In Search of Effective Policies for Foreign Direct Investment: Alternatives to Tax Incentive Policies
Alternatives to Tax Incentive Policies
In Search of Effective Policies for Foreign Direct Investment: Alternatives to Tax Incentive Policies
0 This Article is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted for inclusion in Northwestern Journal of International Law & Business by an authorized administrator of Northwestern University School of Law Scholarly Commons
Recommended Citation Kojo Yelpaala, In Search of Effective Policies for Foreign Direct Investment: Alternatives to Tax Incentive Policies, 7 Nw. J. Int'l L. & Bus. 208 (1985-1986)
Part of the International Law Commons; International Trade Commons; and the Tax Law
In Search of Effective Policies for Foreign
Alternatives to Tax
I. INTRODUCTION ...........................................
The Need for Reevaluation ...........................
THE MNE SYSTEM AND INCENTIVES .....................
INTANGIBLE ASSETS THEORY OF FDI .....................
Market Imperfections and FDI .......................
B. Cost Factors in FDI ..................................
C. Intangible Assets-Incentive Policy Implications .......
1. Market Imperfections and Incentive Policies ........
2. Cost and Incentive Policies ........................
Summ ary ............................................
versity of Ghana, 1970; M.B.A. Bowling Green State University, Bowling Green, Ohio, 1972; Msc.
1979; S.J.D. University of Wisconsin-Madison, 1985. I am immensely grateful to Professors Irish,
Bilder, and Strasma ofthe University of Wisconsin-Madison who were first exposed to some of the
views expressed in this article in my dissertation and gave me very helpful comments. I am also
grateful to the University of Wisconsin-Madison Law School for supporting the research on my
dissertation and the McGeorge School of Law for the assistance that facilitated the completion of
INTERNALIZATION THEORY OF FDI .......................
A. Incentive Policy Implications .....................
1. Tax Incentive Policy Implications ..................
3. Other Policy Implications..........................
It is now largely recognized that the multinational enterprise
("MNE") can play a significant role in the industrialization of a number
of different countries.1 The major way in which the MNE can contribute
toward the industrialization of a country is through foreign direct
investment ("FDI"). To induce such MNE investment, several host countries
1 Several studies have identified the potential ofthe MNE in the development of host countries.
See, e.g., UNITED NATIONS CENTRE ON TRANSNATIONAL CORPORATIONS ("UNCTC"),
TRANSNATIONAL CORPORATIONS IN WORLD DEVELOPMENT: A RE-EXAMINATION (1978); UNITED
NATIONS CENTRE ON TRANSNATIONAL CORPORATIONS, TRANSNATIONAL CORPORATIONS IN WORLD
DEVELOPMENT: THIRD SURVEY, U.N. Doc. ST/CTC/46 (1983) [hereinafter UNCTC, THIRD
SURVEY]; Hymer, The MultinationalCorporationand the Law of Unevendevelopment, in ECONOMIcS
AND WORLD ORDER: FROM THE 1970s TO THE 1990s (J. Bhagwati ed. 1972). See also THE
MULTINATIONAL CORPORATION IN THE 1980s (C. Kindleberger & D. Audretsch eds. 1983); R.
VERNON, STORM OVER THE MULTINATIONALS: THE REAL ISSUES (1977) [hereinafter R.
VERNON, STORM]; G. REUBER, PRIVATE FOREIGN INVESTMENT IN DEVELOPMENT (1973). More
recent indications of the belief in the potential of MNEs to affect the development of host countries
can be found in the United States' encouragement of foreign direct investment ("FDI") in the
Caribbean Basin Region.
The commitment of the Reagan Administration to encouraging United States foreign investors
to invest in the Caribbean Basin Region was well articulated by President Reagan; see Caribbean
Basin Initiative, Address Before the Permanent Council of the Organization of American States
("OAS"), 18 WEEKLY COMP. PRES. Doc. 217 (Feb. 24, 1982). See also Caribbean Basin Initiative
Legislation, Statement by the President, 19 WEEKLY COMP. PRES. Doc. 240 (Feb. 16, 1983);
Caribbean Basin Initiative, Remarks to Ambassadors of Member Nations of the OAS, 18 WEEKLY COMP.
PRES. Doc. 1655 (Dec. 22, 1982). See also H.R. 6715, 97th Cong., 2d Sess., 128 CONG. REc. 3968
(1982)-a bill to amend the Internal Revenue Code of 1954 to allow a credit against income tax for
the use of certain forms of business insurance for new business activities in designated Caribbean
have relied significantly on fiscal incentives in general and tax incentives
in particular for over half a century. However, after several decades, the
effects of these tax incentives on the motivation of MNEs to invest in
incentive granting countries continues to be the focus of a heated debate.
While some believe that tax incentives do actually motivate MNE
investment behavior,4 others believe that such motivational effects are either
2 Several studies have discussed host countries' tax incentives laws and policies. See, eg.,
Kopits, Effects of Tax Changes on Direct Investment Abroad, in UNITED STATES TAXATION AND
DEVELOPING COUNTRIES 223 (R. Hellawell ed. 1980); Patrick, U.S. Tax Treaties With Developing
Countries, in UNITED STATES TAXATION AND DEVELOPING COUNTRIES 307 (R. Hellawell, ed.
1980) (There are several other interesting discussions of this subject in this book.) See also R.
GROSSE, FOREIGN INVESTMENT CODES AND THE LOCATION OF DIRECT INVESTMENT (1980); R.
ANTHOINE, TAX INCENTIVES FOR PRIVATE INVESTMENT IN DEVELOPING COUNTRIES (1979);
TAX INSTITUTE OF AMERICA, TAX INCENTIVES (1971); Adler, U.S. Taxation of US. Multinational
Corporations:A Manualof Computation Techniques andManagerialDecision Rules, in 2
INTERNATIONAL FINANCE AND TRADE 157 (M. Sarnat & G. Szeg6 eds. 1979); Surrey, InternationalTax
Conventions: How They Operate and What They Accomplish, 23 J. TAX'N 364 (1965); Surrey, Tax
Incentives as a Devicefor Implementing Government Policy: A Comparison with Direct Government
Expenditures, 83 HARV. L. REV. 705 (1970); Patrick, U.S. Tax Policy and Foreign
InvestmentsLegislative and Treaty Issues, 5 DEN. J. INT'L L. & POL'Y 1 (1975). For other studies on national
tax legislation and the FDI process, see also UNITED NATIONS CENTRE ON TRANSNATIONAL
CORPORATIONS, NATIONAL LEGISLATION AND REGULATIONS RELATING TO TRANSNATIONAL
CORPORATIONS, U.N. Doc. ST/CTC/26 (1983) [hereinafter UNCTC, NATIONAL LEGISLATION].
3 See, e.g., D. HARTMAN, DOMESTIC TAX POLICY AND FOREIGN INVESTMENT: SOME
EVIDENCE (National Bureau of Economic Research, Inc. Working Papers Series, Working Paper No.
784, 1981); Hartman, The Effects of Taxing Foreign Investment Income, 13 J. PUB. ECON. 213
(1980); Hartman, Tax Policy and Foreign Direct Investment in the United States, 37 NAT'L TAX J.
475 (1984); Batra & Ramachandran, MultinationalFirmsand the Theory of InternationalTrade and
Investment, 70 AM. ECON. REV. 278 (1980); Shah & Toye, FiscalIncentives for Firms in Some
Developing Countries: Survey and Critique,in TAXATION AND ECONOMIC DEVELOPMENT 279-80
(J. Toye ed. 1978). For a recent study on the determinants of FDI, see Schneider & Frey, Economic
and PolicitalDeterminantsof Foreign Direct Investments, 13 WORLD DEV. 161 (1985).
4 See, e.g., Hartke, A Foreign Trade and Investment Policyfor the 1970s, in AMERICAN LABOR
AND THE MULTINATIONAL CORPORATION 54 (D. Kujawa ed. 1973), where he had the following to
say about U.S. tax policies:
Tax Policies-U.S.tax policies have been designed to encourage the outflow of U.S. capital at
the expense of domestic investment....
Our present tax laws create an attractive tax shelter for any income earned abroad by a
U.S. corporation. Income earned in the United States is ordinarily taxed in the year when
earned. Under our tax laws, however, income earned abroad is never taxed until it is brought
home to the States. The unregulated U.S. transnational is in the position of earning capital in
the United States, investing that capital overseas, and never paying a dime in U.S. taxes on any
income earned abroad.
Even if the profits are brought back to the United States, existing tax law permits the U.S.
corporation to take a dollar for dollar creditagainst domestic federal tax liability for any income
taxes paid abroad. The domestic corporation that has chosen to invest in Indiana must pay
state taxes and cannot take them as a credit against its federal income tax liability.
Id. at 59 (emphasis in original).
See also Goldfinger, An American Trade Union View ofInternationalTrade and Investment, in
AMERICAN LABOR AND THE MULTINATIONAL CORPORATION, supra, at 28, where he presents the
proposals of the AFL-CIO to include the following:
negligible or nonexistent.5 This debate is hardly a trivial matter because
whether MNEs are motivated or not by tax incentives has important
implications for labor, business, and government policy of home countries.
The current debate over the runaway plant in the United States6
assumes the efficacy of tax incentives. If on the other hand, tax
incentives are not an effective means of attracting FDI, then countries
interested in attracting MNE involvement in their local economies need to
adopt more appropriate policies. A policy is appropriate for these
purposes if it addresses the actual motivational factors in MNE foreign
operations. In other words, an appropriate FDI incentive policy is one that is
closely tied to or derived from the theory of the MNE. The question,
then, is whether current tax incentive policies are appropriate.
For years, tax incentive policies have been based on the capital
arbitrage theory of FDI.7 The capital arbitrage theory, however, is no longer
considered an adequate explanation of the FDI process. Consequently,
any incentive policies based on this theory are likely to be both misplaced
and ineffective. It is the purpose of this article to investigate possible
The U.S. government must stop helping and subsidizing U.S. companies in setting up and
operating foreign subsidiaries.
Taxes on U.S. corporations' overseas operations are made to more closely relate to the tax
Foreign operations of U.S. firms should carry the same load as their domestic operations.
Current U.S. tax provisions encourage movement abroad, even though $15 billion was spent in
1971 by U.S.-based firms in new plant and equipment outside the United States and a host of
technology transfers took place. These foreign investments have increased 60 per cent faster
than U.S. domestic investments in the past decade.
Id. at 49.
5 See, e-g., G. REUBER, supra note 1, at 120-28. For a general review and discussion on
motivational affects, see also Stewart, Taxation and Technology Transfer, in CONTROLLING
INTERNATIONAL TECHNOLOGY TRANSFER 137 (T. Sagafi-nejad, R. Moxon, & H. Perlmutter, eds. 1981)
[hereinafter CONTROLLING INT'L TECH. TRANSFER]; C. IRISH, INTERNATIONAL INCOME
TAXATION AND AFRICAN DEVELOPING COUNTRIES (4 Stud. Tax'n & Econ. Dev. 1978).
6 See Goldfinger, supra note 4, at 49. For a more recent study of the concerns of American
labor unions, see Helfgott, American Unions and Multinational Companies: A Case of Misplaced
Emphasis, 18 COLUM. J. WORLD Bus. 81 (1983).
