1 Introduction

View the most recent version.

The traditional view of the manufacturing sector is that it was a weak partner in the chain of Canadian economic development. The growth of the manufacturing sector was fostered by a development policy that involved the encouragement of western settlement, expansion of the east-west railway system and the imposition of tariffs to protect the fledgling Canadian manufacturing sector. The manufacturing sector, rather than being regarded as the engine of growth, was seen to hold back an otherwise rapidly growing economy. Emerging from traditional accounts is the view that tariffs provided the infant manufacturing sector with protection and that it became a perpetual child that never matured, continuing instead to operate at a low level of efficiency. Dale's (1966) work on the tariff reinforced this view with the argument that the large Canadian secondary manufacturing sector, which emerged behind this wall of protection, was inefficient when compared to its United States' counterpart.²

As appropriate as the characterization of the Canadian manufacturing sector as an infant might have been at Confederation in 1869, it was no longer the case by 1929. The Canadian economy and its manufacturing sector had grown markedly through the first three decades of the twentieth century. It is therefore appropriate to try and quantify the extent to which Canadian industry had matured before the dramatic events of the thirties and forties. There are a number of dimensions to the concept—the manufacturing sector's ability to compete in export markets, its ability to exploit plant or firm scale economies, and its relative efficiency. In this paper, we examine the basis for the conventional view by focusing on the efficiency of Canadian manufacturing industries relative to those of the United States in 1929.

Productivity failures have been categorized as technical or allocative (Farrel, 1957). Technical efficiency measures the relationship between outputs and inputs. Allocative inefficiency measures the extent to which factors are not combined in the proportions justified by factor price ratios. This paper focuses on technical efficiency and uses a total³ rather than a partial labour factor productivity measure, as has been the general practice.⁴

Labour productivity (LP) is a partial measure that captures the efficiency with which labour is transformed into output; total factor productivity (TFP) measures are more comprehensive measures that capture the efficiency with which multiple factors (i.e., both labour and capital) are transformed into output. TFP measures capture allocative efficiency better than do partial productivity measures.⁵

While the relative TFP is a more comprehensive measure of the production efficiency difference in the two countries than the relative labour productivity level, the two are not independent of one another. Within the growth accounting framework, the relative labour productivity level can be decomposed into two components: one due to the difference in TFP and the other arising from differences in the capital intensity in Canada and the United States (Baldwin, Gu and Yan, 2008):

Where TFP and LP represent total factor and labour productivity, K and L capital and labour, s represents the average share of GDP accruing to capital and C and U are superscripts referring to Canada and the United States.

Relative Canada/U.S. total factor productivity (a measure of technical efficiency) can be the same in the two counties and relative labour productivity can be less than one when capital intensity in Canada is less than the United States. Thus, a country may have lower labour productivity not because of lower technical efficiency (as measured by total factor productivity) but because it has lower capital intensity. And the latter may arise from a response to differences in factor price ratios across countries.

In this paper, we find that the gap between Canadian and U.S. levels of efficiency in the manufacturing sector that has been inferred to exist when partial labour productivity measures are compared virtually disappears when a total factor productivity measure is used, that labour costs relative to capital costs were considerably lower in Canada than in the U.S., and that Canada employed considerably more labour relative to capital (or less capital per worker) than did the United States. Most of the differences in labour productivity arose not from differences in technical efficiency but from lower capital intensity.

The first section reviews the findings of earlier studies on the extent and reasons given for the gap in manufacturing productivity between Canada and the United States. The second section reviews the evolution of the Canadian manufacturing sector between 1900 and 1929. The third section compares certain salient characteristics of the Canadian and American economies. The fourth section contains the detailed total factor productivity estimates.

 

^2. The theme that Canadian industry is inefficient is not confined to economic historians. The traditional industrial organization literature (Eastman and Stykolt, 1967) argued that not only has Canadian tariff policy expanded sectors where Canada has a comparative disadvantage, but also that those industries receiving tariff protection did not operate as efficiently as they might have. The inefficiency, it was argued, existed because of smaller plant scale and shorter production runs that arose out of the small size of the Canadian market (Fullerton and Hampson, 1957; Daly et al., 1986, Economic Council of Canada, 1961).

^3. Statistical agencies refer to these as multi factor rather than total factor productivity measures since many factors contribute to the production process that at the moment are unmeasured and existing measures will therefore be less than complete—they will involve multiple factor but not all factors. Nevertheless, we adopt the more common phrase used in academic circles in this document for ease of communication. Other papers in the Canadian Productivity Review published by Statistics Canada make use of the phrase 'multifactor productivity measure'. The two are synonymous.

^4. Keay (2000) and Inwood and Keay (2006) also focus on total factor productivity.

^5. In practice, TFP estimates may also capture the extent of economies of scale and other organizational differences (Hulten, 2001); but so too do simple labour productivity measures.