One indication of the success of Canada's monetary policy is that inflation—the rate of change of consumer prices as reflected in the consumer price index (CPI)—is much less newsworthy today than it was during the 1970s, when it was often a headline issue. Chart 2 traces the path of Canadian inflation from 1966 to 2004, showing what are often referred to as the "twin peaks" of Canadian inflation, when inflation rose to over 14 per cent in 1973 and to almost 13 per cent in 1979. 1 The chart also shows the sharp decline in inflation during the early 1980s, from over 12 per cent to 4 per cent, the modest increase from 4 per cent to 6 per cent over the balance of the 1980s, and then the reduction from just over 6 per cent in 1990 to roughly 2 per cent over the following few years. Since 1992, the rate of CPI inflation in Canada has fluctuated around 2 per cent.

For many young Canadians today, who do not remember the high rates of inflation in the 1970s and early 1980s, it is difficult to appreciate why inflation matters; they have only experienced a world of low and stable inflation—inflation so low and so stable that most probably ignore it completely. Even for those older individuals who experienced first-hand some of the inflation-related economic disruptions in the 1970s and 1980s, inflation remains a bit of a puzzle; while they may have an instinctive appreciation of why inflation is undesirable, they find it difficult to be precise about why inflation is a problem. Why does inflation matter?

2.1 Inflation and uncertainty

For many years, the academic literature on inflation emphasized the distinction between anticipated and unanticipated inflation, stressing that it was only the latter that created problems. The argument held, for example, that if everyone anticipated inflation of 5 per cent in the coming year, all contracts could be modified with relative ease to incorporate 5 per cent higher wages or a 5 per cent higher interest rate. In such a setting, inflation would not present much of a concern.

Among central bankers, however, this distinction has little relevance, for two reasons. First, real-world institutions are very slow to incorporate the kind of adjustment to inflation that this academic argument requires (Friedman 1977). Second, and more important, the real world rarely, if ever, displays a stable and predictable rate of inflation. Actual inflation tends to be quite unstable, as Chart 2 shows in the case of Canada, and it tends to be even less stable when its average level is higher. High inflation is typically volatile and therefore difficult to predict, and this uncertainty generated by inflation is the real problem because it leads households and firms to make decisions that they would be unlikely to make in a more certain, low-inflation, environment (Stuber 2001).

Chart 3 shows, for the same collection of countries as in Chart 1, the relationship between a country's average rate of inflation and inflation's variability over time. 2 Countries with higher average inflation rates also tend to be countries with more volatile inflation rates. Canada appears at the lower left in the chart, with average annual inflation of 4.6 per cent and the standard deviation of inflation equal to 3.2 per cent. 3

Given the connection between inflation and uncertainty, the only effective way to avoid the uncertainty associated with inflation is to reduce inflation itself. By reducing the average rate of inflation, the associated volatility will be reduced and so, too, will be the costs of the uncertainty.

2.2 Why inflation uncertainty matters

In an economy where the vast majority of transactions are made in private, decentralized markets, and prices are determined by the interaction between buyers and sellers, market prices play a key role in transmitting information and guiding the economy's allocation of resources. By resource allocation economists refer to the overall pattern of production and consumption—which firms produce which goods, how much of various goods are produced, what technologies are employed to produce them, and how consumers divide their scarce purchasing power among the many available goods and services. In such an economy, market prices are a signal of the relative scarcity of individual products. Increases in individual prices are a signal that specific goods are becoming scarcer. Producers are then encouraged to increase their supplies of these goods, while consumers are encouraged to economize on their use, thus alleviating the scarcity. Conversely, decreases in individual prices are a signal that specific products are becoming more abundant. In this case, firms are led to reduce their supplies of these products, and consumers are encouraged to increase their demands. The information conveyed by prices in a market economy is therefore central to the market's ability to allocate resources in an efficient manner—leading to more production of those products that are highly valued and less production of those goods on which society places less value. 4

