Milton Friedman and Bob Mundell are widely—and rightly—regarded as the best monetary economists alive today, and all those who have read their debate will readily appreciate why. Economists are of course famous for their propensity to argue with each other, but what is particularly striking from the Friedman-Mundell debate is actually the extent to which they agree on certain points. And one of the most important points on which they agree is that the record of central banks has been very poor indeed. Over the last century the number of central banks has grown from 21 to nearly 200, and their power has increased enormously, and yet that same century—the century of central banking—also experienced historically unprecedented inflation. The effects of this inflation have been economically and socially catastrophic.
The rise of central banking was thus associated with unparalleled monetary instability, and we can only conclude that this association reflects cause and effect: that central banking leads to monetary instability. If monetary stability is important for the well being of the economy, as it surely is, and if central banking is incompatible with monetary stability, as the experience of the last century suggests, then we are forced to the conclusion that central banking must be abolished: to quote Professor Friedman’s elegant phrase, “money is far too serious to be left to central bankers.”
But what are the alternatives? One option is for Canada to abandon its own monetary policy and fix the loonie to the US dollar; alternatively, Canada could scrap its own dollar and adopt the US dollar as its currency. However, both these alternatives would make Canada dependent on the Federal Reserve instead of the Bank of Canada, and as such would merely replace one set of central bankers with another. For what it is worth, I tend to agree with Professor Mundell that the historical record suggests that Canada would be better served by the Fed than by the Bank of Canada, but Friedman is certainly right when he says that “the history of US monetary policy since the establishment of the Fed has had many more periods of poor than of good policy.” The Fed might have a better record than the Bank, but it has a poor record nonetheless. In any case, if money is too important to be left to central bankers, then replacing Canadian central bankers with American ones clearly does not solve the problem.
The same goes for another alternative—setting up a new currency unit jointly with the US and possibly other countries as well. For Canadians, this is not much different from adopting the US dollar outright, and would only make Canada dependent on the monetary policy of a new set of central bankers. A North American version of the euro is therefore not the answer.
I don’t think the European euro is much use either. The euro has a poor record and very uncertain future, and lacks credibility because it is issued by a badly designed issuer, the European Central Bank, under the dubious auspices of the European Union. The euro is also inferior to the US dollar or the deutschmark by any reasonable criterion. If the EU has to have a common currency—which of course it does not—then it would have been much better off adopting the mark as its common currency.
Then there is a gold standard. In theory, a gold standard offers the big attraction of a fully automatic monetary system—a monetary system that does not rely on Canadian or foreign central bankers to make good monetary policy decisions—but I agree with Professor Friedman that its drawbacks should rule it out. As he points out, its resource costs are a waste and the belief-system that was needed to underpin the historical gold standard no longer exists. In addition, its historical record is not that good and a new gold standard could make the Canadian price level quite unstable, because it would make Canadian prices dependent on demand and supply factors in the gold market, and these can be quite volatile. The gold standard has had its day, and we should let it rest in peace.
So what’s left? If central bankers really aren’t to be trusted to manage the currency, and they aren’t, then what we want is a fully automatic monetary system that does not depend on central bankers to manage it.
The objective of this (or any other) monetary system should be to maintain price-level stability, and the first step in designing a new monetary system should be to decide on an operational definition of what this objective actually means—a choice of price index to target, and the target itself. The most obvious price index would be the CPI, but in selecting our CPI target we would have to take account of the fact that most indices suffer from a “quality bias”—they fail to take full account of quality improvements in the goods and services they cover. In practice, the measured CPI tends to grow at an annual rate that is perhaps 1-2 per cent more than a “true” CPI that fully adjusts for quality improvements. This means that “true” price-level stability—what’s often called “zero inflation”—actually occurs at a measured CPI rate of inflation of 1-2 per cent a year. The objective of our monetary system should therefore be to ensure that inflation—as measured by the rate of change of the CPI—is as close to this range as possible.
Perhaps the best way to achieve this target is to convert the Bank of Canada into a currency board that issues currency against a reserve ratio of at least 100 per cent. The amount of central bank currency in circulation would then depend entirely on the public demand for it, and the board would have no scope for discretionary decision-making: The currency board system would be fully automatic.
However, the new Canadian Currency Board would differ from previous currency boards in one important respect. Instead of fixing the exchange rate of Canada’s currency against a particular foreign currency such as the dollar, it would peg its currency to the value of some appropriately chosen financial derivatives contract. This contract would be a futures or option contract written on the CPI itself. Several possible CPI-based contracts have been discussed in recent economic literature.
This new monetary system would produce price-level stability because it would rely on private-sector arbitrage to ensure that the amount of central-bank currency in circulation is consistent with the target rate of inflation. If inflation was expected to exceed the target, then private-sector arbitragers would expect to make profits by selling central-bank money back to the issuer, and the amount of such money in circulation would fall until private-sector inflation expectations had been brought back down. Similarly, if the inflation rate was expected to be too low, arbitragers would expect to profit by buying central bank currency, and their arbitrage operations would lead to a rise in the expected inflation rate. The system thus exhibits stabilizing or negative feedback, in which incipient departures of the expected inflation rate from the target rate are automatically corrected by private sector parties exploiting the arbitrage opportunities created by such departures.
This new system would therefore have the big attraction of the gold standard, its automaticity, whilst avoiding the drawbacks of the gold standard and ensuring a stable price level.
It is also important to ensure that there are no legal or regulatory restrictions for or against the use of particular currency units. Professor Mundell lays great emphasis on the notion of ideal (or optimal) currency areas, and with good reason, but the fact is that we have no real idea what these areas might actually be in practice. We can of course easily see that the border between the Canadian and US dollar currency areas is much the same as the political border between the two countries, but the reason for this is almost certainly not that the political border happens to coincide with the border between two ideal currency areas. The reason is surely obvious: Governments north of the Canadian-US border have promoted the use of the Canadian dollar as the basic currency unit, and governments south of the border have done the same for the US dollar.
