated—sometimes quite the reverse. In the past two decades or so we have seen a number of countries pursue wholesale economic policy reforms that have ultimately proved disappointing—because of weaknesses in design or implementation; or because of political opposition that led to their reversal before benefits could be realized. The consequence was often slow or non-existent growth, high and often rising inflation, debt burdens that soared out of control—often cumulating in a major financial crisis, and yet another reform effort.

Some reforms are, at best, misguided. When duties on intermediate goods are removed or lowered, but those on finished goods left untouched, effective protection increases and makes later reform more difficult.

Or take another example. When duties are removed but the exchange rate is left unchanged, the demand for imports will rise but there will be no offsetting pressure to increase exports. Hence, economic activity in import-competing sectors will fall, but that in exportables would remain largely unchanged, thus raising unemployment in most cases.

Context and credibility

Crises force significant policy reforms on a government. Hence my reference to a dangerous moment. In a crisis, there is little enough time to act, let alone think. The exact nature of the crisis will determine what reforms are needed and in what order. In the Asian crisis countries in 1997, a key factor in the crisis was chronic weaknesses in the banking sectors of many countries. Reform of the financial sector subsequently became a high priority in those countries, although some made more progress than others.

But the immediate need was to halt the drain on foreign exchange reserves that resulted from huge capital outflows in a fixed rate regime. Korean reserves had been virtually exhausted in 1997, for example. The IMF-supported programs in the Asian countries were intended to stabilize the situation as quickly as possible. Only then could the governments involved start to implement the longer-term reforms needed to prevent a repetition of the crises.

Other types of crisis require similarly rapid responses. Foreign exchange crises, usually under a fixed rate regime, force governments to take unpalatable decisions rapidly. Britain in 1967 and 1992, Mexico's tequila crisis in 1994, Argentina in 2001—all involved significant devaluations. But devaluation by itself does not usually end the crisis. Confidence will only be restored by tackling the underlying economic problems that led to the crisis; and tackling them in a credible way. The debt crisis in Mexico in 1982, when the government announced it could no longer service its debt obligations, was also a symptom of economic weaknesses that went well beyond the problem of the sovereign debts to commercial banks that it could not repay.

Loss of confidence in a government's ability to deliver economic growth and prosperity can precipitate a different sort of crisis situation. This is arguably what happened in 1984 in New Zealand when the Labor government came to power aiming to reverse decades of relative economic decline or when Margaret Thatcher won the 1979 election in the U.K.; or in France when "cohabitation" was forced on President Mitterrand in 1986 after the first years of his socialist government in France.

Governments acting in a crisis situation appear to have a better chance of implementing reforms: that was certainly true of the Callaghan government here in the UK when it introduced reforms as part of a Fund-supported program in 1976. The sense of crisis made it more difficult for the opponents of the reform program to marshal support. Similarly, the Indian reforms implemented in response to the crisis in 1991; those introduced in Mexico in 1994 and in Asia in 1997-98. It is easier for governments to marshal support for reforms in a crisis, in part because there are fewer obvious alternatives.

In fact, crisis seems sometimes to lead to a suspension of "politics as usual" and provides a government with considerable freedom—more than is usual in politics—to undertake reforms.

New governments may enjoy something of an advantage, especially those in democracies that enter office with a mandate for change.

But reforms are painful (or they would have been undertaken already!) and the deeper the reforms, the greater the opposition they are likely to arouse. So politicians have to weigh the magnitude of reform (the size of tariffs cuts, the magnitude of devaluation, the cut in government expenditures) and its projected benefits against the likely opposition they will encounter.

Building credibility for a reform program is essential. The economic actors, whose behavior the reforms are intended to influence, will respond to the altered incentive structure more rapidly if they are persuaded by the speed with which a program of reforms is introduced, by the comprehensiveness of the program, and if they are convinced that the reforms will be adhered to. International credibility is important too. Access to the international capital markets will be easier. And support from institutions like the IMF always depends on a government being able to demonstrate credible commitment to a reform plan.

But the more opposition there is and hence the greater likelihood that reforms will be reversed, the more slowly will economic actors respond to changed incentives.

And the global economic environment in which reforms are introduced will have a crucial impact on their success. Unfortunately, in many cases crises erupt when economic vulnerabilities become apparent in a global downturn. In such cases, governments are not only pressed for time: they also have little room for maneuver.

