What Is Foreign Direct Investment?

The first thing that should be said about foreign direct investment is that the term is a misnomer. Foreign direct investment is not, either in an accounting sense, whether one is talking of financial accounting or of balance of payments accounting, or in an economic sense, truly “investment.” For accounting purposes, foreign direct investment takes place when the book value of the net worth of an investment controlled by investors in a country other than the country in which the investment is legally domiciled increases.² As noted above, in most cases, the “investment” is in fact a subsidiary of a multinational corporation.³ The subsidiary itself is typically an ongoing business under the managerial control of the parent firm. On the balance sheet of the subsidiary, the subsidiary’s net worth is simply the value of assets of the business minus the liabilities owed to entities other than the business’s owners; by the fundamental financial accounting identity the net worth must be equal to the book value of owner’s equity, which, in turn, consists of paid-in capital of the owner plus retained earnings.⁴

(assets) ≡ (liabilities) + (owner’s equity),

Thus, from a balance of payments perspective, foreign direct investment from one country to another consists (mostly) of any net increases in the paid-in capital of investors in the first country to their subsidiaries domiciled in the second country, plus any increase in the retained earnings of these subsidiaries. These increases are recorded in the long-term capital accounts of the two countries as a long-term capital outflow of the investors’ country and as a long-term capital inflow of the subsidiaries’ country. To count increases in retained earnings thusly might appear a bit odd, as no transaction actually takes place between the two countries. The dilemma is resolved by IMF balance of payments accounting standards by considering all earnings reported by the subsidiary (whether retained by the subsidiary or repatriated to the parent) as being transferred to the investors in the home country. (It should be remembered that in almost all cases the “investor” is in fact the parent organization of a multinational corporation.) The portion of these earnings retained in the subsidiary is then classified as a long-term capital flow back to the host country.

From the perspective of a subsidiary, then, foreign direct investment is a source of funds and not a use of funds. Furthermore, it is but one possible such source. Others include borrowing from local or international lenders other than the subsidiary’s foreign owners and, if the foreign parent does not hold 100 percent ownership, raising equity capital from local minority shareholders. Capital expenditures of the subsidiary, which correspond roughly to the economists’ concept of real investment, are uses of funds. It follows that foreign direct investment flows do not necessarily correspond to real capital formation generated by subsidiaries of multinational corporations.⁵

U.S. Direct Investment Abroad and Capital Expenditures by Majority-Owned Foreign Affiliates of U.S. Firms

(In billions of dollars)

Source: U.S. Department of Commerce.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

U.S. Direct Investment Abroad and Capital Expenditures by Majority-Owned Foreign Affiliates of U.S. Firms

(In billions of dollars)

Source: U.S. Department of Commerce.

• Download Figure
• Download figure as PowerPoint slide

U.S. Direct Investment Abroad and Capital Expenditures by Majority-Owned Foreign Affiliates of U.S. Firms

(In billions of dollars)

Source: U.S. Department of Commerce.

• Download Figure
• Download figure as PowerPoint slide

Indeed, there may be no correspondence whatsoever. Consider two polar cases. In the first, a multinational corporation acquires an ongoing firm in a country other than its home country, paying the current owners of that firm cash. In the year that this transaction takes place, the acquired firm engages in no capital investment whatsoever. The transaction represents a direct investment, but no economic investment takes place. In the second case, a subsidiary already under the control of a multinational corporation embarks upon a large capital expansion program but reports zero earnings and zero dividends in that same year, and no change takes place in the parent firm’s paid-in capital; the capital expansion thus is financed via borrowing from financial lenders. No direct investment between the home and host countries is attributable to this parent-subsidiary pair, but the subsidiary has contributed to real capital formation in the host country.

An interesting question then is, How much of a discrepancy is there between measured foreign direct investment flows and economic investment? Unfortunately, global data do not exist that would enable this question to be answered. The question can be answered with respect to U.S. outward investment, because the U.S. government does keep data on the new capital expenditures of majority-owned overseas subsidiaries of U.S. firms (Chart 1).

It is immediately apparent from Chart 1 that capital expenditures of majority-owned affiliates of U.S. firms have consistently been greater than outflows of direct investment from the United States. That is, direct investment from the United States significantly understates the economic investment worldwide (out-side the United States) of U.S.-based multinational corporations.⁶

Would a similar statement hold for direct investment from other industrial countries? The answer is “probably not” because of vintage effects specific to direct investment of the United States: the average age of foreign affiliates of U.S. firms is somewhat greater than that of foreign affiliates of multinational corporations from other countries; hence, the contribution of retained earnings to U.S. direct investment outflows likely is higher than for other countries. (The qualifier “likely” is used because data limitations prevent testing of whether this last statement is really true.)

Chart 1 also suggests that new capital expenditures of majority-owned subsidiaries of U.S. firms do not move in lockstep with outflows of U.S. direct investment.⁷ The likely reasons for the discrepancies are that (1) U.S. outward investment is, as already suggested, only one source of financing of the capital expenditures of foreign subsidiaries of U.S.-based firms and (2) the mix of financing is sensitive to such variables as interest rates in differing capital markets and other financial factors. Thus, following 1979, when U.S. interest rates rose relative to those elsewhere, the ratio of capital expenditures by these subsidiaries to U.S. outward direct investment fell sharply, indicating that these subsidiaries were turning to other sources of finance. Indeed, the sharp drop in U.S. direct investment abroad during the early 1980s was likely caused in part by U.S. firms borrowing funds or receiving dividends from their foreign subsidiaries in response to very high costs of raising financial capital within the United States.

