The choice between fixed and flexible exchange rates is one of the most important in macroeconomic policy, but the debate that surrounds it seems deceptively simple. Though non-economists seldom claim to understand most macroeconomic issues, such as the cause of unemployment, the role of central banks, or the importance of government deficits and debt, there is a surprising level of agreement among them that, for Canada, a fixed exchange rate would be preferable to the status quo of a flexible, sometimes even volatile, exchange rate. But that’s only because the benefits of a flexible exchange rate are not widely understood. The real policy choice is not nearly so simple, for both theoretical and empirical reasons. This is one area of research where professional economists could help resolve the debate.
Before laying out the debate in its simplest form, it is worth emphasising an important point of agreement between Robert Mundell and Milton Friedman. As Professor Mundell puts it:
The choice between fixed and flexible exchange rates is an oxymoron. The alternatives are incomparable. A fixed exchange rate system is a monetary rule. A flexible exchange rate system is the absence of that particular monetary rule and is consistent with price stability or anything at all, including hyperinflation. The real choice is between a fixed exchange rate monetary rule and [a flexible exchange rate with] alternative monetary rules such as inflation targeting or monetary targeting.
Some economists argue that a country benefits from fixing the value of its currency to a stable currency, such as the US dollar, because by doing so it gains more credibility in its monetary policy. For countries that lack monetary-policy credibility, this argument is surely correct. But it is worth separating the benefit of greater credibility in monetary policy from the other benefits of a fixed exchange rate. In many countries, especially small developing countries, enhanced monetary-policy credibility may be most easily achieved by fixing the exchange rate. In Canada’s case, however, a formal system of inflation targeting has been in place since 1991 and it has been very successful in keeping inflation low and stable. In fact, if there has been any criticism about the Bank of Canada’s monetary policy in recent years, it is that it has achieved “too much” low-inflation credibility, that it has been too aggressive in its determination to keep inflation low.
So we shouldn’t be misled by the red herring that fixing the exchange rate will improve Canada’s monetary policy. The Bank of Canada gets top marks for its commitment to low and stable inflation, and the past decade has shown that it is prepared to do whatever is necessary to keep inflation within the target bands. In short, there is little evidence based on policy performance over the past decade—or on a reasonable forecast of future policy—that Canada’s monetary policy would be improved by tying its currency to the US dollar.
The relevant policy choice, especially considering domestic political realities, is therefore between maintaining the status quo of flexible exchange rates and inflation targeting, on the one hand, and, on the other, fixing the value of the Canadian dollar to the US dollar (probably through a currency board that limits the number of C-dollars issued to a fixed percentage, possibly 100, of the country’s stock of US dollar reserves). Given the Bank of Canada’s success with inflation targeting, the debate over this choice should focus on questions other than monetary policy credibility.
Leaving credibility aside, the debate between fixed and flexible exchange rates then boils down to a weighing of the microeconomic benefits of fixed rates against the macroeconomic benefits of flexible rates. This contrast between micro and macro effects explains why the debate is viewed by non-economists as being so one-sided: Most non-economists can easily appreciate the benefits to firms and individuals that come from fixed exchange rates but they have much less appreciation for the role of flexible exchange rates in the macroeconomic adjustment process.
The microeconomic benefit of a fixed exchange rate is the greater certainty that it delivers to individual firms and consumers who are involved in international transactions. Exchange-rate volatility creates uncertainty about the prices that must be paid (or will be received) for goods and services, especially for people and businesses involved in long-term contractual arrangements. It also creates uncertainty about the rate of return investors will earn from holding bonds or equity denominated in foreign currencies. Assuming, reasonably, that firms and consumers are risk-averse, it is obvious that greater exchange-rate uncertainty will lead to fewer international transactions. The adoption of fixed exchange rates will reduce this uncertainty, increase the flow of international transactions in goods, services, and assets, and therefore increase the economic gains from trade.