Uncertainty

Crisis by definition is a dangerous moment and requires a response. But when profound changes in the structure of an economy are desperately needed if long-term prospects are to improve—as has been true in most emerging market economies as the shortcomings of earlier policies were recognized—then considerable uncertainty surrounds the outcome.

The authorities do not know how much opposition there will be to reforms. They do not know how much change it will take to convince economic agents that the new policies are firmly in place and that there is no point on betting against their reversal. Nor do the authorities know how quickly speculative pressures will subside, how some economic actors will respond to the realigned incentive structure and how quickly.

There is equal uncertainty for economic decision-makers. They do not know how likely it is that reforms will stick. They cannot judge, in many instances, how reforms will affect them, their competitors, and the economic environment in which they operate. Economic actors often decide to fight against reforms rather than adjusting to the new incentive structure. They often delay their response until their likely impact becomes clearer.

We can make judgments about the likely economic response when we do certain things, based on what we know about how economies work, and on past experience. But though we can often predict the direction of change, when we alter incentives in an economy we cannot predict the timing or degree of the response with any great certainty.

In 1966, in the midst of a foreign exchange crisis, India devalued the rupee by a little over 35% but exports did not appear to respond. This was largely because the removal of export subsidies at the same time as the devaluation had, in effect, acted as a disguised appreciation of approximately the same magnitude as the devaluation, so the rupee price of exports had not increased significantly, and in some cases had fallen. To make matters worse, a bad harvest in a then predominantly agricultural economy led to a drop in real GDP. The outcome was a political backlash which gave reform a bad name and resulted in a fifteen year period before reforms could be tried again.

It is clear when we start trying to put some numbers on these variables—the magnitude of a devaluation, the scale and speed of tariff reduction, the timeline for reducing the fiscal deficit be reduced—that the challenge of getting reforms right becomes truly formidable. There are no guarantees, but there are ways in which the risks inherent in the reform process can be reduced, though not eliminated.

But there are things we do know. Without reforms things will get worse and, while we may not know exactly how a program of trade liberalization will affect a particular economy, there is plenty of evidence across countries and over time to support the view that opening up economies to trade is beneficial. The competition it brings is a powerful force for increased economic efficiency because it helps ensure that resources are allocated in the best possible way; it helps eliminate domestic monopolies; it drives down prices both for domestic consumers—as well as producers in import-consuming industries—and in the international marketplace.

And trade liberalization generally leads to higher growth rates than would otherwise have been the case. Warcziarg and Welch have shown that of 133 countries they surveyed between 1950 and 1988 those that liberalized their trade regimes enjoyed annual growth rates of about one half of one percentage point higher after liberalization. And opening up to international trade seems to have become increasingly important: the same study showed that removal of trade barriers during the 1990s raised growth rates by an estimated 2.5 percentage points a year.

Studies of trade reforms in individual countries also confirm the magnitude of gains that can be had. Korea's sustained focus on trade liberalization as part of the reforms introduced from the late 1950s onwards played a major role in that country's spectacular growth performance. Over a long period Chile also reduced its tariffs and opened up its economy. In 1974, tariffs often exceeded 100%; the World Trade Organization estimates the simple average tariff for 2003 to have been 5.9%. The Chilean has experienced sustained growth over a long period.

Similarly with labor market reform. Rigid labor market regulation stifles employment growth in the formal sector while increasing the size of the informal sector. We know that the less rigid and more flexible labor markets are, the better economies perform and the higher growth rates will be.

We may not be able to predict how the labor market in any one economy will respond to changes in labor laws—how rapidly unemployment will fall; how employment will rise and/or shift to the formal sector; and the impact of the changes on overall economic performance. But we have plenty of evidence to support the argument that such changes will result from greater flexibility.

There is always a fortuitous element in reform programs. This is partly because of the domestic context in which they take place: a good or bad harvest can make a big difference to the extent of opposition to difficult reforms. It will be easier to tighten fiscal policy in more buoyant domestic conditions.

The global economic environment is also important. It is much easier to introduce large-scale reform when the global outlook is more benign. Trade liberalization is likely to bring economic benefits more rapidly at a time of global economic expansion, when trade is growing. As a general rule, governments are wise to seize the opportunity to press ahead with reforms in the context of growth—one reason why we at the IMF have been urging emerging market economies to reduce their public debt burdens now, during the current upturn. But there the sense of urgency that provides the main impetus for reform is likely to be lacking when the global outlook is brighter and politicians will be understandably—but mistakenly—tempted to postpone uncomfortable decisions.

