Should Countries Promote Foreign Direct Investment?
Gordon H. Hanson
Department of Economics and School of Business Administration
University of Michigan
And National Bureau of Economic Research
September 2000
Abstract. In this paper, we ask whether policies to promote foreign direct
investment make economic sense. To contain our enterprise, we focus on the specific issue
of whether spillovers related to production by multinational firms are sufficient to justify
subsidies to FDI. Many developing countries offer a wide variety of tax breaks to FDI. The
available evidence suggests that FDI is quite sensitive to host-country characteristics.
Higher taxes deter foreign investment and a more educated work force and larger consumer
and industrial markets attract FDI. There is weak evidence, at best, that FDI generates
positive spillovers for host economies. Plants in industries with a larger multinational
presence tend to experience lower rates of productivity growth. Using a simple theoretical
model, we show that subsidies to FDI are likely to be warranted where multinationals are
intensive in the use of elastically-supplied factors, the arrival of multinationals to a market
does not lower the market share of domestic firms, and FDI generates strong positive
productivity spillovers for domestic agents. Existing empirical research suggests the first
and third conditions are unlikely to hold. In three cases of FDI promotion – Ford and GM in
Brazil and Intel in Costa Rica – there appears to be little basis for subsidizing FDI.
______________
I thank James Hines, Mustafa Mohaterem, and Dani Rodrik for helpful comments and
discussions. Poh Boon Ung provided excellent research assistance. This paper has been
prepared for the G24 research program.

1Executive Summary
In this paper, we examine whether policies to promote foreign direct investment
(FDI) make economic sense. While eliminating barriers to foreign investment help achieve
global market integration, promoting FDI goes one step further by favoring one form of
integration – expanded foreign control of productive assets – over others, such as increased
trade in goods, more international licensing of technology, or larger cross-border flows of
portfolio capital. Assessing the consequences of promoting FDI for national welfare is a big
task and one we in no way pretend to complete in full. To contain our enterprise, we focus
on whether existing academic research suggests that the benefits of FDI are sufficient to
justify the kind of policy interventions seen in practice. This will help identify a set of
practical guidelines for when and where promoting FDI might be welfare enhancing.
In the introduction, we frame the discussion by outlining the conditions under which
economic theory suggests that government policies favoring foreign over domestic capital
are justified. In Section 2, we review the types of policy incentives that G-24 and other
countries offer to multinational firms. In section 3, we survey the theoretical and empirical
literature on FDI, with an emphasis on research which examines whether FDI is a source of
positive externalities for host countries. In section 4, we develop a simple theoretical model
of FDI, which we then use to evaluate three cases in which developing-country governments
have offered special incentives to multinational firms – Ford and General Motors in Brazil
and Intel in Costa Rica. The point of these case studies is to see whether economic theory
and relevant empirical literature would suggest that policy intervention in favor of FDI was
justified. Finally, in section 5 we offer concluding remarks on factors to consider when
trying to determine whether promoting FDI will raise host country welfare.
For small open economies, efficient taxation of foreign and domestic capital depends
on their relative mobility (or, more generally, on their elasticity in supply). If foreign and
domestic capital are equally mobile internationally, it will be optimal (under standard
assumptions) for countries to subject each to equal tax treatment; if foreign capital is more
mobile internationally, it will be optimal to have lower taxes on capital owned by foreign
residents than on capital owned by domestic residents. Absent market failure of some kind,
there is no justification for favoring FDI over foreign portfolio investment. In practice,
countries appear to tax income from foreign capital at positive rates (if rates that are lower
than those for domestic capital) and to subject different forms of foreign investment to very
different tax treatment. FDI appears to be sensitive to host country characteristics. Higher
taxes deter foreign investment and a more educated work force and larger consumer and
industrial markets attract FDI. There is also some evidence that multinationals tend to
agglomerate in manner consistent with location-specific externalities.
There is weak evidence, at best, that FDI generates positive spillovers for host
economies. A commonly-mentioned possibility is that FDI raises the productivity of
domestic agents. While multinationals are attracted to high-productivity countries, and to
high-productivity industries within these countries, there is little evidence at the firm or plant
level that FDI raises the productivity of domestic enterprises. Indeed, it appears that plants
in industries with a larger multinational presence tend to enjoy lower rates of productivity

2growth over time. Empirical research thus provides little support for the idea that promoting
FDI is warranted on welfare grounds.
Using a simple theoretical model, we derive conditions under which subsidies to
FDI would be likely to raise host-country welfare. Subsidies to FDI are more likely to be
warranted where multinationals are intensive in the use of elastically-supplied factors, the
arrival of multinationals to a market does not lower the market share of domestic firms, and
FDI generates strong positive productivity spillovers for domestic agents. Empirical
research suggests the first and third conditions are unlikely to hold. In the three cases we
examine – Ford and GM in Brazil and Intel in Costa Rica – it appears that the second
condition holds but not the first or third conditions. This suggests that Brazil’s subsidies to
foreign auto manufacturers may have lowered national welfare. Costa Rica appears to have
been prudent in not offering subsidies to Intel.
There clearly is a need for much more research on the host-economy consequences
of FDI. The impression from existing academic literature is that countries should be
skeptical about claims that promoting FDI will raise national welfare. A sensible approach
for policy makes in host countries is to presume that subsidizing FDI is unwarranted, unless
clear evidence is presented (on a case by case basis) to support the argument that the social
returns to FDI exceed the private returns.