7 Traditionally, international capital mobility is explained by differential rates of return on
capital as between countries. The theory states that capital will move from a capital rich country to a
capital scarce country in response to higher rates of return until such rates are equalized. The MNE
as an exporter of international capital is in a sense simply an arbitrager of capital. It pursues profits
by moving equity capital from countries with low rates of return to countries with high rates of
return. Thus, the profits made result from arbitrage activity. This is essentially a portfolio
investment theory; it has been used to explain the motivation behind FDI. For a more extensive
explanation, see R. CAVES, MULTINATIONAL ENTERPRISE AND ECONOMIC ANALYSIS 31-35 (1982)
[hereinafter R. CAVES, MULTNATIONAL ENTERPRISE]. The connection between differential rates of
return and differential tax rates as a motivation for FDI has been extensively explored elsewhere by
the author. See Yelpaala, The Efficacy of Tax Incentives Within the Framework of the Neoclassical
Theory ofForeign DirectInvestment: A LegislativePolicy Analysis, 19 TEx. INT'L L.J. 365, 371-73
(1984) [hereinafter Yelpaala, Efficacy of Tax Incentives].
alternative incentive policies based on different theories of the FDI
process. Specifically, this article will analyze possible alternative and
complementary incentive policies under the intangible assets hypothesis,8 the
industrial organization theory, 9 and the internalization theory of FDI.' 0
The Need for Reevaluation
The controversy over the MNE and the appropriate home and host
country policies has produced an interesting collection of opposing
views."1 Yet, to advocate a shift of focus from the current tax incentive
policies and laws to other policies requires some explanation. The utility
of tax incentive policies for attracting FDI has been doubted for years.
In a recent article this author questioned the theoretical basis of current
tax incentives, reviewed various studies on the subject, and concluded
that the theoretical basis for current tax incentive policies is flawed.12
Several reasons were offered for this conclusion. First, incentive granting
countries assume either explicitly or implicitly that there is some causal
relationship between differential tax rates and FDI. Several existing
policies appear to have been induced by deductive logic from the neoclassical
international trade theory which inspired the capital arbitrage theory of
FDI.13 This theory established a spurious causal link between
differential rates of return, differential tax rates, and the FDI process. 4 Second,
incentive granting countries assume that all FDI investors have a
shortterm profit motive which can be induced by tax rate reductions. Third,
there is an assumption that the MNE parent and its units for foreign
operations would independently seek to maximize profits and deal with
each other at market prices.15 This article will demonstrate that these
assumptions cannot easily be supported by our knowledge of the
characteristic attributes and distinctive behavior of the MNE. Fourth, even if
tax incentives would motivate FDI activities, the existence of competitive
tax concessions given by host countries would tend to undermine their
effectiveness. 6 Fifth, tax incentives might not only be ineffective but
might also induce a reverse subsidy to capital-rich developed countries. 17
Sixth, several studies have stressed that tax incentives have been
disregarded by MNEs because these incentives tend to be plagued by
bureaucratic uncertainties. Finally, tax incentives are inefficient and costly.",
Notwithstanding these problems with tax incentives, several
countries have continued to rely on the capital arbitrage theory of FDI as the
basis for FDI inducement policies. Indeed, there has been, in very recent
times, a proliferation of very liberal revisions of tax incentive laws by
several host developing countries. 9 Determined to maintain their
industrialization drive, newly-industrialized countries such as South Korea
and Taiwan have engaged in competitive liberal tax policies directed at
MNEs.2 0 In addition, middle- and low-income countries faced with the
growing need for technology transfer, capital, and managerial expertise
for industrial expansion, manufactured exports, and mineral extraction,
find themselves in strong competition for FDI with the more
economically successful industrialized countries. As a result, they have also
found it necessary to further liberalize their tax incentives. 21 In a
remarkable reversal of economic policies, even the People's Republic of
China has hopped on the bandwagon of countries seeking foreign capital
and MNE involvement through tax incentives.2 2 The theoretical basis
17 See Yelpaala, Efficacy of Tax Incentives, supra note 7, at 388.
18 See G. REUBER, supra note 1.
19 The United Nations Centre on Transnational Corporations has, for years, been collecting
legislation affecting foreign investment in several countries. Such legislation includes incentive
legislation for attracting foreign investment. The list of countries includes several developing countries.
See UNCTC, NATIONAL LEGISLATION, supra note 2.
20 See Yelpaala, The Impact ofIndustrialLegislation on the Behavior ofMNEs and Laborin the
IndustrializingCountriesof the East and SoutheastAsia, 6 MICH. Y.B. INT'L LEGAL STUD. (1985)
(forthcoming) [hereinafter Yelpaala, Impact of IndustrialLegislation].
21 UNCTC, NATIONAL LEGISLATION, supra note 2, at 29-39.
22 See, eg., The Law of the People'sRepublic of China on Joint Ventures Using Chinese and
Foreign Investment, 18 INT'L LEGAL MATERIALS 1163 (1979); Regulationsfor the Implementation
of the Law ofthe People'sRepublic of Chinaon Joint Ventures Using Chineseand ForeignInvestment,
22 INT'L LEGAL MATERIALS 1033 (1983); Rulesfor the Implementation
ofExchangeControlRegulations,22 INT'L LEGAL MATERIALS 1049 (1983); Income Tax Law ConcerningJoint Ventures with
Chinese and ForeignInvestment, 19 INT'L LEGAL MATERIALS 1452 (1980); Regulations on the
Registration of Joint Ventures Using Chinese and Foreign Investment, 19 INT'L LEGAL MATERIALS
1453 (1980). There has been a proliferation ofarticles written on the People's Republic of China and
foreign capital, joint ventures and similar issues. Only a few are noted here. See Note, The New
LegalFrameworkforJoint Ventures in China: Guidelinesfor Investors, 16 LAW & PoL'Y INT'LBus.
1005 (1984); Klitgaard & Rasmussen, PreferentialTreatmentfor ForeignInvestment in the People's
Republic of China: Special EconomicZones andIndustrialDevelopmentDistricts,7 HASTINGS INT'L
& COMP. L. REv. 377 (1984); Comment, Chinese Legal Inducementsfor the Development of a
Domestic PetroleumIndustry, 20 TEx. INT'L L.J 189 (1985); Current Development, EvaluatingChina's
SpecialEconomic Zones, 2 INT'L TAX & BUS. LAW. 376 (1984); Fenwick, Equity Joint Ventures in
the People'sRepublic of China: An Assessment of the FirstFive Years, 40 Bus. LAW. 839 (1985).
for their incentives continues to be the capital arbitrage theory.
Reliance on tax incentives by developing countries as instruments
for attracting FDI continues to be baffling in the face of the known
patterns of behavior of the MNE. Despite their sometimes liberal tax
incentive packages, developing countries have been the least attractive to
MNEs. 23 FDI tends to originate from a few developed market
economies and go to a few developed countries.24 Current statistics show that
MNEs tend to locate their foreign operations in the richer of the richest
and the richer of the poorest countries of the world.2 5
These patterns or preferences suggest that MNEs are attracted to
countries with the greatest economic potential or with affluent markets.
They also suggest that the tax burdens in these markets have not been a
significant deterrent to investments. They further suggest that, left alone,
MNEs will continue to locate their plants in countries with the highest
economic potential. The implication is that developing countries will
continue to be the least attractive to MNEs unless something is done to
reverse the pattern.
If MNEs are to play a significant role in the impending structural
transformation of the 1980s, home and host countries must develop
appropriate FDI policies which reflect the theory of the MNE. Since the
capital arbitrage theory and its connection to tax incentives has proven to
be dubious at best and ineffective at worst, it is important at this stage to
investigate the possibility of more effective FDI incentive policies based
on other theories of the MNE. Therefore, the focus of this article is on
the incentive policy possibilities of other theories of the FDI process.
II. THE MNE SYSTEM AND INCENTIVES
One significant problem of policymakers is their apparent lack of
complete understanding of the MNE phenomenon. From its inception,
the MNE has been the subject of intensive study and controversy.26 As
23 See UNITED NATIONS CENTRE ON TRANSNATIONAL CORPORATIONS, SALIENT FEATURES
AND TRENDS IN FOREIGN DIRECT INVESTMENT 11-12, 38, U.N. Doc. ST/CTC/14 (1983)
[hereinafter UNCTC, SALIENT FEATURES]. See also ORGANIZATION OF ECONOMIC COOPERATION &
DEVELOPMENT ("OECD"), INVESTING IN DEVELOPING COUNTRIES 20 (5th rev. ed. 1983)
[hereinafter OECD, INVESTING].
24 UNCTC, SALIENT FEATURES, supra note 23; UNCTC, THIRD SURVEY, supra note 1, at 27.
25 See generally UNCTC, THIRD SURVEY, supra note I, at 34-35. See also OECD,
INTERNATIONAL INVESTMENT AND MULTINATIONAL ENTERPRISE: RECENT INTERNATIONAL DIRECT
INVESTMENT TRENDS (1981) [hereinafter OECD, INTERNATIONAL INVESTMENT].
26 The literature on the MNE is so voluminous that it would be impossible to provide a complete
list here. It will be sufficient to list material representing various viewpoints in the debate. See, eg.,
L. TURNER, MULTINATIONAL COMPANIES AND THE THIRD WORLD (1975); N. GIRVAN,
CORPORATE IMPERIALISM: CONFLICT AND EXPROPRIATION (1976); GLOBAL COMPANIES: THE
POLITsuch, the MNE should be a well-known entity to policymakers.
Nonetheless, the MNE is an economic phenomenon that appears to defy
definition. Although several definitions have been suggested, 7 the most
useful definition is one which links the MNE with its characteristic
attributes and distinctive behavior.
The MNE is best understood as a system. 5 It is functionally an
international system that owns or commands globally vast amounts of
technological, financial, managerial, human, or marketing resources.
The MNE represents a network of operations, services, or a global
multiplant system connected through a common resource pool and a
common strategy with all its component parts. The structure is controlled by
a monocentric or polycentric management command system. As such,
the MNE enjoys tremendous flexibility in its operational decision
process. Its decisions are neither bound nor seriously limited by
considerations of distance, time, space, or by regional, national, or cultural
allegiances. Though considerations of plant location, marketing, import,
or export operations might fall to the dictates of the management control
system, such considerations are designed to enhance the flexibility of the
MNEs differ in size, technical capabilities, and degree of product
ICAL ECONOMY OF WORLD BUSINESS (W. Ball ed. 1975); R. BARNET & R. MULLER, GLOBAL
REACH: THE POWER OF THE MULTINATIONAL CORPORATIONS (1974); T. GLADWIN & I.
WALTER, MULTINATIONALS UNDER FIRE (1980); A. FRANK, CRISIS: IN THE WORLD ECONOMY
(1980); S. AMIN, ACCUMULATION ON A WORLD SCALE; A CRITIQUE OF THE THEORY OF
UNDERDEVELOPMENT (1974); C. VIATSOS, INTERCOUNTRY INCOME DISTRIBUTION AND
TRANSNATIONAL ENTERPRISES (1974); R. VERNON, SOVEREIGNTY AT BAY: THE MULTINATIONAL SPREAD
OF U.S. ENTERPRISES (1971) [hereinafter R. VERNON, SOVEREIGNTY AT BAY]; R. VERNON,
STORM, supra note 1; C. BERGSTEN, T. HORST, & T. MORAN, AMERICAN MULTINATIONALS AND
AMERICAN INTERESTS (1978); MULTINATIONAL MANAGERS AND POVERTY IN THE THIRD
WORLD (L. Tavis ed. 1982); S. LALL & P. STREETEN, FOREIGN INVESTMENT, TRANSNATIONAL
AND DEVELOPING COUNTRIES (1977).
27 The Multinational Enterprise as a phenomenon has been characterized by definitional
problems from its inception. Some definitions stress multinationality, geographic dispersion, and
management orientation, while others stress control. The following are examples of the attempts at
defining the MNE. See generally UNITED NATIONS DEPARTMENT OF ECONOMIC AND SOCIAL
AFFAIRS, MULTINATIONAL CORPORATIONS IN WORLD DEVELOPMENT 5 (1974). Some definitions
emphasize several different criteria. For instance, for criteria such as: 1). organizational structure
and ownership, see Aharoni, On the Definition of a Multinational Corporation,in THE
MULTINATIONAL ENTERPRISE IN TRANSITION 4-5 (A. Kapoor & P. Grub eds. 1972); 2). the number of
countries in which operations are carried on, see Dunning, The MultinationalEnterprise: The
Background, in THE MULTINATIONAL ENTERPRISE 15, 16 (J. Dunning ed. 1971); 3). size of total
operations, see R. VERNON, SOVEREIGNTY AT BAY, supranote 22, at 4 (1971); 4). attitude of management
(ethnocentric, polycentric or geocentric), see Perlmutter, The Tortuous Evolution of the
MultinationalCorporation,in THE MULTINATIONAL ENTERPRISE IN TRANSITION, supra, at 53; Ball, Cosmo
Corp: The Importance ofBeing Stateless, in WORLD BUSINESS 337 (C. Brown ed. 1970).
28 See Yelpaala, Impact ofIndustrialLegislation, supra note 16.
and geographic diversification. For example, MNEs range from the
world's largest grocery products company, Unilever, with operations in
seventy-five countries and sales in 1980 of $24 billion,2 9 to Wrigley
Chewing Gum Company, with a sales of $500 million.3" It has been
estimated that there are today about 10,000 MNEs with a total of about
100,000 subsidiaries.3 Most of these MNEs are small and have
operations in two or three countries. Nonetheless, the MNEs that are
responsible for a substantial portion of FDI are the largest 500 companies in the
world. Each of the MNEs, small or large, constitutes a separate system
varying in size, complexity, and flexibility depending upon its total
available resources and the number of countries in which it operates. Hence,
the impact of any policy designed for the MNE would vary depending
upon the MNE's degree of flexibility.
This flexibility is inherent in the MNE control and management
command system and permits the MNE to manipulate its global
operations in response to various economic stimuli, costs, government attitude,
and geo-political considerations.3" Thus, the same MNE, through the
establishment of foreign subsidiaries or global affiliates, may extract raw
material in one country, to be manufactured as semi-finished products in
a second country, to be used in the manufacture of a finished product in a
third country, to be marketed in yet a fourth country.3 3 Therefore,
through this system of linkages between parent MNE and its foreign
affiliates and between the affiliates themselves, the MNE is afforded the
opportunity to use each affiliate as a conduit for shuttling resources from
one country to another or to shift profits from high-tax regimes to
lowtax regimes according to the dictates of the management command
An important aspect of the MNE system is the degree of isolation
between MNEs. Each MNE system constitutes a separate system,
suggesting that an important industry in a country might be fragmented
along oligopolistic lines thereby resulting in certain allocative
inefficiencies and the loss of control by the host government. This is illustrated by
the experience of Ghana in dealing with some of the largest MNEs in the
aluminum industry. These MNEs operate as isolated units, with most of
their linkages created within their own systems. Accordingly, the
aluminum industry in Ghana is fragmented and so are its connections with the
rest of the global aluminum industry. As a result, Ghana exports
bauxite, imports alumina, exports primary aluminum, and imports rolled
aluminum to be fabricated into aluminum products.3 4
One may characterize the typical large MNE as potentially "foot
loose"'35 or as an acquisitive and gigantic octopus, its tentacles sprawling
all over the globe. This analogy is borne out by the high degree of MNE
concentration by country of origin, industry, growth, diversification, and
internal operational dynamics. The bulk of the FDI in the world comes
from a very small number of industrialized countries. Only six such
countries-United States, United Kingdom, Germany, Japan,
Switzerland, and France-account for more than eighty percent of the total
global stock of FDI.3 6 From this small number of countries, the MNE
system spreads out globally. In 1976 there were about 11,000 companies
that operated 82,600 foreign affiliates worldwide.3 7 Out of this total,
about 371 MNEs operated in at least twenty or more countries.3 8
For some of the world's largest MNEs, the geographic
diversification of their operations is even more outstanding. In 1973, it was
estimated that IBM operated in as many as eighty countries, while Siemens
and ITT operated in fifty-two and forty countries, respectively.3 9 The
corresponding figures for MNEs in the oil industry are equally
impressive. Mobil, Gulf, and Shell operated in sixty-two, sixty-one, and
fortythree countries, respectively.' ° These foreign operations have always
34 D. HART, THE VOLTA RIVER PROJECT 59 (1980).
35 P. BUCKLEY & M. CASSON, THE FUTURE OF THE MULTINATIONAL ENTERPRISE 13 (1976).
36 OECD, INVESTING, supra note 23, at 20.
37 J. DUNNING, INTERNATIONAL PRODUCTION AND THE MULTINATIONAL ENTERPRISE 3
(1981) [hereinafter J.DUNNING, INT'L PROD.].
39 P. BUCKLEY & M. CASSON, supra note 35, at 12-13.
been very important in MNE total operations.4 1 For example, in 1977
the sale of the foreign affiliates of 866 of the world's largest industrial
enterprises accounted for twenty-seven percent of their total sales
volume.42 The foreign content ratio of these enterprises in manufacturing,
as calculated by John Dunning and Robert Pearce, was also quite
The relationship between these foreign affiliates and their parent
MNEs is not one of isolated and independent existence. They are linked
together through the necessity of horizontal and vertical diversification,
common resource pool, common international marketing and
distribution strategy, common technological needs, and management control.'
These connections are certainly obvious in the manufacturing industry
where the incidence of horizontal diversification has been adequately
documented.4 5 Subsidiaries of MNE trading companies, such as the
Japanese trading houses, have always needed their parents for access to
world markets.4 6 Evidence of vertical diversification can be found in the
frequency and size of intra-firm transactions between MNEs and their
subsidiaries. Though the statistics and data on these transactions are still
incomplete, the information which does exist suggests that the size of
intra-firm transactions within the MNE system is fairly significant. For
instance, in 1974, forty-six percent of all imports to the United States
were intra-firm.4 7 It is also estimated that roughly fifty percent of the
exports from the United States in 1970 and thirty-nine percent of
Canadian exports in 1971 were all within the MNE system.4 8
In 1977, United States MNEs accounted for about eighty-four
percent and fifty-eight percent of the value of United States exports and
imports, respectively.4 9 Intra-MNE imports accounted for forty-eight
percent of the total value.5 ° These MNEs have shown a strong
preference for intra-firm transactions with higher degrees of processing.
Moreover, MNEs no longer consider locating their production facilities near
their markets as important. Large MNEs in the automobile and
elec41 J. DUNNING, INT'L PROD., supra note 37, at 3.
42 Id. See also UNCTC, THIRD SURVEY, supra note 1, at 18.
43 See J. DUNNING & R. PEARCE, THE WORLD'S LARGEST INDUSTRIAL ENTERPRISES 110
(1981). See also J. DUNNING, INT'L PROD., supra note 37, at 4.
44 R. VERNON, STORM, supranote 1, at 19-37. See also P. BUCKLEY & M. CASSON, supra note
35, at 20.
45 P. BUCKLEY & M. CASSON, supra note 35.
46 T. OZAWA, MULTINATIONALISM, JAPANESE STYLE 45 (1979).
47 UNCTC, THIRD SURVEY, supra note 1, at 161.
49 J. CONNOR, supra note 29, at 13.
tronics industries, for example, now prefer to set up networks of
integrated production units located in different countries but all involved in
the processing of different stages or parts of the final product." Such
production is generally required to serve the needs of the parent or other
members of the group in third countries. A significant amount of MNE
investments in certain cheap labor havens such as Taiwan and South
Korea are of this character.5 2 Bruce Kogut 53 has suggested that the
dynamism and flexibility of the MNE system constitutes its primary
advantage because it provides several options to the MNE.5 4 Viewed in
this way, the MNE system has operational value to its managers since it
increases and links together options and opportunities which otherwise
would have been less attractive. . I
It is clear from this analysis that the MNE is a very complex
phenomenon which, in several cases, has developed a global mind set. The
MNE views its own national economy as too limited for its potential in
production, sales, and resource utilization. This global orientation,
coupled with the existence of foreign affiliates, means the MNE can expand
its economic space at will. If this is the type of entity that current tax
incentive legislation is addressing, then serious questions are raised about
the appropriateness of any policy that does not focus on these attributes
of the MNE. Given the tremendous flexibility in the behavior of the
MNE, one must ask: what types of policies would induce MNE
investment and where would its operational effects be felt? Since the MNE can
structure its manufacturing and distribution system by choosing the
countries with policies which suit it best, the MNE system may
purposely burden some countries with negative externalties and bless others
with substantial welfare benefits.
The operations of the MNE can affect any or all of the nations
included in this system at both the micro- and macro-economic levels in
terms of contributions to gross national product, prices, profitability, and
national revenue. For those countries linked together by this process, the
ability of each to benefit from MNE operations will not be a simple
function of any monolithic tax incentive legislation, but how each country by
design fits into the MNE system. Can incentive legislation which does
not recognize the inherent flexibility of MNE behavior effectively induce
51 OECD, INTERNATIONAL INVESTMENT, supra note 25, at 31.
52 Yelpaala, Impact of IndustrialLegislation, supra note 20.
53 Kogut, ForeignDirectInvestment as a SequentialProcess, in THE MULTINATIONAL
CORPORATION IN THE 1980s, supra note 1, at 38, 42-43.
54 Id. at 46-47.
Northwestern Journal of
International Law & Business
Thus the question of what are the appropriate policies for attracting
FDI requires at least a thorough understanding of the MNE system. In
addition, such policies should be based on those theories of the MNE
that most closely reflect the MNE system and its distinctive behavior.
The task of the following sections is to explore, within the context of the
MNE, those policy options which appear to incorporate some of the
III. INTANGIBLE ASSETS THEORY OF FDI
A useful place to start an exploration of alternative incentive
policies is the intangible assets theory. Critics of the capital arbitrage theory
argue that central to the MNE's motivation to engage in foreign
operations is the possession of some proprietary knowledge or intangible
assets.55 (As used here, the two terms are interchangeable.) This theory
holds that the most successful firms in any industry possess, in some
form, intangible assets to the exclusion of others. The nature or
character of these intangible assets may take different forms. They may
represent technology, knowledge of cost-minimizing productive efficiency,
patented processes, registered trademarks, designs, or brand names.
They may even rest on production or trade secrets known to and shared
by the employees of a particular firm or skills in styling or promoting
products.5 6 What all these intangible assets have in common is that they
are thought of as having the character of a "public good" in the sense
that they can be exploited over and over by several different firms
without depleting their usefulness.5 According to Harry Johnson, the
acquisition of such assets involves some initial investment outlays, but once
the knowledge or intangible asset is acquired, the investment costs are
considered sunk costs.58 After these developmental costs, the marginal
cost of exploiting such an asset abroad is practically nil for the MNE.
The marginal cost of such assets is lower than that of the local firm or
other firms that do not own similar intangible assets. Thus, according to
this theory, the local firm and others which do not own intangible assets
are at a competitive disadvantage with MNEs in the same industry.
-5 See, e.g., Gruber, Mehta, & Vernon, The R & D Factorin InternationalTrade and
InternationalInvestment of United States Industries, 75 J. POL. ECON. 20 (1967); Horst, Firm and Industry
Determinantsof the Decision to Invest Abroad: An EmpiricalStudy, 54 REV. ECON. & STAT. 258
(1972) [hereinafter Horst, An EmpiricalStudy]; Johnson, The Efficiency and Welfare Implicationsof
the InternationalCorporation, in THE INTERNATIONAL CORPORATION 35 (C. Kindleberger ed.
56 R. CAVES, MULTINATIONAL ENTERPRISE, supra note 7, at 3-4.
57 Johnson, supra note 55, at 36.
The possession of intangible assets or similar proprietary knowledge
alone cannot explain the MNE motivation for FDI because MNEs have
other alternatives to FDI. They may choose to service foreign markets
through exports from their production at home. They may even grant an
assignment or license of their know-how or technology to a local
producer. The choice to engage in foreign operations, therefore, must have
some additional advantage: it must offer the highest return in
comparison with the other alternatives. For instance, the direct exports option
might be saddled with heavy transportation costs, excessive customs
duties, high tariff and nontariff barriers, thereby making the product
uncompetitive at prevailing market prices in the importing country.
Similarly, several obstacles to licensing may exist. The host government
may control and supervise licensing agreements in terms of their
substantive content or royalties agreed upon by the parties. Moreover, the local
licensee may be both unwilling and unable to pay the appropriate rent.
Under either of these circumstances the owner of the intangible assets
cannot reap the desired maximum returns.
Two general explanations for FDI are offered by the intangible
assets theory.5 9 First, MNEs engage in foreign operations because of the
problems of market imperfections associated with arms-length
contractual transactions. Second, MNEs engage in foreign operations for cost
reasons." MNEs have firm-specific revenue-producing assets which
cannot be employed efficiently from their home base because of information,
communication, and transportation costs associated with the distance.6 1
To understand fully the intangible assets theory, it is necessary to
examine these two explanations in greater detail.
Market Imperfections and FDI
According to the market imperfection hypothesis, MNEs with
intangible assets cannot sell or license them efficiently in conventional
markets because of certain factors related to the assets' infirmities. First,
since intangible assets are treated as "public goods" from the point of
view of resource allocative efficiency, the rental price of such intangible
assets should be close to zero.6 2 While such a societal goal might be most
59 See R. CAVES, supra note 7, at 3-4; Coase, The Natureof the Firm,4 ECONOMICA 386 (1937);
S. HYMER, THE INTERNATIONAL OPERATIONS OF NATIONAL FIRMS: A STUDY OF DIRECT
FOREIGN INVESTMENT 48 (1976); Johnson, supra note 55, at 37; P. BUCKLEY & M. CASSON, supra note
35, at 22-23.
60 Hirsch, An InternationalTrade and Investment Theory of the Firm, 28 OXFORD ECON.
PAPERS 258 (1976).
61 Id. at 259.
62 Johnson, supra note 55, at 41.
acceptable to the renters of technology, it is not in the interests of
profitmaximizing MNE owners of technology. This situation is further
complicated by the perceptions and realities of technology transfer. For
example, what is the appropriate rental price of obsolete technology to a
licensee in a developing country? Should the rental price be controlled
by its market value in the home market, in which it is obsolete, or in the
host country? It appears equally difficult to determine the rental price of
frontline technology when the foreign licensee has been induced into
questioning the claimed value of technology in general. Under these
circumstances the intangible assets theory suggests that MNEs would prefer
to invest in the use of their technology abroad.
In any arms-length transaction for technology, there are likely to be
several problems related to uncertainty in the terms and distrust among
the parties. For example, the seller must prove the quality, operational
capabilities, and value of the technology, but disclose only as much as is
necessary to strike a deal. The buyer may suspect the seller of
opportunism and overstating the value. There might also be problems relating to
disparities in bargaining power. In view of these contract law related
problems, profit-maximizing MNE owners of technology would prefer to
invest in their use abroad. 63
While these arguments appear relevant to horizontal FDI, they may
be equally applicable to vertical investments.' MNEs relying on foreign
suppliers for primary factor inputs have to work the spot and futures
markets efficiently or develop long-term stable contractual relationships
with their suppliers. However, both of these markets, either in the short
or long run, are riddled with costly and significant contract and
performance uncertainties.6 5 Without control over the production and
transportation, MNEs would face uncertainties in the levels of output, delivery
dates, and prices over time. Consequently, firms seek to own or control
vertically their sources of primary or intermediate factor inputs through
The relationship between costs and FDI in the use of intangible
assets was developed most elaborately by Seev Hirsch.66 According to
Hirsch, given the possession of some intangible assets ("K"), the
question whether a profit-maximizing firm will service its foreign markets
63 R. CAVES, supra note 7, at 4-6.
64 Id. at 16.
65 M. CASSON, ALTERNATIVES TO THE MULTINATIONAL ENTERPRISE 48-49 (1979).
66 Hirsch, supra note 60, at 260.
through exports or engage in foreign operations is determined by cost
First, the firm will be concerned with the production costs in both
the host ("Pb") and home ("Pa") countries. Second, there are costs
associated with controlling foreign operations ("C") from the home base.
Some of these costs relate to the internal structures, dynamics,
management, and coordination of decentralized global production operations.
Other cost items relate to the host country's environment, such as its tax
system, labor codes, and other regulatory institutions and machinery.
Finally, the firm must face the export marketing cost differential ("M")
which is the positive difference between export and domestic marketing
costs. 68 The former exceeds the latter because it is always more costly to
export than to produce and sell within the host country.
Given these cost items, Hirsch concluded that FDI will take place
under the following conditions:
(1) Pb + C < Pb + K
) Pb±C<Pa M.6 9
Since Pb is the same on both sides of Equation 1, that equation merely
states that FDI will take place when there are net positive returns on the
use of intangible assets in the host country, taking into account all the
costs associated with controlling foreign operations. In comparison,
Equation 2 takes into consideration comparative total costs. In other
words, FDI will take place if the cost of production plus the cost of
controlling foreign operations in the host country is less than the
alternative-the total of the costs of production and the export marketing cost
The intangible assets hypothesis has some empirical support. The
boom in United States MNEs and their foreign operations, particularly
after the Second World War, is said to have been closely associated with
their adjustments to technological possibilities.70 In fact, this
relationship dates back to the period before the First World War. In a historical
analysis of United States MNEs, Mira Wilkins found that, even in the
early days of the MNE, superiority in knowledge was very important in
the FDI process. 71 Superior know-how explains the FDI operations of
such firms as Singer and American Radiator. A wide variety of
68 Id. at 261.
69 Id. at 264.
70 Hufbauer, The MultinationalCorporationand DirectInvestment, in INTERNATIONAL TRADE
AND FINANCE: FRONTIERS FOR RESEARCH 253 (P. Kenen ed. 1975).
71 M. WILKINS, THE EMERGENCE OF MULTINATIONAL ENTERPRISE: AMERICAN BUSINESS
ABROAD FROM THE COLONIAL ERA TO 1914, at 77 (1970).
cal studies conducted to test the validity of the intangible assets theory
tend to show that the theory explains the foreign investment behavior of
several United States MNEs in the manufacturing industry. Several
studies by economists such as William Gruber, Dileep Mehta, Raymond
Vernon,7 2 Thomas Horst,7" Thomas Pugel,74 and Sanjaya Lall,7 5 all
found a positive link between research and development ("R & D")
outlays and FDI. Others found a significant relationship between
managerial skills and talent and FDI.7 6 This theory has also found empirical
support in the investment behavior of United States firms involved in the
extractive industries,77 banking, and other service industries.7 8
C. Intangible Assets-Incentive Policy Implications
An analysis of the policy implications of this theory requires some
understanding of its basic underlying assumptions. The intangible assets
theory assumes implicitly that all intangible assets constitute
legally-protectible property with universally-recognized attributes. It assumes the
existence of an international legal regime and a uniform set of national
rules adequately protecting the property interest in these assets,
including the grant of monopoly or semi-monopoly rights to owners of
industrial property limited in time to the period of protection. As a
technology owner, the MNE is one of the most powerful phenomena in
the capitalist system; nonetheless it insists on these protections at home
and abroad against new entrants, imitators, and industrial pirates. This
protection becomes an economic imperative if the MNE is to maintain its
industrial characteristics and reap monopoly profits on the use of its
Thus, at the international level, adequate legal protection requires
some participation by host countries in various multilateral, bilateral, or
regional conventions and treaties which recognize and protect industrial
property rights similar to those enjoyed in the home country. Despite
the current controversy, the Paris Convention of 1883 is an example of a
multilateral system of industrial property protection to which several
developed and developing countries have subscribed. 79 There appears to be
little or no conceptual difficulty with the recognition of industrial
property. The problem lies with the nature and extent of protection. 80
Therefore, some regional groups have sought to harmonize their treatment of
industrial property in the FDI process."1 At the national level, adequate
protection for technology owners would include not just statutes
protecting patents, trademarks, and know-how, but effective judicial systems to
stop infringers and provide remedies.
Given these implicit assumptions about the legal regimes of host
countries, the theory postulates the conditions for FDI to take place.
MNEs, the theory states, are very sensitive to instability and uncertainty
in their technology transfer and supply contracts. Licensing contracts
are fraught with many uncertainties relating to royalties or fees and other
factors. Long-term raw material or intermediate supply contracts are
similarly fraught with uncertainties relating to the vagaries of price
fluctuations, quality, non-performance, and the inability to meet delivery
dates. To bypass these contract-related problems, MNEs engage in
horizontal and vertical FDI so that they can control the manufacturing
process using their own technology or maintain complete control over their
If these assumptions are true, then this theory has implications for
countries that possess exploitable natural resources or are interested in
expanding their manufacturing industries. Several developing countries
are heavily dependent on their primary resources for their economic
survival. The policies of these countries have often also stressed the need for
manufactured export expansion or import substitution. Countries
relying on such policies hope that they will have some positive effects on
employment, balance of payments, and intersectoral linkages. The
theory, therefore, opens up certain policy possibilities if the host countries'
objectives can be achieved.
Market Imperfections and Incentive Policies
It should be noted that the contract-related market imperfections
described above do not immediately suggest any tax policy instruments
as a means of inducing FDI. They are not the type of factors which
generally concern tax policymakers. Moreover, if these market
imperfections exist, then MNEs have the necessary conditions for FDI.
Therefore, these imperfections should not be removed if the host country
wishes to entice FDI. If, on the other hand, the market imperfections do
not exist, the question is whether the host country should create them to
The experience of several host countries tends to show that these
imperfections already exist. The practice of several host countries of
intervening in contractual transactions between MNEs and local licensees
is not to create distortions within the markets, but to correct certain
actual or perceived distortions, inequities, and negative effects of
technology transfer transactions not addressed by the theory. 2 For instance, the
theory does not question the appropriateness of the technology
transferred nor its price and general impact on the national economy. It is the
belief of several developing countries that the technology transferred by
MNEs is generally more capital intensive than their needs, exploitative in
price, restrictive in use, and marginal in its developmental impact on the
local economy. Moreover, the relationship between local licensees and
MNE technology owners is characterized as being unequal in terms of
bargaining power. Large MNEs, in particular, are perceived as having
excessive bargaining power-because of their oligopolistic position in
these markets and their monopoly rights in technology-with which they
can exploit local licensees.
Countries such as Brazil, Mexico, and South Korea have all
intervened in various ways to control licensing fees, royalties for patents, and
trademarks transactions.8 3 Others have sought to control the type and
82 Correa, Transferof Technology in LatinAmerica: A Decade of Control,15 J. WORLD TRADE
L. 388, 391-92 (1981); Ozawa, Technology Transferand ControlSystems: The JapaneseExperience,
in CONTROLLING INT'L TECH. TRANSFER, supra note 5, at 376; Mason, Technology Transfer
ControlSystems: The Case ofEast and Southeast Asian Developing Countries,in CONTROLLING INT'L
TECH. TRANSFER, supra note 5, at 430; Juncadella, The ForeignInvestment Laws of Latin America:
Present and Future, 16 INT'L LAW. 463 (1982); Wilner, The Transfer of Technology to Latin
America, 14 VAND. J. TRANSNAT'L L. 269 (1981); Bledel, The Latin American Development Process
and the New Legislative Trends, 10 GA. J. INT'L & COMP. L. 325 (1980); French, Brazil's Profit
Remittance Law: Reconciling Goals in Foreign Investments, 14 LAW & POL'Y INT'L Bus. 399
(1982); Kryzda, Joint Ventures and Technology Transfers, 12 CASE W. RES. J. INT'L L. 549 (1980);
Radway & Hoet-Linares, Venezuela Revisited: Foreign Investment, Technology and Related Issues,
15 VAND. J. TRANSNAT'L L. 1 (1982); Colloquiumon CertainLegalAspects ofForeignInvestment in
Mexico: Regulation of CapitalInvestment, Patentsand Trademarks,and Transferof Technology, 7
GA. J. INT'L & CoMP. L. 1 (1977).
83 Rosenn, Regulation of Foreign Investment in Brazil: A CriticalAnalysis, 15 LAW. AM. 307,
321-25 (1983); see also sources cited supra note 82.
economy of the MNE and its operational dynamics, countries in the first
category appear to draw the important distinction between ownership of
the operational resources and control over their use. Ownership of the
operations by MNEs Perse does not raise questions about allocative and
distributive efficiencies in an economy. It is how these operations are
controlled or managed that might raise serious questions about the
efficient utilization of resources and distributive equities. The extent to
which management is responsive to contradictory home country policies
intensifies the concern over sovereignty. Some countries, such as Brazil
and others in Latin America, have introduced very complex and
sophisticated regulatory schemes designed to monitor transactions within the
MNE system and control the channels of transmission and flow of
profits, dividends, fees, and other charges within the system.142 A number of
studies on these regulatory schemes point out that, even though they
might have been motivated by a sophisticated understanding of the MNE
system, they were ad hoc and uncoordinated in nature. 143 Consequently,
they were ineffective. Moreover, their efficiency was further undermined
by the unavailability of sufficient enforcement resources. 144
The second category of regulatory schemes appears to be most
prevalent in developing countries. Countries such as Ghana and Nigeria
introduced various regulatory schemes starting with controls over transfer
pricing, remittances, and requiring local participation by statute. 145
Local participation statutes were motivated by many factors. Several of the
142 For examples from Brazil, see Foreign Capital Inducement Law No. 4.131 of 3 Sept. 1962 (as
amended by Law No. 4.390 of 29 Aug. 1964); Executive Act No. 55.762 of 17 Feb. 1965
(consolidating Law No. 4.131 of 3 Sept. 1962 and Law No. 4.390 of Aug. 1964); Normative Act No. 15 of Sept.
11, 1975; and Normative Act No. 64 of Sept. 16, 1983. For examples from Columbia, see Decree
1900 of 1972, (implementing Decision 24 of the Andean Common Market concerning the treatment
of FDI, patents, trademarks and royalties); Law 55 of 1975 (restricting new investments in banking);
and Finance Technology transfer laws-to be found in Decree 444 of 1967, Decree 688 of 1967, and
Decree 1234 of 1972. For examples from Mexico, see Law for the Promotion of Mexican
Investment and Regulation of Foreign Investment, 1973; Law Concerning the Control and Registration of
the Transfer of Technology and the Use and Exploitation of Patents and Trademarks, 1981; Law on
Inventions and Trademarks, 1976. The United Nations Center on Transnational Corporations has
recently compiled a list of the appropriate legislation aftfecting FDI in these and other countries. See
UNCTC, NATIONAL LEGISLATION, supra note 2.
143 See Rosenn, Regulation ofForeignInvestment in Brazil: A CriticalAnalysis,15 LAw.
AMERICAS 307 (1983); Correa, Transfer of Technology in Latin America, 16 J. WORLD TRADE L. 388
(1981); Inneadella, The Foreign Investment Laws ofLatin America: Present and Future, 16 INT'L
LAw. 463 (1982); Wilner, The Transferof Technology to Latin America, 14 VAND. J. TRANSNAT'L
L. 269 (1981); Bledel, The Latin America Development Process and the New Legislative Trends, 10
GA. J. INT'L & COMp. L. 352 (1980).
144 Rosenn, supra note 143, at 352-54; Correa, supra note 143, at 408.
145 For Ghana, see the Investment Policy Decree, 1975 (NRCD329). Since this Decree, Ghana
has enacted other laws designed to encourage FDI. The most recent legislation is the Investment
Code, 1985 (P.N.D.C.L. 116) which repealed the Investment Code, 1981 (Act 437). For Nigeria, see
host governments seriously concerned about the nature, quality, and
impact of MNE activities on their economies decided to induce greater and
more efficient integration of MNE operations into their local economies
through legislation. MNE operations, they thought, would be most
beneficial to them if they generated greater linkage effects and resulted in
better and more efficient local resource utilization.
To ensure that MNE operational decisions would take into account
national industrial and economic policies, output targets, export levels,
and even the balance of payment positions, those host governments
enacted local participation laws which demanded majority local ownership
of MNE operations within their countries. These types of laws have
become commonplace. They exist within several countries in Southeast
Asia, Africa, and Latin America.'4 6 Such host governments seem to
believe that the simple fact of local majority ownership, dictated by law,
will translate into effective control. According to the theory, those laws
should discourage FDI. However, the real question is: what is their
impact on the behavior of MNEs? Do these legislated local majority
ownerships translate into effective control by local individuals, and hence the
achievement of host government objectives?
MNE Responses to Ownership and Control Laws
The extent to which ownership and control laws of the type
described above have been effective depends upon the behavior of MNEs.
Until about the last decade, there appeared to have been a definite
preference of most MNEs (particularly United States MNEs) for wholly- or
majority-owned operations in developing countries. MNEs in more
recent times have demonstrated greater flexibility with regard to ownership
and control. There is a growing trend toward accepting minority
ownership positions of FDI in developing countries. However, that an MNE
holds a minority interest is not always dispositive of the locus of control.
A United Nations report on MNEs indicates that, notwithstanding their
minority equity positions, MNEs still exercise substantial de facto
control over their operations.14 7
In a recent survey on this issue, it was found that the experience of
several countries in legislating control has been unsuccessful.' 4 8 MNEs
from Europe, Japan, and the United States all seemed to find a way
The Nigerian Enterprises Promotion Act (1977) (A/63), (Decree No. 3), generally known as the
1977 Indigenization Act.
146 See supra notes 84-87.
147 UNCTC, THIRD SURVEY, supra note 1, at 62-64, 163.
148 See Yelpaala, Impact of IndustrialLegislation, supra note 20.
around local majority participation statutes to maintain effective control,
at least over very critical aspects of their operations. 149 Moreover, the
local majority owner does not always care about legal right of control.
Most often the local owner still needs the MNE because it has the
resources, access to markets, and an international or global network. For
profitable operations, the local majority owner usually relies on the
MNE, resulting in relinquishment of effective control. Several other
studies have confirmed the importance of the issue of control to MNEs
even though legal ownership may be declining in importance. 150 MNEs
operating in Nigeria, for example, have adopted schemes, structural
reorganizations, and several other methods to maintain control which were
violative of the spirit and letter of the Nigerian Indigenization Decree. 5 '
For instance, an MNE manufacturing for local distribution would divide
its operations into two parts: (1) manufacturing, which would be
whollyowned by the MNE; and (
) marketing and distributing, a joint venture
in which the MNE would take a minority interest but find local partners
who are known to be incompatible with one another. The two companies
would be in the same building and share the same board of directors and
officers.1 2 As such, the MNE would maintain an effective control over
its entire operations notwithstanding the law.
What is obvious from the Nigerian experience is that majority local
participation statutes, by their nature, stress ownership and not effective
control. They confer only the legal right of control whereas the
effectiveness of the right is determined by other extraneous and non-legal factors.
Thus, the local majority owner might not have the know-how,
managerial expertise, or operational experience to run the operation efficiently.
Moreover, the industries normally targeted for local majority ownership
are the same industries in which the MNEs enjoy a decided monopolistic
advantage over the local partner. Such an advantage would often be used
subtly to control the operations effectively. Therefore, several of the host
country's desires for economic independence and policy autonomy
remain mere wishes and somewhat elusive.
The response of MNEs to various regulatory schemes of host
150 T. BIERSTEKER, DISTORTION OR DEVELOPMENT?: CONTENDING PERSPECTIVES ON THE
MULTINATIONAL CORPORATION (1978); Biersteker, The IllusionofState Power Transnational
Corporationsand NeutralizationofHost-CountryLegislation, 17 J. PEACE RESEARCH 207 (1980)
[hereinafter Biersteker, Illusion].
151 Biersteker, Illusion, supra note 150, at 215-18.
152 This is what Biersteker calls the two company strategy. IaL at 216.
tries indicates quite clearly how adaptable MNEs can be in the face of a
changing legal environment. It also suggests that MNEs attach a great
deal of importance to the issue of control. They quickly recognize the
distinction between legal control and effective or operational control and
tend to accept situations where the effectiveness of control is maintained
by them. In terms of developing incentives for FDI, the apparentpolicy
prescription for host countries would then be to denationalize FDI
operations by repealing all the MNE control legislation or by setting up
ownership laws which do not interfere with the desired MNE effective
This policy prescription is dubious for a number of reasons. First,
one wonders whether control and FDI are necessarily inseparable
concepts. Control, according to Stephen Hymer, is only a means to an
end-maximizing returns on firm-specific advantages within certain
imperfect market structures. Control might well be a means to monopoly
profits. It might also be a statement of the level of uncertainty in a
particular environment in which the MNEs operate. If MNEs control their
operations to maximize profits, the lack of control need not result in
lower levels of FDI since the loss of control does not necessarily imply
lesser returns. Moreover, as discussed under the intangible assets
hypothesis, controlling MNE operations in industries characterized by
several market imperfections should result in optimal levels of FDI and
allocative efficiencies in host countries. Besides, the growing number of
joint venture operations between different MNEs in the developed
market economies seems to suggest that control is not essential to foreign
operations of MNEs since at least some of them normally would take a
minority position in those operations. The desire for control might then
be a manifestation of some uncertainty about targeted profit levels or
some other problem within the environment of the host country.
Whether MNE control would work as an incentive for FDI would,
therefore, depend upon the nature of the markets within which MNEs
operate. The more imperfect the markets, the more effective control
would act as an incentive in the FDI process. Market imperfections
would tend to increase the level of uncertainty reflected in the cost of the
business environment, thereby requiring MNE control for efficiency and
profitability. In addition, the existence of certain forms of uncertainty
would require at least effective control to induce FDI. Where, for
example, the operational utility of FDI is dependent upon uncertain levels of
local capabilities, scale, efficiency, and quality for the product to be
internationally competitive, effective control might operate to induce FDI.
Such a policy could then be complementary to the technology policy
discussed earlier. However, market imperfections in certain industries may
also require local control. This is most evident in the extractive
industries. The MNE might be encouraged to invest without control when
most of the risk in the long-run is borne by the host country. That seems
to explain production sharing, service contracts, and staging in mineral
Since MNEs have shown a remarkable adaptability in the area of
ownership and control, and host governments have continued to express
concern over the degree of asymmetrical vulnerability dependence on
foreign participation in their local economies, breaking down the FDI
package could lead to an accommodation on the issue of control with
respect to both the host governments and MNEs. As demonstrated
earlier, host governments' policy objectives are multiple. A government
might want ownership and control over a manufacturing industry
because it wants to encourage the acquisition of technology or managerial
expertise. It might want local control in an industry that produces an
important manufactured export product. A country might also want to
develop competent local capability in an industry requiring a
capital-intensive and complex technology as in the case of the chemical and
petrochemical industries. The policy and strategy for ownership and
control in these cases could depend upon the availability of suitably
trained technicians and management either locally or internationally for
hire. On the other hand, the MNE might see control as essential for
monitoring a new technology, maintaining product quality, or even for
ensuring that the particular operations fit into the MNE system. These
goals are not all necessarily inconsistent with those of the host
It appears that the issue of control could be used as an incentive for
FDI by host governments. For instance, a policy encouraging foreign
joint venture operations in an industry where local control is desired
could be achieved in different ways. The host government could use a
fade-out, production sharing, service, or management contracts
effectively as incentives. In the case of manufactured exports, the MNE may
only invest if it has control over the manufacturing process for quality
reasons. A fade-out policy can accommodate the MNE concerns where
it is allowed to control the operations, but to phase-out such control to
suitably-trained local management within a definite amount of time. In
the alternative, the policy could allow MNE control over only the
manufacturing process subject to a fade-out scheme. This approach should
result in allowing MNE temporary control over crucial aspects of FDI
operations where productivity, efficiency, and quality will be enhanced in
the short- and long-term. The host government, the local partner, and
the MNE all stand to benefit from his policy. In fact, for this approach
to be effective, it would require a well-developed industrial policy which
either targets or ranks various industries for development and foreign
participation. It should result in some pragmatic response to the type,
timing, and process of effective control by local entrepreneurs. Because
the terms will be negotiated and will constitute part of the government's
policy for encouraging FDI, the approach should decrease the
uncertainty to MNEs, stabilize their FDI expectations, and induce FDI as a
Finally, any attempt to address squarely the issue of control,
certainly one of the most important concerns of MNEs, will be beneficial to
both the host government and the MNE. At least, the host government
would again be actively involved in the FDI process. The experience of
several countries suggests that active involvement might be very
important for the type, size, timing, and control over FDI.1" This policy
prescription is likely to be unpopular because of the issue of monopoly rents
and the politico-economic ramifications of denationalization. However,
an efficient use of this policy option must be put within the context of a
host country's short- and long-term economic objectives. It might be
effectively used where the host country develops some capability to
effectively monitor aspects of MNE operations such as profits and intra-firm
transactions. On efficiency grounds, therefore, various local statutes
seeking to reduce inefficiencies in the market by controlling MNE
intrafirm transactions are well-motivated. However, these regulatory schemes
should be such that they do not interfere with host governments'
industrial policies and operational efficiencies. Ownership and control
structures should then encourage maximum and efficient resource utilization
and output levels within the context of host countries' distribution goals.