All of this information simply reaffirms the opening sentence of this section, to the effect that the term “foreign direct investment” is a misnomer. Indeed, from a national income perspective-as well as from the perspective of a firm-foreign direct investment is also a source of resources for investment (where net outward direct investment implies a depletion of these resources), rather than investment itself. This is evident from the national income identity I ≡ S − ΔA_I, where I is gross domestic investment, S is gross domestic savings, and ΔA_I is net change in the international assets held by domestic residents (including official reserves). Note also that −ΔA_I / ≡ X −M, where X equals exports and M equals imports.⁸ Foreign direct investment is just one component of ΔA_I; an increase in inward direct investment does not automatically imply an increase in I, nor does outward direct investment imply an automatic decrease in I. It has already been shown that in some cases there can be no effect. For example, in the first case above, the acquisition of a domestic firm by a foreign one, the inward direct investment flow generates a decrease in ΔA_I (implying an increase in funds available for domestic investment). But this is exactly offset by claims by domestic residents on foreigners, because the original owners of the acquired firm now hold some claim on the new owners (and this shows up in the income identity as an offsetting increase in some other component of ΔA_I).

Thus, foreign direct investment has been defined in this section with an emphasis placed on what it is not. But this is far from the end of the story. If foreign direct investment were nothing more than an international transfer of financial capital, the end of the story would in fact be close. But foreign direct investment, being a measurable manifestation of the international spread of multinational corporations, entails much more than a financial transfer. Associated with foreign direct investment are transfer of technology and other so-called intangible assets, the stuff of which long-term economic growth is largely made. These transfers, and why they take place, shall be examined shortly. Before doing so, however, it will be useful to look at some historic facts about foreign direct investment.

Foreign Direct Investment in the Long Reach of History

An unfortunate fact is that very little data pertaining to foreign direct investment exist for the years prior to World War II, and much of what data do exist are unreliable. Nonetheless, economic historians have been able to build a convincing case that foreign direct investment-and multinational corporations-played important roles in economic development at least as far back as the industrial revolution of the late nineteenth century. Indeed, the history of foreign direct investment goes back even further: the East India Company, virtually from its start a multinational corporation, was chartered in London in 1600, but little analysis of its role has been attempted.

During the late nineteenth and early twentieth centuries, large amounts of long-term financial capital flowed across national boundaries, financing projects as diverse as building the railroads that opened the American West and the construction of great tea plantations on the Indian subcontinent. Fairly reliable-records of these flows exist, but it is difficult to say how much of them represented direct investment. In a pioneering study published in 1938, Cleona Lewis, working mostly with U.S. data, estimated that the vast bulk of the flows represented portfolio, rather than direct, investment, and this was accepted as fact for about four decades (Lewis (1938)). However, beginning in the middle 1970s, a number of economic historians have perused in fine detail the international accounts of the United Kingdom, by far the largest international investor nation of the period, and concluded that at least one third and perhaps more of the stock of the U.K.’s overseas investments were direct in 1914.⁹ Studies of other European countries’ accounts have yielded similar results.¹⁰ By the first decade of this century, significant numbers of U.S. direct investors were active in Europe, European direct investors were active in North America, and a number of what are today’s largest multinational corporations from both continents can trace their first international business activities to this era (Wilkins (1970) and (1989)). This “north-north” foreign direct investment covered a spectrum of sectors, including petroleum and other natural-resource-based industries, manufacturing, and services, especially transport (for example, railroads) and financial services (for example, insurance).

However, the bulk of foreign direct investment early in this century was “north-south” in nature, that is, the home countries were those of Europe, the United States, Canada, and Japan, and the host countries were the less-developed countries of Asia, Africa, and Latin America. This north-south foreign direct investment was heavily concentrated in resource-based industries and in transport and utilities. Thus, it is estimated that in 1914 the shares of British outward direct investment of these two sectors were 53 percent and 31 percent respectively, and that the share of both sectors combined of U.S. outward foreign direct investment was 68 percent, with manufacturing accounting for another 18 percent (Houston and Dunning (1976), and Wilkins (1970)).

World War I put a brake on the international expansion of multinational corporations, and, indeed, international business activity as a percentage of all economic activity almost surely shrank in the interwar years 1918-38. (But, again, data limitations prevent full exploration of this issue.) However, during these years, some sectors witnessed large amounts of new direct investment. These included manufacturing (this was the period, for example, when Ford and General Motors established operations in Europe, Latin America, and Asia) and, most especially, petroleum, this latter sector seeing enormous expansion in Latin America (especially Venezuela) and the Middle East. During the Great Depression years of the 1930s in particular, there was very little expansion of the international activity of multinational corporations (and, in some sectors and regions, this activity contracted), with the exception of those in the petroleum industry.

Foreign direct investment picked up sharply after World War II, especially after 1960. U.S.-based firms in the manufacturing sector especially expanded their international activities. Most of this expansion was focused on developed countries, especially those of Europe, where U.S. firms sought to establish local subsidiaries in response to the formation of the European Common Market, now the European Union (EU). As a result, between 1950 and 1970, the stock of U.S. manufacturing direct investment in Europe increased almost 15-fold, while the share of the stock of U.S. foreign direct investment in developing countries declined to less than 40 percent in 1970. By this year, the share of the stock of U.S. outward foreign direct investment in manufacturing had grown to 41 percent.

In 1970, total flows of foreign direct investment on an outward basis, as reported in IMF statistics, were slightly less than $13 billion; flows from the United States were about $7½ billion, or about 60 percent of the total; and flows from the seven largest industrial countries (the United States, Japan, Germany, France, Italy, the United Kingdom, and Canada) were a bit less than $12 billion, or about 91 percent of the total. Foreign direct investment flows from all of the industrialized countries equaled over 99 percent of the total. International data on foreign direct investment for years prior to 1970 are incomplete, but it is probable that the U.S. share of total foreign direct investment flows throughout the period 1950-70 was well over 60 percent.

Recent Trends in Foreign Direct Investment

By 1993, of the reported total outward foreign direct investment flow of over $186 billion, the U.S. share had fallen to 31 percent, the seven largest industrial countries’ share to 79 percent, and the share of all industrialized countries to 94 percent. Of the 6 percent share that was not accounted for by industrial countries, the majority (3.5 percent of the total) was accounted for by Asian countries, and the remainder was spread over a number of developing countries (including but not limited to several of the major oil exporters).