The presence of inflation in a market economy—and its associated uncertainty—means that prices no longer convey this valuable information so clearly, with the result that market outcomes lack the efficiency that would be achieved in a non-inflationary world. With inflation, market prices still contain information about scarcity, but they also convey other information that has little to do with scarcity—in particular, they carry information about perceptions of the overall inflation rate. The result is that, in the presence of inflation, buyers and sellers are never quite sure what a high price means: does it mean that a specific good is becoming scarcer, or does it simply mean that all prices are rising as part of a widespread inflation? If the first is true, buyers should economize and sellers should strive to sell more. But if the second is true, this specific price is not rising relative to most other prices, and thus buying and selling behaviour should not change. The problem in the real world is that, when inflation is present, both types of changes are occurring simultaneously, and it is very difficult for households and firms to figure out what is going on. Inflation causes market participants to make mistakes—transactions they would not have made if inflation had not been clouding the environment—with the result that market outcomes are not as efficient as they would be in the absence of inflation.

The argument that inflation reduces the efficiency of the price system has been made by many central bankers over the years. Thomas Melzer (1996), the former president of the St. Louis Federal Reserve, put it this way:

In my judgment, the number-one cost of inflation, even low inflation, is unnecessary confusion in the relative price system. It is important for the sake of efficient functioning of markets to let relative prices give the clearest possible signals in terms of credit markets, commodity markets, labor markets and in general any market where today's prices depend on perceptions of tomorrow's value of money.

We referred earlier to the idea that inflation alters the allocation of resources and reduces the efficiency of the economy. But what does this really mean? One example of a misallocation of resources caused by inflation is the shift in resources towards socially wasteful activities and away from producing "real things" that we care about. Axel Leijonhufvud (1977, 280-81) makes the point when he says:

[In the presence of inflation,] being efficient and competitive at the production and distribution of "real" goods and services becomes less important to the real outcome of socio-economic activity. Forecasting inflation and coping with its consequences becomes more important. People will reallocate their effort and ingenuity accordingly. . . .

In short, being good at "real" productive activities—being competitive in the ordinary sense—no longer has the same priority. Playing the inflation right is vital.

Imagine all the "real things" that could be done by all the clever people who, in a world of even moderate inflation, are focused instead on forecasting inflation, creating and trading financial instruments designed to retain their value in an inflationary world, and modifying or interpreting a tax system affected in significant ways by inflation. In a world of low and stable inflation, these valuable resources can instead be used to do the things we really care about—teach piano lessons, drill for oil, design computer programs, or write books.

A world without inflation does not mean a world of complete certainty or unchanging relative prices. Far from it. Relative prices are constantly adjusting—sometimes suddenly—in a world in which changes in consumer tastes and the development of new technologies are constantly occurring. But this volatility is unavoidable and is standard fare in market economies. In such a dynamic world, firms and households have a difficult enough time making their best decisions regarding the allocation of scarce resources. The problem with inflation is that it makes what is already a confusing world even more difficult to understand. Monetary policy aimed at maintaining low and stable inflation can make a real contribution to our quality of life by making the decision-making environment clearer for everyone.

Nothing has been said so far about whether low inflation contributes to higher long-run rates of real economic growth. The empirical evidence on the relationship between inflation and long-run growth is quite sensitive to changes in sampling details, and thus, economists are currently unable to provide a compelling case for the existence of any such relationship. See the Appendix for a brief discussion.

  1. Several measures of inflation are commonly used. The one in Chart 2 is the rate of change in the total CPI. The Bank of Canada also emphasizes the path of "core" inflation, which strips out the eight most volatile elements of the CPI (fruit, vegetables, gasoline, fuel oil, natural gas, intercity transportation, tobacco, and mortgage-interest costs). But for our purposes here, either measure is suitable. []
  2. The measure of volatility is the standard deviation of inflation. We have removed Sudan, Chile, Mexico, and Israel from the sample. Given their very high inflation and volatility, their inclusion in the same chart would make it difficult to distinguish among the other countries. []
  3. Again, correlation does not imply causality. See Longworth (2002) for a discussion of other measures of inflation uncertainty in Canada. []
  4. For a classic and elegant discussion of the importance of information to the operation of the price system, see Hayek (1945), especially pp. 524-28. []