Canada: Selected Issues
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
Close

This Selected Issues paper reviews Canada’s business tax system, looking at the incentive effects of the country’s business tax regime and their implications for output and employment. It presents estimates of marginal effective tax rates on corporate-source income in Canada and comparator countries across sectors, asset classes, means of finance, and asset ownership. The paper also examines labor markets in Canada. It notes that unemployment rates in Canada have risen across all demographic groups, industries, and regions, although young and less-educated workers and workers in agriculture and primary industries have been most severely affected.

Abstract

This Selected Issues paper reviews Canada’s business tax system, looking at the incentive effects of the country’s business tax regime and their implications for output and employment. It presents estimates of marginal effective tax rates on corporate-source income in Canada and comparator countries across sectors, asset classes, means of finance, and asset ownership. The paper also examines labor markets in Canada. It notes that unemployment rates in Canada have risen across all demographic groups, industries, and regions, although young and less-educated workers and workers in agriculture and primary industries have been most severely affected.

V. Measurement Bias in the CPI in Canada

1. The consumer price index (CPI) is commonly used as a measure of inflation and is often used to adjust tax rates, government spending, and private sector wage contracts for changes in the cost of living. Given the CPI’s central importance in this regard, considerable attention recently has been directed at gauging its accuracy in measuring the cost of living. This paper presents an overview of research in this area and a comparison of the potential biases in the CPIs for Canada and the United States.

2. The CPI in Canada measures the cost to households of purchasing a representative basket of goods and services. The composition and the relative weights of the goods and services in this representative basket are revised every four years using information on consumer expenditures from the Family Expenditure Surveys. Prices for the goods and services in the representative basket are derived by pricing 600 items at various retail outlets across Canada.

3. By pricing a fixed-weighted basket of goods and services, the CPI is able to isolate the impact of price changes. However, the use of fixed weights in the index means that the CPI over time is likely to become less representative of the “true” cost of living as individuals respond to price changes and new choices. Differences between the CPI and the cost of living, typically referred to as measurement biases, arise from several sources. Major sources of bias include: (1) commodity substitution bias; (2) outlet substitution bias; (3) quality bias; and (4) new goods bias.

A. Commodity Substitution Bias

4. Commodity substitution bias arises because, as relative prices change, consumers substitute cheaper goods and services for more expensive ones in their purchases. Thus, the commodity composition of actual purchases of goods and services may differ significantly from their relative shares in the basket priced by the CPI.1 For Canada, Généreux (1983) estimates that the average annual commodity substitution bias between 1957 and 1978 was about 0.2 percent. Fortin (1990) and Crawford (1993) have argued that this bias has declined over time as a result of more frequent CPI basket revisions and less dramatic changes in relative prices. For the United States, the Advisory Commission (1996) estimates that the substitution bias is about 0.4 percent annually, roughly comparable to Lebow et al.’s (1994) estimate in the range of 0.4 to 0.5 percent.

B. Outlet Substitution Bias

5. As prices change, consumers tend to shift their purchases from one type of retail outlet to another.2 At the same time, innovations in retailing will lead to the creation of new types of retail outlets that will not be covered (at least initially) in the CPI’s sample of retailers.3 Analysts have derived estimates for this bias by calculating the shift in the market share toward low-price outlets and the percentage price differential between high and low price outlets. Crawford (1993) estimates this bias in Canada at 0.1 percent annually, based on an annual shift of 2 percentage points in the market share of lower-priced retailers and assuming a price differential of 10 percent between higher- and lower-priced outlets. This estimate is comparable to that of Lebow et al. (1994) for the United States.

C. Quality Bias

6. The quality of the goods and services in the market basket may change over time. To the extent that the CPI fails to pick up these changes in quality, the index will not reflect the “true” change in prices. This source of bias may have the effect of either overstating or understating changes in the cost of living. Gordon (1990) estimates that the positive bias in the CPI for durable goods in the United States was 1 percentage point per year over the 1973–83 period.4 Crawford (1993) assumes that all of the quality bias for Canada occurs in consumer durables and estimates its average annual magnitude at about 0.2 percent. For the United States, Lebow et al. (1994) identify those categories where year-to-year quality adjustment difficulties appear to be most acute, and they attribute Gordon’s estimate of the bias in consumer durables to these categories. These assumptions yield an upper bound for the bias at 0.3 percent per year.