Chart 2 indicates flows of foreign direct investment from the seven largest industrial countries from 1970-93, with flows from the United States and the remaining six countries separately indicated.¹¹ Several features stand out. First, the U.S. share of these flows was 50 percent or more from 1967 to 1979. After 1979, however, the U.S. share began to fall, reaching a low of just over 17 percent in 1990, the year of peak flows worldwide. Following 1990, however, the U.S. share rebounded sharply, reaching over 39 percent in 1993, a phenomenon examined in more detail below.

Foreign Direct Investment Flows from the Seven Major Industrial Countries

(In billions of dollars)

Source: International Monetary Fund.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Foreign Direct Investment Flows from the Seven Major Industrial Countries

(In billions of dollars)

Source: International Monetary Fund.

• Download Figure
• Download figure as PowerPoint slide

Foreign Direct Investment Flows from the Seven Major Industrial Countries

(In billions of dollars)

Source: International Monetary Fund.

• Download Figure
• Download figure as PowerPoint slide

Second, total flows of foreign direct investment from the seven largest industrial countries experienced a “minisurge” during the late 1970s and early 1980s but fell sharply following 1981. From the trough year of 1982 onward, however, foreign direct investment flows increased steadily, and the growth became spectacular following 1985 until, as noted in the first section of the paper, these flows peaked in 1990. Exactly what caused this surge of foreign direct investment growth is not entirely clear. If the magnitude of the 1985-90 surge is ignored, the trends in foreign direct investment flows from the seven largest industrial countries during the past twenty-five years or so correlate quite closely with trends in national income growth (that is, these flows seem to wax and wane more or less contemporaneously with first differences in annual income), but the income elasticity of foreign direct investment flows appears to be highly unstable.¹²

What is most clear about the 1985-90 surge is that, by a number of measures, foreign direct investment became more diversified during this period. As already noted, the predominance of the seven largest industrial countries—-and, within this group, the predominance of the United States—as home countries to foreign direct investment decreased. The sectoral diversity of foreign direct investment increased; in particular, the amount in the service sectors rose sharply relative to manufacturing.¹³ However, by certain other measures, foreign direct investment became less diversified. In particular, the share of flows going to the industrial countries actually increased during the 1980s, and, correspondingly, the share going to the developing countries fell.

Indeed, one of the most notable features of the period was the share of foreign direct investment received by the United States, which became the largest host country, as well as the largest home country, to foreign direct investment. Chart 3 traces the cumulative stocks of foreign direct investment in the United States, as well as the stocks of U.S. direct investment abroad, from 1970 to 1993. Whereas the stock of foreign direct investment in the United States ($13.3 billion) was only 17½ percent of that of the stock of U.S. direct investment abroad ($75½billion) in 1970, the stock of the former ($369 billion) was 96½ percent that of the latter ($382 billion) in 1989. Whereas foreign-controlled business operations accounted for a negligible portion of U.S. economic activity in 1970, they grew to account for a very substantial portion by 1989 (Graham and Krugman (1995)).

However, from 1989 to 1993, U.S. direct investment abroad grew much faster than foreign investment in the United States, so that, by the end of 1993, the Stock of the former ($445½ billion) was only 81 percent of that of the latter ($548½ billion). What happened? The first thing to be noted is that some aspects of the trends are more apparent than real; in particular, data on U.S. direct investment abroad have been affected by recalibrations periodically made by the U.S. Department of Commerce (for example, the drop in the stock of U.S. direct investment abroad shown in Chart 3 as occurring in 1981 was largely the result of such an adjustment, rather than actual disinvestment by U.S. firms).

Chart 4 indicates the U.S. flows of outward foreign direct investment from 1982 through 1992, broken down by major components: new equity, retained earnings, intrafirm debt, and valuation adjustments.¹⁴ Several points emerge from the rather confusing picture posed by the chart. First, U.S. firms largely were absent from the foreign direct investment surge of the mid-to-late 1980s (new equity flows were a trickle in the years 1983-90, and there was net disinvestment in 1985, 1988, and 1989), but new equity flows accelerated sharply following 1990. Second, intrafirm debt flows were negative during the U.S. high-interest years of 1982-84, lending credence to the earlier hypothesis that low U.S. foreign direct investment outflows in those years were in part a result of interest rate differentials. Third, retained earnings were strongly positive throughout most of the period. Fourth, valuation adjustments were significant (and, because these are largely generated by currency fluctuations, the effect of adding valuation adjustments to reported foreign direct investment flows is probably to mislead slightly the analyst who does not take these adjustments into account).

Stocks of Foreign Direct Investment in the United States and U.S. Direct Investment Abroad

(In billions of dollars)

Source: U.S. Department of Commerce

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Stocks of Foreign Direct Investment in the United States and U.S. Direct Investment Abroad

(In billions of dollars)

Source: U.S. Department of Commerce

• Download Figure
• Download figure as PowerPoint slide

Stocks of Foreign Direct Investment in the United States and U.S. Direct Investment Abroad

(In billions of dollars)

Source: U.S. Department of Commerce

• Download Figure
• Download figure as PowerPoint slide

The Determinants of Foreign Direct Investment

Over the years, there have been many efforts to explain why firms engage in foreign direct investment. Perhaps the place to start is to note that, in a world of perfect competition, most foreign direct investment simply would not occur: multinational corporations would incur transactions costs not incurred by their domestic rivals, and these costs would drive returns to subnormal levels. It should be emphasized as well that foreign direct investment is not simply an international flow of financial capital, as might occur if arm’s-length investors in capital-rich countries simply sought higher returns in capital-scarce countries: while such flow does occur, it is in the form of portfolio investment rather than direct investment. Given these considerations, analysts have largely looked to the internal characteristics of multinational corporations in order to explain foreign direct investment.

