Lessons from Europe's experience with exchange rate regimes

Comments on Charles Wyplosz' paper by William R White, Economic Adviser, Bank for International Settlements, at a conference on 'The role of regional financial arrangements in crisis prevention and management: the experiences of Europe, Asia, Africa and Latin America', organised by the Forum on Debt and Development, Prague, 21 June 2001.

Growth in the global economy, and especially the United States, has recently been decelerating sharply. In this light, it is somewhat amusing to hear many previous advocates of New Era economics now say "Well, we always knew that the stock market was overvalued ..., we always knew that investment cycles were still possible ...". I think this kind of historical revisionism is too easy. We should rather be asking ourselves how we came to believe certain things at certain times, and what factors caused us to revise our views when we did.

Regarding Charles' paper, I like his overview of what went on in continental Europe, particularly in the postwar period until the mid-1990s. It has some very intriguing results, not least of which is that financial repression was good for Europe, as indicated by his statistical estimates. While I am more than a little suspicious about this, as I will argue later, I do agree with the associated proposition that governments should be very careful about how they go about financial liberalisation. Sequencing, particularly in so far as it affects external capital flows, can be very important. Moreover, the period of transition seems inherently to be a dangerous one. This applies even in cases where the target state of affairs is judged to be totally desirable.

My principal criticism of this paper, and perhaps I am actually inviting Charles to write another paper, reflects the fact that this is a conference on the role of regional financial arrangements in crisis prevention and management. Unfortunately, what I did not see in the paper was exactly how the analysis it contains actually relates to this topic. A related criticism about focus is that the second part of the paper, in which the lessons are drawn, seems to me to have very little to do with the first part, in which a rather straightforward historical description of European economic developments is presented.

Focusing then on the lessons themselves, Charles seems to be saying that continental Europeans have always valued fixed exchange rate systems. This is because the benefits have been thought to exceed the costs, even when these costs include a significant degree of financial repression to make the fixed rate systems functional. In effect, recognising the reality of the "impossible trinity", Europeans have opted for financial controls to square the circle.

The benefits of fixed exchange rates, as argued in the paper, seem to be of two sorts. First, they encourage trade, competition and growth. Second, the fixing of exchange rates can lead on to a single currency, which may be both economically and politically desirable. What about the costs of financial repression? Basically, the paper says that the costs are not large. Indeed, repression may even be good for you in its own right. While Charles does not say this explicitly, he seems to suggest that other regions might want to go down the European path. That is, they might wish to institute regional fixed exchange rate systems and then gradually evolve the institutions needed to foster the movement to a single currency. If this is the basic lesson for other regions to be drawn from this paper, based on European experience, I would argue that it may not be generally valid. Let me explain why. In effect, I question in turn the analysis of each of the benefits and costs just described.

I do not dispute the fact that continental Europeans have traditionally preferred fixed exchange systems. Where I do disagree with Charles is his suggestion that the principal reason for this was the belief (and I will return to this) that fixed exchange rates encourage trade and growth. Other reasons suggested by Charles for this preference could also be plausible. A first possibility is that fixed rate systems were chosen to foster stability given that financial markets were shallow. This could be historically true, but there is a logical problem with this argument. If Europeans consciously kept financial markets shallow in order to fix the exchange rate, there must have been some other more fundamental motivation than the absence of well developed financial markets.

However, there remains a second and more plausible reason for this preference. It is that Europeans preferred fixed exchange rates in order to encourage discipline and price stability. With Germany as the European anchor, this view seems very much in accordance with conventional wisdom. Charles, however, disagrees with this conclusion on the following grounds. He argues that, if Europeans were after discipline via fixed exchange rates, they ought to have had lower inflation rates than other countries. I think this logic is incorrect. Fixing an exchange rate amongst themselves in no way provides an inflation anchor relative to some other group of countries. And, in any event, fixed exchange rates only give the participating countries similar rates of inflation over the long run. Nigel Lawson, it will be recalled, tried at the end of the 1980s to peg the pound to the Deutsche Mark. His experience was that, before you get to price convergence, it can be a very long run indeed

In spite of all these counterarguments, Charles continues to conclude that the real motive behind the decision of Europeans to fix the exchange rate was to promote trade. But this is where I have another problem. As even Charles himself says in the paper, there is absolutely no empirical evidence that fixed exchange rates encourage trade. I invite you to take a look at the broadest set of data. What happened after the collapse of Bretton Woods? The answer is that trade exploded when currencies started to float. And what has happened in recent years since many smaller countries have chosen to float their currencies? Again, the growth of trade volumes has been dramatic.

I also found unconvincing Charles' discussion of why the literature on this topic might be wrong. In the very short run, exchange rate volatility might make cross-border trade less attractive. However, this risk can be easily hedged in modern markets. In the medium term, exchange rate misalignments could conceivably hollow out the trading sector of the country with the overvalued currency. However, this effect would be only temporary and would be matched by the stimulus to trade given to others.

So to sum up, I simply do not believe that fixed exchange rates are needed to promote trade and growth. Nor do I accept that this belief, in fact, provided the principal motivation behind the revealed preference of continental Europeans for fixed exchange rate systems.

The second postulated benefit of fixed exchange rate systems, as argued in Charles' paper, is that they may lead on to a single currency. The question of course is whether such a single currency would be economically and politically desirable for regions other than Europe. As for economic desirability, one has to go back to the literature on optimum currency areas. In this context it is not so obvious that there are many such regions already out there, with the possible exception of those Gulf States already pegged to the dollar. As for the political desirability of a single currency, the impetus provided to the concept by the two world wars in Europe can hardly be underestimated. In contrast, were Canadians and Mexicans to be told that accepting a single currency with the United States would lead to political unity, I doubt there would be great popular enthusiasm.

Finally, let me turn to the costs of the financial repression which might be needed to make a fixed exchange system work properly. Charles recognizes that internal financial repression does have costs in that it leads to a lower quality of financial services and inefficient resource allocation. In particular, such a system often leads to a wasteful government getting enhanced access to national savings. However, he argues (and I agree) that liberalised domestic financial systems also have costs associated with the booms and busts caused by alternating waves of optimism and pessimism in a credit based economy. But the former set of costs seems to me, although perhaps not to Charles, to dominate the latter. If so, this makes internal repression broadly undesirable and explains why the vast majority of countries in the world are now moving towards financial liberalisation. As for financial repression with respect to the external accounts, I would again argue that there are significant costs, even if liberalisation also holds certain dangers. These latter concerns have been attested to by a number of recent crises in emerging market economies, which has led countries to be more careful about how and how quickly they go about removing capital controls. It has not, however, led to any kind of a significant move in the direction of reimposing controls.

What is the bottom line for me? Fixed rate systems in themselves need not enhance trade or growth, although in some cases they might do so. Nor do they necessarily lead on to the development of a single currency. And even if they did have this tendency, it is not clear that a single currency would be desirable on either economic or political grounds. As for the costs of financial repression, I believe they could be significant. For all these reasons, I conclude that the positive aspects of the post war experience with fixed exchange rate systems in continental Europe may not be easily replicated elsewhere.