D. New Goods Bias

7. Over time, items included in the market basket may cease to be produced and/or new products are introduced. Matching discontinued items in the market basket with new products replacing them can introduce a bias, since the two products are not likely to be identical. In addition, new products generally are brought into the CPI sample in a way that guarantees that their prices have no effect on the level of the index at the time they are introduced. Given that the prices of some types of new goods (especially large ticket items, such as consumer electronics) may fall shortly after their introduction to the market, the CPI could overstate changes in the cost of living, since the CPI market basket may not be updated to include these new products until after these price declines have taken place. In attempting to measure this bias, analysts have tended to take a fairly narrow view of the share of spending on new goods in total consumer expenditure. In particular, Crawford (1993) argues that new goods are limited to household appliances and electronic equipment and therefore his estimate of the new goods bias in Canada at 0.1 percent annually is at the low end of the range. For the United States, Lebow et al. (1994) find that the new goods bias has an upper limit of 0.5 percent.

E. Overall Bias

8. It is believed that the separate biases have a cumulative effect on the overall CPI. Thus, aggregating the estimates of the various types of biases yields a lower average range for the estimate of an overall bias in the CPI for Canada than for the United States. The average overall bias in Canada’s CPI is estimated to be in the range of 0.5–1 percent per year, whereas the average range for the bias in the U.S. CPI is estimated to be 0.6–1.5 percent per year (tabulation below). A lower aggregate bias in the CPI would be expected in Canada than in the United States because the CPI basket of goods and services is revised more frequently in Canada and Statistics Canada has taken steps to reduce lower-level substitution bias.

Magnitude of Bias in the Change in the CPI

(Percent per year)

article image

Plausible range of estimates.

References

  • Advisory Commission to Study the Consumer Price Index 1996, “Toward a More Accurate Measure of the Cost of Living: Final Report to the Senate Finance Committee,”

    • Search Google Scholar
    • Export Citation
  • Backa, L., 1996, “Measurement Bias in the Consumer Price Index,” (unpublished; Ottawa: Department of Finance).

  • Crawford, A., 1993, “Measurement Biases in the Canadian CPI,” Technical Report No. 64 (Ottawa: Bank of Canada).

  • Diewert, E., Axiomatic and Economic Approaches to Elementary Price Indexes,” Working Paper 95–01 (University of British Columbia Department of Economics).

    • Search Google Scholar
    • Export Citation
  • Fortin, P., 1990, “Do We Measure Inflation Correctly?,” Zero Inflation: The Goal of Price Stability ed. by Lipsey R. (Toronto: C.D. Howe Institute).

    • Search Google Scholar
    • Export Citation
  • Genereux, P., 1983, “Impact of the Choice of Formulae on the Canadian Consumer Price Index,” proceedings of a conference of Statistics Canada “Price Level Measurement.”

    • Search Google Scholar
    • Export Citation
  • Gordon, R., 1990, “The Measurement of Durable Goods Prices,” (Chicago: University of Chicago Press).

  • Lebow, D., and J. Roberts D. Stockton, 1994, “Monetary Policy and the Price Level,” (unpublished; Washington: Federal Reserve Board).

    • Search Google Scholar
    • Export Citation
  • Shapiro, M. and D. Wilcox, 1996, “Mismeasurement in the Consumer Price Index: An Evaluation,” Working Paper No. 5590 (Cambridge Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
1

The substitution bias often is separated into upper and lower level substitution bias, reflecting the way in which the CPI is derived in practice. Upper level bias refers to the bias arising from substitution by consumers between the major spending categories in the CPI. Lower level substitution bias arises from substitution among the goods and services that comprise spending categories at a more disaggregated level. In January 1995, Statistics Canada began calculating its lower level price indices using the geometric mean in order to minimize lower level substitution bias.

2

For example, in recent years, there has been a significant shift in retail purchases in Canada and the United States away from traditional higher-priced outlets (such as department stores) to lower-priced ones (such as discount stores and mail-order catalogues).

3

The creation in Canada and the United States of so called “warehouse” stores that sell items in bulk at low margins over wholesale prices and offer limited customer services is a good example.

4

For example, innovations in durable goods generally are assumed to have improved the quality of these commodities and is not adequately captured in the CPI, thus overstating the rate of price increase. In contrast, a service like rent may result in a bias that understates the cost of living, because the index is obtained by tracking a progressively older rental stock and may fail to correct for declining quality which comes with increasing age.

  • Collapse
  • Expand