Relative factor abundance and FDI factor intensity in developed countries

Abstract

This study looks at the link between the patterns of trade-revealed comparative advantage and net inward foreign direct investment in five developed countries: the United Kingdom, the United States, Japan, France, and Italy. It thus extends earlier work by Maskus and Webster (1995) who analyzed two countries, the United Kingdom and South Korea. Despite assertions in the literature that market access is the primary motive for foreign direct investment flows among developed countries, this study shows that there is a significant role for comparative advantage in determining inflows of foreign direct investment in developed countries, especially in the services industry.

Keywords:

• foreign direct investment
• comparative advantage
• Heckscher-Ohlin-Vanek

JEL Codes :

• F21
• F14

Notes

Two other members of the G7, Canada, and Germany, are excluded since comparable data are unavailable.

The years of 1999 and, more significantly, 2000 have been characterized as years with exceptionally high cross-border Merger and Acquisition activities. The total value of world FDI inflows in 2000, for example, is more than twice the value in 1998 or 2001 (United Nations 2002).

An early alternative approach such as Mundell (1957) argues that when there are significant barriers to trade based on the Heckscher-Ohlin-Vanek principle, FDI would rise in order to ‘jump’ the tariff barrier. This implies that the relationship between comparative advantage and the location of FDI is the reverse of the firm-specific advantage argument.

Rearranging equation (3) yields . The left-hand side still measures the trade-revealed factor abundance for country c in terms of factor k and provides the basis for the inequality condition for the relative abundance of factor k with respect to k′.

United Nations (1993) reports stocks of net inward FDI for each country. The actual definition of the ‘stock’ data varies from country to country. Some countries such as the US and France obtained their stock data from surveys of companies using book value and/or market value of the capital stock. Other countries such as Japan and Italy use long-run accumulated flows with different accumulation periods and depreciation allowances. Due to these differences in the stock definition, we also construct the second (medium-run) measure of FDI flows by accumulating four years of flows on or before the year of the export data. See the data appendix for more details on FDI stock.

There are 36 sectors based on the International Standard Industrial Classification (ISIC, Rev 2) provided in the Input–Output tables. ‘Services’ contain the following sectors: Electricity, Gas and Water, Construction, Distributive Trade, Transport and Storage, Communication, Finance and Insurance, Real Estate, Restaurants and Hotel, and Other Services.

Davis and Weinstein (2001) summarize the reasons why we should use an individual country's technology matrix rather than adopting the assumption of identical technologies and explain some desirable properties of more uniformly defined data.

We generally use four and 11 years of accumulated GFCF to construct medium- and long-run domestic capital stock. For France, however, the available GFCF series at the desired aggregation level dictates the use of three years of data (1987–1989).

Since aggregate skill levels change only very slowly over time, our results should be unaffected by the slight differences in the time of data coverage.

We emphasize, as do Maskus and Webster (1995), that this is not a full analysis of the determinants of factor contents and that a regression analysis is well beyond the scope of this paper. We leave that for future research.

Unfortunately, the STAN database we use only provides UK technology and trade data for 1990, while the United Nations FDI statistics for the UK are only available for 1989.

It is also possible for the patterns of inward FDI to dominate completely the outward patterns, resulting in Maskus and Webster's (1995) findings.

Bayoumi and Lipworth (1998) also argue that Japanese FDI outflows led to a short-run increase in Japanese exports since the new facilities in the foreign locations usually required Japanese capital goods.

We thank an anonymous referee for pointing out this explanation to us.

See Table A3. Notice also that for the service sector, France is a net source of FDI in all but one sector for stocks, but is a net recipient in all sectors of the single year FDI flow. This illustrates the high variability of FDI flows and provides an indication of the possible inconsistencies in factor content analysis based on short-run data.

This is made worse by the fact that around 60% of developed countries’ FDI flows are in the services sector.