The industrial organization theory also posits that certain industrial
traits at home and abroad induce FDI.' 54 With regard to developing
host country policies, it is important to determine whether the industrial
traits work as a pull from or a push toward the host country. In other
words, which market structures are most active in generating FDI
behavior, those of the home or host countries? Are market imperfections of
the home needed first to serve as a push of FDI to the host country?
From the literature surveyed so far, it would appear that the
indus153 Yelpaala, Impact of IndustrialLegislation, supra note 20.
154 Caves, supra note 98, at 24.
try specific characteristics of the home country serve as a push while
those of the host country serve as a pull for FDI. According to the
theory, MNEs seek to replicate their home country industrial traits and
market structures abroad. By accomplishing this, they can then easily
exploit the monopolistic advantages which were the basis for their
market power at home. Sanjaya Lall has suggested that MNEs have
monopolistic advantages at home which may be transferable to the host
country.155 Therefore, it would appear that an oligopolistic market
structure in the home country would be a necessary, but not sufficient,
condition for FDI. Where the host country has no policy of restricting
FDI, and the opportunity to replicate the industrial characteristics and
exploit the monopolistic advantages exist, the market structure of the
host country will serve as a pull for FDI. As long as the opportunities
exist in the host country for MNEs to reap monopoly profits or to
maintain their relative market positions, why should the host country also be
called upon to induce FDI through tax incentives or other subsidies?
With the transferable monopolistic advantages at home and the
profitable opportunity to exploit them abroad (along the lines discussed in the
theory) the necessary push and pull exist and no incentives are required,
at least not tax incentives.
In fact, intuitively, the basic thesis of the theory militates against tax
incentives. Seller concentration generally is indicative of the market
power of MNEs. Given their dominant position in the host country's
economy or the potential for it, they may charge monopoly prices and
reap excess profits. Take, for instance, the several and different types of
intra-firm transactions that could be used to increase the returns to a
parent company, as discussed earlier. When these are taken together
with the potential for excess profits, there is no reason to offer incentives
to increase these returns to the MNEs. In fact, Richard Caves suggested
that, when excess profits are earned, host countries should impose an
excess profits tax.156 Since the MNE and its activities are a result of
market imperfections generated at home which may lead to economic
and political power abroad, it is doubtful whether host countries should
create a permissive environment that would complicate their policy
options. Based on the basic tenets of the theory, tax incentives would be
both misplaced and redundant.
The argument is further strengthened by Fredrick Knickerbocker's
findings of oligopolistic reaction.157 If foreign investors have a kind of a
155 Lall, MonopolisticAdvantages, supra note 75, at 102.
156 Caves, supra note 98, at 24.
157 F. KNICKERBOCKER, supra note 106, at 5.
"herd mentality," all that is required is to induce the first MNE and the
rest would follow. The first MNE will not move, however, unless it can
replicate and exploit its monopolistic advantages abroad. Tax incentives
merely increase the monopoly returns and speed up the process.
Therefore, tax incentives are not effective as an inducement but are only
relevant, perhaps, as to the timing.
It would appear that certain host country policies could induce FDI
under this theory even if the desirability of doing so is in doubt. The
possession of technology, trademarks, brand names, and other industrial
properties protected by law give MNEs the opportunity to control their
entrepreneurial resources and create an effective barrier to entry. This
implies that both the home and host countries should have effective
industrial property protection laws which can give MNEs some monopoly
power to exploit their know-how and effectively differentiate their
products from others. These laws would assist MNEs in the erection of the
necessary barriers to entry and hence give them the opportunity to reap
monopoly or near-monopoly rents. Where effective entry barriers exist,
either through inducement or not, MNEs would be inclined to exploit
their monopolistic advantages in the host country. However, from a
cost-benefit perspective the host government might consider
countervailing disincentives to monopoly prices both in the goods and rental
markets. The host country may set limits on rents, royalties, and fees
chargeable on the transfer of know-how, management, and technical
service agreements to subsidiaries as has been done, for example, in Brazil.
It may also institute effective accounting control systems and subject
monopoly profits to excess profits taxes.
Another area of possible host government policy is in the creation of
technology. As shown earlier, there is a positive and strong relationship
between technology intensity and FDI. Industries that are characterized
by heavy expenditures in knowledge creation are also very active in
foreign investments. One then wonders whether host countries should not
play an active and aggressive role in the R & D component of FDI.
Given the growing concern of several host developing countries over the
appropriateness of technology, host countries should seek to influence
the technology that enters the FDI at the R & D, or adaptation stage.
Government subsidies on specific technology development based upon
host country or regional needs might be appropriate. These incentives
would relate to the cost of technology development. That is the most
appropriate stage to influence the nature of the technology that enters the
host country's industry.
It should be stressed that these types of host government subsidies
alone would not be sufficient for FDI since the theory suggests several
other elements. In addition, the financial resources of MNEs are
comparatively much larger than certain host countries, thereby raising the
question of the appropriateness of these subsidies in relation to the
financial ability of host countries. The quest for monopoly profits, however, is
not likely to take MNEs to such countries. Besides, any attempt by host
countries to actively participate in the direction and nature of their
economic fate is required. Even if such an attempt is ineffective in the short
run, it will nevertheless help to focus attention on the most important
policy issues and areas of conflict.
Finally, in trying to determine the incentive policy implications, it is
also important to keep in mind another crucial element: the flow of FDI
occurs between similar industries even though diversification across the
industry is not impossible. This implies that any incentive policies
should be industry specific. Host countries should focus their FDI
inducement efforts at specific characteristics of industries seen as important
to their industrial development. This approach provides some flexibility
to the host government in the choice of its policy instruments and the
selection of beneficiary firms. It would also reduce the cost of the
incentives and possibly eliminate redundancy.
The industrial organization theory permits a serious investigation of
the policy dilemmas of host countries seeking economic development and
independence through FDI in a world of MNEs dictated by free
enterprise and perfect competition. For the purposes of incentive policies, the
fundamental tenets of the industrial organization theory do not a priori
suggest any tax incentive policies for FDI. In fact, they suggest a
possible excess profit tax on FDI monopoly income. However, under this
theory, certain industrial traits may present some policy options for host
If MNEs actually prefer to invest in industries characterized by
certain market imperfections and also to own and control their foreign
operations because of these market imperfections, then policies designed to
encourage FDI should focus on either the market imperfections or
ownership and control. Market conditions in the world are generally
characterized by these imperfections, thereby suggesting that host countries
need not do anything. However, in the case of ownership and control,
the theory suggests denationalization or policies that would encourage
foreign ownership and control. The issue of ownership and control is,
nonetheless, so complex that this policy prescription is not necessarily
appropriate. In view of the political economic heat that the MNE often
generates in the host country, and various host country industrial and
distribution policies, ownership and control can be used as a means of
encouraging certain types of FDI. In addition, the host country might
try to encourage certain initial technology creation activities and improve
its protection of industrial property rights.
V. INTERNALIZATION THEORY OF FDI
Internalization is the most ambitious recent attempt to develop a
comprehensive, complementary, or eclectic theory of the FDI process.
As a concept, internalization was first developed by R.H. Coase in 1937
to explain the process in the domestic context.1 58 The theory, however,
provides some basis for explaining the motivation for FDI and the
growth of the MNE. It was synthesized by Peter Buckley and Mark
Casson, 159 used by John Dunning to explain his eclectic approach to
FDI,1 0 and systematically developed into a general theory of FDI by
Alan Rugman. 161 According to Dunning, no theory standing alone can
explain such a complex phenomenon as the FDI process. 162 All the
major theories of FDI must be seen as complementary to one another under
the unifying theoretical theme of internalization.
Internalization may be defined as bringing under the same
ownership and control, the same activities linked together by markets. The
firm, according to the internalization theory, is "an organization for
allocating resources without the exchange of ownership. ' 163 There are two
sets of markets within which a firm may transact its business: the
external markets and the internal markets. External markets refer to market
transactions involving arms-length deals which result in the exchange of
ownership. On the other hand, internal markets concern deals which are
merely notional and in which ownership does not change. 164
Internalization will take place in response to any type of distortion or market
imperfections in the external goods and factor markets. 165 The basic
motivation behind internalization is efficiency. The operational efficiency
of any business requires that it be able to coordinate various aspects of its
158 Coase, supra note 59, at 386.
159 p.BUCKLEY & M. CASSON, supra note 35.
160 J. DUNNING, INT'L PROD., supra note 37.
161 Rugman, Internalizationas a General Theory ofForeignDirectInvestment: A Re-appraisalof
the Literature, 116 WELTWIRTSCHAFrLICHEs ARCHIV. 365 (1980).
162 J.DUNNING, INT'L PROD., supra note 37, at 33.
163 See M. CASSON, supra note 65, at 45.
165 P. BUCKLEY & M. CASSON, supra note 35, at 36-37.
activities through a set of efficient external markets-in the goods and
factor markets. However, the markets for several intermediate products,
defined broadly to include intangible assets, human skills, knowledge,
and semi-finished products, are either inefficient or difficult to organize.
Internalization permits a firm to by-pass these inefficiencies or replace
them with its own internal efficient set of markets. It is, therefore, the
internalization across national boundaries that explains foreign
production, sales, and other operations of MNEs. 16 6 The result of
internalization is that each market is fragmented, since each MNE constitutes a
separate and independent system from other participants in the same
market. Thus, there are few if any intersectoral linkages with
The internalization paradigm suggests that the MNE is the efficient
alternative to the lack of free trade.167 In other words, if the world were
characterized by free trade there would be no need for the MNE.
However, since the underlying conditions for a perfect competitive model in
the goods and factor markets do not hold in the normal world, free trade
between nations is destroyed. The use of the MN-E and the FDI process
is then an efficient second-best response to the existence of market
failure. In the normal Herscher-Ohlin world, location-specific endowments
lead to international trade. Under the internationalization paradigm,
however, firm specific endowments or advantages internal to the MNE,
together with industry specific factors, lead to FDI. These firm specific
advantages are the same as those described earlier as intangible assets.
The intangible assets hypothesis and the internalization theory both view
the exclusive possession of some proprietary knowledge, know-how, or
trade secrets as very important in the FDI process. They also consider
these assets to have the character of public goods and transactions in
them are characterized by market imperfections. These imperfections
relate to the industry specific factors.
The two theories provide different explanations of the FDI process
emerging from the same sources of market imperfections. The intangible
assets theory, its critics claim, ignores the costs associated with the
creation of the firm-specific advantages. 168 Moreover, it predisposes one to
think of a single firm-specific asset, innovation, or endowment, in which
case leasing would be most appropriate. Consequently, the theory fails
to focus on the methods of the transmission of innovative capability
167 Rugman, supra note 161, at 366, 368.
168 See P. BUCKLEY & M. CAssoN, supra note 35, at 37-44; J. DUNNING, INT'L PROD., supra
which requires a set of efficient markets. FDI takes place under the
intangible assets theory to make some returns on a costless firm-specific
advantage. On the other hand, under the internalization theory, FDI is
motivated by the efficiency considerations of MNEs. Considerations that
induce them to attempt to own or control markets of their specific
endowments so that they can appropriate a fair return for their costs in
knowledge creation expenditures.