31. Introduction
There is a presumption among many academics and policy makers that foreign
direct investment (FDI) is somehow special.1 One common view is that FDI helps
accelerate the process of economic development in host countries. Optimism about the
economic consequences of foreign investment, coupled with heightened awareness about
the importance of new technologies for economic growth, have contributed to wide-reaching
changes in national policies towards FDI. During the last two decades, many emerging
economies have dramatically reduced barriers to FDI, and countries at all levels of
development have created a policy infrastructure to attract multinational firms.2 Standard
tactics to promote FDI include the extension of tax holidays, exemptions from import duties,
and the offer of direct subsidies. As of 1998, 103 countries offered special tax concessions
to foreign corporations that set up production or administrative facilities within their borders
(Avi-Yonah, 1999). Typically, these concessions are applied to multinational enterprises
but not to local firms in the same lines of activity.
In this paper, we examine whether policies to promote FDI make economic sense.
While eliminating barriers to foreign investment are a means of achieving global market
integration, promoting FDI goes one step further by favoring one form of integration –
expanded foreign control of productive assets – over others, such as increased trade in
goods, more international licensing of technology, or larger cross-border flows of portfolio
capital. Assessing the consequences of promoting FDI for national welfare is a big task and
one we in no way pretend to complete in full. To contain our enterprise, we focus on
whether existing academic research suggests that the benefits of FDI are sufficient to justify

1 Throughout the paper foreign investment refers to inward foreign investment.

4the kind of policy interventions seen in practice. This will help identify a set of practical
guidelines for when and where promoting FDI might be welfare enhancing.
In the remainder of the introduction, we frame the discussion by outlining the
conditions under which economic theory suggests that government policies favoring foreign
over domestic capital are justified. In Section 2, we briefly review the types of policy
incentives that G-24 and other countries offer to multinational firms.3 This will help fix
notions of standard practice. In section 3, we survey the theoretical and empirical literature
on FDI, with an emphasis on research which examines whether FDI is a source of positive
externalities for host countries. In section 4, we develop a simple theoretical model of FDI,
which we then use to evaluate three cases in which developing-country governments have
offered special incentives to multinational firms -- Ford and General Motors in Brazil and
Intel in Costa Rica. The point of these case studies is to see whether economic theory and
relevant empirical literature would suggest that policy intervention in favor of FDI was
justified. Finally, in section 5 we offer concluding remarks on factors to consider when
trying to determine whether promoting FDI will raise host country welfare.
Setting the Stage: Foreign versus Domestic Investment
To begin, it is helpful to specify what we mean by promoting foreign direct
investment. The benchmark one adopts depends on whether it is optimal for countries to
subject foreign and domestic capital to equal tax treatment. Countries may have reason to

2 A recent UN study shows that during the period 1990-1998, over 135 countries reduced regulatory restrictions
on FDI (UNCTAD, 1999c).
3 The G-24 consists of Algeria, Côte d'Ivoire, Egypt, Ethiopia, Gabon, Ghana, Nigeria, the Democratic
Republic of Congo, Argentina, Brazil, Colombia, Guatemala, Mexico, Peru, Trinidad and Tobago, Venezuela,
India, Iran, Lebanon, Pakistan, Philippines, Sri Lanka, and the Syrian Arab Republic.

5tax domestic and foreign capital differently. For a small open economy facing an immobile
supply of labor and an internationally mobile supply of capital, the optimal factor income
tax falls entirely on labor (see, e.g., Gordon, 1986; Razin and Sadka, 1991). Not taxing
capital is sensible because the immobile factor bears the incidence of any tax on factor
incomes, making it more efficient to tax the immobile factor directly.4 By extension, if
foreign capital is perfectly mobile but domestic capital is not, then the optimal tax on factor
incomes falls on domestic labor and capital but not on foreign capital. More generally, if
foreign capital is elastic in its supply relative to domestic capital, then it is optimal for
countries to tax income from domestic capital at higher rates (where the optimal tax rate on
income from foreign capital may be positive if its supply elasticity is less than infinite).5
If one presumes that domestic and foreign capital are equally elastic in supply (e.g.,
they are equally mobile internationally), then promoting FDI means any policy which
subjects foreign direct investment to favorable tax treatment relative to domestic investment
and foreign portfolio investment (whether in the form of debt or equity). If, on the other
hand, one presumes that foreign capital is more elastic in supply than domestic capital, then
promoting FDI means any policy which favors direct investment inflows over portfolio
investment inflows, holding constant a country's relative tax treatment of domestic
investment income and foreign investment income.

4 That small open economies do tax corporate income at rates comparable to or higher than large economies is
something of a puzzle. See Gordon (1992) and Gordon and MacKie-Mason (1995) for a discussion. Relatedly,
Rodrik (1997) discusses how globalization contributes to conflicts between an electorate demanding more
social insurance and a government able to tax fewer factors of production.
5 From the perspective of global welfare, the residence principle provides the basis for efficient international
taxation. The income of all residents, whether originating from domestic or foreign sources, is taxed at equal
rates and nonresident income is untaxed. However, if all governments were to grant foreign tax credits (and
many do), a small open economy would lose little by taxing nonresident income up to the rates at which income
is taxed in the home countries of nonresidents (see Avi-Yonah (1999) for a discussion).