It follows that we don’t really know where the border between the two currency areas should ideally be. Would the citizens of Windsor, Ontario be better off using US dollars instead of Canadian dollars? Maybe, but I don’t know and I doubt that anyone else does, either. For that matter, perhaps their neighbours in Detroit just across the border would be better off using Canadian dollars, but again no one knows for sure. All we can say with any confidence is that if we had ideal currency areas, it is most unlikely that people would be switching currencies just as they crossed the political border.
The best the Canadian government can do is to permit individuals the maximum possible freedom in their choice of currency unit, and then allow currency areas to evolve naturally in response to market forces. If currency areas subsequently change, as they presumably would, such changes should reflect underlying economic forces, and we might reasonably suppose that actual currency areas would then gradually move towards their ideal equivalents. Whatever the ideal currency areas might be, there is only one way to find out and achieve them in practice — to encourage full currency competition.
Another issue raised in the Friedman-Mundell debate is the “wild card,” as Friedman calls it, of the Internet and electronic currency. In this context, some recent research—in particular, by Ben Friedman at Harvard University, but also, if I may say so, by David Cronin from the Central Bank of Ireland and myself—suggests that the implications of electronic currency are likely to be more profound than is generally appreciated. The key point is that improvements in IT will undermine the demand for central-bank money, and in so doing create various problems for central banks. As electronic currency continues to improve, it will become an increasingly effective substitute for conventional cash issued by the central bank. The demand for the latter is therefore likely to fall significantly as people switch over to electronic currency. At the same time, increased competition and improvements in settlement technology are also likely to produce a drastic fall in the demand for the other component of central-bank money, commercial bank deposits at the central bank. Consequently, the demand for both types of central-bank money is likely to fall very considerably in the foreseeable future. Indeed, it is possible—and we would say, highly probable—that the demand for centralbank money will disappear altogether.
The erosion of demand for central-bank money will cause a number of major problems for central banks. To begin with, if central-bank money supplies are maintained at their current levels, then the falling demand for such money will inevitably produce inflation: If the demand for central-bank money falls and its supply remains the same, the market for central-bank money can only clear if the price level rises. If they are to avert such an outcome, central banks would have to reduce the supply of their money in line with the falling demand for it—which is a difficult and historically unprecedented task.
However, such a policy would impose potentially crippling financial costs on central banks. Since they would be buying their own currency back, their revenue from money creation—their “seigniorage”—would become negative, and they would suffer major losses. (Seigniorage occurs when governments exchange money they have printed for real goods and services.) If they are to avoid insolvency, many central banks would then need bailing out by their respective governments—a fitting irony, given that many central banks were first set up to raise seigniorage revenue for the governments they served. In turn, the need to finance central bank losses is bound to damage the much vaunted “independence” that many central banks currently enjoy, not to mention their moral authority when they next lecture private-sector institutions on the need to bear their own losses or go out of business.
A falling demand for central-bank money also has another adverse consequence. As the demand for central-bank money continued to fall, any given shock—such as a particular improvement in e-currency technology—would have an increasingly large effect on the price level. This means that the price level would become increasingly vulnerable to shocks, and therefore become increasingly volatile. As technological progress continued, there would come a point where discretionary monetary policy would no longer be able to maintain any reasonable degree of price-level stability. By that point, the central bank would have had to replace discretionary monetary policy with an appropriate rule to govern the issue of central-bank money— either that or face escalating monetary instability. This rule might be a gold standard, for example, or—better still—a price-index derivatives rule such as that suggested earlier. But whatever precise form it might take, the rule would put the issue of currency on a purely automatic basis and do away with monetary policy as we currently know it. In short, electronic currency will eventually make discretionary monetary policy unworkable. The days of discretionary fiat money appear to be numbered, whether we like it or not.
The 20th century saw a huge increase in the number of central banks and a vast expansion of their powers. Thanks to these banks’ licenses to print money, it also experienced historically unprecedented inflation. It is therefore not surprising that “all the great classical economists [without exception] … abhorred the idea of inconvertible currencies,” as Professor Mundell points out. The experience of central banking over the last century amply confirmed how right they were.
There is no good reason why Canada must continue to have a central bank. After all, Canada didn’t have one till 1935, and the Canadian monetary system before then was in many ways much better than it has been since. For most of its history, Canada effectively operated a system of free banking on a gold standard, and that system served the country much better than the inflation-prone fiat money regime that replaced it. Of course, defenders of the Bank of Canada will argue that the improved inflation record of the last decade shows that the Bank has now got its act together. However, there is a difference between a genuine teetotaler and an alcoholic who goes without a drink for a while, and the fact remains that if one takes a long-run historical perspective, the Bank’s record is still very poor. In any case, since the development of emoney still poses a major danger to the Bank’s future ability to manage the currency, the improvement in the Bank’s performance is unlikely to last.
Even central bankers are coming to realise that the continued existence of central banks should not be taken for granted. To quote a recent article in the Financial Times: “Central banks may be at the peak of their power [but] their extinction cannot be ruled out. Societies have managed without central banks in the past. They may well do so again in the future.”
The author of this article, Mervyn King, is in a good position to judge—he is the Deputy Governor of the Bank of England. At the end of the day, I suspect that it will be the combination of market forces and technological progress that will have the decisive say, and that the debate between free bankers and central bankers will ultimately be settled in the market place. If the defenders of central banking are not careful, the invisible hand of the market will sweep central banks into the dustbin of history—even as they continue to insist we cannot do without them.