Canada and U.S. Outward FDI and Exports: Are China and India Special?

Abstract

Using cross-country time series data for the period 1989–2001 we analyze the Canadian and United States' outward FDI and export performances, particularly to China and India. Casual examination of data may suggest that Canada is underperforming in its exports and FDI to China, but results from our econometric model do not support that conclusion. We found that Canada's FDI in India is lower than that predicted by the model. Interestingly, while the evidence that investors from the United States tend to invest more in the growing economies is quite strong, it is weak in the case of Canada. Although in-depth research is necessary to understand these differences, it is plausible that there are mismatches between the areas where investment opportunities are available in the fastest-growing parts of the world such as China and India, and the areas in which Canadians have comparative advantage. For example, financial services and mining constitute a big share in Canadian FDI abroad but these sectors are yet to be opened up in China and India. The U.S. FDI base is more diversified and better able to take advantage of the increased opportunities in fast-growing countries such as China and India.

Keywords:

• FDI
• exports
• gravity model
• Canada
• United States
• China
• India

ACKNOWLEDGMENTS

Authors thank Thomas Rawsky, Daniel Boothby, Jianmin Tang, Someshwar Rao, Brian Fileds, Claudiu Tunea, Jean McCardle and participants at the Statistics Canada Socio-economic Conference and the CEA Meetings for comments. We also like to thank Dipak Roy and Biswajit Nandy for sharing their thoughts and experiences on Canadian investments in India, and the editor and the two anonymous referees for constructive comments and suggestions.

Notes

¹The global stock of FDI almost tripled during the 1980s followed by a twofold increase during the 1990s reaching to US$15 trillion in 2008 (UNCTAD 2009a).

²China's isolation has been more profound. Its policy reversal came a decade earlier than that of India and was more radical and complete at least with respect to FDI.

This study was undertaken when both the authors were with Industry Canada. Views expressed here are their own and do not necessarily reflect those of Environment Canada or Industry Canada, or Statistics Canada.

³The comparatively low FDI statistics in India, however, raises some issues around how they are measured in India and China and their comparability (Pfefferman 2002). International Financial Corporation (IFC) notes that India's FDI statistics exclude reinvestment earnings, subordinated debt, and overseas commercial borrowings. These are included in FDI in other countries. IFC also estimates that ‘round-tripping’ in China could be 50% of total FDI inflows. After accounting for these differences IFC concludes that there is not a huge difference between China and India in FDI inflows as percent of GDP (Gordon 2002). See also (Srivastava 2003).

⁴Measured in purchasing power parity (PPP) term China's share of global output has risen from close to 11% in 2000 to over 13% in 2004 (Krueger 2005).

⁵The output share in computed from World Development Indicators Database and the trade shares are derived from UNCTAD's trade database available online.

⁶The choice of study period was driven by the availability of comparable data for the set of major economies Canadian and U.S. FDI and exports are destined for.

⁷Hong Kong's share in Chinese FDI stock has fallen from 68% in 1992 to 38% in 2000.

⁸A rise in investment in the services industry in China recent years is in line with the global trend.

⁹Many foreign countries invest in India through Mauritius for tax savings.

¹⁰Also see Fredricksson (2003) for a review of UNCTAD's research on transnational corporations.

¹¹See McCallum (1995) for its application in explaining Canada-U.S. trade.

¹²The law of diminishing returns implies that since the marginal product of capital is high in poor countries, capital should flow from rich to poor countries. But empirically it is not true. Lucas (1990) reconciles this paradox.

¹³ Anderson and Wincoop (2003) developed a model that has solid theoretic foundation. The interesting feature of their model is the inclusion of the “multilateral resistance” indices as explanatory variables, which can avoid omitted variable bias facing earlier gravity models.

¹⁴The only difference is that the trade-to-GDP ratio is used for the exports equations while the FDI-to-GDP ratio is used for the FDI outflow equations.

¹⁵Some researches use GDP per capita and population separately instead of GDP. For example, see Razin et al. (2003, 2004).

¹⁶Due to Globalization of the world economy, manufacturers can produce components in one country and ship them into another country for processing into final goods, which implies that FDI may induce exports.

¹⁷Tariff rate is a more direct measure of a country's trade policy. It is not used as an explanatory variable in this paper simply because we could not obtain bilateral tariff data between Canada (or the U.S.) and it's trading partners.

¹⁸ Yamarik and Ghosh (2005) examined the robustness of 47 variables used in the gravity model literature and found 20 The test results are available upon request.

¹⁹of them are robustly linked to trade including the level of development, trade agreement and policies, common language, historical ties, geographic factors and so on.

²⁰See Helpman (1984).

²¹Table 6 shows that manufacturing accounted for about 70% of total FDI inward in China in 2003. Similarly, Table 10 shows that about 47% of FDI inflows in India during the period of 1991 to 2004 are taken place in ICT and transportation equipments.