6For FDI to merit special treatment, there needs to be market failure that is specific to
production by multinational firms.6 Asymmetric information between domestic and foreign
investors is one commonly-mentioned source of market failure, though one that is not
specific to FDI. If domestic investors are better informed about domestic investment
opportunities than foreign investors, then, all else equal, a capital-importing country would
raise welfare by subsidizing foreign capital inflows (Gordon and Bovenberg, 1996). Could
asymmetric information justify favoring FDI over foreign portfolio investment? Razin,
Sadka, and Yuen (1998) suggest the answer is no. Since FDI, but not portfolio investment,
gives a foreign investor a controlling interest in a domestic firm, multinational firms are
likely to be at a informational advantage relative to foreign portfolio investors (though not
relative to domestic investors). In this case, the optimal tax policy is to subsidize foreign
portfolio investment and to leave foreign direct investment untaxed.7
There are other sources of market failure which could justify special treatment of
FDI (Caves, 1995). A much-cited possibility is that FDI generates productivity spillovers
for the host economy (Blomstrom and Kokko, 1998). One idea is that multinational
enterprises possess superior production technology and management techniques, some of
which are captured by local firms when multinationals locate in a particular economy.8 A

6 One potential source of market failure we do not consider is that related imperfections in financial markets,
which some suggest may make FDI preferable to foreign portfolio investment in that it may be less volatile (see
Rodrik and Velasco (1999) on short-term capital flows). Hausmann and Fernandez-Arias (2000) and
Fernandez-Arias and Hausmann (2000) cast doubt on this hypothesis (and also discuss recent literature on
currency crises and foreign capital infows).
7 Another argument for promoting FDI is that multinationals force governments to "bid" for the right to host
their new facilities. If governments fail to offer sufficiently generous tax breaks, they may not attract FDI. A
multinational considering a large investment project may be able to extract tax concessions because countries
have in effect granted it market power by allowing the firm to hold a one-sided auction. The auction allows the
multinational to extract all benefits associated with its presence in a country. A preferred solution for host
countries in this case is multilateral cooperation to avoid one-sided bidding for FDI. See Bond and Samuelson
(1986), Black and Hoyt (1989), and Janeba (1998) for alternative views on the subject.
8 That multinationals may be technologically advanced relative to local firms (Davies, 1977; Teece, 1977) is
not in and of itself a justification for promoting FDI. If multinationals do possess superior technology, then we

7related source of spillovers is forward and backward linkages between multinationals and
host-economy firms (Rodriguez-Clare, 1996), which may result from multinationals
providing inputs at lower cost to local downstream buyers or by their increasing demand for
inputs produced by local upstream suppliers. A further possibility is that FDI shifts rents
earned by multinationals to the host economy (Glass and Saggi, 1999a; Janeba, 1996).
Multinationals may have global market power and may share monopoly rents with
managers and workers in their various operational units. By attracting multinational firms,
the host economy may capture a portion of the rents that these firms generate.9
While these and other types of market failure are plausible, each is also subject to
controversy. Spillovers associated with FDI are supported by casual evidence from many
countries, but their existence and magnitude are, as we shall see, difficult to establish
empirically. Indeed, micro evidence from large samples of manufacturing plants in
developing countries fails to support the existence of positive productivity spillovers related
to FDI. There is also reason to believe that multinationals tend to have market power in
their respective industries. Whether or not they share rents with employees in their foreign
affiliates is an empirical question. Attracting FDI may shift a portion of the rents that
multinationals earn to the host economy, but it may also reduce the profits of local firms that
compete with multinationals at home or abroad.10

expect the rate of return on their investments to be higher than that for local firms, in which case multinationals
require no artificial inducement to choose the optimal level of FDI.
9 Yet another possibility is that FDI reduces informational barriers to trade and investment between the host
economy and the rest of the world (Rhee, 1990). Agents elsewhere may lack complete information about factor
productivity, government policy, or the general business climate in the host economy. By attracting FDI, the
host economy may signal to the rest of the world that it has a positive environment in which to do business. In
this case, there are likely to be diminishing returns to promoting FDI. After the first several multinationals have
established themselves in a country, any signal from additional FDI is likely to be uninformative.
10 Other research suggests FDI may be detrimental to the host economy. Recent variants of this argument
include the idea that, once established in a country, multinationals favor high trade barriers and will lobby host-
country government to raise tariffs (Bhagwati et al., 1987; Blonigen and Ohno, 1998). There is some empirical

8Arguments for promoting FDI are based on claims about the economic environment,
which can and should be subject to empirical verification. Before deciding to promote FDI,
it is essential to evaluate possible sources of market failure associated with multinational
firms. It is this task to which we devote much of the paper.
2. Promotion of FDI in Practice
In this section, we summarize current government policies to promote foreign direct
investment in G-24 and other countries. We begin with a brief review of corporate taxation
at the national level and then discuss the range of tax and other incentives which countries
offer to multinational enterprises. The source for all data, except where noted, are annual
editions of Corporate Taxes: A Worldwide Summary by PriceWaterhouse.
Policies to promote foreign direct investment take a variety of forms. The most
common are partial or complete exemptions from corporate taxes and import duties. These
policies are typically the result of formal legislation or presidential decree, which apply to
all foreign corporations that meet certain restrictions. These restrictions vary considerably
across countries. In many cases they require that multinationals establish production
facilities in the host country in specified lines of activities or in designated regions, such as
export processing zones, and that multinationals export output that embodies inputs which
were imported duty free. Direct subsidies, and other types of concessions, are often
negotiated between multinationals and host governments on a case-by-case basis. Such
individualized subsidies appear to be common, but are hard to document systematically.

support for this view (Blonigen and Figlio, 1998). The difficulty of evaluating the economic impact of
potential policy changes resulting from FDI leaves these issues beyond the scope of this paper.

