A conventional view of Canada-U.S. economic integration is one of the two countries "making things together" rather than simply selling final goods and services to each other. Indeed, one of the major concerns expressed by business leaders and some politicians about security-related border initiatives implemented post-9/11 was that the initiatives would disrupt cross-border supply chains, particularly for industries where just-in-time parts delivery was critical for efficient production. The development and implementation of programs to facilitate faster cross-border commodity flows, such as Free and Secure Trade (FAST), were promoted by multinational companies, especially the automobile companies, to reduce delays of inter-affiliate shipments with their attendant costs.
Despite its prominence in the public policy debate surrounding border security and clearance initiatives, there has been relatively little research on the magnitude and nature of bilateral trade in intermediate goods. In particular, there is virtually no statistical evidence on the behavior of intermediate goods trade in the post-9/11 period. This study attempts to fill in the knowledge gap by creating and analyzing time series data on U.S. intermediate goods trade with Canada over the time period 1990-2011. It also provides some analysis of the determinants of the changes observed in bilateral intermediate goods trade.
Contrary to the popular premise that bilateral supply chains are becoming more integrated, we find that bilateral trade in intermediate goods as a percentage of total bilateral trade was lower in 2011 than in 1990. Most of the decline occurred between 1990 and 2002. This was followed by a modest recovery in the percentage from 2003 to 2008. The severe recession of 2008-2009 was associated with another decrease in the percentage followed by a recovery. In short, there has certainly been no increase in the intermediate goods intensity of bilateral trade pursuant to the implementation of the Canada-U.S. Free Trade Agreement in 1989. This result is broadly consistent with studies done by the Conference Board of Canada which examined a shorter time period.
Our analysis also identifies divergent time series patterns for intermediate goods intensities of imports and exports. Specifically, the two time series are mirror images. U.S. imports of intermediate goods from Canada as a share of total U.S. imports from Canada decreased from 1990 to 2002 and then increased for most of the subsequent time period. The opposite pattern is observed for U.S. intermediate goods exports as a share of total exports to Canada. We identify the extent to which changes in overall intermediate goods import and export trade intensities are functions of changes in the relative importance of individual industries in bilateral trade versus changes in the intermediate goods trade intensities of individual industries. The relative importance of each phenomenon varies over the sample time period. Specifically, both phenomena contribute to observed changes in import and export intermediate goods trade intensities for approximately the first half of the sample time period. In the second half, changes in intermediate goods trade intensity for imports are almost entirely due to changes in industry mix, while changes in the intermediate goods trade intensity for exports are virtually entirely due to decreases in the intermediate goods trade intensities of individual export industries.
The different time series patterns for intermediate goods exports and import intensities, as well as the varying contributions of changes in industry trade mix versus changes in the intermediate goods trade intensities of individual industries, suggest that there is no single explanation for the overall behavior of bilateral intermediate goods trade. We find some evidence of bilateral intermediate goods trade intensities being reduced by increased vertical production integration between the U.S. and Mexico, as well as between the U.S. and China. There is also evidence of changes in the Canada-U.S. exchange rate affecting intermediate goods import and export intensities which is consistent with a hypothesis that "pass through" of exchange rate change is weaker in the case of intermediate goods than final goods.
It has been argued that trade integration between Canada and the United States is an increasingly important policy issue as bilateral integration becomes more about "making things together" than "selling things to each other." (1) An equivalent way of describing the phenomenon of making things together is that bilateral trade is primarily characterized by exports and imports of intermediate goods used in further production, rather than by exports and imports of final goods. Hence, the thickening of the Canada-U.S. border in the wake of post-9/11 enhanced security initiatives raised particular concerns about disruptions to trade in intermediate goods ( Goldfarb, 2007; Hart, 2010). Indeed, the design and implementation of customs pre-clearance programs, such as Free and Secure Trade (FAST), largely reflects political priorities in the two countries to address border security-related disruptions to vertically integrated supply chains, particularly in the motor vehicle industry, where timely cross-border shipments of intermediate goods is critical to efficient production (Globerman and Storer, 2011).
Despite the importance policy-makers have placed on improving the efficiency of cross-border supply chains, particularly in the aftermath of border-thickening, security-related developments post-9/11, there has been relatively little published research documenting changes over time in the relative importance of intermediate goods trade in the overall bilateral trade relationship. Nor has much attention been paid to the possible determinants of intermediate goods trade as a share of total bilateral trade. The purpose of this study is to provide some additional evidence and insights into these issues.
The paper proceeds as follows. In the second section of the paper, we review a number of relatively recent empirical studies of trade in intermediate goods. A particular focus is on studies of intermediate goods trade between Canada and the U.S. The bilateral trade experience is then compared with the trade experiences of other regions of the world. Section Three discusses a set of factors that have been identified as influencing trade in intermediate goods. In Section Four, we present and discuss our own estimates of the intermediate goods trade intensity of bilateral trade. One main finding is that the behavior of total bilateral intermediate goods trade as a share of total bilateral trade obscures prominent differences between imports and exports. Indeed, intermediate goods imports as a share of total imports follow a mirror-image pattern to intermediate goods exports as a share of total exports over our sample time period 1990-2011. Specifically, the import share decreases over the decade of the 1990s and then increases post-2002. The opposite pattern is observed for the export share.
Section Five presents evidence on the contribution of changes in the industrial composition of exports and imports to the observed changes in intermediate goods' share of bilateral trade. Changing industry trade weights indirectly contribute to changes in the share of bilateral trade accounted for by intermediate goods trade. We also report changes in the share of trade accounted for by intermediate goods holding industry-level trade weights constant. This is a direct measure of the changing importance of intermediate goods trade at the industry level. We find that the behavior of bilateral intermediate goods trade for imports over the period 1990-2001 reflects the combined impacts of changes in the relative importance of individual importing industries, as well as changes in the ratio of intermediate goods imports as a share of total imports within industries. On the other hand, the behavior of overall intermediate goods imports as a share of total imports primarily reflects changes in the industrial mix of imports over the period 2001-2011. Changes in industrial mix are relatively unimportant as an explanation of the observed changes in the behavior of intermediate goods exports as a share of total exports, particularly over the period 2001-2011.
In Section Six, we consider possible explanations of the time series patterns described in Section Five. The most distinctive attribute of the intermediate goods bilateral trade experience over the sample period is the marked decline in U.S. exports of intermediate goods as a share of total U.S. exports to Canada from 2001 to 2011. While post-9/11 border security developments may have influenced this variable in the short run, the main factor at work seems to be the substantial growth in motor vehicle assembly activity in Mexico, with some possible secondary influence from shifts of machinery assembly to China. The primary result is that shipments of vehicle parts made in the United States were redirected towards Mexico and away from Canada. Another impact of increased auto assembly activity in Mexico is reduced U.S. vehicle imports from Canada, which indirectly increased the intermediate goods share of total imports. We also find evidence that exchange rate changes may have influenced the share of bilateral trade accounted for by intermediate goods. The final section of the paper provides a short summary and a brief discussion of policy implications.
TRADE IN INTERMEDIATE GOODS: EMPIRICAL STUDIES
Two relatively recent studies by the Conference Board of Canada identify and assess patterns of trade for different types of goods in the North American context (Goldfarb and Chu, 2008; and Hodgson, 2008). Since the two studies are very similar in their methodologies and findings, it is sufficient to summarize only one of those studies.