When reforms are introduced, specific measures might work more quickly or more slowly than anticipated; and that could in turn affect the ability of policymakers to sustain support for further reforms.

For politicians there is a trade-off—between implementing tough reforms that will bring considerable benefits down the road and confronting opposition to unpopular measures. There will always be a temptation to do the minimum necessary. But that inevitably increases the risk that the reforms will not deliver the hoped-for results, and in some cases will fail, only to be followed by another crisis and another reform package—and that in turn will meet yet greater opposition.

The stronger a reform program—in terms of its scope and the speed with which it is implemented—the greater will be its chances of success. And the more signs of success there are, the more the government will have to build on and the better it will be able to cope with opposition to change.

Winners and losers

Building support for change is an important element in ensuring a reform program can be implemented successfully. Ideally, one would be able to identify the winners and losers of any reform process; make sure the winners realized who they were; and then seek ways of countering the opposition.

But it is usually difficult to identify winners and losers with any great certainty. Some short-term losers might recognize that they are vulnerable—protected domestic industries facing the prospect of greater foreign competition might have good reason to worry. But not always. One of my favorite stories concerns Mexican refrigerators. A major refrigerator manufacturer in Mexico was strongly opposed to NAFTA because he reckoned that opening up trade would enable American manufacturers to make significant inroads into the Mexican market.

This turned out to be a misjudgment. A big weakness with Mexican-made fridges was the poor quality of compressors previously available to the fridge-makers. NAFTA made it possible to import and use higher quality US-made compressors in Mexican fridges and so enabled the manufacturer in question to become a leading player in the American refrigerator market for smaller models.

In the short-term, though, predicting such outcomes is difficult and beneficiaries do not recognize their prospective gains. It takes time for the full effect of reform measures to work through the economy. Removing subsidies and moving to a flexible exchange rate regime will help export-oriented firms increase their overseas markets, for example: but it is unlikely to do so overnight.

And in the meantime, people are uncertain and afraid. You cannot know that you will be hired by an expanding exporting firm: but you can guess that your job—or your profits—are threatened by greater import competition.

Better understanding of the reform process can dampen some opposition. The removal of agricultural export subsidies is often opposed without recognition that exchange rate change can confer bigger benefits.

Winners will become evident as a reform program progresses; and then they can be enlisted to support further reforms. In the short-term, though, it is not possible to identify the firms and workers who will start up or expand businesses as reforms take hold. Ending protection and freeing up trade makes it easier for new firms to compete successfully—but the identity of those firms and those workers who get the jobs they create will only become apparent later.

Losers are similarly difficult to identify in advance, but their identity might become apparent more quickly as firms lay off workers, or go out of business, in response to more competitive pressures.

It is clearly important that reforms that are introduced do not create vested interests that might then oppose future reforms. Establishing preferential trade agreements that create trade diversion can make further trade liberalization more difficult. A mis-guided regulatory framework could also make it difficult to deepen financial markets at a later stage.

Comprehensiveness

The main focus here has been on trade and labor market reforms—in deference to the audience here this evening. But similar considerations apply in many other areas. A successful reform program means tackling problems on a wide front.

Major areas of reform include governance issues and the establishment of secure property rights and predictable legal systems; better government expenditure management; the move to less distortionary tax systems; pension system reform; and the privatization of state owned enterprises, among others.

Commitment and follow-through

I've talked about the difficulty of framing a reform program because of the huge uncertainties involved; and I've mentioned some of the ways in which those uncertainties about the outcome can be reduced.

The importance of commitment and perseverance cannot be overstated. Too many reform programs have lost momentum because policymakers lost their appetite for change, or were overwhelmed by opposition, or because uncertainty on the part of economic actors led to a slow response that helped engender political opposition. The lack of follow-through can have a significant economic cost, and it can also be politically painful for governments whose enthusiasm for reform faded at the wrong moment. The cost of failed or partially implemented reforms can be high.

Look at Argentina, an economy still recovering from the impact of the crisis in 2001. Yet in the 1990s, Argentina's reform program was, initially, seen as a model.

