In 1960, the Conference Board of Canada launched its series of quarterly surveys of consumer attitudes. Since then, the results of these surveys have provided policy makers, economic forecasters, and business managers with useful information on future economic trends. The increasing demand for the survey results both by the public and the business sectors as well as the expanding coverage given to them in the media appear to confirm their popularity and usefulness.
To better gauge consumer sentiment, the Conference Board of Canada also designed the Index of Consumer Attitudes (ICA) for the entire nation as well as for the five Canadian regions (the Atlantic Provinces, Quebec, Ontario, the Prairies, and British Columbia). (1) The ICA is derived from answers to four attitudinal questions contained in the survey of consumer attitudes. Because two of these four questions relate to the respondent family's economic prospects over the next six months and to the employment prospects for the community over the next six months, the index is widely thought to reflect consumer expectations with respect to future economic conditions, which in turn would influence consumption behavior.
A number of empirical studies have investigated the effectiveness of indices of consumer confidence as leading indicators of consumer spending in Canada. An early paper by Shapiro and Angevine (1969) finds that the composite index of consumer mood derived from the Financial Post's Quarterly Survey of Consumer Buying Intentions is a very useful tool for predicting consumer spending on durable goods. (2) In a follow-up study, Angevine (1974) also finds that measures of consumer mood based on the Financial Post's Survey are statistically significant explanatory variables for three consumption categories: nondurable goods, semidurable goods, and services. Rayfuse (1982) reports that consumer sentiment as captured by the ICA tends to lead movements in consumer expenditures on durable goods by one quarter. (3) More recently, Cote and Johnson (1998) examine whether the ICA improves the fit of a consumption equation that includes other macroeconomic variables. They find that inclusion of the ICA improves the explained variation in growth in consumer expenditures per capita by 18 percentage points. (4)
The purpose of our article is to provide a fresh empirical perspective on the effect of consumer confidence on household spending in Canada. Although past studies confine their attention to national data, the present investigation examines the ability of the ICA as well as the individual attitudinal questions used in the survey to forecast Canadian household spending not only nationwide but also at a regional level. To our knowledge, our study represents the first endeavor in the literature to assess the usefulness of regional consumer confidence indices in forecasting household spending. Our work also employs significantly extended time series for the various components of consumer expenditure used in previous Canadian studies.
Employing the results of the Conference Board of Canada's survey of consumer attitudes for the sample period 1979Q4-2001Q4, our empirical findings indicate that at the national level, the ICA is a reliable predictor of total personal consumption expenditures as well as of various subcategories of consumer spending in the presence of control variables. On the other hand, our regional results are quite diverse, suggesting that economic fore casters and government policy makers should be cautious about using regional confidence indicators to predict consumption growth.
The remainder of this article is structured as follows. Section II describes the four attitudinal questions used in the construction of the ICA at both the national and the regional levels. Section III discusses the econometric methodology and the data employed in our empirical analysis. Section IV reports the correlations between growth in total personal consumption expenditures (PCE) and the contemporaneous ICA, as well as the different lags of the ICA. …