6 Conclusions

View the most recent version.

Over the past three decades, Canada's manufacturing sector has become increasingly integrated into world markets. The effect ofintegration through increased exports has not been felt equally across all geographic regions or in the same way.

This paper has provided a cross-regional comparison of the impact of trade intensity on the structural characteristics of the manufacturing sector—characteristics like plant size, product line diversity and production-run length—that are associated with higher productivity. It has been argued in the past that Canadian plants were too small and produced too many products, resulting in short production runs and high unit costs, and that greater integration into North American markets would partially overcome these deficiencies. We find that across regions, higher levels of export intensity are positively associated with longer production runs, larger plant sizes and a smaller number of product varieties per plant.

Throughout the paper, cross-regional differences in the industrial structure stand out. Regions that are smaller and further from the centre of the North American market have smaller plants and have shorter production runs. In addition, these regions are less integrated into North American markets as indicated by their export intensities, and this has an additional impact on their plant size, specialization and length of production runs. Export intensity is directly related to plant size, product specialization and production-run length. Plants are more likely to be larger where there is more exporting taking place—in Ontario and Quebec, the industrial heartland of Canada. These are the regions that are best situated with respect to the U.S. market and, therefore, are best able to take advantage of its large market. These were also the markets that started with the highest level of product diversity at the beginning of the period—where product packing to achieve economies of scale was most evident.

It was in these regions where the response to increases in export intensity was largest. It was in Central Canada (primarily Ontario) where plant scale increased the most, where product diversity declined the most and where production-run length increased the most on average. It was here that the relationship between higher levels of export intensity and plant scale, or production-run length was also the strongest. The reason for this comes from the fact that some aspects of industrial structure reacted more to export change (increased integration into North American markets) than elsewhere. Changes in plant size over time reacted more to changes in export intensity over time in Ontario than elsewhere. Not all aspects of industrial structure that are studied here reacted similarly. The impact of export intensity is not different for the number of products when the time dimension is considered, but it is different for the average plant size and the length of the production run. There is a greater impact of change in export intensity on these two variables in Ontario than in Quebec and an often lesser one in the two outlying regions—Atlantic and Western Canada.

The changes in plant size, number of products and length of production runs come from two quite different groups of firms whose reaction were not expected to be the same—non-exporters and exporters. In each case, the group reacted much as the model predicted; generally the differences reinforced regional disparities. While non-exporters were expected to see a characteristic like plant size decrease in response to integration, the response was larger in the outlying regions. While exporters were expected to see a characteristic like plant size increase in response to integration, the response was largest in the central heartland.

In summary, the regional perspective adopted in this paper provides new insights into the underlying effects of trade. The effects of trade are not the same in all locations. And the variations that we observe over space are not random. Instead, we observe patterns that point to the importance of location relative to markets as key variables determining how trade affects regional economies. Given the geographic diversity of the Canadian economy, taking into account differences related to geography substantially improves our insights on the effects of trade on the economy.