Speech

Perspectives and Lessons from Country Experiences with Inflation Targeting, Remarks at a Panel on Inflation Targeting, by Mr. Murilo Portugal, Deputy Managing Director, IMF

May 17, 2007

Ladies and Gentlemen:

I am pleased to be participating in today's panel on the issues of inflation targeting, employment creation, and economic development, three very important issues; and I would like to thank the G-24 and the Political Economy Research Institute at the University of Massachusetts for taking the initiative in organizing this workshop.

The published agenda for this workshop raises a number of important questions regarding inflation targeting. One is whether inflation targeting is compatible with growth, employment generation and poverty reduction. Another is how inflation targeting can be adapted or modified to take account of concerns for employment, competitiveness, and growth. A third question is whether there are reasonable alternatives to inflation targeting in a world with increasing international integration both through trade in goods and through financial capital flows and with ever growing financial innovation.

In my remarks today, I will try to address these questions drawing on the experience with inflation targeting in emerging market economies over the past decade or so. But I will also comment on some key features of the macroeconomic experience and lessons raised by the experience of emerging market economies with inflation targeting.

The spread of inflation targeting

Since it was first introduced by New Zealand in 1990, inflation targeting has been adopted by 24 countries, of which 16 are emerging market and developing economies. In our discussions with IMF members, it is also apparent that more emerging market and developing countries are likely to move to inflation targeting within the next 10 years.

So it is important first to understand why is inflation targeting becoming increasingly popular? I would point to four main factors.

First, the experiences of developed and developing countries alike during the 1970s and 1980s showed that higher inflation could not promote stronger growth, external competitiveness, or employment, on a sustainable basis. On the contrary, high and unpredictable inflation was shown to be bad for growth, for employment, and for equitable distribution of real income in the long run. In retrospect, this perhaps should have been obvious: high and uncertain inflation deters productive investment and employment generation, and the people best able to protect themselves against inflation are generally those with high incomes or wealth.

These experiences have led governments almost everywhere to the conclusion that the main contribution that central banks can make to good economic performance over the long term is to provide an environment in which inflation and inflation expectations are anchored at a low level.

This experience partly addresses one of the questions of this workshop, namely whether inflation targeting is compatible with growth and employment creation. To the extent that inflation is inimical to long-term growth and enduring employment creation and to the extent that inflation targeting is an effective policy framework to bring inflation down, there would seem to be no incompatibility between inflation targeting and growth and employment.

The second reason why inflation targeting became popular is that innovations in financial products and financial markets inevitably affect the relationships between financial activity and the real economy. As a result, money and credit aggregates have often become less reliable or useful as intermediate targets for inflation. In many countries the relationship between money and inflation became unstable in the short run. In fact, this was the main reason why inflation targeting was invented in advanced countries in the 1990s. For many countries, moving to inflation targeting was seen as a more practical, or less risky, alternative than targeting a monetary aggregate.

This problem also occurs in emerging market countries that may undergo major structural changes and credit expansion due to the development process, which make monetary aggregates less reliable predictors of future growth and inflation.

A third consideration in the adoption of inflation targeting is that the increasing international integration of goods and financial markets, which has been an important common factor behind changes in monetary policy regimes. With more open capital accounts, many countries have decided that a flexible exchange rate framework is better suited to cushioning domestic economic performance from external disturbances than fixed nominal exchange rates. However, when a peg is replaced with a flexible exchange rate regime a new nominal anchor is needed. In principle, either monetary targeting, aimed at anchoring nominal GDP, or inflation targeting could provide such a nominal anchor. But as I mentioned earlier, financial innovation may make monetary aggregates less reliable as intermediate policy targets.

Fourth, the experiences of the industrial countries that have adopted inflation targeting are generally considered as having been successful in providing both policy credibility and flexibility, resulting in better inflation and growth performances, together with increased resilience to shocks. Closely associated with this success has been the apparent benefit of policy transparency for the credibility of inflation targeting frameworks and, ultimately, for macroeconomic performance. This experience has also influenced emerging markets to adopt inflation targeting.

Although the adoption of inflation targeting by advanced economies has generally been regarded as a success, its adoption in emerging market and developing economies has been more controversial. Debate has focused on three main questions some of which were also identified to be discussed in this seminar. The first is whether achieving better inflation performance is likely to come at the cost of poorer growth performance. The second question is what pre-conditions need to be met for the adoption of inflation targeting, and whether these preconditions are likely to be too demanding for most emerging market and developing countries. The third question is whether an inflation targeting framework is more robust than alternative monetary approaches to the kinds of financial market disturbances to which emerging market and developing countries may be particularly vulnerable.

The experiences of countries like Brazil, the Czech Republic, Korea, Mexico, and South Africa are helpful in starting to answer these questions that have been debated, sometimes passionately, for nearly a decade, in academic circles, among central banks, and also inside the IMF.

