Canada’s changing FDI patterns are well known " its share of inward FDI has fallen steadily over the past 30 years, whereas it has maintained its share of the rapidly growing global stocks of outward FDI (fig- ure 1). To put this into a domestic context, in 1970, for every four dollars of FDI in Canada, Canada had only one dollar of FDI abroad. Today, for every one dollar of inward FDI, Canada has $1.20 of outward FDI (figure 2). Furthermore, this outward FDI is increasingly outside the United States.

The first big question to ask is, why should we care? That is, why should it be of concern to Canadian policy makers and to Canadians in general where multinational enterprises (MNEs) locate? Why should it be of concern to Canadians that foreign MNEs are deciding to locate in Canada less often relative to global trends and at the same time Canadian MNEs are locating abroad more often? The literature is very clear that Canadians should care for many reasons. There is a rich discussion of the costs and benefits of both inward and outward FDI to home and host countries. First, inward FDI is an impor- tant source of R&D diffusion " that is, Canadian firms are more productive as a result of the presence of foreign firms in Canada. Second, foreign firms have both higher levels of productivity and trade propensities than do Canadian firms. Given the benefits associated with trade, both exports and imports, the presence of foreign firms is an important factor in Canada’s economic success. Third, inward FDI contributes to domestic capital formation. Therefore, if there are more Canadian firms locating abroad and less foreign firms locat- ing in Canada, there will be short run and long run implications for capital formation in the Canadian economy. Finally, many studies have found a complementarity between international trade and FDI, thus indicating that the changing FDI patterns will have a significant impact on Canada’s trade patterns. In short, FDI is important in many dimensions for the Canadian econo- my. As a result, therefore, the FDI trends that have emerged for Canada are important for both Canadian pol- icy makers and the private sector.

The second big question to ask relates to the decomposition of Canada’s FDI by industry: What indus- tries are playing increasing or decreas- ing roles on the outward and inward sides? Relatively little is known about the sectoral composition of Canadian inward and outward FDI. What I pres- ent here are some of the trends which were identified in a research project undertaken for Industry Canada. The surge in Canada’s outward FDI is in large part attributable to the surge in services FDI " and this is true whether we consider Canada’s FDI in the United States, the United Kingdom, or the rest of the world (ROW). On the inward side, FDI into the manufactur- ing sector is larger than in services or natural resources, and the importance of manufacturing has been increasing over the 1980s and 1990s. In short, therefore, services is most important for Canadian outward FDI, whereas manufacturing is most important for Canadian inward FDI.

The changing composition in Canada’s FDI is provided in fig- ure 3. It shows that in 1982, the share of Canada’s inward and out- ward FDI in natural resources, manu- facturing and services were roughly equal, each accounting for about a third of Canada’s total FDI. Over the period 1982 " 2001, Canada’s share of inward FDI in manufacturing has increased steadily, accounting for about half of Canada’s inward FDI stock in 2001. Although in 1999, services were significantly more important than natural resources on the inward side; in the years 2000 and 2001, the importance of natural resources increased and that for serv- ices fell. This increase in the impor- tance of natural resources on the inward side is the result of the surge in mergers and acquisitions of oil and gas firms in Canada that were in large part attributable to the low value of the Canadian dollar. Nevertheless, services remained mar- ginally more important than natural resources in 2001.

In short, manufacturing is the most important sector for Canadian FDI on the inward side. On the outward side, the picture is very different. There, Canada’s share of FDI in servic- es has continued to trend upward, accounting for about 60 percent of Canada’s outward FDI in 2001. On the other hand, both manufacturing and natural resources have seen their shares on the outward side fall, with natural resources falling significantly more than manufacturing.

Breaking these data down into Canada’s FDI with the United States and the ROW (figure 4), we see a strikingly similar pattern on the out- ward side, but a much different one on the inward side. Specifically, in terms of Canada’s outward FDI, a majority of Canada’s outward FDI to the United States and the ROW is in services, followed by manufacturing, with natural resources being the least important. On the other hand, for inward FDI from the United States, manufacturing remains the most important industry, followed by serv- ices " and this has been the case for much of the sample. In contrast, this pattern has only recently emerged for FDI from the ROW. That is, since 1997, much of the surge in Canada’s inward FDI from the ROW has been in manufacturing.