Additional microeconomic benefits are achieved if we go beyond fixing the exchange rate and adopt the US dollar as legal tender—a policy known as “dollarisation.” In this case, in addition to the gain from even greater exchange rate certainty, we would also avoid all costs associated with the conversion of currencies. Set against these gains would be the loss of “seigniorage” revenue that accompanies the sacrifice of a domestic currency. (When governments persuade their citizens to part with real goods and services in exchange for currency, this “seigniorage” is equivalent to taxing them.) But the estimates I have seen suggest that the savings of currency conversion costs are greater than the loss of seigniorage revenue, so that there is a net benefit from dollarisation.
The macroeconomic benefit of having flexible exchange rates is that the exchange rate can change in response to external shocks, thus lessening the need for internal adjustment. Such internal adjustment can occur either through wages and prices or through employment and output. Given the sluggishness of wages and prices, however, a system of fixed exchange rates would force the adjustment onto domestic quantities rather than prices, thus leading to greater volatility of employment and output. This is the well-cited “shock absorber” argument in favour of maintaining flexible exchange rates; at its core is the premise that volatile exchange rates are more desirable than volatile employment and output.
One criticism of this argument in favour of flexible exchange rates is that the exchange rate is only one price, and its flexibility cannot hope to accommodate the widespread changes in international prices that are constantly occurring. This is true enough from the perspective of what economists call “general equilibrium”: An external shock requires that internal prices in many domestic industries adjust, sometimes in opposite ways. And it is too much to expect a single price—the Canadian-dollar price of foreign exchange—to achieve this complex adjustment all on its own. What is required is internal wage and price flexibility. Robert Mundell makes this point in the debate when he says: “Exchange rate changes can never be a substitute for the vast number of changes in individual prices that have to be made in an efficient market.” In other words, Mundell believes the exchange rate is not an effective tool for adjusting an economy to external shocks.
From another perspective, however, a flexible exchange rate is a very effective tool for dealing with external shocks—the perspective of maintaining what is variously referred to as internal balance, full employment, or a zero output gap. In the general equilibrium perspective, a decline in the world’s demand for some range of Canadian products requires, in the full passage of time, a reduction in relative prices and wages in those sectors. But in the span of time over which nominal wages and prices are slow to adjust, domestic output can be kept closer to its potential level by allowing the Canadian dollar to depreciate, thus providing an offsetting stimulus to other export sectors of the economy. There may be more disequilibrium in the short run as a result of the exchange-rate flexibility, but over the long run the wage and price adjustment will take care of this. In the meantime, we will be spared the costs that would have accompanied the larger decline in aggregate output and employment.
A serious problem with this policy debate is the near-absence of empirics. From a theoretical point of view, professional economists more or less agree on the nature of the various costs and benefits, but there is mostly ignorance regarding their magnitude. For example, most theories that involve risk-averse agents predict that an increase in uncertainty will lead to less activity being undertaken, whether the activity in question is buying or selling goods or services, or investing in plant and equipment across an international boundary. It is surprising, therefore, to find such a lack of robust empirical support for the hypothesis that volatility in nominal exchange rates decreases trade flows. One possible explanation is that since exchange rates tend to be very volatile and trade flows much smoother, it should not be too surprising to find no clear evidence of a significant statistical relationship between them. This is similar to the difficulty economists face in pinning down a statistical relationship between investment and real interest rates, despite the very compelling theory that such a relationship exists. Another, more behavioural,explanation is that firms involved in international trade use the forward markets to hedge their currency risks, and thus exchange-rate volatility is successfully circumvented. (Of course, hedging costs money and effort that could be avoided with a fixed exchange rate.)
Economists’ ignorance regarding empirical magnitudes is just as pervasive when it comes to the shock-absorbing quality of flexible exchange rates. One can make a good argument that Canada’s aggregate economic performance following the Asian crisis was better than it otherwise would have been as a result of the 12 per cent depreciation of the Canadian dollar between 1997 and 1998. The export-oriented manufacturing sectors centred in Ontario and Quebec were given a boost that to some extent offset the decline in the resource sector caused by the 30 per cent decline in world commodity prices over the same period. But how much better was Canada’s economic performance because of the depreciation? Would GDP have grown only by two per cent in 1998 when in fact it grew by 3.3 per cent? Or would Canada have experienced a recession—an actual decline in GDP? No solid empirical work yet sheds light on this question.