One of the earliest studies was made by John Dunning, who examined the operations of British affiliates of U.S.-based multinational corporations relative to the British-owned rivals of these affiliates (Dunning (1958)). He found that the U.S.-owned affiliates were more productive than their local rivals, a fact that Dunning ascribed to the abilities of the former to transfer technologies and other “intangible assets” (for example, marketing and other managerial skills) from the U.S.-based parent firms to the United Kingdom and to adapt these to the British environment. But also, over time, the British firms caught up with their U.S.-owned rivals, leading Dunning to conclude that the overall effect of U.S. foreign direct investment was to raise British levels of productivity in all firms in an industry and thus generate net benefits for the British economy. Dunning’s enthusiasm for foreign direct investment in the United Kingdom almost forty years ago is mirrored in the enthusiasm of many countries for foreign direct investment today.

A second pioneering study of the same epoch by Stephen Hymer explored the nature of the internal characteristics of multinational corporations that enabled them to engage in foreign direct investment, identified as “economies of scale” and “special management skills” (Hymer (1960)). Hymer was gloomier than Dunning in his assessment of multinational corporations; he argued that, while these firms were likely to be more efficient than their rivals, they would also gain market power so that the benefits of competition were lost, with the latter effect more than offsetting the former benefits. This last concern came to dominate a wide-ranging debate on the merits of the multinational corporations during the 1970s. At that time, it was very much in fashion, especially in developing countries, to view multinational corporations as powerful, monopolistic institutions that would corrupt governments and reduce host nations’ welfare.¹⁵ As a consequence, policies proliferated in the 1970s to restrict inward foreign direct investment and regulate closely the activities of multinational corporations. Such policies were mostly implemented by developing countries but also by some industrial countries, such as Canada.

Composition of Flows of U.S. Direct Investment Abroad

(In billions of dollars)

Source: U.S. Department of Commerce.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Composition of Flows of U.S. Direct Investment Abroad

(In billions of dollars)

Source: U.S. Department of Commerce.

• Download Figure
• Download figure as PowerPoint slide

Composition of Flows of U.S. Direct Investment Abroad

(In billions of dollars)

Source: U.S. Department of Commerce.

• Download Figure
• Download figure as PowerPoint slide

Some analysts, while also starting from Hymer’s analysis, arrived at quite different conclusions. In a series of influential articles and books, Raymond Vernon emphasized the importance of new product technology as a determinant of both international trade and investment and the role of factors specific to the home markets of firms as a determinant of how this technology gets created and diffused (Vernon (1966), (1971a), and (1971b)). Taking a somewhat different tack, Peter Buckley and Mark Casson noted that international exploitation of firm-specific advantages internationally does not require that a firm actually own and manage international operations (Buckley and Casson (1976)). Rather, the advantages could be exploited via export or licensing agreements with firms based in countries outside the home market (as, for example, many multinational corporations struck with Japanese firms during the 1960s and 1970s, when Japan severely restricted foreign direct investment inflows). Thus, to explain why operations in one country are actually owned and managed by a firm in some other country, Buckley and Casson turned to the organizational theory of the firm, as originally developed by Ronald Coase and expanded by Oliver Williamson (Coase (1937) and Williamson (1975)). According to these authors, there are economies to be realized by internalizing within a single organization production and marketing functions. These economies of internalization result, in essence, from firms’ efforts to avoid the high transactions costs that are associated with attempting to achieve similar outcomes through buying and selling services in external markets. Such transactions costs include opportunity and moral hazard costs. If these economies can continue to be realized by expanding the organization to cross national boundaries, foreign direct investment occurs.

Buckley and Casson triggered a number of subsequent studies applying the organizational theory of the firm (a theory that itself has seen considerable refinement in recent years) to the multinational corporation.¹⁶ For example, John Cantwell posits that rivalry among multinational corporations leads to more rapid development and diffusion of desirable new technologies than would occur in a world where there were no economies of internalization and where technology was transferred by licensing (Cantwell (1989)). This suggests that foreign direct investment has a positive effect on economic efficiency and growth, the conclusion first reached by Dunning. In the following sections, some of the logic and evidence for this position is examined from the perspective of both home and host countries.

Before moving to this examination, however, it should be noted that there are aspects of foreign direct investment and the multinational corporation not adequately explained by Buckley and Casson or subsequent theories based on the organizational theory of the firm. In particular, these theories cannot explain why foreign direct investment has undergone surges. There are economic theories of the determinants of foreign direct investment and the multinational corporation not based on the organizational theory of the firm, for example, ones based on the dynamics of oligopoly and on new theories of economic geography, that have some explanatory power. For reasons of space, these are not reviewed here.¹⁷ For the moment, however, most analysts agree that explanations of foreign direct investment based on the organizational theory of the firm have more power than other genres of theory (see, for example, Graham and Krugman (1995)).

The Economic Consequences of Foreign Direct Investment for Host Countries

Foreign direct investment and multinational corporations can have both positive and negative economic effects on host countries. Positive effects come about largely through transfer of technology and other intangible assets, leading to productivity increases that improve the efficiency of resource utilization and ultimately result in higher per capita income. As has already been suggested, such effects can be direct, for example, if subsidiaries of multinational corporations are more productive than local rivals, or if they transfer technologies or other assets to local suppliers, distributors, or other firms with which the multinational corporations do business and, by doing so, enhance the productivity of these. But these effects can also come about indirectly, for example, if increased interfirm rivalry engendered within a sector by the entry of multinational corporations leads to all firms in that sector becoming more productive.