9Table 1 shows corporate tax rates in 1990 and 1998 for each G-24 country for which
data are available and averaged across regions for selected other countries.11 Since some
countries have progressive tax rates (lower rates for smaller corporations) or rates which
vary across sectors, we report the minimum and maximum tax rates which apply to
corporate income. In 1998, maximum tax rate on corporate income in the G-24 ranged from
a high of 57% in Iran to a low of 25% in Brazil. Several countries, including Argentina,
Columbia, Guatemala, Peru, the Philippines, and Sri Lanka, tax corporate income at a flat
rate, while others, including Ghana, Iran, Mexico, and Trinidad and Tobago, tax income
earned by small corporations at rates much lower than for large corporations. Between 1990
and 1998, most countries reduced their maximum corporate income tax rate, with high-tax
countries undertaking the largest cuts in absolute terms.
Tax rates in individual G-24 countries are roughly comparable to the averages for
other countries in their respective regions. A few outliers are apparent. On a region-by-
region basis, tax rates in 1998 were relatively high in the People’s Republic of the Congo,
India, Iran, and Pakistan. Tax rates in North America, Oceania, and Western Europe are on
average similar to those in Latin America and lower than those in Africa and Asia.
Tables 2-5 give a brief description of how G-24 and a sample of five other countries
treat foreign corporations that operate within their borders.12 Most countries for which data
are available grant corporate income tax exemptions to foreign corporations making inward
direct investments. Typically, these exemptions last for less than a decade from the
initiation of a new project, though in some cases they are long-lived. Most countries also
offer exemptions to foreign corporations on import duties, where these tend to be restricted

11 We include all countries for which data are available.

10
to inputs that are used to produce exports or, in a few cases, capital goods. Exemptions from
value-added taxes are a somewhat less common tax concession countries make available to
multinational firms. Similar tax concessions are also available to domestic firms in some
countries, though these concessions are for the most part tied to participation in export
processing zones (EPZs), export activity outside of such zones, or production in officially-
designated priority sectors or regions. Comparing 1990 and 1998, there is a slight increase
in the number of countries offering exemptions from valued-added taxes and import duties
and supporting EPZs.13
Not included in the tables are details on direct subsidies which host government
offer to multinational firms on a case by case basis. These arrangements are frequently
unpublicized, but the practice appears to be relatively common. Brazil is one country which
actively pursues multinational firms and has offered generous subsidies in a number of
instances (see the GM and Ford examples in section 4). For instance, the country gives
generous tax incentives to firms that locate manufacturing facilities in the Amazon region.
Unspecified government subsidies appeared to be important in luring Multibras, a U.S.-
owned firm, to construct a $400 million plant to manufacture air conditioners and
microwave ovens in Manaus in 1998. Investment subsidies also appeared to be important in
convincing Honda to build a motorcycle plant in the area. In the absence of tax breaks,
there appears to be little reason why multinationals would locate in the region.
Poorer countries in Europe have also been aggressive in pursuing multinational
firms. To give a few examples: in 1991, Portugal provided a lump-sum subsidy and
promised tax breaks on future earnings to Ford and Volkswagen in return for their

12 The non G-24 countries were chosen to represent regional diversity, important sources for FDI (Japan), and
countries with policies that are relatively friendly towards FDI (Costa Rica, Ireland, Thailand).

11
constructing a jointly-operated auto production facility in the country; in 1995, Ireland
granted employment subsidies to IBM and Citibank for locating data processing jobs in the
country (and granted similar subsidies to Berg Electronics the following year); and in 1996,
Turkey lured Honda to build a new auto production facility in the country by easing tax
rules on new plants and relaxing import duties on auto parts. Subsidies are by no means
limited to relatively low-income regions. Germany offered investment subsidies to
Advanced Mirco Devices in 1995, after it decided to build a semiconductor plant in Saxony,
and to Motorola in 1998, after it decided to build a new facility in Bavaria. In the United
States investment subsidies from state governments helped attract Mercedes-Benz to
Alabama and BMW to North Carolina.
It appears in Tables 2-5 that most countries do not follow the residence principle
(see note 5) in setting tax policies.14 While many countries tax domestic income earned by
foreign corporations at lower rates than domestic income earned by domestic corporations,
the former rates are in most cases above zero in the long run (though often not during the
first few years after a project is undertaken). Foreign tax credits, which allow corporations
to deduct taxes paid to foreign governments from their tax liability on foreign income,
complicates the picture. As of the mid 1990s, the United States, the United Kingdom, and
Japan granted foreign tax credits to multinational corporations based within their respective
borders, and many other high-income countries, including Australia, Canada, France,
Germany, the Netherlands, and Switzerland, exempted the foreign earnings of their firms
from domestic taxation (Hines, 1996). Where foreign tax credits apply, and where a
country's tax rate on domestic income earned by foreign corporations does not exceed the

13 See Madani (1999) for more details on EPZs.
14 This statement is based on the presumption that foreign capital is more mobile than domestic capital.

12
home-country tax rate for these firms, taxing foreign corporations merely shifts tax revenue
from FDI source countries to FDI host countries and does not necessarily distort investment.
One important question is whether the concessions offered to foreign corporations
evident in Table 2 represent favorable treatment of inward FDI relative to inward foreign
portfolio investment. A full evaluation of this issue is beyond the scope of this paper.
Income from inward portfolio equity investment, portfolio debt investment, and direct
investment are governed by complicated tax rules which vary considerably across
countries.15 We will, however, hazard a few general comments. Table 2 shows that income
from inward direct investment is subject to myriad tax breaks in G-24 and other countries.
With regards to inward portfolio equity investment, the presence of capital gains taxes,
which vary from country to country, would tend to disfavor this vehicle relative to FDI.16
With regards to portfolio debt investment, recent abolitions of withholding taxes on
portfolio interest income for foreign residents (Avi-Yonah, 1999), which occurred
throughout the OECD and in many developing countries as well, would tend to favor this
vehicle relative to FDI. The absence of portfolio interest income withholding taxes means
that foreigners do not pay tax on income they earn from corporate or government bonds,
bank accounts, or certificates of deposit in a country. To summarize briefly, it appears that
many countries may have tax policies which favor FDI relative to some types of inward
portfolio investment but disfavor it relative to others.
Other policies clearly do favor FDI. A country that offers exemptions to value
added taxes or import duties to foreign but not domestic corporations favors FDI, since

15 A comparison is further complicated by the proliferation of bilateral investment treaties and the indirect
effects of trade polices on FDI. See UNCTAD (1999a, 1999b).
16 However, given widespread evasion of capital gains taxes by nonresident foreigners in emerging economies,
the effective capital gains tax may be zero in many cases.

