Archived — Working Paper Number 32: The Canada-U.S. Productivity Growth Paradox

by Serge Coulombe, University of Ottawa, March 2000


Productivity data on the business sector, which covers around 75 percent of the economy, provide important information on the evolution of living standards.

The data on multifactor productivity (MFP) growth and labour productivity growth produced by the official statistical agencies in Canada (Statistics Canada) and the United States (Bureau of Labor Statistics, or BLS) send mixed signals regarding the comparative evolution of living standards in the two countries. Since the early 1980s, Statistics Canada's MFP growth measures for the business sector indicate that the Canadian economy has outperformed the US economy while labour productivity data produce a reverse picture. This is the so-called Canada-U.S. Productivity Paradox.

In this study, we investigate the productivity paradox with an analysis of Canadian and U.S. business sector productivity data since 1961. The main finding of our analysis is that business sector MFP growth estimates provided by Statistics Canada in March 1999 are neither consistent over time nor comparable to U.S. estimates.

The analysis identifies three significant problems with the methodology used by Statistics Canada for estimating MFP growth. First, Statistics Canada's labour force index is biased. The agency appears to significantly overestimate the contribution of the changes in the labour force composition in the 1960s compared with those in the 1980s and the 1990s. MFP growth in the 1960s is therefore underestimated compared with the 1980s and the 1990s.

Second, the concept of capital used by Statistics Canada appears too narrow for MFP growth measurement. By excluding land and inventories, which seem to grow at a substantially slower pace than the other components of the capital stock, the Canadian statistical agency tends to overestimate the contribution of capital accumulation to output growth.

Third, Statistics Canada appears to systematically underestimate the transitory growth rate and the level of the capital stock in Canada. This underestimation of the capital stock stems from the methodology used by the Canadian statistical agency to account for depreciation. The bias engendered by the underestimation of the growth in the capital stock (third problem) more than offsets the bias generated by the narrow definition of the capital stock (second problem). As a result, Statistics Canada overestimates MFP growth by approximately a quarter of a percentage point annually.

Our conclusion is that Statistics Canada should thoroughly revise its methodology for estimating the capital stock and for measuring changes in labour force composition. The paper proposes methodological changes to address these problems.

Date modified: