 | | Alan Arcand
Principal Economist
Centre for Municipal Studies |
There has been much hand-wringing in Canada over the country’s woeful productivity performance over the last 25 years or so. This concern is warranted because productivity growth is vitally important to raising living standards and enhancing the competitiveness of firms.
The topic of productivity can sometimes sow confusion, not the least of which because there is more than one measure of productivity. One of the more popular measures is multifactor productivity (MFP), sometimes referred to as total factor productivity. MFP accounts for increases in output not explained by increases in the other inputs to production—capital and labour. A gain in MFP results from innovation in some form— technological progress, organizational change, and exploiting scale economies.
A numerical example will help to illustrate the concept of MFP. Suppose that the Canadian economy expands by 3 per cent in a year, while all inputs to production grow by 2 per cent. This means that MFP growth would have added 1 per cent to overall growth. In terms of measurement by economists, MFP is derived from a growth accounting exercise. MFP is calculated as a residual; it is the part of the GDP growth that is left over when the growth contributions of labour and capital inputs have been deducted.
How poor has Canada’s productivity performance, and specifically growth in MFP, been? According to Statistics Canada, MFP actually fell by an average of 0.04 per cent per year from 1988 to 2011. In comparison, MFP in the United States increased by 0.96 per cent annually over the same time frame. Unfortunately, the disparity in productivity growth between the two countries, and specifically growth in MFP, has only widened over the past ten years. MFP growth in Canada fell by 0.45 per cent annually from 2001 to 2011, while it increased by 1.06 per cent per year in the United States. This is a difference of 1.5 percentage points per year! As a result of this wide and growing gap, the competitiveness of Canadian firms vis-à-vis their American counterparts has been eroding year by year.
But a new economic research paper by W. Erwin Diewert and Emily Yu claims that Statistics Canada has been underestimating MFP growth for years, raising hopes that Canada’s productivity performance has been stronger than Statistics Canada suggests. By building new estimates for GDP growth, capital input growth and labour input growth, they were able to deduce new estimates of MFP growth. In fact, Diewert and Yu find that MFP growth in Canada averaged 1.03 per cent per year from 1961 to 2011. This compares with Statistics Canada’s average annual MFP growth estimate of 0.28 per cent per year over the same period, a significant difference of 0.75 basis points per year. If Diewert and Yu’s calculations are indeed closer to reality than those of Statistics Canada, it would help to explain why economists and other analysts have been left scratching their heads over the apparent failure of many productivity-improvement policy changes that have been enacted by the federal and provincial governments over the years.
Their detailed estimates of GDP growth and labour input growth largely aligned with those of Statistics Canada. But their calculation of capital input growth came in much lower than that of Canada’s national statistical agency. Recall that MFP growth is calculated as a residual; it is the difference between GDP growth and capital and labour input growth. If GDP growth and labour input growth are nearly identical using both methodologies, but capital input growth is lower in Diewert and Yu’s calculations, then it follows that they find MFP growth to be stronger than through Statistics Canada’s calculations.
Statistics Canada has quickly jumped to the defence of the methodology it uses to compute productivity data, particularly capital input growth—the main source of the discrepancy. The most compelling part of the rebuttal is that Statistics Canada follows international guidelines and practices adopted by other statistical agencies when computing productivity statistics. Why is this important? By adopting international standards, Statistics Canada can ensure that its estimates are comparable to those produced by other statistical agencies, particularly the U.S. Bureau of Labour Statistics.
Comparability of analysis among countries and specifically between Canada and the U.S. is important, since in many ways, the relative productivity performance of Canada is more important than its absolute performance. As of now, we do not know whether productivity growth in the United States would also be boosted by Diewert and Yu’s alternative productivity growth calculation. That is a limitation of their paper, but also an avenue for future research. Economics, like science in general, moves forward through argument, discussion and debate—the essence of the peer review system. Indeed, these types of debates are part of a healthy scientific process. But we should not let them distract us from what is most important—that more still needs to be done to boost Canada’s productivity performance.
References
Diewert, W. Erwin and Emily Yu, “New Estimates of Real Income and Multifactor Productivity Growth for the Canadian Business Sector, 1961-2011." International Productivity Monitor 24 (2012): 27-48.
Gu, Wulong, “Estimating Capital Input for Measuring Business Sector Multifactor Productivity Growth in Canada: Response to Diewert and Yu." International Productivity Monitor 24 (2012): 49-62.
Related Research