13
domestic corporations which receive foreign bank loans, issue bonds to foreigners, or have a
non-controlling portion of their stock owned by foreigners do not receive comparable tax
breaks. Export processing zones (EPZs) do not favor FDI over foreign portfolio investment,
so long as domestic firms have equal access to EPZs (which is often the case) and are
equally likely to engage in export production as are multinationals (which is less likely to be
the case). Direct subsidies to multinational firms, examples of which we discuss in section
4, also favor FDI relative to other forms of inward foreign investment.
III. FDI and Host Country Economic Performance
There is an immense academic literature on FDI and multinational firms. Since our
interests are rather narrow, we focus on empirical research which studies the impact of FDI
on host economies. Within this literature, we emphasize two strands: one which examines
the determinants of where multinationals locate production facilities and another which
examines sources of market failure related to FDI. We begin with a brief review of theories
of multinational production.
What Explains Multinational Production?
Following Dunning (1981, 1993), it is standard to view multinational enterprises as
arising from three distinct types of advantages. A firm must own or control a unique mobile
asset (e.g., a patent or trademark) it wishes to exploit (the ownership advantage); it must be
cost efficient to exploit the asset abroad in addition to or instead of in the firm’s home
country (the location advantage); and it must be in the firm’s interest to control the asset’s
exploitation itself, rather than contracting out use of the asset to an independent foreign firm

16
offering incentives, and (iii) raises the economic benefits of FDI relative to exporting and
arms length production in the host country.
An emerging body of literature considers how production externalities affect the
behavior of multinationals and the impact of FDI on host economies. The existence of
externalities associated with FDI raises the possibility that promoting FDI may be welfare
enhancing. One line of work suggests that the arrival multinational firms in an economy
may help jump start the process of industrial development by increasing the scale of
operations in domestic upstream and downstream industries – that is, by creating forward
and backward linkages. In Rodriguez-Clare (1996), the arrival of multinationals increases
an economy’s access to specialized intermediate inputs (which are produced in more
developed economies and accessible abroad only through multinationals), which in turn
raises the economy’s total factor productivity. The idea is that multinationals in effect give
less developed economies access to the stock of knowledge capital (which is one
interpretation of specialized intermediate inputs) in more developed economies. This access
makes labor and other factors in the host economy more productive. Multinationals crowd
domestic firms out of production, which raises the possibility that some domestic factors of
production may lose from FDI. Under some conditions, the demand for labor by the
entering multinationals is weaker than that for the exiting domestic firms, in which case the
arrival of multinationals may lower national welfare. Thus, the “linkage” effect of
multinationals on factor demand may be positive or negative.
There are other mechanisms through which multinationals impact industrial
expansion in host countries. Gao (1999) offers a model in which the creation of
multinational firms spreads industry from more- to less-industrialized countries, thereby

17
reducing industrial concentration in the former. By lessening industry agglomeration,
multinationals raise national welfare in less-industrialized regions but may lower it in more-
industrialized regions. Markusen and Venables (1999b) offer a similar model in which the
“catalyst” effect of multinationals on a host economy (through forward and backward
linkages) may spur domestic industry so much as to drive multinationals out of the market.
Additionally, while multinationals create forward and backward linkages, they also increase
competition for local firms and thus may redistribute income away from some groups (e.g.,
specific factors in industrial production).20 If the impact of multinationals on the
profitability of domestic firms is sufficiently negative, FDI may lower host country welfare,
in which case the optimal policy towards FDI is a tax (Glass and Saggi, 1999).
Another mechanism through which multinationals may create spillovers for local
industry is worker training. If multinationals bring new technology into an economy, then
local firms may benefit by being able to hire workers whose training costs have in effect
been paid by the multinational (Motta, Fosfuri, and Ronde, 1999). Of course, multinationals
may bid to retain these workers, in which case domestic labor captures the full benefit of
worker training. In either case, multinationals raise national welfare.
What Determines The Location of Multinational Production?
A large empirical literature examines the factors which determine where
multinational enterprises locate their production facilities. We discuss this literature briefly
in order to identify the potential efficacy of using tax incentives to attract FDI.
One strand of literature applies general equilibrium theories of multinationals to data
to see whether FDI is associated with variation in trade costs or factor costs across countries.

20 See Matouschek (1999).

19
host country. He interprets this result to mean that countries with larger supplies of human
capital are more likely to attract FDI, especially in sectors which are relatively intensive in
the use of skilled labor. In related work, Markusen and Maskus (1999a,b), using data on
aggregate sales by foreign affiliates of U.S. multinationals and aggregate sales by affiliates
of foreign multinationals in the United States, find that affiliate sales are higher when
source and host countries have relatively similar market sizes but are unrelated to
differences in the relative supply of skilled labor in source and host countries. They
interpret these results to mean that multinationals have expanded their global production
operations more through horizontal FDI than through vertical FDI.
The empirical work cited above is general equilibrium in orientation, in that it is
based on estimating a reduced form relationship between exports and/or multinational sales
and measures of industry technology, and country trade costs, size, and factor supplies. All
prices and outputs are implicitly endogenous. Other research takes a partial equilibrium
approach in that it examines the impact of policy or other host-country conditions on FDI,
holding constant at least some prices and sectoral outputs. This line of work focuses on the
country characteristics which appear to attract multinational firms.
In a widely cited paper, Wheeler and Mody (1992) examine outward foreign
investments by U.S. multinational enterprises. They find that U.S. outward FDI is higher in
countries with larger markets, a larger stock of initial FDI, higher quality of infrastructure,
and more industrialized economies. These results are broadly consistent with theory. The
authors interpret the correlation between current and past FDI to indicate that multinationals
are attracted to locations with a larger concentration of industrial firms – that is, that there