The market imperfections that may trigger internalization are
several, but can conveniently be categorized into two groups: (1) market
imperfections which may arise from exogenous government-induced
inefficiencies, and (
) market imperfections which arise from normal market
failure in the goods and factor markets. 169 Government-induced
imperfections will normally arise from government intervention in the
international markets through tariffs and non-tariff barriers, controls, and
restrictions on international capital flows, foreign exchange regulations,
and discrepancies between corporate tax rates.170 These imperfections, it
is argued, provide an incentive for internalization at the international
level. For instance, exchange control restrictions or corporate tax
differentials would induce an MNE to shift its profits form one country to
another through its own internally-controlled transfer pricing system by
the use of a foreign subsidiary.
Similarly, several non-government market failures will provide the
incentive for internalization. For instance, imperfections may exist in the
intermediate goods markets where the transactions are characterized by
significant time lags without any efficient futures markets.
Discriminatory pricing may be required to exploit market power.17 1 There may also
be significant disparities in the bargaining power between buyers and
sellers of intangible assets. Finally, in the case of bilateral concentration of
market power, the costs of sanctions inherent in the bargain process
create some uncertainty and, therefore, may require some joint control or
ownership. Any or all of these market imperfections which might be
structural or cognitive would induce an MNE to replace these
imperfections with its own internal efficient markets for these goods.
Given the advantages of internalization outlined above, MNEs will
internalize their international operations if and only if the benefits
outweigh the costs of doing so." Of importance to the cost function of an
internalizing MNE are those associated with generating its firm specific
169 P. BUCKLEY & M. CASSON, supra note 35. See also Rugman, supra note 161, at 371.
170 P. BUCKLEY & M. CASSON, supra note 35, at 44.
171 Id. at 38.
172 Id. at 37-40.
advantages described as "public goods." Internalization should be seen
as an attempt by the MNE to make some positive net returns on its
costly prior outlays in knowledge-generating investments. These costs
are not sunk costs. Internalization will also involve several other cost
which have to be minimized. First, internalization will increase the
resource costs resulting from market fragmentation. Second, since
internationalization might involve a monocentric or polycentric control system
of geographically-dispersed units of internalized operations, both the
administrative and communication costs of coordinating these units will
increase and, thereby, increase the general overhead costs. Finally, host
governments might also increase the costs of internalization by various
types of overt and covert discriminatory actions against foreign-owned or
-controlled subsidiaries. These types of politically-induced costs of
foreign operations will be positively correlated to the political stability of
the host country. When these cost items, weighed against all the benefits,
are not prohibitive, internalization of foreign operations will take
Internalization, its proponents assert, is a general theory of FDI
since it is eclectic and seeks to incorporate other theories of FDI as
consistent and complementary explanations of the FDI process. It is
claimed that internationalization incorporates profit maximization
behavior in the real world of market imperfections-an element not
considered by the neo-classical Herscher-Ohlin profit maximization theories. 174
As shown above, it incorporates and explains further the intangible assets
hypothesis. Furthermore, it also explains the industrial organization
theory as a complementary part of internalization. Oligopolistic market
structures, collusion, and barriers to entry all manifest certain
imperfections in the external market which explain the desire for ownership and
control, thus internalization. 175 Finally, internalization incorporates
Raymond Vernon's product cycle hypothesis which explains FDI as
being sequential and determined by the life cycle of the product:
innovation, spread, maturity, and senescence. 176 FDI takes place at the
173 Id. at 44.
174 See Rugman, supra note 161, at 371-73, and J. DUNNING, INT'L PROD., supra note 37.
175 See Rugman, supra note 161, at 371-73, and J. DUNNING, INT'L PROD., supra note 37.
176 The product cycle hypothesis was one of the most important explanations of the FDI process
of United States manufacturing enterprises. It has, however, been partially rejected by its author
and other economists as an important explanation of FDI. See Vernon, InternationalInvestment
and InternationalTradein the ProductCycle, 80 QUART. J. ECON. 190 (1966). For Vernon's
reconsideration ofthe product cycle hypothesis, see Vernon, The ProductCycle Hypothesis in a New
InternationalEnvironment, 41 OxF. BULL. ECON. STATIST. 255 (1979). For a critic's view of it, see
Giddy, The Demiseof the ProductCycle Model in InternationaBlusiness, 13 COLUM. J. WORLD BUs.
maturity stage of the product. The connection between internalization
and the product cycle hypothesis is the initial motivation for research
and knowledge creation which lie at the center of internalization. Once
the initial outlays in and motivation behind R & D are explained,
everything about the product cycle hypothesis follows. 177
A. Incentive Policy Implications
A discussion of the incentive policy implications of the
internalization theory requires an understanding of the importance of this theory to
the definition of the MNE adopted in this study. According to the
theory, the MNE is motivated by the desire for internal operational
efficiency which can more easily be achieved through ownership and
control. As such, the internalization theory helps to explain why,
functionally, the MNE constitutes a multinational system. The MNE is more
like an organism responding to various internal and external stimuli
given certain internal capabilities. It seems to function best when it is
within an environment where it controls its operations. Thus, the theory
seems to explain the reasons for the preference of MNEs for shuttling
their resources within their systems. It also explains the desire of MNEs
to create options and flexibility so that they can respond to various
economic stimuli internationally. In addition, it offers important insights
into transfer pricing and market fragmentation.
One of the benefits to be derived from MNE internal control systems
is the ability to set prices for transactions or to allocate costs within the
system without reference to prevailing prices or cost structures.
However, the MNE system represents market fragmentation. Each individual
MNE constitutes a separate internal market with very little or no
integration with others within the same industry. Thus, an industry in a host
country, such as the aluminum industry in Ghana, could be fragmented
along oligopolistic lines and yet would be connected in an undesirable
way with other countries within the MNE system. The internal
operational dynamism of MNEs explains the concerns of several home and
host countries over the effects of MNEs' activities on national economic
policies and development. Therefore, several questions arise as to the
appropriate host country policies towards FDI.
An important factor in determining the possible policy prescriptions
of the internalization theory is the issue of costs and benefits.
Speciflcally, how are the costs and benefits from MNE operations distributed?
According to the theory there are two types of costs and benefits involved
177 See Rugman, supra note 161, at 371-75.
in internalization: 1) private costs and benefits relating to the firm,1 78
and 2) social costs and benefits relating to the national welfare
implications of MNE operations.1 79 FDI takes place when the benefits to the
MNE outweigh the costs.
The fact that it is beneficial for the MNE to internalize does not
necessarily mean that the host country or society at large benefits from
the operations of the MNE. The operational costs of MNEs might
understate the social costs. Therefore, the private benefits might exceed the
social benefits. In an explanation of the policy implications of the MNE,
Mark Casson suggests that, in several areas, MNEs might be benefiting
more than host governments since the private benefits exceed the social
benefits. 8 0 In this respect Casson made the startling assertion that there
might be too much FDI in the world.18 In his view, that MNEs benefit
more than governments from internalizing their operations implies that
there could have been more arms-length licensing arrangements,
presumably between MNEs and local licensees. In fact, he suggests that host
governments should improve the external market environment to make
ltieccehnnsoi nloggy.1ea82sier by providing better protection to owners of
Taken to its logical conclusion, the argument advanced above
suggests that host governments should not encourage FDI. In fact, they
should discourage it. Any incentives that would increase internalization
would be misplaced on the basis of the current costs and benefits
involved. However, an incentive policy depends upon the validity of the
assertion that there might be too much FDI. It has been shown that very
little of the FDI goes to the poorest of the developing countries. Even
among the richer of the developing countries there is a concentration of
FDI in the richest of them. Taken as a whole, the developing countries
have continued to attract only a small fraction of the FDI in the world,
accounting on the average for only one-third of the total flows of FDI
between 1960 and 1979.83 Therefore, it is doubtful whether, as a general
statement, Casson's suggestion is applicable to several developing
coun178 See generally J. DUNNING, INT'L PROD., supra note 37; P. BUCKLEY & M. CASSON, supra
note 35; M. CASSON, supra note 65; Giddy, supra note 176. See also Gamier, Context and Decision
Making Autonomy in ForeignAffiliates of U.S. MultinationalCorporations,25 ACAD. MGMT. J. 893
(1982) (an interesting article that bears on the internalization theory).
179 See J. DUNNING, INT'L PROD., supra note 37, at 33.
180 M. CASSON, supra note 65, at 83.
181 Iae at 84.
182 Id at 85.
183 Id at 50, 89-91.
tries. What Casson has uncovered is a distribution problem not unique
These countries seem to recognize the distribution problem and the
need for improved relations in arms-length technology transfer
transactions. Yet it is doubtful whether the distribution problem is an
outgrowth of the insufficient number of licensing contracts. It might also be
explainable by the industrial traits and the general environment of the
MNE. Increasing the number of licensing contracts might have very
little practical utility to host countries without the necessary local
entrepreneurial group to engage in licensing operations. Moreover, there is
an implicit assumption that licensing will improve the host country's tax
revenue picture. This does not follow. Any increased protection to
industrial property as suggested by Casson might increase the market
imperfections and the disparities in bargaining power between licensors and
licensees, thereby encouraging oppressive terms. For several host
developing countries, the question still remains as to how to entice and
channel sufficient amounts of FDI for their economic development.
It is useful to examine the nature of the conditions for FDI to see
whether they are susceptible to inducement policies. It should be noted
that the policy discussion of the intangible assets hypothesis and the
industrial organization theories are relevant since they also relate to the
same market imperfections discussed here. The internalization theory
draws a distinction between two market imperfections that explain
FDI-those that are germane to market failure and those that are
induced by government policies. Government-induced imperfections
would include tariff barriers, tax rate differentials, and exchange
restrictions. If either or both of these sets of imperfections exist, the necessary
incentive for FDI operations will exist.
It is, once again, possible to have redundancy and costly tax
incentive policies. If market failure exists, then the government-induced
imperfections need not be generated. If market failure does not exist, then
tariff and non-tariff barriers might provide the necessary incentive for
defensive investments, thereby making tax incentives redundant under
those circumstances."8 4 The theory also asserts that the world is
imperfect and that the information markets are almost always imperfect,
thereby suggesting that the necessary conditions for FDI exist through
184 See generally Horst, OptimalBehavior, supra note 32; Mundell, supra note 91.
the malfunctioning of the markets. Hence, host governments need not
generate more imperfections, particularly in the form of tax incentives.
The argument against government-induced market imperfections is
even stronger given that they generally tend to increase the social costs of
FDI which might not be compensated for in the distribution of MNE
generated wealth. The efficacy of tax incentives might be in doubt for
other reasons. One of the main arguments of the internalization theory is
the possibility that the MNE may engage in market fragmentation
whereby it closes its internal markets to outsiders. Through this process,
it can erect some barriers to entry and engage in transfer pricing to take
advantage of tax rate differentials between countries. The importance of
transfer pricing as an incentive for sourcing investments in several
industrializing countries cannot be doubted from the previous discusison of
the MNE system. The extent to which this will continue to be an
incentive for FDI will depend on the response of host countries to transfer
There appears to be greater host country awareness of the problems
of transfer pricing and the need to control intra-firm transactions.
Several countries have sought to regulate such transactions in technology
transfer and other commodities trade. 85 However, the efficacy (but not
the desirability) of such regulations is in doubt. Mark Casson suggests
that one of the best policies towards redressing the distribution problem
in FDI requires tightening up host countries' regulations on transfer
pricing so as to close all loopholes and effectively enforce the
regulations. 86 This would eliminate tax differentials and price manipulation as
an incentive for FDI. It would also increase the social benefits to the
With better host country accounting controls and better information
on prices and costing practices, the efficacy of tax incentives as an
encouragement for related party transactions is very doubtful. This is
particularly so in industrialized countries. Even in the case of developing
countries, the sensitivity of host governments to accounting malpractices
is such that transfer pricing will most likely not continue be a very strong
incentive in the FDI process.