After the experience of the 1980s, when the Argentine economy had contracted by about half a per cent a year, while inflation had soared—to a peak in the late 1980s of 3,000 per cent—the 1991 Convertibility Plan was seen as the centerpiece of reform. The aim was to deliver high growth and low inflation, based on disciplined macroeconomic policies and market-oriented structural reform. The Argentine government wanted to escape from past inflationary failures: and at the start the government showed every sign of a readiness to take the difficult steps needed to deliver macroeconomic stability.

Central to the convertibility plan was its guarantee of peso convertibility with the dollar at parity. Having a fixed exchange rate system was seen as crucial for building up the anti-inflationary credibility that the government needed. A quasi-currency board was established in order to underpin the system and so reinforce the government's commitment. The early results were promising. For such a plan to work in the longer term, however, fiscal restraint is required.

In the first two years of the plan, Argentina experienced real annual GDP growth of over 10%, and more than 5% in 1993-94. Inflation was down to single digits by 1993. There was a huge surge in capital inflows. The loss of confidence in emerging markets that followed Mexico's so-called Tequila crisis of 1995 seemed to do little more than interrupt performance improvements temporarily. Argentine growth rebounded in 1996-97.

But this apparently impressive performance masked structural weaknesses that weren't confronted. The convertibility plan was only sustainable if there was fiscal discipline and sufficient flexibility in the economy to adjust to shocks. However, fiscal control was undermined by off-budget expenditures; and it was too weak to prevent growing reliance on private capital flows to finance government borrowing. The estimated structural fiscal position went from rough balance in 1992-93 to a deficit of about 2.75% of GDP in 1998. In addition, there was persistent off-budget spending, mainly as a result of court-ordered compensation payments after the social security reforms of the 1990s, and arrears to suppliers. This raised the government's average new borrowing requirements to more than 3% of GDP a year over this period. In 1996, for example, when off-budget spending is included, the total deficit was 4% of GDP.

The problem was made worse by overoptimistic assessments of the economy's growth potential. There were more temporary factors at play than was realized at the time. This, of course, meant that the authorities had less room for maneuver on the fiscal front than they believed. The decentralized system of government in Argentina made fiscal control particularly difficult. The provinces, for example, had little incentive to improve revenue performance or constrain expenditure growth: and this important weakness in the fiscal structure was not tackled.

Exports grew, but nowhere nearly as rapidly as imports. By the late 1990s, there was an uncomfortably high debt to export ratio: it was 455% in 1998, and jumped, unexpectedly, to 530% in 1999.

Added to all this was the fact that the pace of structural reforms lost momentum in the mid -90s; indeed, some reforms were reversed. Perhaps the biggest weakness here was the labor market which was heavily regulated in Argentina. Individual workers have long enjoyed considerable protection, with high barriers to dismissal and the guarantee of generous fringe benefits. Yet a fixed exchange rate regime required labor market flexibility to enable the economy to respond to shocks.

Some labor market reforms were introduced in the 1990s. In 1991, for example, there was some modest improvement in flexibility. In 1995 came equally modest steps that made it easier to hire temporary workers and introduced more flexible working hours. But these improvements did not go far, and reform in this area quickly lost momentum. To make matters worse, Congress diluted a further government attempt to increase labor market flexibility—so much so, in fact, that the final outcome promoted further centralization of collective bargaining. The result of all these lukewarm changes was predictable: unemployment rose, to 12% in 1994; and productivity growth fell to zero in the second half of the 1990s.

These structural rigidities were all the more problematic in the light of the fixed exchange rate regime. It became increasingly clear during the 1990s that successful operation of the quasi-currency board required much better fiscal control than the government was able to deliver. Since the collapse, of course, there has been much debate about whether the pegged exchange rate regime was ever appropriate as a long-term policy instrument.

The arguments over the exchange rate regime have ensured a plethora of different explanations of the eventual crisis in 2001. Both the then government and the IMF have been criticized, most recently by the Fund's own Independent Evaluation Office, in a report published last month. But it is clear that the impact of the crisis would at least have been mitigated if the original reform program had been more ambitious—specifically if it had included serious reform of the labor market and if it had tackled the problem of fiscal relations between central and provincial governments; and if this more ambitious program had been fulfilled.