21
the impression given by Wheeler and Mody (1992), Brainard (1997), Yeaple (1999), and
others, a growing tax literature finds that FDI is lower in regions with higher corporate
taxes. The elasticity of FDI with respect to the after-tax rate of return is approximately
unity. One study which controls for omitted variables in a particularly convincing manner is
Hines (1996), who examines the allocation of inward FDI across U.S. states. He compares
investment from countries which grant foreign tax credits with countries that do not in high-
tax versus low-tax U.S. states. Relative to investors from countries which grant foreign tax
credits, investors from countries which grant no tax credits should be more sensitive to
cross-state differences in tax rates. Hines’ approach allows him to control for unobserved
factors which influence the attractiveness of a state to foreign investors (and which are
common to investors from different countries). His results imply an elasticity of capital
ownership with respect to state taxes of –0.6. This suggests that multinationals are
influenced by cross-country or cross-region differences in tax rates.
With regards to agglomeration effects, Head, Ries, and Swenson (1995) examine the
location decisions of new Japanese manufacturing plants in the United States. They find
that, controlling for the local concentration of U.S. plants in the same industry (among other
factors), Japanese plants are more likely to choose a location the higher is the existing local
concentration of Japanese manufacturing plants in the same industry. This finding is similar
to Wheeler and Mody’s finding that firms are attracted to large concentrations of other
industrial firms. But, by focusing on Japanese plants and controlling for the location of
overall U.S. manufacturing activity, this approach goes further than previous studies in
controlling for the effects of unobserved, site-specific characteristics.23 This suggests that

22 See Hanson (2000a, 2000b) for surveys of the literature and discussion of these issues.
23 Aitken, Hanson, and Harrison (1997) take a similar approach.

22
multinationals (at least the Japanese firms in Head, Ries, and Swenson’s data) are attracted
to locations with other firms in their own or related lines of activity.24
Does FDI Generate Positive Spillovers for the Host Economy?
That multinational firms are different from purely domestic firms is abundantly
clear. Across countries and time several empirical regularities are apparent. Relative to
their domestic counterparts, multinationals are larger, pay their workers higher wages, have
higher factor productivity, are more intensive in capital, skilled labor, and intellectual
property, are more profitable, and are more likely to export (Haddad and Harrison, 1993;
Aitken, Hanson, and Harrison, 1997; Aitken, Harrison, and Lipsey, 1997; Aitken and
Harrison, 1999; Blomstrom and Sjoholm, 1999).25 That multinationals possess these
attributes is not surprising, given that to become a viable multinational a firm must have
outperformed domestic and foreign rivals in some dimension. The relative technological
superiority of multinationals also makes it possible that they would be a direct or indirect
source of technological advancement for domestic firms in host countries, especially where
these countries are relatively far from the technological frontier.
Theory identifies several channels through which multinationals generate
externalities that raise the productivity of host-country factors of production. It is entirely
possible, however, that the net effect of such linkages on host-country welfare is negative,

24 Many Japanese firms which invest in the United States are part of industrial groups, which may influence
their responsive to host-region economic conditions (Belderbos and Sleuwaegen, 1996).
25 For a detailed discussion of manufacturing plants in developing countries see Tybout (2000). In the United
States, Figlio and Blonigen (1999) find that counties in South Carolina with more employment in multinational
firms experience higher wage growth, lower growth in public expenditure, and shifts in the composition of
public spending away from education towards transportation and public safety.

23
once we take into account the impact of FDI on the profitability of domestic firms. Whether
spillovers from multinationals raise host-country welfare is an empirical question.
Early literature is optimistic about the impact of multinationals on host-country
productivity. 26 Caves (1974) finds a positive correlation between industry average value-
added per worker and the share of industry employment in foreign firms for Australian
manufacturing in 1966.27 More recent work confirms this basic finding in a wide array of
environments. A partial list of studies which find a positive correlation between average
industry productivity and the presence of foreign firms in the industry include Globerman
(1979) for Canada in 1972; Blomstrom and Persson (1983), Blomstrom (1986), and Kokko
(1994) for Mexico in the 1970s; and Blomstrom and Sjoholm (1999) for Indonesia in 1991.
In related work, Borensztein et al. (1998) find a weak positive correlation between FDI
inflows and per capita GDP growth for a panel of countries in the 1970s and 1980s (though
when they interact FDI and the level of schooling FDI has a negative “direct effect” on
growth and a positive “indirect effect” through schooling).
Taking a slightly different approach, Mansfield and Romeo (1980) use a sample of
U.S.-based multinationals to examine which sorts of technology these firms transfer abroad
and whether there is leakage of these technologies to non-U.S. firms in host countries. They
report that in 20 of 26 cases transferred technologies became known to foreign rivals within
six years. In nine of the cases, access to U.S. technology appeared to accelerate foreign
firms’ introduction of competing products or processes by two years or more.28

26 For a survey of the literature see Blomstrom and Kokko (1998).
27 Caves also finds that for Canadian manufacturing (1965-1967) industry average profitability of domestic
plants is positively correlated with industry average profitability of foreign plants and negatively correlated with
the average share of foreign plants in industry sales (which he interprets as a pro-competitive effect of FDI).
28 See Veugelers and Cassiman (1999) for evidence that the propensity to transfer technology abroad is not
linked to multinationality per se but rather to access to international markets.

