1 Introduction

View the most recent version.

Over the past three decades, the Canadian economy has become increasingly integrated into world markets as a result of successive rounds of tariff cuts under the General Agreement on Tariffs and Trade and the implementation of the Canada–United States Free Trade Agreement (FTA) and its successor, the North American Free Trade Agreement (NAFTA).

Focusing on the impacts of the FTA and NAFTA, many studies have analyzed how trade has affected employment, plant size and productivity, among other factors (see Schwanen 1997; Head and Reis 1999a, 1999b and 2001; McCallum 1999; Trefler 2004; Baldwin and Gu 2008; Brown 2007). Most studies, however, have focused on only the national impact, which belies the fact that Canada is composed of a distinct set of regional economies whose ties to the North American and global economies are often quantitatively and qualitatively different.

An examination of the response at the national level misses differences that may exist at the regional level. The ability of regions to take advantage of trade liberalization and the nature of their trade with the rest of the world depend, at least in part, on their location relative to North American and world markets. Location is important because distance remains an important determinant of market access (Brown and Anderson 2002). Since distances from main North American markets differ by region, so too might responses to increasing trade liberalization. Gu and Sawchuk (2006) show that the two-way trade at the commodity level with the United States has increased more rapidly for Ontario and Quebec than for Atlantic Canada and the West.

In this paper, we seek to determine how Canada's integration with the world through trade has differentially affected regional economies. To do so, we ask how plant specialization, plant size and production-run length—three characteristics of industrial structure that affect productivity— have changed as manufacturing plants in different Canadian regions have integrated themselves into world markets. The data that are used for the analysis cover the period from 1973 to 1999. Over this period, tariff rates declined steadily, first with the Kennedy Round, then the Tokyo Round, and then with the FTA and NAFTA.

Trade liberalization is expected to affect the industrial organization of production (Horstman and Markusen 1992). Trade barriers have long been held to have a detrimental effect on Canada's industrial structure (Economic Council of Canada 1967, 1975). Some have argued that trade barriers have led to suboptimal plant size, or to excessive product differentiation, or to production runs that are too short to fully exploit the economies of production-run length (Daly, Keys and Spence 1968; Spence 1977; Caves et al. 1980). Trade liberalization was seen to be a solution to these perceived problems (Harris 1984). Increased access to U.S. markets was seen to be a way of more fully exploiting scale economies—either through increased plant size, or fewer product lines and the concomitant increase in production-run length. These changes were expected to have productivity-enhancing effects.

Models of imperfect competition suggest that the impact of increased trade liberalization will be felt differently across regions. In particular, those that face inherently larger markets—those that are larger and closer to U.S. markets—will respond differently to declines in trade barriers.

Therefore, we examine whether each of these industrial characteristics has responded differently across Canadian regions to being integrated into world markets.

In order to address these questions, we focus on that part of Canada's industrial sector where there has been particular concern about efficiency—the Canadian manufacturing sector. It is here that Canada is said to suffer a particularly large productivity disadvantage (Bernstein, Harris and Sharpe 2002). And it is here that advocates of trade liberalization saw the possibility that trade liberalization would effect structural changes that would enhance productivity.

The remainder of the paper is organized as follows. In Section 2, we outline the framework that is used to inform our investigation. In Section 3, we describe the estimating framework and the data that are used in the analysis. Section 4 examines average changes that have occurred across Canadian regions in each of these areas—in plant scale, in product specialization and in the length of the production run. Section 5 provides a multivariate analysis that examines the relationship between plant scale, product specialization, and production-run length and the export intensity of a region—holding other characteristics of the region constant.