The experience of emerging market and developing economies

A growing number of studies—including one published in September 2005 IMF World Economic Outlook—suggest that inflation targeting in emerging market and developing countries has been associated with better macroeconomic performance than under alternative monetary policy frameworks in otherwise similar economies over the same period.¹

It is true that most emerging market and developing countries—whether they targeted inflation or not—performed much better in terms of growth and inflation since 2000 than during the 1990s. But between these two periods, the inflation targeters typically succeeded in cutting their inflation rates from over 10 percent per annum to around 4 percent, roughly by twice as much as in non-inflation targeters. Both groups of countries also enjoyed increases in real GDP growth rates, typically around 2/3 of a percentage point, but differences between the two groups in terms of growth were not statistically significant.²

But there are a number of reasons why one might think that inflation targeting should be favorable to growth. The fact that it has been successfully adopted by countries with very diverse structures is an indication that it is a flexible policy framework that can be adapted to country circumstances and can be adapted over time.

If a country pursues "strict" inflation targeting relying heavily on a relatively rapid impact to keep inflation close to target this approach would tend to increase the volatility of output and employment. However, a "flexible" approach that would rely more on the impact of excess demand pressures on inflation, involving longer policy lags would reduce output and employment volatility.

An inflation targeting framework provides several elements of flexibility that can be used by policy makers to adapt to country specific circumstances. One is the choice of the price index used to target inflation. In some countries core inflation may be a better predictor of future inflation, while in others headline inflation may be the better predictor and may dominate inflation expectations. Hence the choice of the price index can be made accordingly. The choice of the numerical value of the target, as well as the width of the band around the central target provide another element of flexibility. Countries that are more likely to be subject to external and supply shocks may, for instance, opt for a wider band around the central target.

Another choice to be made is over which period the inflation target is to be achieved. This period should match the length of the lags in the monetary policy transmission mechanism, and the length of these lags varies from country to country. Given the forward looking nature of inflation targeting, there is flexibility for the policy maker to choose to which shocks to respond and over which period of time. If inflationary expectations are well anchored, a country can accommodate temporary supply shocks that may have a large short-term impact on inflation, but may not require an overreaction from monetary policy, because with well-anchored expectations, price shocks are less likely to have second-round effects. Of course doing that would depend critically on inflation expectations being well anchored, which in turn depends critically on the credibility that the Central Bank might have already gained. In practice, the real anchor of the system is the credibility of the commitment of the Central Bank to achieve the inflation target.

Stable inflation expectations also allow for an extension of maturities of fixed-rate instruments for which a secondary market exists, and such a lengthening of the yield curve can have positive effects to long term financing and to growth. And this is another channel by which stable inflation expectations help growth.

Research at the IMF suggests that neither industrial nor emerging market inflation targeters pursue inflation targets in a fashion that ignores the short-term consequences for growth.³ Both groups of countries miss their targets far more often—around one-third of the time—and for too long to be consistent with strict inflation targeting. In other words, inflation targeters typically appear to be willing to accept significant deviations of inflation from target in order to avoid inducing excessive volatility in growth or unemployment, while at the same time ensuring that inflation is brought back on track over the medium term. Consistent with this finding, the evidence also shows that emerging market economies with inflation targets typically experience lower variability of both growth and inflation than emerging markets with other monetary policy regimes.⁴

Overall, these findings are generally favorable to inflation targeting. In particular, they suggest that, when inflation targeting is pursued in a flexible manner—that is, taking into consideration the short-term consequences for output—the policy framework can achieve substantial reductions in inflation without significant cost in terms of growth, and that the framework is entirely compatible with low volatility of both inflation and growth. Against this background, it is not so surprising to note that no country that has adopted inflation targeting so far has abandoned it in favor of some other framework.

In short, the emerging countries that adopted inflation targeting managed to achieve substantial reductions in inflation in the 2000-2005 period without any significant sacrifice in terms of growth performance.

What is still not clear is the extent to which the better macroeconomic performance in inflation targeting countries can be credited exclusively to the change in the monetary policy regime. This is because in most countries adopting inflation targeting, the change in the monetary framework has been part of a more wide-ranging set of structural and policy reforms, including substantial fiscal consolidation. At the same time, many emerging market economies that have not yet adopted inflation targeting have also introduced wide-ranging reforms. Consequently, it is also hard to argue that all of the improvement in macroeconomic performance of the inflation targeters relative to other emerging markets can be attributed to factors other than monetary policy. Or put more plainly, even when all other considerations are taken into account, the adoption of inflation targeting appears to have helped emerging market economies to get inflation down to low levels without any significant cost in terms of growth.