Effective implementation of fiscal, labor market and structural reforms could have meant that the economy was robust and flexible enough to cope with unanticipated shocks and thus could have avoided the economic collapse that resulted from the abandonment of the convertibility plan.

Instead of Argentina I could equally well have pointed to Turkey in the 1980s, when an ambitious, and initially very successful, reform program petered out. Some reforms were enduring. Quantitative restrictions on imports were removed, the exchange rate became more realistic, and the structure of the economy was transformed—exports went from 5-6% of GDP to about 20%.

But by the late 1980s, the then Turkish government displayed a reluctance to confront the re-emerging problem of persistent high inflation and to impose fiscal discipline. In Turkey's case, of course, that meant the reform program was left unfinished and that the potential benefits were therefore not fully realized. This ultimately led to a series of financial crises.

The current Turkish government has shown itself to be committed to a tough reform program that should bring Turkey's debt down to sustainable levels and deliver rapid economic growth over the medium and long term. But the cost of adjustment this time is higher than it would have been if the 1980s reforms had been adhered to and if those reforms had encompassed the financial sector which was the proximate cause of the most recent crisis in 2000.

But there are examples of countries where experience illustrates what benefits commitment can bring. Korea was a poor, rural peasant economy in the 1950s. There was a widespread view that it was not a viable economy without significant injections of foreign aid. The realization that financial aid from America was going to decline was part of the motivation for embarking on a wide-ranging series of reforms starting in the late 1950s.

These reforms had the objective of turning Korea into a fully-fledged market economy. They systematically addressed fundamental problems. There was a determination to tackle structural problems in the economy, and to stick with a reform program. Sound macroeconomic policies were put in place.

In 1960, there was a large but necessary devaluation, and export incentives were adjusted to relative constancy in the real returns to exporters over time. Over the next few years, almost all quantitative restrictions on imports were converted to tariffs. In 1964, a major fiscal reform was introduced, which greatly reduced the government's budget deficit; at the same time interest rate ceilings were relaxed and the exchange rate regime was changed to a crawling peg.

Tax policy was reformed and tax collection improved. Public spending was brought under control and high tariffs reduced. The huge budget deficits of earlier years were virtually eliminated in just a few years. At an early stage, the importance of infrastructure investment as an aid to exporters and import-competing firms was recognized-and appropriate measures taken.

Later in the reform process, greater emphasis was put on further liberalization of trade and the financial sector. And the reform process has continued as Korean policymakers adapt to deal with the problems that come with further growth. But the main thrust of economic policy has remained largely constant-an outward orientation with strong incentives for exporters, and a commitment to growth through trade. This has meant that the country has been well-placed to cope with the fresh challenges that economic success brings.

The rewards Korea reaped from these ambitious reforms are well-known. Economic growth was spectacular, especially in the 1960s and early 1970s. GDP per capita rose seven fold in the three decades to 1995. The third poorest country in Asia in 1960 became one of the region's richest by the mid-1990s. It is a performance of which any country would, and should, be proud.

Korea's successful performance was reinforced by policymakers' ability-and willingness-to try to anticipate bottlenecks and potential crisis points. They were arguably successful in this until the 1990s when they failed to anticipate the problems that weaknesses in the financial sector could bring for the wider economy.

Roger Douglas, New Zealand's finance minister in the 1980s when that country's far-reaching reforms were undertaken, is clear that ambitious reforms bring great rewards. He has described how in the decade before 1984, New Zealand's economic growth rate was half the OECD average, with unemployment rising apparently inexorably and, with it, government spending and government debt and debt servicing costs. In the quarter of a century up to 1984, New Zealand's standard of living had fallen from being third-highest in the world, to a ranking somewhere in the mid-twenties.

The economy Roger Douglas found when he became finance minister in 1984 was highly regulated, with quantitative import restrictions and subsidies for exporters; high tariffs, distorted price controls, expensive social policies and low workforce skill levels.

The reform program introduced by the Labor Government of which Roger Douglas was a member was sweeping. The exchange rate was floated, industrial subsidies were removed, the tax system was reformed, financial markets were deregulated as were prices and incomes, quantitative import restrictions were scrapped, the labor market was reformed. A long list of changes, in fact, and all introduced at a breathless pace.

Douglas says it is impossible to go too fast, that speed is essential, in part because many structural reforms will anyway take several years to implement and have their full impact.