“External benefits” associated with multinational corporations can also boost productivity. For example, a multinational corporation might employ local workers (including at technical and managerial levels), who, as a result of this employment, upgrade their own knowledge and skills and subsequently leave the multinational corporation and become employed elsewhere. To the extent that their new knowledge and skills can be utilized in their new positions, such knowledge and skills must be counted as external benefits associated with the multinational corporation, that is, benefits that are captured neither by the multinational corporation itself nor the users of its products or services. Both direct benefits, brought about by linkages between multinational corporations and local firms, and indirect benefits, whether created via increased rivalry or the generation of external benefits, are typically termed “spillover effects.”

Negative economic effects can also be both direct and indirect. Direct negative effects, from a purely economic perspective, can arise from the market power of the multinational corporation and its ability to use this power to generate supranormal profits and transfer these to its shareholders, who presumably are not residents of the host country. In addition to negative economic effects, the multinational corporation might be capable of indirectly creating negative economic effects for the host country. For example, multinational corporations might be able to influence the local political process to the economic detriment of the host country’s economy (for example, by inducing politicians to grant to the multinational corporation direct or indirect subsidies, such as investment incentives or protection from imports in the local market).

There is no reason why, in principle, the positive effects should be dominated by the negative effects or vice versa. This indeterminacy is perhaps why debate over multinational corporations has long been lively and subject to “sea change.” As suggested in the previous section, it was the fashion during the 1970s, especially in developing countries, to view the negative effects as dominant, whereas in most of those countries today the positive effects are largely viewed as dominant. Interestingly, in the United States, where official policy had been to stress the positive aspects of foreign direct investment throughout the 1950s, 1960s, and 1970s, when foreign direct investment began rapidly to rise during the late 1980s, many politicians began to question the benefits of this influx, and some restrictive measures have in fact been passed (Graham and Krugman (1995)). Thus, to an extent, trends in official policy of the United States have moved against the overall world trend.

Given that, in principle, the economic effects of foreign direct investment can be in net positive or negative, the issue becomes an empirical one: as events actually transpire, which effects dominate? The first thing to be said on this issue is that it is truly difficult to measure the effects of foreign direct investment and multinational corporations. Does participation by multinational corporations, for example, increase or decrease firm concentration in the affected industries? Several points can be made in this regard: (1) there is an association between global industry concentration and participation by multinational corporations, but the correlation is not exact (the R² is far from 1.0);¹⁸ (2) however, over time, the trends in industry concentration have been nonuniform, as concentration in industries with heavy multinational corporation participation generally fell from 1962 to 1982 (Dunning and Pearce (1985)) but rose in the later 1980s; and (3) at the level of individual countries, the trend in most such industries over the past twenty years has been for concentration to rise.¹⁹

However, as has been noted by numerous analysts, a rise in industry concentration does not necessarily imply an increase in monopoly power within an industry, especially at the level of individual countries. Effective competition within a country might actually be negatively correlated with domestic industrial concentration. If, for example, the concentration were to be the consequence of an opening of the country to international competition, followed by the rationalization of the domestic industry to become more efficient, domestic industrial concentration could rise even though effective competition certainly increased. Likewise, a fall in industry concentration does not necessarily imply a reduction in monopoly power; for example, a trend within an industry toward greater differentiation of products that were imperfect substitutes might lead to an increase in production of individual product varieties by just one or a small number of firms that monopolize individual market segments. Whether overall competition were to be effectively rising or falling would then depend upon the degree of intervariety competition.

In a word, while there is some information to be had in the study of firm concentration, it ultimately does not shed much light on whether multinational corporations do or do not achieve monopoly rents in host countries. To determine this, one might think that studies of firm profitability would be fruitful. But this route, too, leads to indeterminant results. This outcome is particularly true if one seeks to investigate the profitability of multinational corporations’ operations in individual host countries: because multinational corporations can (and apparently do) use transfer prices to shift reported profits from one locale to another (largely for reasons of tax minimization), it is impossibly difficult to determine what really is the return of a multinational corporation in any particular location.²⁰ But even if one looks to the worldwide profitability of multinational corporations, one must adjust for such variables as differences in intensity of factor usage (multinational corporations, over a range of industries, tend be more capital intensive in their factor utilization than do nonmultinational rivals; hence, the former might be expected to report higher profit rates than the latter). From the many studies of the profitability of multinational corporations, the weight of the evidence seems to be that these firms are marginally more profitable than their nonmultinational rivals, but that the difference largely disappears (or becomes statistically insignificant) when factor intensity is controlled for.

Thus, the evidence on the existence, let alone the magnitude, of negative economic consequences of foreign direct investment and multinational corporations is inconclusive but tending toward dismissal of the idea that these consequences are weighty. What about the evidence for positive economic consequences?

The evidence is overwhelmingly positive with respect to direct effects. Most of this evidence is based on studies of technical efficiency. In terms of technical efficiency, a mass of empirical data supports the hypothesis that multinational corporations significantly outperform domestic rivals in host countries.²¹ One major qualifying comment that must be added, however, is that most of the empirical studies on this issue have focused on labor productivity, which can be increased (without an overall increase in multifactor productivity) by substituting capital for labor. Because, as just noted, multinational corporations typically utilize more capital-intensive production methods than do their nonmultinational rivals, some of the measured differences in labor productivity might be ascribable to differing factor intensities. However, a limited number of empirical investigations suggest that the multifactor—as well as the labor—productivity of multinational corporations tends to be higher than that of nonmultinational rivals.

Likewise, numerous empirical studies confirm the presence of spillover effects from multinational corporations in host countries that create benefits to the local economy. As might be expected, the evidence is very strong that linkages between locally owned firms that act as suppliers or distributors for local affiliates of multinational corporations and these affiliates tend to increase the efficiency of the former. However, some studies have suggested that in the manufacturing sector foreign-owned firms do not necessarily provide more technical assistance to their suppliers than do efficient locally owned firms (see. for example, Goncalves (1986)).

Similarly, there seem to be positive spillovers on competitor firms generated by the participation of multinational corporations in specific sectors. Dunning defines two categories of studies of the effects of multinational corporations upon locally owned competitor firms: those that are based on field studies and those that are based on econometric studies (Dunning (1993)). Both categories tend to support the contention that inward foreign direct investment tends to act as a stimulus to enhance the technical efficiency of local firms that must compete against multinational corporations. Two qualifying statements must be made, however. First, such studies (of both categories) have tended to examine the medium-term impact of the entry of multinational corporations on labor productivity of local rivals and have not considered the long-term impact on, say, local research and development capability. Multinational corporations have historically concentrated research and development activity in their home countries; thus, if the entry of multinational corporations causes local firms to reduce research and development (and if this research and development produces benefits captured in the local economy), some local benefit-—likely long-term in nature—might be lost,²² Second, most of the studies have been cross-sectional in nature; they show that domestic firms that must compete in sectors in which multinational corporations participate tend to have higher technical efficiencies than domestic firms that do not face such competition. These studies thus do not show whether the technical efficiencies of the domestic competitors to multinational corporations improved following the entry of the latter. A limited number of time-series studies of this issue have been conducted, with mixed results (see. for example, Globerman (1985) and Dunning (1985)).

These empirical observations are roughly consistent with a model introduced by Grossman and Helpman dealing with endogenous technological innovation and trade (Grossman and Helpman (1991, Chap. 7)). The formal model is a standard two-country model, in which one country is labor rich in its factor endowment and the other rich in human capital. The principal conclusion of this model is that, when technological innovation is endogenous (that is, the result of private entrepreneurs who engage in research and development as a profit-maximizing activity), and where knowledge diffuses internationally, a variant of the Heckscher-Ohlin result holds, notably that research and development takes place in the country where factor configurations make this a relatively low-cost activity (namely, the country that is rich in human capital). Under certain world factor configurations, however, the cost of manufacture of goods embodying the technology might be lower in the other country, giving firms an incentive to become multinational by manufacturing in locations other than where the research and development is performed. The labor-rich country in effect trades commodities for technology.

Overall, the evidence seems to support those who maintain that foreign direct investment tends to have net positive effects on host country economies. It would seem that, to the extent that there are negative effects associated with the monopoly power of multinational corporations, these are rather minimal and do not manifest themselves in the form of significant monopoly rents. Also, it is probably true that if foreign direct investment continues to proliferate, entry by new multinational corporations from a multitude of home countries will reduce the monopoly power of incumbent multinational corporations. The evidence that foreign direct investment does convey positive effects through technology transfer and transfer of other intangible assets, in contrast to the evidence on negative effects, is very convincing.

This last point is very important to developing countries. As is explained in the next section, inward foreign direct investment likely reduces in net the amount of international savings available to these countries. This effect, taken by itself, would reduce capital formation in these countries and their long-term growth prospects. However, this effect is likely to be more than offset by growth in productivity owing to technology transfer, such that the net effect on economic growth is positive. Furthermore, increases in growth seem to be correlated with increases in the domestic savings rates of dynamic developing countries, such that any reduction in international savings might be more than offset by increases in domestic savings.²³ The net effect would be a “virtuous circle” of increased domestic savings leading to increased domestic capital formation and, hence, additional growth. In all likelihood, the contribution of foreign direct investment to growth in developing countries is thus strongly positive. Recent empirical work by Borensztein, De Gregorio, and Lee (1994) largely supports these conclusions. Their work indicates that foreign direct investment figures importantly in the transfer of technology to developing countries and thus contributes more to growth than domestic investment. However, there is a qualifying element: this result holds only when the host country has a “minimum threshold stock of human capital,” that is, it does not seem to hold for very poor countries with relatively few educated citizens. Also, their results show that foreign direct investment has the effect of increasing total investment in the economy more than one for one, suggesting complementarity between foreign direct investment and fixed domestic capital formation by domestically owned firms.

The Economic Consequences of Foreign Direct Investment for Home Countries

Concern over the effects of foreign direct investment on home countries has long been focused on employment, that is, on whether outward foreign direct investment leads to some form of job loss in the home country.²⁴ To the professional economist, such concern is largely misplaced, because overall levels of employment are determined by macroeconomic factors on which foreign direct investment has little bearing. The concern is more legitimate, however, if one talks about the quality of employment. The following questions are then relevant:

(i) Does foreign direct investment lead to sectoral reallocation of employment, such that workers are released from labor-intensive sectors in greater numbers than they can be reabsorbed at prevailing wages in other sectors, and such that the labor share of national income must fall in order for labor markets to clear? In other words, does the Stolper-Samuelson effect hold? Alternatively, does outward foreign direct investment reduce the demand for high-skill workers in the affected industries, such that these workers must seek jobs in other industries, where they are not fully compensated for the skills that they possess?²⁵ That is, does outward foreign direct investment cause “bad” (low-paying) jobs to be substituted for “good” (high-paying) ones?

(ii) Does outward foreign direct investment reduce capital formation at home, such that productivity (and, hence, per capita income) grows at a lower rate than it would have, had the investment not taken place?

(iii) Does outward foreign direct investment have any effect on the rate of technological innovation in the home country?

Closely related to issue (i) is whether or not foreign direct investment and international trade are complements or substitutes, that is, whether increases in outward foreign direct investment are related to increases or decreases in exports or to increases in imports of related goods. (It is implausible that increases in foreign direct investment would be related to decreases in imports.) If, say, increases in foreign direct investment were characteristically to lead to decreases in exports (that is, foreign direct investment and exports were to be substitutes) or to increases in imports of goods or services in the same sector (that is, foreign direct investment and imports were to be complements), the case could be made that outward foreign direct investment had similar effects on the home-country economy as import liberalization in the relevant sector (as well as similar effects on labor markets, namely, that the labor share of national income could be reduced or the sectoral composition of demand for labor altered—or both—-to the detriment of labor income). However, if increases in outward foreign direct investment were to lead to increases in exports (that is, foreign direct investment and exports were to be complements). It would be difficult to construct a case that foreign direct investment had harmful effects on labor, assuming that jobs in the affected sector were considered to be “good” jobs to begin with!

The issue of whether foreign direct investment and exports are substitutes or complements has been rather extensively investigated empirically, and the bulk of the evidence points toward a complementary relationship rather than a substitutive one.²⁶ Less research has been performed on the relationship between foreign direct investment and imports, but the available evidence suggests that this relationship, too, is complementary (Bergsten and Graham (forthcoming, App. A)). Thus, outward foreign direct investment seems to be positively associated with trade expansion. Both exports and imports seem to expand as foreign direct investment grows. It should be noted that the trade expansion is likely to be intraindustry, that is, exports and imports will tend to grow within sectors and not necessarily across sectors, suggesting that much of the trade is of differentiated products.

What are the implications for labor? Whether the effect of foreign direct investment is to reduce labor’s share of national income cannot be determined from the data, but such an effect seems highly unlikely (there is no evidence that foreign direct investment causes imports to expand in labor-intensive sectors). The effect on the sectoral composition of labor is, however, likely to be positive, because of the complementarity between foreign direct investment and exports and the evidence supporting the contention that export-generating sectors also generate “good” jobs (that is, ones that pay higher-than-average wages) in the advanced countries.

The effects of foreign direct investment on capital formation in home countries depend largely upon the international capital flows that it creates. As noted earlier, these flows effectively add to or subtract from the pool of savings available to finance capital formation. Given that I ≡ S, where I is domestic investment (which includes new capital investment but also net additions to inventories) and S is the total savings pool (including international sources), it follows that Δi = ΔS, and, ignoring inventory changes, it would seem that a net increase in capital outflows from a host country will displace domestic investment on a one-to-one basis.²⁷ However, foreign direct investment generates a variety of international capital flows, including various payments to the parent firms of multinational corporations located in the home countries, as well as the foreign direct investment flows themselves. (Moreover, as Chart 4 indicates quite clearly for the United States, new equity outflows are only a part of U.S. foreign direct investment outflows in recent years.) In addition, foreign direct investment can affect the payment of taxes by multinational corporations to their home governments (from a national income accounting perspective, taxes are part of national savings) and, perhaps, the profitability of home operations of multinational corporations (affecting their retained earnings, which is another component of national savings).

The upshot of all of this is that it is truly difficult to assess the effect of foreign direct investment on capital formation in home countries. There is, in fact, a case to be made for asserting that in the long run outward direct investment increases the capital stock of a country, because direct investors will require that their investments have net positive present value and, hence, that future capital inflows must in magnitude exceed current outflows. But this assertion begs the question of what exactly is the counterfactual. If the alternative to foreign direct investment is additional current consumption, obviously a different consequence will follow than if the alternative is additional current investment.

A corollary would be that inward direct investment in the long run reduces the international savings of a host country. While this is almost surely true, it is not at all clear that the effect is either reduced savings available to the host country or reduced capital formation. Indeed, at the conclusion of the previous section, the case was made that the overall effect would be increased growth and increased capital formation resulting initially from an increased efficiency of use of capital and sustained by a virtuous circle.

If the effects of foreign direct investment on domestic capital formation are, at root, indeterminant, its effects on domestic research and development activity are more clearly likely to be favorable. This result follows from the following observations.

First, there is a divergence between the total social returns generated by new technology creation and the returns appropriated by the innovator (because of spillover effects). This is a classical result, and one consequence is that private agents, left to their own devices, will tend to underfund research and development relative to what would be the socially optimal level of such funding.

Second, foreign direct investment increases the total size of the market available to the innovating firms and, hence, the appropriable returns available to them. The consequence is that they should be willing to invest more in research and development. Furthermore, as an empirical point rather than a theoretical one, firms tend to do this in their home countries, although research and development by multinational corporations (that is. research and development not performed in their home countries) has in recent years grown rapidly. The unequivocal result is that foreign direct investment should, all else being equal, increase the amount of research and development that is performed. It has also been argued that increased rivalry among multinational firms occasioned by foreign direct investment will accelerate the rate at which these firms introduce the fruits of research and development (new or improved products, or more efficient ways of producing them) to the market (Graham (1985)).

That multinational activity should boost innovation in the home country thus seems a plausible hypothesis, A contrary hypothesis can, however, be offered. For example, Porter (1990) suggests that, in some sectors, U.S.-based firms have used foreign direct investment as a means to switch from higher-cost to lower-cost areas, and that this has actually suppressed incentives to invest in product and process innovation. The question then posed is, Which of these hypotheses is, in a real world context, the dominant one on the basis of empirical evidence? Does, in fact, multinational activity boost innovation in the home country? The answer, unfortunately, is that there is a dearth of hard empirical evidence bearing on this issue. The general empirical evidence that does exist, however, tends to support the former hypothesis that outward foreign direct investment stimulates technological innovation in the home country (see, for example, Mansfield, Romeo, and Wagner (1979)). Evidence for the latter hypothesis tends to be in the form of case studies of particular industries, and one might conclude that, while the hypothesis might ring true for certain specific cases, as a general proposition it is not on the mark.

The New Policy Environment for Foreign Direct Investment

Although multinational corporations carry out both a large portion of the world’s trade and most of its direct investment, the international legal framework for trade is very detailed, whereas that for direct-investment-related activity (other than trade) is quite limited. This asymmetry has long been observed by certain scholars, who have suggested that there might be benefits to extending the multilateral trade rules to cover direct investment and allied activities.²⁸ Some progress was made on that front in the Uruguay Round of multilateral trade negotiations. Three of the Uruguay Round agreements have significant bearing on foreign direct investment: the Agreement on Trade-Related Investment Measures, which proscribes governmental mandating of certain performance requirements (specifically, local content and trade-balancing requirements) that were determined to be inconsistent with obligations under the General Agreement on Tariffs and Trade; the General Agreement on Trade in Services, which, interalia, binds governments to certain right-of-establishment and national treatment standards for foreign-controlled enterprises in specified industries; and the Agreement on Trade-Related Intellectual Property, which obligates World Trade Organization (WTO) members to certain standards on intellectual property.

Despite this progress, many trade policy officials believe that more language should be incorporated into the world’s multilateral trading rules pertaining to foreign direct investment, perhaps in some future multilateral agreement on investment.²⁹ As of the time of this writing, two very general issues in this regard were outstanding, notably (i) the venue in which such an agreement would be negotiated and (ii) the substantive content of the agreement.

On the issue of venue, the choice at the moment seems to boil down to the Organization for Economic Cooperation and Development (OECD) or the WTO. The case for the OECD as venue is that there is more consensus on substantive issues among its members, the industrial countries, than among a larger grouping and, hence, that progress could be faster, and the result more profound, in this venue than in the WTO. The case for the WTO as a venue is that almost all countries are members or prospective members and that a future multilateral agreement on investment would be effective only if a larger group of countries than the OECD member countries were to agree to abide by its rules.

On the issue of substance, there seems to be widespread consensus as to the broad substantive issues that a future multilateral agreement on investment should cover. Included in such an agreement would be principles pertaining to transparency (host and home countries’ laws and policies regarding direct investment and related activities should be public knowledge and as unambiguous as possible), right of establishment, national treatment (investments, including direct investments, of foreigners should be subject to treatment under law that is no less favorable than that accorded to equivalent domestically owned investments), and investor protection (including principles pertaining to expropriation). Substantial disagreement among countries, however, lurks beneath the surface at the level of detail. In particular, how to deal with the likelihood that all nations would claim exceptions to the general principles is problematic.

There is also considerable consensus that an agreement should deal with investment measures that have the potential to distort trade flows or otherwise have effects equivalent to trade barriers. These measures would include performance requirements not currently covered by the Agreement on Trade-Related Investment Measures and perhaps also investment incentives and other subsidies granted by governments to induce foreign direct investors to locate facilities within their jurisdictions. There is much sympathy for the idea that a multilateral agreement on investment should establish procedures to settle disputes between investors and states, in order to supplement existing state-state dispute settlement procedures embodied in the WTO.

A potential model for a future investment agreement is Chapter 11 of the Agreement Establishing the North American Free Trade Association (NAFTA Chapter 11), which contains all of the elements listed above. However, some critics have noted that NAFTA Chapter 11 has done little except to codify the existing investment rules of the three NAFTA countries (Canada, Mexico, and the United States). These critics believe that, without liberalization of existing investment restrictions, negotiation of a multilateral agreement would be a sterile exercise. Other critics, however, believe that NAFTA already goes too far and that, in particular, the investor-state dispute settlement mechanisms in NAFTA Chapter 11 encroach unduly upon national government sovereignty.

These difficulties notwithstanding, it was agreed at the meeting of the OECD countries at the ministerial level in May 1995 to begin negotiations on an investment agreement. The WTO will likely consider any new investment initiative before its ministerial-level meeting to be held in Singapore in December 1996. Thus, by the time that the WTO meeting is held, there might be some progress within the OECD.

The view of this author on the issue of venue is that, in light of the timing of the two ministerial meetings, substance might actually be allowed to drive venue. If substantive progress is being made at the OECD, meaning progress toward liberalization, it would then be advisable to allow the OECD effort to continue without a parallel effort being launched at the WTO. Then, at some later time, the fruits of the OECD work could be the basis for negotiations at the WTO. If no such progress is made at the OECD, by contrast, it would seem advisable that the whole effort be allowed to migrate to the WTO. As has been noted by a number of trade policy experts, oftentimes trade liberalization is more easily achieved in a “big package” than in a small one. The reason is that political constituencies in favor of liberalization can be counted on more reliably to provide political support for a big package than a small one, whereas often a small package galvanizes opposition from special interests more effectively than the liberalization constituency.³⁰ Thus, failure to achieve liberalization at the OECD might not necessarily imply automatic failure at the WTO.

If neither the OECD nor the WTO proves to be fertile ground for a multilateral agreement on investment, there are other possible venues. Prime among these would be the Bretton Woods Institutions, the World Bank or the IMF. The World Bank already contains much expertise in the field of direct investment policy, notably within the International Center for the Settlement of Investment Disputes (ICSID) and the Multilateral Investment Guarantee Agency (MIGA). The ICSID was created to facilitate settlements of disputes between investor firms and host countries. At present, 130 nations are signatories to the ICSID. However, the ICSID has not been frequently used as a facility actually to settle investment disputes. But the ICSID could become much more active in the future if it were used, as envisaged, as the principal arbitral body for the settlement of investment disputes under Chapter 11, Part B of NAFTA.

The MIGA is an institution designed to supplement and perhaps eventually supplant national investment insurance programs. The MIGA was designed to encourage foreign direct investment specifically in developing economies. Like the ICSID, the MIGA to date appears to be somewhat underutilized. In 1994, 147 countries were signatories to the MIGA, but only about one hundred contracts were outstanding, with contingent liabilities of about $ 1 billion. However, the demand for MIGA guarantees has been rising, especially for insurance associated with investment in the formerly socialist countries.

The main advantage of the World Bank thus is that many of the facilities that might be attached to a multilateral investment agreement already exist within it. If the venue chosen for such an agreement were to be the World Bank’s sister organization, the IMF, these facilities would be located close by. The advantage of the WTO is that trade and investment policy issues are complementary and, thus, it would be desirable for the secretariats associated with a new investment agreement to work closely with the existing WTO staff. However, the overriding objective would be to achieve the best possible agreement; if choice of venue were to affect outcome, the venue that enabled this objective to be achieved should be the one chosen.