Sensitivity of Capital Stock and Multifactor Productivity Estimates to Depreciation Assumptions: A Canada-U.S. Comparison

Article excerpt

THE UNITED STATES IS CANADA'S dominant trading partner and competes head to head for foreign direct investment (FDI), innovation activities, and skilled labour within the continent as well from outside of the region. Hence, Canada needs to be highly competitive in terms of productivity, costs, and business and market framework policies and programs.

In this context, there has been a great deal of research and policy discussion over the past 20 years about Canada's relative productivity performance vis-a-vis the United States. Estimates suggest that Canada's labour productivity level is considerably below the U.S. level and that the gap has widened since 2000 (Lee and Tang, 2000; Rao et al., 2004, 2008; Baldwin, Gu and Yan, 2008). The research also suggests that the Canada-U.S. labour productivity gap is broadly-based across major Canadian industries, and is the result of a multifactor productivity (MFP) (innovation) gap.

Comparable and consistent estimates of capital stock by industry are needed for accurate Canada-U.S. MFP comparisons and to understand the reasons behind Canada's relative MFP performance on an industry basis. (2) Current Statistics Canada geometric depreciation rates, which are key parameters underlying capital stock estimates, are in general higher than those used by the Bureau of Economic Analysis (BEA) for the United States (Table 1) (3), particularly for building and engineering structures. (4) Given similar economic organization and production in the two countries, the large variation in certain asset depreciation rates between the two countries cannot be explained by country specifics such as differences in climate. (5)

If we use the "official" capital stock data from Statistics Canada and the BEA for comparing MFP levels between the two countries, we would underestimate Canada's capital intensity relative to the U.S. level, which in turn would overestimate Canada's relative MFP levels.

The primary objective of this article is to develop capital stock estimates for Canadian and U.S. industries using the same depreciation rates for the two countries. In particular, we address the following three policy research questions or issues:

* Should current official capital stock data be used for Canada-U.S. industry capital intensity and MFP comparisons?

* How are the Canada-U.S. industry capital intensity comparisons affected when we use either Canadian or U.S. depreciation rates for estimating capital stock data in both countries?

* How do Canadian industries perform relative to their U.S. counterparts in terms of MFP (growth rates and levels) when consistent estimates of capital stock data are used?

Note that this article does not make any attempt to justify favoring either Statistics Canada or BEA depreciation rates for estimating capital stock series. This is beyond the scope of our research. Instead, it simply compares Canada's performance in capital intensity as well as multifactor productivity when either Statistics Canada or BEA depreciation rates are used for both countries.

In the first section, we use either Statistics Canada or BEA depreciation rates in estimating capital stock at the industry level in both Canada and the United States, and then compare them to the official estimates. In sections 2 and 3, we estimate Canada's MFP growth and levels at the industry level, and compare the results with the United States. The final section summarizes the key findings and discusses their research nd policy implications.

Estimating Capital Stock

In this section, we compare and discuss capital stock estimates based on Statistics Canada and BEA geometric depreciation rates. At the outset, it should be noted that only non-residential machinery and equipment (M&E) capital and structures capital (building and engineering construction) are included. owner-occupied dwellings, inventories, and non-depreciable land are excluded from the estimates. …