Province-Level Impacts of Canada’s Trade Agreements

Abstract

Given Canada’s extended geography and regional economic diversity, individual provinces have differing exposures to particular international trade agreements. We demonstrate this by estimating the impacts of the Canada-Korea free trade agreement on the province of Ontario, using a dynamic general equilibrium model to generate Canada-level impacts, which are then decomposed on the basis of partial equilibrium model simulations on a trade dataset in which Ontario is represented as a separate international trading entity. We show that geography and sectoral specialization matter and that general equilibrium effects must be taken into account in partial equilibrium assessments of sectoral impacts of major trade agreements.

KEYWORDS:

• Canada
• CGE
• CPE
• GSIM
• GTAP
• Ontario
• provincial trade
• trade liberalization

ACKNOWLEDGMENTS

This paper has benefited from the comments and suggestions of an anonymous reviewer and discussions of the paper at presentations at the Canadian Economics Association Annual Meetings, Toronto, May 2015, and at the Western Hemispheric Trade Conference, Laredo, April 2015. The authors thank Rahel Aichele for detailed comments on the paper and Natassia Ciuriak for editing the manuscript. All remaining errors of fact or interpretation are the sole responsibility of the authors.

Notes

¹ This approach complements that in Lysenko, Ciuriak, and Xiao (2015), which involves replacing Canada by its provinces in the GTAP data set. This latter approach allows the generation of full CGE impacts at the province level, but data constraints result in less sectoral detail than is possible under the present approach.

² Using distance between capitals or major centers of commerce (Toronto, in the case of Ontario) as our measure of economic distance, we calculate Ontario’s distance from Korea to be 10,590 km, 12% greater than the weighted distance for the ROC to Korea of 9,442 km. Ontario’s GDP is about 60% the size of the ROC’s. Assuming unitary size and distance elasticities, Ontario’s trade with Korea should be 0.53 times as intensive as the ROC’s (0.53 = 0.6/1.12).

³ These are simple average tariffs that take into account tariff equivalents of tariff rate quotas. See the International Trade Centre’s Market Access Map (2014a). On a trade-weighted basis, the average tariff faced by Canada falls to 9.7%.

⁴ The interested reader is referred to Ciuriak and Xiao (2014) for the details of the CGE model used to generate the Canada-level results.

⁵ The detailed assumptions underlying the policy shocks and a fuller discussion of the impacts are set out in Ciuriak and Xiao (2014) and Ciuriak, Xiao, and Dadkhah (2015).

⁶ See Statistics Canada (2013). Projections reported here are based on the medium-growth scenario in Statistics Canada (June 2010) applied to the 2011 census benchmarks.

⁷ A detailed description of the sources of information and methods applied by Statistics Canada to develop the provincial trade data is provided by Généreux and Langen (2002).

⁸ The actual first-round impact in the GTAP model will vary from the GSIM result because of different solution algorithms and differences in the implicit demand and supply elasticities. However, the order of magnitude of the impacts will be the same.

⁹ A detailed description of the sources of information and methods applied by Statistics Canada to develop the provincial trade data is provided by Généreux and Langen (2002).

¹⁰ In several cases, there are no HS international imports that correspond to the positive CSNA-consistent I-O international imports. In this case, Canadian HS international imports shares are used in the allocation.

¹¹ In several cases, when a province ships a product inter-provincially without corresponding shipments to the United States, other shares are used: e.g., the province’s shipment shares to the world in 2010 or another year. In several cases, a province ships a product to other provinces, but it never exports it internationally. In this case, HS shares of other provinces’ within product data are used. Specifically, Yukon shares are used to allocate one of the products of the Northwest Territories and Nova Scotia shares are used to allocate one of the products of PEI.

¹² Some estimates of domestic shipments turn out to be negative (perhaps because we do not take into account shipments out of inventories). There are only a few cases of this and the negative numbers are small. We set them equal to zero.