1 Introduction

View the most recent version.

The central role of capital and labour in explaining the process of economic growth has never been in doubt; the debate has centered rather on how capital and labour, separately or in combination, contribute to economic growth. Having internationally comparable measures of capital and labour allows researchers to examine the underlying economic process and to choose the analytical model that best fits the empirical evidence. Furthermore, once the specific model is chosen, a detailed study of the data can reveal areas where one country may be underperforming others, thereby allowing for the identification of potential policy prescriptions that will improve economic performance. However, in the absence of accurate and internationally comparable capital and labour data, suggestions for improvements in productivity are problematic.

Much work has recently been done in producing comparable estimates of labour inputs for both Canada and the United States. Recent work by Baldwin et al. (2005) has found that headline labour inputs for Canada and the United States are not comparable. When the necessary adjustments are made to improve the comparability between the two measures, the level of hours worked in the United States increases considerably. As a result, Baldwin et al. (2005) find that the magnitude of the labour productivity level gap between Canada and the United States is smaller than widely believed.

While much work has been done to improve the comparability of capital stocks between the two countries, measurement issues still remain, as capital stocks can vary based on the underlying assumptions of service lives and asset-specific decay patterns. Blades (1983) pointed to the need to be cautious about different assumptions being employed across countries as to the length of life of capital assets. Baldwin and Gorecki (1986) pointed out the need to use similar methodologies when comparing Canadian and U.S. capital stocks¹—a message that seems to have been forgotten in the ensuing years.²

While differences in methodology may not impact greatly on estimates of growth in multifactor productivity (MFP), they can have a significant impact on comparisons of levels of productivity and levels of capital intensity. The objective of this study is to apply common depreciation rates to both Canadian and U.S. assets in order to produce more comparable capital stock estimates between the two countries. These estimates of capital stocks are then used to compare capital intensities across industries and asset types between the two countries. An accompanying paper (Baldwin, Gu and Yan 2007) examines the level of MFP in Canada and the United States when both comparable labour and capital data are used.

The following section briefly discusses methods and data sources used for capital stock estimates for the business sector in Canada and the United States. Section 3 presents the decomposition methodology used in this study and Section 4 presents results. Section 5 concludes.

 

^1. For a study of Canadian-U.S. productivity levels, they estimate Canadian and U.S. capital stock in the manufacturing sector using similar assumptions about length of life and discard patterns (see Baldwin and Gorecki 1986: 109–118).

^2. For a recent exception, see Schreyer (2005).