24
What does it mean for industry productivity or industry technology adoption to be
positively correlated with the presence of multinational firms in an industry? A common
interpretation of this finding is that FDI creates positive productivity spillovers for domestic
firms in host countries. This interpretation, however, is subject to the same concerns about
omitted variables and endogeneity bias that we encountered in empirical work on the impact
of taxes and agglomeration effects on firm location decisions. Most empirical studies in the
literature use cross-section data on average industry characteristics. A positive simple
correlation between industry productivity and the presence of multinationals is, in principle,
just as likely to mean that multinationals are attracted to high-productivity industries as it is
to mean that multinationals raise host-country productivity. Though most empirical studies
introduce additional controls in estimating the correlation between industry productivity and
multinational presence, the included variables surely do not exhaust the set of factors which
are likely to influence industry productivity and multinationality.29
Recent work attempts to address these identification problems through using micro-
level, time-series data on individual manufacturing plants. By looking at how the
productivity of domestic plants changes over time in response to the presence of
multinationals, it is possible to control for the presence of unobserved factors which
influence both the productivity of domestic plants and the behavior of multinationals.
Haddad and Harrison (1993), using data on Moroccan manufacturing plants for the period
1985-1989, find a weak negative correlation between plant total factor productivity growth

29 Other work cites less direct evidence of spillovers related with FDI. In many contexts, multinational firms
develop backward linkages with local industry (e.g., Behrman and Wallender, 1976; Lall, 1980) or provide
training to local workers (e.g., Chen, 1983; Gehrschenberg, 1987). These activities may create channels
through which spillovers could flow from multinationals to host economies, but their occurrence in no way
establishes that such spillovers actually exist. Anecdotal evidence of FDI “demonstration effects”, in which
local firms learn about modern technology or management techniques by watching multinationals (see
Blomstrom and Kokko, 1999), are plausible but pure conjecture in the absence of formal statistical analysis.

26
interviews, Rhee (1990) and Rhee and Belot (1990) find that the arrival of multinational
firms contributed to the eventual export success of one or more labor-intensive industries in
11 developing economies. Using data on Mexican manufacturing plants in the 1980s,
Aitken, Hanson, and Harrison (1997) find that Mexican manufacturing firms are more likely
to export the larger is the concentration of multinational firms in their region and industry.
This correlation is robust to controlling for industry agglomeration in the region and for the
endogeneity of multinational location. In interpreting these results, it is important to
recognize that these studies examine economies in the aftermath of liberalization episodes,
in which barriers to trade and FDI fell considerably. Hence, the information spillovers these
studies detect may be confined to post-reform transition periods and so short-lived.32
IV. Evaluation FDI in Practice
To summarize key results of the literature on FDI, we have seen that:
(i) the only justification for favoring FDI over both foreign portfolio investment and
domestic investment is the existence of market failure that is specific to
multinational production;
(ii) G-24 and other countries offer myriad tax concessions to FDI, which violate the
residence principle (by taxing nonresident income) and subject FDI and foreign
portfolio investment to unequal tax treatment;
(iii) in theory, FDI raises national welfare by bringing foreign technology and other
foreign resources into an economy, which raises the productivity of domestic
factors, but in the absence of externalities there is no justification for taxes or
subsidies which are specific to FDI;

31 It is important to note that these results are for the direct impact of FDI on domestic enterprises in the same
lines of activity. It is possible that FDI raises the productivity of domestic agents through indirect, general-
equilibrium effects, such as backward-forward linkages or productivity spillovers common to all industries.
32 There are other potential spillovers from multinationals, which in the interest of space we do not discuss. See
Blomstrom and Kokko (1998) for a discussion of empirical research on the impact of multinationals on worker
training, industry linkages, industry competition, and source-country economies.

27
(iv) in theory, externalities associated with FDI may raise or lower national welfare,
depending on whether positive productivity spillovers from multinationals more than
offset the loss in profits due to crowding domestic firms out of the market;
(v) empirical research suggests that FDI is sensitive to both host-country tax polices and
economic conditions, including the education level of the labor force, overall market
size, and the size of the local industrial base;
(vi) empirical research shows mixed support for the idea that FDI generates positive
spillovers for domestic industry – while multinationals tend to be high-productivity
firms which pay relatively high wages, on average (in at least some countries) their
presence appears to depress the productivity of domestic plants (perhaps by driving
them into less profitable market segments).
In this section, we apply insights from the literature to examine several cases of FDI
promotion policies. First, we build a simple theoretical model of FDI. In the model, FDI
raises the productivity of domestic factors and possibly domestic firms but also increases
competition with these firms. The effect of FDI on national welfare depends on the relative
magnitude of increased domestic factor income versus the reduced profitability of domestic
industry. To show these effects in as transparent a manner as possible, we identify some but
not all general equilibrium effects of FDI. In particular, we assume that subsidies to FDI are
financed by lump-sum taxes and we ignore other forms of foreign investment. Given tax
policies in many countries may be inefficient (see (i) and (ii) above), we are in a sense
giving FDI promotion policies the maximum benefit of the doubt by ignoring some of their
possibly more adverse distortionary consequences.
Second, we describe three cases of FDI promotion, projects by GM and Ford to
build auto production plants in Brazil and a project by Intel to build a semiconductor factory
in Costa Rica. After presenting the relevant details of the cases, we examine whether the
theory developed in the first part of this section and estimates of key behavioral parameters
culled from empirical literature would suggest that these policies were justified on welfare

