Archived — Occasional Paper Number 27: The Location of Higher Value-Added Activities

by Steven Globerman, Western Washington University, April 2001


Executive Summary

The primary purpose of this report is to identify and assess the implications of industrial clustering for the future location of technology- or knowledge-intensive activities in North America. A related purpose is to identify and discuss potential initiatives that might be pursued in Canada to blunt or reverse the advantages that specific regions in the United States enjoy as a consequence of hosting already established clusters of innovative firms and skilled and entrepreneurial individuals.

The concern of policymakers in small, developed countries like Canada that trade liberalization will encourage productive resources to migrate to larger economies has been somewhat diminished by evidence showing that trade liberalization increases intra-industry trade and international integration instead of reducing the overall level of economic activity. However, as firms rationalize production across geographic locations to take advantage of the horizontal and vertical value chains, policymakers have become more concerned about the nature of economic activity encouraged by specialization and "agglomeration economies". Since knowledge-intensive activities draw more heavily upon human capital than physical capital and social infrastructure that can be created anywhere, the challenge for Canada, competing against the inherent size advantage of the United States, is to create attractive opportunities for industrial clusters. The increased ease of doing business across borders might facilitate the relocation of existing technology-intensive activities from Canada to the United States, especially if agglomeration economies in those activities favour locating in the latter country.

Closer economic integration may affect the mix of value-added activities in Canada by altering the size of industries in Canada and the strength of firms with different mixes of value-added activities within industries, or by altering the optimal mix of value added within firms in response to international comparative advantage, competitive pressures and the changing economic environment. That is, firms will alter their geographical location in response to closer economic integration, holding constant their industrial focus, size and so forth. Whether this will reinforce the U.S. advantage in products that are human capital-intensive and the Canadian advantage in resource-intensive industries will depend on the degree to which inter-industry trade patterns are affected by closer economic integration and whether existing industrial clusters are characterized by economies or diseconomies of scale.

The location of clusters is not fixed; they expand and contract geographically, and can emerge in locations distinct from the areas encompassing older clusters. While economists are in disagreement on whether historical accidents or the antecedent conditions of a region play a larger role in determining the geographical location of a cluster, several causes have been suggested to explain the benefits of agglomeration arising through external economies of scale. Firstly, a large industrial centre offers a pooled market to workers with specialized skills, creating liquidity in the labour market, which benefits both workers and firms. Secondly, a large industrial centre provides specialized non-traded inputs in greater variety and at lower cost. Thirdly, clusters promote technological transfers and spillovers as closer geographical proximity improves communication. However, too dense a cluster of economic activity creates congestion and diminishing returns.

As well, more research is required to determine whether a region's institutional characteristics impact on the formation of agglomeration economies, and what role government policy can play in promoting clusters. The evidence on whether clusters benefit more from a large number of small firms, which are more likely to contract out, or from the hub-and-spoke model associated with a small number of large firms, is unclear. What is clear is that vertical disintegration of economic activity contributes to the critical mass of specialized business and technical services required to encourage and sustain industrial clusters. But it is not clear that foreign participation in a local economy discourages the formation of vertical and horizontal linkages locally by centralizing innovative activity in the home country, or that foreign affiliates are increasingly dispersing value-added activities to exploit differences in location advantage and local technical expertise. It may be that the forces influencing clusters are dependent on the type of economic activity and industry involved. Low taxes and generous subsidies are obviously preferable for firms and high-skill workers when choosing a location. Fiscal incentives aimed at individual firms are inefficient; a more promising route is to use tax breaks and subsidies to make a region more attractive to a variety of technology-intensive firms. Of course, a satisfactory level of telecommunications and transportation infrastructure, public utilities and other social infrastructures are necessary to sustain an industrial cluster, but not in and of themselves sufficient.

The strength of local competition, including openness to foreign ownership, and the presence of sophisticated customers in a region might improve the nature and importance of external economies of scale. Public policy can nurture a competitive industrial environment and facilitate the migration of labour and skills. Making knowledge-intensive sectors relatively free of regulatory barriers should be the starting point of any coordinated set of government policies aimed at making Canadian regions attractive as locations for knowledge-intensive clusters. Secondly, government policies should promote international labour mobility, especially for skilled professional and technical workers, and for Canada's major urban areas in order to attract industrial clusters.

More importantly, the evidence shows a positive relationship between university research, centres-of-excellence, and a region's innovative performance. This relationship is strongest in larger metropolitan centres with a concentration of high-technology production and established linkages between researchers and the business and financial communities. Thus, government policy will be most effective when focused on "pre-competitive" research, as private firms are easily able to draw on non-local experts for specific, codifiable functions.

Finally, it has been hypothesized that electronic commerce will reduce some costs dependent on physical proximity, and thus reduce the importance of clusters, though this outcome is far from clear.

In conclusion, the paper states that governments should focus less on industrial policy, whereby they target "desirable" industries or "national champions," and instead encourage clusters by promoting conditions within regions that contribute to the realization of external economies. Governments may need to be involved in rationalizing the competing claims of regions for public support. It may be ineffective for a small, open economy to encourage more than one cluster to develop in specific industrial areas. Cooperation would mean allowing and encouraging patterns of regional specialization that maximize the nation's welfare, rather than that of individual provinces. The federal government might justifiably see its role as assisting provincial governments to enhance the specific environment of regional clusters in policy areas where the federal government is dominant. It may well be that the effective promotion of knowledge-intensive clusters in Canada requires a substantial reorganization of government responsibilities and financing arrangements.

Date modified: