FACTORS IN CANADA-UNITED STATES REAL INCOME DIFFERENCES

This paper is a part of a larger study of economic growth in Canada, following the methods developed by Edward Denison in his book The Sources of Economic Growth in the United States and the publication Why Growth Rates Differ. The new material in this paper relates to Canada and the Canadian/U.S. comparison, while the material on Northwest Europe is drawn from the Brookings study. The present paper sets out the results to date on the differences in real output per employed person between Canada and the United States for one year, 1960. At this stage in our research the results indicate that the level of real output per employed person in Canada was about 20 per cent lower than in the United States in that year. On the basis of historical output data, it would appear that this margin of difference in Canadian/U.S. product levels has persisted throughout the present century. The central part of this paper examines the significance of differences in factor inputs in Canada and the United States and their contribution to the difference in income. The level of inputs per employed person in Canada accounts for only about 2 percentage points of the income difference between Canada and the United States. These results indicate that the overwhelming part of the difference in output per employed person between the two countries reflects the differences in output in relation to total factor inputs, rather than the magnitude of other factor inputs used in combination with labour. This result is consistent with earlier studies by Denison and others which have indicated the crucial importance of output in relation to total factor inputs, both in output growth over time and intercountry comparisons of output level. The body of the paper can give only brief attention to the numerous conceptual and statistical questions that arise in such a wide-ranging study, and the authors do not pretend to have tackled, let alone resolved, all of the wide range of problems related to this study. Nor do they claim any high degree of precision for the results, especially in the light of the statistical limitations of the basic data.