These data clearly indicate that on the inward side, manufacturing remains an important sector for Canadian FDI, whereas on the outward side, services are the most important sector.

We now turn to a finer level of industrial disaggregation. Data have been obtained on Canada’s inward and outward FDI at the SIC-C indus- trial classification (table 1). Figures 5 to 7 break FDI down into these indus- tries. Each figure has three panels. The first gives the share of FDI in each sector in 1983 and the second in 2001. The bars in each panel sum to 100 percent. The third panel gives the change in the share of FDI in each industry. The bars in the third panels therefore sum to zero.

What is immediately apparent in figure 5 is the relative importance of industries A to H in the year 2001 (below the horizontal line in the fig- ure). These are both natural resource and manufacturing industries. We do not see these industries playing such an important role on the outward side. This observation points to a sharp difference between Canadian outward and inward FDI. That is, Canada’s inward FDI is much more concentrated in manufacturing and natural resources in comparison to Canada’s outward FDI. As for the service industries (above the hori- zontal line in the figure), the big industry for both outward and inward FDI is L (finance and insur- ance). This industry plays a far larger role on the outward side " in 2001, almost 40 percent of Canada’s out- ward FDI was in finance and insur- ance, but only 13 percent on the inward side.

The two industries that saw their shares increase most for Canadian inward FDI are H (electrical and elec- tronic products) and A (food, beverage and tobacco). Industry C (energy) saw its share fall the most. Although indus- tries G (transportation equipment), K (communications), B (wood and paper), and L (finance and insurance) saw their shares increase, these increases were relatively small. As for the outward side, industry L (finance and insurance) is far and away the most important sector, followed by industry E (metallic minerals and metal products), industry H (electrical and electronic products) and industry C (energy). The industries that saw the largest increase in their importance are industries L (finance and insurance) and MNO (general services to business, government services, education, health and social services). The two industries that saw their shares fall the most were industry I (construction and related activities) followed by industry C (energy).

Figure 6 provides the industrial distribution in Canada’s FDI with the United States, and figure 7 for the ROW. There are two dramatic similar- ities in the change in the distribution to the United States and the row. Specifically, the share of Canadian inward FDI into industry C (energy) from both the United States and the ROW fell dramatically. Another simi- larity occurs on the outward side, where industry L (finance and insur- ance) saw its share increase both for Canadian FDI into the United States and the ROW.

There are also sharp differences in the change in distribution to the United States in comparison to the ROW. Specifically, on the inward side, the United States saw its share in industry A (food, beverage and tobacco) fall, whereas the share from the ROW into Canada’s industry A increased dramatically. Also, the share of US FDI in Canada into industries G (transportation equipment) and H (electrical and electronic products) increased sharply, whereas there was a much smaller increase for H and a fall for G for the ROW.

On the outward side, the share of Canada’s FDI in the United States in industry C (energy) fell more dra- matically than in the case for the ROW. In contrast, industry I (construc- tion and related activities) saw its share fall far more dramatically for the ROW than for the United States. There are many other differences as well, with only the most significant ones highlighted by this discussion.

These data indicate that Canada’s outward stock is predominantly in services, followed by manufactur- ing, with natural resources playing a relatively small role. In contrast, on the inward side, manufacturing is the most important sector, followed by services and then natural resources. In addition, although this trend is similar for the United States and the ROW on the outward side, this is not the case on the inward side. That is, a majority of Canada’s inward FDI from the United States remains in manu- facturing, whereas this has only recently been the case for inward FDI from the ROW.

What explains these changing FDI patterns?

In a study I undertook for Industry Canada, a full model for FDI was developed and estimated. That analysis allowed for the identification of the major factors that help explain these changing FDI trends. Only the major results of that study are summa- rized here.

Corporate taxes paid were found to be strongly positively related to Canada’s outward FDI " that is, the higher are corporate taxes paid, meas- ured at the industry level, the more likely it is that Canadian MNE firms will locate outside of Canada. As for the inward side, the results were mixed. Although inward FDI from Europe and the ROW were deterred by Canadian corporate tax rates, such taxes were not found to be a deterrent to US and Japanese MNEs locating in Canada. The impact of corporate taxes on the decisions of MNEs on where to locate is offset by the generosity of Canada’s capital consumption allowances " the more capital con- sumption allowances that a firm is allowed to claim for tax purposes, the less Canadian outward FDI there is, but more inward.