"When an economy has been driven down a blind alley and ends up facing a brick wall, what matters is to back it out as soon as possible and get it back onto the high road to a better future".

What, when and how

So what have we learned about the best way to implement reforms, and when? Of course, the specific reform measures will depend on the circumstances of individual countries, and so too will the most appropriate timing. But it is clear that some general principles appear to hold good.

Roger Douglas is emphatic that reforms should be pursued on a broad front: that clearly brought results in New Zealand and in Korea; I think Chile's experience supports that thesis too.

But it is also clear that trade and exchange rate reforms should be an early priority. The more rapidly an economy is opened up to the rest of the world, the better. This is partly because opening brings considerable economic benefits, as I mentioned earlier; and partly because the more open the economy the more difficult it will be to reverse the reforms.

Fiscal discipline has to be a key component of any sound economic reform program. Its absence helped undercut the Turkish reforms of the 1980s and was clearly a major factor in Argentina's more recent crisis.

Fiscal rectitude is difficult in countries that have pressing social needs and serious shortcomings in infrastructure that can impede growth. But better targeted spending-to those most in need-can help. So can more efficient tax collection which will boost tax revenues, and a widening of the tax base.

Beyond identifying winners and losers and building support for change, it seems clear that political leaders need to be ready to confront their opponents. How this is done can be critical. In Britain, Margaret Thatcher successfully took on the miners in 1984: but the Heath government failed when it confronted the same group in 1973-74. Confronting opponents is partly a matter of picking fights that are winnable-and luck plays an important part here-and it is partly a matter of having the courage to see the battle through to the end.

Effective reform also requires good communication, so that citizens know what the aims are and what the potential ramifications are. Transparency is also important because it is an important means of building support.

The IMF's role

The IMF has an important role in all this. The Fund's principal objective, as it always has been, is the maintenance of international financial stability. That gives the Fund an important role in crisis prevention and resolution. Reform is, in a very broad sense, the Fund's business. Article IV surveillance on our member countries provides an annual assessment of economic policies, along with recommendations for reform, as appropriate. This applies equally to the largest industrial economies, emerging market countries and the poorest countries. This evening I have referred mainly to the reform process in emerging market countries. But the Fund also regularly recommends reform in our richest members: we recently made recommendations relating to the fiscal situation in the United States.

The Fund's role in the reform process under surveillance is an advisory one. This is partly in order to respect the sovereignty of member governments. But it also reflects the recognition that national policymakers are those who have to adopt reforms and try to ensure their success. It is they who have to secure the support of civil society and the various economic actors involved and whose response will contribute to the success or otherwise of the reforms.

But for countries receiving financial assistance from the Fund as they try to resolve crises, such assistance is provided on a conditional basis, with payments provided in installments against agreed benchmarks. The terms of such assistance are agreed with member countries, the government must "own" the reform programs that are undertaken. The Fund's role is to be assured that the program has a reasonable prospect of success.

There is a third way in which the Fund assists countries implementing reforms: technical assistance. This can be more important than is often realized. On a wide range of policy issues, from tax administration to budget accounting, from financial sector regulation to the development of commercial codes, the Fund has amassed considerable expertise and experience that can benefit governments seeking advice on formulating and implement such measures. It is clear from the wide range of governments that seek technical assistance from the Fund that this expertise has become a valuable and valued resource.

Currently, for example, the Fund is actively supporting banking reform in Turkey, providing technical advice for budgetary reforms in China and encouraging trade reform in Africa. The Fund has actively supported countries adopting floating exchange rate regimes and pension reforms.

Conclusion

Let me briefly conclude by re-iterating that reform is obviously important-and often urgently needed. Many countries need to implement change if they are to continue to grow, or are to achieve sustained growth. But well-judged and effectively implemented reform is also important-and difficult. Mis-guided or overly timid reforms can undermine the reform process and make it difficult to muster support for future reform programs.

Perseverance is crucial. Reforms have to be followed through. There is no scope for talk of `reform fatigue'. Reforms bring the flexibility that is essential for economies to adapt successfully to changes in the global economy.

Finally, there is one ingredient alien to economists-luck. We do not like admitting that there are things that we do not know and, more importantly, can not know. But the reform process is, inevitably, something of a mystery. We cannot predict the exact outcome. But by proper and effective implementation we can increase the prospects of success.

Thank you.