This does not, of course, mean that inflation targeting has proven to be entirely satisfactory in all situations. Let me highlight two particular issues that have been problematic in several countries:

First, a perennial issue in any open economy, including an inflation targeter, is the role of the exchange rate. This has been a particularly difficult issue for inflation targeters that have previously relied on the exchange rate as the nominal anchor for policy. In some cases, it has been quite difficult either for the central bank to subordinate exchange rate and competitiveness concerns to the inflation objective, or for the central bank to convince markets that it has done so. In Chile, Hungary, and Romania for example, markets have tested the central bank's commitment to the inflation target as opposed to an exchange rate objective.

Eventually, however, the private sector and policy makers need to accept that international cost competitiveness is more fundamentally an issue of productivity—that is, using labor and capital efficiently to keep costs down. A weak exchange rate may be able to boost competitiveness temporarily, essentially by cutting wages and profits, but it does not address the underlying issue of the need to raise productivity.

In other cases, domestic factors can play a key role in triggering shifts in market sentiment. For example, Brazil in 2002 and the Philippines have experienced episodes of turbulence in which political developments and related concerns over future fiscal policy appear to have played a key role. Last year, Iceland also experienced some exchange market pressure, reflecting concerns over the large current account imbalance and banking sector vulnerabilities associated with rapid credit growth.

In no case, however, have episodes of either transient or more long-lived exchange rate pressures led to a breakdown of the policy framework, as can occur when an exchange rate peg becomes unsustainable. Indeed, such episodes, however painful at the time, can also strengthen policy credibility if the central bank keeps its focus on the inflation objective. Nonetheless, inflation targeting frameworks in many emerging market and developing economies are yet to be tested by a more volatile and trying global macroeconomic and financial market environment.

Lessons from inflation targeting experiences

So what lessons can we draw from the experiences of inflation targeting countries? I will highlight a few that I think may be particularly relevant in answering the questions posed in the agenda for this seminar:

• First, I would say that the evidence suggests that there is in fact no real trade-off over the medium- to long-term between growth and low inflation. Indeed, the evidence points the other way, that a credible commitment to maintain low inflation is good for long-term growth. Moreover, the evidence points to this in emerging market as well as industrial countries.
• Second, no country appears to pursue a strict form of inflation targeting that focuses single-mindedly on inflation even in the short-term. Instead, a flexible form of inflation targeting is practiced in which short-term conflicts between growth and inflation resulting from adverse supply shocks are typically resolved in favor of reducing the volatility of employment and output.
• Third, inflation targeting does not offer a magic formula for insulating a country from external financial disturbances or other influences on the real exchange rate and external competitiveness. In that respect, inflation targeting is not different from alternative policy regimes. However, what inflation targeting can do is to ensure that when such disturbances do happen, they do not trigger an inflationary process that undermines long-term growth.

The experience of emerging market countries and, indeed, the industrialized nations that have adopted inflation targeting, also suggests that inflation targeting can be successful even if not all of the desirable conditions are in place. The minimum requirements appear to be three-fold:

• The central bank needs to have the autonomy—and associated accountability—required to pursue a clear mandate, with the policy objectives supported by the government, both in words and in deeds. In particular, fiscal discipline can contribute significantly to the credibility and achievement of the inflation objective.
• The central bank also needs to have effective instruments for influencing domestic spending and savings behavior, which generally requires functioning financial markets, as well as a reasonably stable financial system.
• Lastly, the central bank needs to have adequate economic and financial data, analytical capacity, and a reasonable understanding of how monetary policy affects inflation, in order to respond in a timely manner to inflation pressures.

Although one can debate what constitutes a minimum required standard for each of the different elements I have referred to, there is little doubt that the further along a country is in developing these elements prior to adopting inflation targeting, the more credibility the policy framework will have, and the better its macroeconomic performance is likely to be. Nonetheless, the fact that countries have been able to undertake inflation targeting successfully with widely differing initial conditions illustrates one of the main strengths and attractions of the framework—its flexibility and adaptability to different circumstances.

Thank you.

¹ See, in particular: F. Mishkin and K. Schmidt-Hebbel, 2005, "Does Inflation Targeting Make a Difference?" forthcoming in K. Schmidt-Hebbel and F. Mishkin, 2005, Monetary Policy Under Inflation Targeting, (Santiago: Central Bank of Chile), and M. Vega and D. Winkelried, 2005, "Inflation Targeting and Inflation Behavior: A Successful Story?" International Journal of Central Banking, Vol. 1(3), 153-175..

² See, for example, S. Roger, 2006, "An Overview of Inflation Targeting in Emerging Market Economies," forthcoming in Bank of Thailand, 2007, Challenges to Inflation Targeting in Emerging Countries, (Bangkok, Bank of Thailand).

³ See S. Roger and M. Stone, 2005, "On Target? The International Experience with Achieving Inflation Targets", IMF Working Paper 05/163.

⁴ See N. Batini, P. Breuer, K. Kochhar, and S. Roger, 2006, "Inflation Targeting and the IMF," IMF Board Paper SM/06/33.