Our ignorance about the empirical importance of these relationships makes the resolution of the exchange-rate debate understandably difficult. And we should probably not expect a consensus about these magnitudes anytime soon. To make things worse, the costs and benefits of the two policy alternatives are likely to change in the future. As Canada and the United States become more integrated economically, and trade in both goods and assets becomes even more important than it is today, the microeconomic gains from fixed exchange rates will increase. Moreover, if the two economies become more similar in their structure and pattern of production and trade—a hypothesis that admittedly is debatable—there will be less need for a flexible exchange rate to act as a shock absorber. Therefore, it may well be the case that over the next few decades the balance will eventually tip in favour of adopting fixed exchange rates. It is not a sure thing, but it is certainly possible.
There is a final issue worth highlighting. The macroeconomic benefit of a flexible exchange rate is tied to the slow adjustment of domestic wages and prices. As Friedman argued in his classic 1953 paper and says again in the debate with Mundell:
If internal prices were as flexible as exchange rates, it would make little economic difference whether adjustments were brought about by changes in exchange rates or by equivalent change in internal prices. But that condition is clearly not fulfilled … At least in the modern world, internal prices are highly inflexible.
For Friedman, internal price flexibility is clearly not the norm, and so there is a benefit from having flexible exchange rates act as a shock absorber.
Wage and price rigidity is the result of the economic behaviour of firms and consumers, and that behaviour is to some extent reflected by and entrenched in our economic institutions. But economic behaviour, in turn, is usually thought to depend on the institutional environment in which economic actors find themselves. This leads us to the question: If Canada fixed its exchange rate, would wages and prices become more flexible, thus reducing the need for flexible exchange rates as a means of easing macroeconomic adjustments?
Mundell recognises this “endogeneity” of wage and price flexibility when he comments on the likely success of the euro: “I believe that flexibility of individual prices will be fostered by the euro area and that, with exchange rate changes ruled out, policy makers will increasingly turn to deregulation and fewer controls.” For Mundell, price rigidity seems to exist mainly because of price controls and regulations. Less government interference would therefore significantly increase the degree of price flexibility. Friedman, in contrast, believes that price (and wage) inflexibility is more fundamental, more intrinsic to economic behaviour, and thus he is much more sceptical than Mundell that Europe’s common currency will lead to enough price flexibility as to make flexible exchange rates undesirable.
Knowledge about this relationship is crucial to a proper evaluation of the exchange-rate debate. If the adoption of fixed exchange rates does not significantly increase internal price flexibility, there is still a solid argument in favour of flexible exchange rates. But if adopting a fixed exchange rate can indeed be expected to change behaviour in a way that significantly increases wage and price flexibility, the macro benefits of exchange-rate flexibility largely disappear and we are left only with the micro benefits of fixed exchange rates.
What evidence could we examine in order to determine the likely effect of the adoption of fixed exchange rates on the degree of wage and price flexibility? Only a few natural experiments shed light on this relationship. The most important is probably the creation of the European Monetary System in 1979 that, over the subsequent 20 years, embodied several versions of fixed exchange rates between the major European countries. It is difficult to think of European labour-market flexibility without recalling “Eurosclerosis,” the term coined in the mid 1980s to describe how European unemployment rates remained stubbornly high after the recession in the early 1980s, especially in comparison with the United States. But the fact that Europe’s labour markets remained more rigid than North America’s is not especially relevant. What is relevant is whether European labour market rigidity declined in any significant manner after—and as a result of—the creation of the European Monetary System.
I wish I knew the answer to this question, but I know of no research that has examined this relationship. I am convinced, however, that it is a crucial part of the overall policy puzzle: If we are to resolve the exchange-rate debate, we have to know how the exchange rate regime impacts on domestic wage and price flexibility. Robert Mundell seems confident that countries with fixed exchange rates will soon come to have more flexible internal wages and prices. Milton Friedman is not so confident. The rest of the economics profession would do well to spend more of our research time examining this issue.
I certainly plan to.