2. Controlas an Incentive
Although tax incentives might not appeal to the policy analyst
under this theory, other policy instruments are available. First among
these is the issue of control. Like the industrial organization theory,
in185 UNCTC, SALIENT FEATURES, supra note 23, at 39. See also Rosenn, supra note 143.
186 See M. CAssoN, supra note 65.
ternalization centers around the desire to control the operations of the
firms. As explained earlier, if the basic conditions for internalization
generally exist, then the question is whether host governments should
permit and encourage ownership structures that would place the locus of
control both dejure or defacto in the MNE. An extensive analysis of the
issue of control as a possible incentive policy under the industrial
organization theory has already been developed. Several of the arguments
advanced there are equally applicable here. However, one should note the
difference between the two theories. The industrial organization theory
raises several questions about economic distortions, the loss of policy
autonomy, or denationalization that might arise from MNE control over
FDI. The internalization theory seems to associate MNE control with
efficiency. Therefore, it fails to recognize several of the problems
inherent in the political economy of the MNE and FDI. Under this theory,
internalization raises only distribution problems, not issues relating to
the path of industrialization. Internalization does not even raise
questions about the choice of technique as a related question of ownership
and control. Thus, for the host governments, control over operations is
not as important as controlling prices within the internal markets to
eliminate transfer pricing and market fragmentation.
The implication is that foreign ownership legislation could be an
incentive for FDI. Local ownership and control is not essential as long as
foreign ownership and control is accompanied by the appropriate policies
resulting in distributive equities in the private sector and social benefits
from FDI. Though this argument is similar to others advanced under
the industrial organization theory, it should be stressed that price-related
solutions alone cannot correct the problems of market fragmentation and
inappropriate technology. First, the host government policies should
include prescribed technology and local content requirements for various
FDI activities particularly in manufacturing. This type of policy would
tend to generate the use of appropriate technology and the necessary
intersectoral linkages. Then, policies controlling intra-firm prices would
ensure that the internal markets of the MNE are open and the prices of
their transactions with third parties are fair. Hence, there would be a
policy of encouraging FDI through foreign ownership and control
statutes, but ensuring that such ownership and control does not result in
monopoly profits, manipulation of host countries, or denationalization.
Besides the issue of control, there are other possible policy
instruments relating to costs associated with the political environment in the
host country. Internalization may involve costs associated with host
government's discriminatory practices against foreign firms. It may also
involve costs related to political instability and uncertainty. These may,
in turn, induce the firm to underinvest in the collection of information
about the host country's environment, making investment decisions
difficult. It might even involve costs related to the regulatory functions of
the host government.
Under these circumstances, it is possible to introduce policies to
encourage FDI subject to the issue of distribution. For example, the host
country might reduce or eliminate any discriminatory practices against
foreign firms. It may also eliminate irrelevant regulatory functions and
adopt the appropriate policies towards controllable political instability.
The host government might even subsidize the search costs involved in
FDI by investing in information gathering and dissemination. Though
these cost reduction policies alone will not act as incentives for
internalization, they are nevertheless important in the internalization calculus.
Together with the MNE-related costs, they constitute sufficient, but not
necessary, conditions for internalization. Here again cost-related policies
re-emerge as possible incentive instruments, as opposed to the current
emphasis on tax reductions.
Finally, of crucial importance to the internalization theory is the
foreign economic policies of a host country. The MNE, according to this
theory, thrives on a market environment that permits the MNE's
functional dynamism and its ability to shuttle resources between various
points within its system. This suggests quite strongly that a host country
bent on attracting FDI must adapt its foreign economic policies to
respond to these factors. To create the appropriate business and economic
environment for FDI, its economic policies must include free trade and
free enterprise. Nonetheless, since the theory was designed for an
imperfect world, the least that would be required of a host country is a specific
policy that permits the MNE to engage in its distinctive behavior even if
the general policy is one of restriction.
Although tax incentives have remained significant policy
instruments for attracting FDI, they have been largely ineffective. It was the
objective of this article to develop alternative policies for encouraging
FDI based upon different theories of the MINE. Such a task could not be
carried without a thorough understanding of the MNE.
If functional dynamism, operational flexibility, and a "foot-loose"
mentality in the use of its financial, human, technological, and other
sources are the hallmarks of the distinctive behavior and characteristic
attributes of the MNE, then a host country determined to attract
significant amounts of MNE investments must first develop a sophisticated
understanding of the MNE system. This understanding will permit host
countries to develop a more comprehensive and consistent policy
towards the MNE and its investments. This is of fundamental importance
when one considers that the MNE can sometimes evoke an adverse
reaction even among those who seek to entice FDI.
Consistent and meaningful policies require, in addition, some
theoretical basis. In the past, various theories of the MNE were canvassed as
alternative and competing explanations of the MNE and the FDI
process. This article has explored two such theories: the intangible assets
hypothesis and the industrial organization theory. What is interesting
about them is their similarity. For a host country trying to develop
effective policies for attracting FDI, it is neither desirable nor advisable to
treat theories of the MNE as alternatives. As suggested by the
internationalization theory, an eclectic approach is required. The MNE and its
environment constitute a very complex phenomenon requiring such an
eclectic approach for the development of appropriate host country's
policies. For these reasons, the policy prescriptions canvassed in this article
should be seen as complementary to each other.
Three major areas of incentive policies were explored in this article:
technology, labor, and ownership and control. Since FDI generally
constitutes a bundle of resources, capital, technology, and labor, an effective
way to develop policies for encouraging FDI is to evolve separate but
complementary policies towards each component of the FDI bundle of
resources. One such policy would relate to the technology that enters the
MNE system. The technology chosen by the MNE for a particular FDI
might well suit the needs for the MNE's system but not those of the host
country. The technology and its products might be inappropriate not
because of the level of development of the host country, but because they
do not fit into the host country's industrial policies nor foster certain
policy objectives in either the short or long run. Thus, given certain host
country industrialization policies, some specific technology transfer
incentive policies can be adopted to ensure the appropriate technology in
FDI. Suppose the host country is determined to develop a complex
industry, that would require complex technology and the appropriate
corresponding technology transfer policy. The host country may subsidize
the cost of the required complex technology. It may subsidize the
training of the required skilled local labor for such technology or encourage
the efficiency of labor. It may also stabilize the rental rates for
technology transfer. To encourage local adaptation, R & D and the
development of significant local capability in a particular industry, the host
country may subsidize R & D and technology that fits the adaptation
needs through direct grants. It may also subsidize and actively
participate in the training of researchers. In the medium- and long-term, these
incentives could have a significant and positive impact on attracting FDI.
Though the host country can encourage the use of the appropriate
MNE-owned technology in the FDI process, there appears to be even a
more fundamental policy for technology. Since FDI generally involves
the use of MNE-owned technology developed through R & D, it should
be possible for host countries to influence the initial development of
specific appropriate technology. This could be achieved through joint
research ventures between MNEs and host governments. It could involve
just one MNE and a host government, or several MNEs and several host
countries with similar technology and FDI needs. The rights of the
parties and the form of host country involvement would all be a matter of
Given the technology input policies, the host government can then
develop other complementary labor policies to enhance its total incentive
package. It is generally believed that cheap labor policies are required to
entice FDI. However, the evidence seems to suggest that the efficacy of
cheap labor policies might depend on several factors including, in
particular, the skills and efficiency of labor. Therefore, the appropriate labor
policies for attracting FDI must be task directed. Various industries
require various forms of labor input. Since the approach examined here
calls for industry-specific policies within a coherent general industrial
policy, different labor policies might be appropriate for various
industries. For instance, labor policies may be directed at labor costs, its
efficiency, the acquisition of skills, or any other factor depending upon the
industry, the labor situation, and the host government's objectives.
Whatever policies are adopted, the host country should be sensitive to
and develop complementary policies to address the issue of distributive
equities and ensure that the burden of industrialization is not borne by
Ownership and control over foreign operations, some theories have
stressed, go to the very root of the motivation of the MNEs to invest
abroad. If that is the case, then logic demands that any policy designed
to entice the MNE to invest abroad must address the question of
ownership and control within the context of a host country's industrial policy,
its level of technical competence, and the strategic objectives of MNEs.
The question of who owns and manages the resources in a particular
industry might be irrelevant, important, or even crucial to the attainment
of certain host country's objectives. On the other hand, ownership and
control might be crucial to the MNE for a specific FDI or an aspect of it,
but not for other types of FDI. A blanket policy approach could lump
together all forms of FDI under the same ownership and control policies.
However, that is likely to be ineffective. An approach that recognizes the
need for discriminatory policies towards the issue of ownership and
control is required. Ownership and control could be used as incentives for
FDI where they are crucial to either or both the host country and the
MNE. The MNE has over the years demonstrated a remarkable
adaptability to its environment. Ownership and control policies that are
sufficiently responsible to the MNE environment, its strategic goals, and
combined with well-established international contract forms such as joint
ventures, production sharing agreements, service contracts, and fade-out
agreements, might be an effective way to entice FDI. For several host
countries, new and more dynamic arrangements could be even more
effective in enticing FDI. However, given the political economy of the
MNE and its tendency to fragment markets, the ownership and control
policies, should, when appropriate, include local content and intra-firm
pricing policies to minimize the negative impact of the MNE system on
distribution and intersectoral linkages.
In addition to these policy options, host countries may consider
information gathering and dissemination on FDI opportunities. They may
consider subsidizing research costs, improving controllable political
risks, and providing an adequate regime of industrial property protection
Finally, it is important to note that the mere passage of incentive
statutes is hardly sufficient to encourage significant amounts of FDI.
Apart from such laws and policies being coherent and task-oriented, an
active participation of the host government is important. Whatever the
host country does, it should keep in mind the nature of the MNE system.
It must also understand the characteristic attributes of the MNE, its
politico-economic implications, and the subsequent reaction that the MNE
sometimes evokes. It would appear that the more permissive the host
country's economic environment is, the better the climate for FDI. Even
where the host country is involved in an active management of its
economy, specific policies could still create the appropriate climate for FDI.
C. Incentive Policy Implications ......................... 1. Control and FDIPolicy ........................... 2. Attitudes of Host Countries ........................ 3. MNE Responses to Ownership and Control Laws ....
2. Cost as an Incentive ............................... 72 Gruber , Mehta & Vernon, supra note 55 . 73 Horst, The IndustrialCompositionof U.S. Exportsand Subsidiary Sales to the CanadianMar-
ket, 62 AM. ECON. REV . 37 - 45 ( 1972 ). 74 T. PUGEL, INTERNATIONAL MARKET LINKAGES AND U.S. MANUFACTURING: PRICES ,
PROFITS AND PATTERNS 69 ( 1978 ). 75 Lall, MonopolisticAdvantages and Foreign Involvement by U.S. ManufacturingIndustry, 32
OXFORD ECON. PAPERS 102 ( 1980 ) [hereinafter Lall , MonopolisticAdvantages]. 76 See, e.g., T. PUGEL, supra note 74 , at 197-98; J. Jarrett , Offshore Assembly and Production
and the Internalization of International Trade Within the Multinational Corporation ( 1979 ) (Ph .D.
ANALYSIS OF DETERMINANTS AND EFFECTS 22 ( 1979 ). 77 T. PUGEL, supra note 74, at 18- 19 . 78 R. CAVES , supra note 7, at 11. 79 Convention for the Protection of Industrial Property , Mar. 20 , 1883 , 161 Parry's T.S. (Fr.)
409, revised July 14 , 1967 , 21 U.S.T. 1583, T.I.A.S. No . 6923 , 828 U.N.T.S. 305 . 80 An examination of the current draft on the proposed changes in the Paris Union would con-
for the DiplomaticConference on the Revision of the ParisConvention, INDUS . PROP., Nov . 1979 , at
243. 81 W. WHITE & B. RAVENSCROFT , PATENTS THROUGHOUT THE WORLD 571 ( 1982 ).