Opening Statement before the Senate Standing Committee on National Finance

Good evening. Tiff and I are pleased to be here with you today to discuss retail price differentials between Canada and the United States.

The mandate of the Bank of Canada is to enhance the well-being of Canadians by contributing to sustained economic growth. The single, most direct contribution that monetary policy can make to sound economic performance is to provide Canadians with confidence that their money will retain its purchasing power. In Canada, this is achieved through an inflation-control targeting framework.

Inflation-control targeting has been a cornerstone of monetary policy in Canada since its introduction in 1991. Since 1995, the target range has been 1 to 3 per cent, with the Bank’s monetary policy aimed at keeping inflation at the 2 per cent target midpoint.

It is important to stress that the target is for total CPI inflation, which is the best measure of the overall cost of living for Canadians. We do not target or try to control prices of individual goods and services. With 2 per cent inflation, there will always be some prices rising more quickly and others more slowly. Central to a market-based economy is the role of relative price signals in allocating resources.

To understand the evolution of CPI inflation, the Bank looks at the forces shaping various prices, one of which is consumer price differentials between Canada and the United States. We also examine the role played by the exchange rate in price gaps, about which I will speak in a moment. We look at both of these, and the factors that drive them, to understand how prices are set in Canada which, in turn, contributes to our analysis of the inflation expectations of Canadian households and businesses and, ultimately, to our ability to fulfill the Bank’s mandate.

At the Bank, we use two sources of information to estimate the size and evolution of retail price differentials between Canada and the United States. The first is the consumer price index information from Statistics Canada and the U.S. Bureau of Labor Statistics. In particular, we examine the ratio of the Canadian CPI to the U.S. CPI for three price categories: autos, other durables and clothing. These ratios provide information about the evolution of prices in Canada relative to prices in the United States. On their own, however, they do not tell us the absolute dollar amount of price gaps between the two countries.

To assess the price-gap level, we use information about the prices of specific goods in both countries gathered from an informal Internet survey conducted by Bank of Canada staff. We record the prices of a list of identical goods in Canada and the United States at the same retail outlet. U.S. dollar prices are converted into Canadian dollars using the bilateral exchange rate for that day.

Using these two sources of information, the Bank of Canada’s best estimate is that the Canada-U.S. retail consumer-goods price differential was 11 per cent in September, down from 18 per cent in April. In other words, prices of a set of comparable goods in Canada are an average of 11 per cent higher than the prices of these same goods in the United States. I caution that these are estimates and there is some uncertainty around them.

An important question is why has the gap not narrowed more for similar goods, given the increase in the value of the Canadian dollar. For your consideration, allow me to discuss some of the possible factors, both macroeconomic and structural, that influence the prices of goods and services in both countries.

First, cyclical issues may be responsible for some differences between Canadian and U.S. prices. Unexpected economic weakness in one country could lead to an undesirable buildup of inventories and result in local discounting of prices. In addition, as long as markets are partially segmented across international borders, market power may enable firms to respond to local demand conditions—possibly resulting in lower prices in areas with weaker demand.

Secondly, from a macro perspective, prices take time to adjust. Importantly, this is not only true in response to the exchange rate, but also to anything that changes the economic environment (and this is why there are lags in the effect of monetary policy).

In addition to the macro elements, there are structural factors, including profit margins and underlying costs, that influence the prices of goods and services and that can also contribute to the gap in prices in Canada and the United States.

Operating profit margins for both the wholesale and retail sectors will have an impact on the prices of consumer goods in retail outlets. One possible reason that margins have remained elevated in Canada may be that the retail environment is more concentrated in Canada than in the United States. In Canada, the top four retailers have a 28 per cent market share, compared with only 12 per cent in the United States.

On the cost side, there are a number of factors that are important for understanding the differences in prices in Canada and the United States.

Labour costs tend to be higher in Canada than in the United States. For instance, in the retail sector, total compensation per hour is higher in Canada. Despite this, retail- and wholesale-sector employment is higher as a share of total employment in Canada, even though retail and wholesale trade in Canada each represent a smaller share of real GDP. The latter reflects Canada’s productivity performance in the retail sector, which has lagged that of the United States. This is partly a result of a much lower level of capital intensity and lower total factor productivity growth.

Total factor productivity refers to how labour and capital are combined to create efficiency gains. Here, industry structure appears to matter, with larger productivity gains from investment at chain stores than individual stores. Following 1995, productivity surged in the United States, with investment in new equipment and organizational change. In Canada, these effects do not seem to have been experienced to the same degree, perhaps due to the more localized nature of Canadian markets.

Transportation is another factor on the cost side. While fuel costs are only a portion of total transportation costs, the difference in gasoline taxes between the two countries makes gas more expensive in Canada. Lower population density in Canada may also contribute to higher transportation costs on a per-item basis.

All of these factors and other domestic value-added services, such as marketing, are priced in local currency. Thus, when the value of the exchange rate moves, these costs to Canadian retailers, which must be paid in Canadian dollars, do not change.

For imported goods, reflecting the smaller size of the Canadian market, the price paid by a Canadian importer of a good may be higher than that paid by an American importer. The Canadian market is roughly one tenth the size of the U.S. market, suggesting a much smaller potential for economies of scale. While this may not be as much of an issue for large retail chains with presence in both markets, smaller mostly domestic retail firms could be affected.

Finally, given all of these factors just discussed, allow me to turn to the role played by the exchange rate in price differentials between Canada and the United States.

As mentioned earlier, consumer prices tend to be sticky—this is true on both sides of the border. More broadly, the segmentation of international markets, in part due to the higher costs of shopping abroad, creates limits to competitive pressures.

In addition, paying attention and adjusting to fluctuations in the exchange rate requires time and energy. It is costly. When fluctuations in the exchange rate are relatively small, not all firms and consumers tend to pay close attention. It makes sense to focus attention where it matters most. That is one reason why prices are only adjusted over time and price gaps persist.

However, when the movements in the exchange rate are particularly large, they attract more attention and the adjustments are much faster, which we saw in the fall of 2007.

The most useful illustration of these phenomena is the behaviour of online book prices. These are often completely homogeneous goods on both side of the border. The cost of changing prices for an online retailer should be minimal. In addition, it is difficult to think of an easier case for consumers to take advantage of price differentials: comparing prices is just a matter of visiting different websites, there is no constraint or tariff involved when ordering from abroad, and transportation costs are known. Despite all this, while online book prices do change frequently, they typically do not respond quickly to fluctuations in the exchange rate. However, in the fall of 2007, they did adjust to the strong Canadian-dollar appreciation.

Typically, exchange rate movements tend to be reflected quickly in prices for only a narrow range of goods that are homogeneous in nature, e.g. fruits, vegetables, gasoline and meat prices. For other goods and services, there is no apparent short-run pass-through. The greater the value added in Canada to a good or service, the smaller the role played by the exchange rate in its price. However, estimates of pass-through at a more aggregated level, such as core CPI or total CPI, are quite low. Estimated pass-through is about 3 per cent for core inflation and 4 per cent for total inflation. To elaborate further, this means that a 10 per cent rise in the dollar would be expected to lower the level of total CPI by 0.4 per cent.

Similar to the experience of many countries, exchange rate pass-through has declined in Canada in the last 20 years. Some possible explanations include:

• better-anchored inflation expectations associated with the conduct of monetary policy and increased credibility, and
• the changing composition of trade (e.g. switching to different goods as relative prices change).

I trust these comments provide some insights to the underlying dynamics of Canadian-U.S. price differentials.

With that, Tiff and I would be pleased to take your questions.

See Appendix.