Canada’s poor performance on R&D spending has been well documented. Despite the relative gen- erosity of Canada’s tax system to encourage R&D spending, Canadian firms have elected not to undertake as much R&D (relative to GDP) as our major trading partners. Nevertheless, R&D spending is a very important fac- tor in explaining Canada’s increasing FDI abroad. Although finance and insurance is by far the most impor- tant industry in terms of FDI on the outward side both in levels and in growth rates, it is not an important industry in terms of R&D. On the other hand, two other industries (transportation equipment [G] and electrical and electronic products [H]) are the most important sectors vis-à- vis R&D and both of these industries have seen their share in total FDI abroad increase. This explains why R&D expenditure turns out to be a statistically significant factor in explaining Canadian outward FDI. This is in sharp contrast to the inward side where R&D is found to be statis- tically insignificant.

The asymmetry in the surge in Canada’s outward FDI in finance and insurance (L) with no similar surge on the inward side is explained by differences in policy towards FDI. Canada has ownership restrictions that are much more onerous than is the case in other G7 countries, and hence the foreign presence in Canada is not nearly as large as is its importance on the out- ward side. Although Canadian banks are allowed to exploit their strong, firm-specific advantages abroad, for- eign firms in finance and insurance are not allowed to do the same in Canada (at least not to the same extent). As a result, we get the asym- metry with finance and insurance being far more important on the outward side than inward.

The cost of labour was also shown to be an important factor in explaining these changing FDI pat- terns. The results indicate that as the price of labour increased, there was more Canadian FDI abroad and less inward FDI into Canada. Conversely, relatively lower costs in Canada in comparison to the United States and Europe encourages foreign MNEs to locate in Canada and deters Canadian MNEs from moving abroad.

Trade policy was also found to be a very important factor in explaining Canada’s changing FDI patterns. NAFTA is estimated to have reduced Canada’s inward and outward FDI with the United States. This is exact- ly as one would expect: as trade bar- riers between Canada and the United States fell, US firms had less of an incentive to locate in Canada, as they could simply supply Canada through exports from the United States. A similar story obtains for Canadian MNEs. This explains why NAFTA is associated with Canada and the United States each receiving a reduced share of each other’s FDI. Since NAFTA, Canada’s FDI, both out- ward and inward, has increasingly been with countries other than the United States.

Canada has been transformed from primarily a host economy for FDI in the 1970s to an important home country " by 1997, Canada had more outward FDI than inward. Although services FDI is by far the most impor- tant on the outward side, manufactur- ing remains the most important for inward FDI.

The results presented here establish clearly that corporate taxes paid in Canada are an important factor in explaining the surge in Canada’s out- ward FDI. Offsetting this is the gen- erosity in capital consumption allowances, which were related not only to less outward FDI but also to increased inward FDI into Canada. These results imply therefore that reducing taxes may reverse to some extent the trends described above.

But herein lies a difficult policy dilemma. Reducing taxes on MNEs may encourage foreign MNEs to locate in Canada more, but it is unclear from this analysis whether the benefits " and there will be many " will exceed the costs vis-à-vis reduced government tax revenue. Nevertheless, it is clear from the discussion that one cost of higher corporate taxes is reduced inward FDI and increased out- ward. The policy implication is that the benefits associated with increased inward FDI must be factored into any government’s decision to lower taxes; as well, the costs associated with reduced inward FDI must be factored in when governments decide to raise taxes.

This evidence fills an important void in terms of understanding Canada’s industrial distribution of FDI as well as the economic factors that have contributed to these FDI trends. As we know, there is a large literature which indicates that FDI brings with it many benefits " that is, inward FDI tends to be thought of in net positive terms for an economy. The reduced attractiveness of Canada to foreign MNEs is therefore thought of as a net negative development for the Canadian economy. There is less of a consensus vis-à-vis the benefits of outward FDI on the home country.

The evidence presented here indicates that R&D is an important factor in explaining Canada’s out- ward FDI, which is a positive devel- opment for Canada. What also encourages more outward and less inward FDI is high levels of corporate taxation in Canada, which is a nega- tive. In other words, the surge in Canada’s outward FDI is driven by both positive and negative factors. The evidence is therefore mixed as to whether the surge in Canada’s out- ward FDI is a net positive for the Canadian economy.