31
welfare it must be the case that FDI generates positive spillovers and that FDI would not
occur in the absence of a subsidy. In most such cases, a per-unit subsidy will be the optimal
policy, not a lump-sum subsidy (since the former equates the private and social return to
FDI while the later does not). In light of these issues, we consider the effect of a per unit
production subsidy by the host economy on domestic output of foreign firm 1. This allows
us to examine whether a small increase in the subsidy raises or lowers national welfare, and
so determine whether the laissez-faire level of FDI is too high or too low from the
perspective of the host economy. A per unit production subsidy roughly approximates many
types of actual FDI promotion policies.33
A related issue is how we treat production outside the host economy by foreign firm
1. Since the firm is producing for the world market, it would only maintain production in
multiple countries if its unit factor costs were equalized in these countries. Glass and Saggi
(1999) consider the case where FDI equalizes wages between host and source countries for
multinational firms. We consider this outcome to be unrealistic, but our model could be
easily extended to treat this case. For simplicity, we assume that the foreign firm moves its
entire production of good 1 to the host economy. This would occur in the event that FDI by
foreign firm 1 did not equalize factor prices between the host economy and the foreign
firm’s home country. Foreign firm 1 may of course have operations devoted to other
products and industries at home or in other countries. We ignore these, as they do not
impact the host economy’s welfare.

33 Though subsidies for FDI are typically negotiated before MNEs begin production, making them appear
lump-sum in nature, the facts that subsidies are often proportional to plant capacity and delivered in portions as
a project moves towards completion make them similar in effect to per-unit production subsidies (e.g., imagine
capacity subsidies in a world in which firms choose capacities and then engage in Bertrand price competition).

32
Given the simple setup of the model, we need focus on just two sets of equilibrium
conditions, those for factor-market clearing and those for profit maximization of firms
producing in the host economy. Factor-market equilibrium requires that supply equals
demand in the market for unskilled labor,
∑ ++=
i
11ii YAyaxl (1)
and in the market for skilled labor,
∑ +=
i
11i YDyk (2)
By assumption each domestic firm chooses output to maximize profit, taking as given the
output of its rival foreign firm. For domestic firm i, profit maximization implies the
following first-order condition,
0zar ii1 =−− (3)
Second-order conditions are standard. There is a similar first-order condition for the foreign
firm with whom firm i competes. Except possibly for industry 1, these foreign firms are
located abroad. If foreign firm 1 chooses to manufacture in the host economy, it’s output
choice is implicitly defined by the first-order condition,
0szDAR 11i1 =+−− (4)
which reflects the fact that foreign firm 1 may be given a per unit subsidy for producing in
the host economy. Since pairs of domestic and foreign firms compete in a single world
market, we can summarize their profit-maximizing output choices in terms of Cournot Best-
Response Functions, yi=bi(Yi) and Yi=Bi(yi), which are subject to standard conditions.34

34 These are that bi’ ≤ 0, Bi’≤ 0, and bi’>1/Bi’. By the second-order conditions to profit maximization, Cournot
stability conditions require that ri11 + ri12Bi’≤ 0 and Ri11 + Ri12Bi’≤ 0.

33
The arrival of foreign firm 1 in the host economy has three effects. First, an increase
in production by foreign firm 1 raises the demand for unskilled and skilled labor in the host
economy. Since manufacturing is relatively intensive in the use of skilled labor, the relative
demand for and the relative wage of skilled labor rises. For domestic firms in the host
economy, the rise in z increases their marginal costs. All else equal, higher marginal costs
mean lower output and lower profits. This effect applies to all firms in manufacturing and
not just the domestic firm which competes directly with foreign firm 1. Second, an increase
in production by foreign firm 1 generates a productivity spillover for domestic firms. If this
spillover is positive, it will, all else equal, cause domestic firms to raise their output and earn
higher profits. The first and second effects work in opposite directions, and so have an
ambiguous impact on domestic firm output and profits. The third effect is isolated to
domestic firm 1: as foreign firm 1 raises its output, the price for domestic firm 1’s output
falls, causing it to reduce output and thereby earn lower profits.
To consider these effects in more detail, we examine the impact of a change in the
production subsidy to foreign firm 1 on host-economy welfare. Since we have assumed that
manufacturing firms produce for the world market, we can examine the welfare effects of a
subsidy to foreign firm 1 without taking consumer surplus in the host economy into account.
As long as the final consumers of the output that foreign firm 1 produces are located abroad,
this simplification is reasonable.35 For our purposes, the relevant components of host
economy welfare are incomes to unskilled and skilled labor, profits to domestic firms, and
the subsidy to foreign firm 1,36

35 Industry 1 may, for instance, produce an input that is consumed by domestic firms and then exported in the
form of a final good (e.g., hard disk drives which are assembled into personal computers).
36 We assume that foreign firm 1 repatriates any profits it earns in the host economy to its home country.

34
∑ −+−++=
i
1ii
i sY]y)za(r[zklW (5)
Using the factor-market clearing condition for skilled labor in equation (2), we can rewrite
host-economy welfare as,
∑ −−++=
i
1ii
i
11 sY]yar[YzDlW (6)
We turn now to the thought experiment that is the motivation for the model: what is the
impact of an increase in the production subsidy to foreign firm 1 on host-economy welfare?
The base case from which we begin is laissez-faire (zero subsidy). Determining the welfare
consequences of a subsidy to the multinational firm will then also determine whether the
social return to FDI exceeds the private return.
Totally differentiating equation (6), we obtain,
∑ ∑ −λ+−+++=
i
11i
i
3iii
i
2
i
11111 dsYdY'rdy]a'Brr[dYzDYdzDdW (7)
The first two terms in equation (7) represent the change in factor income, which are positive
as long as the subsidy causes foreign firm 1 to increase its output. The third and fourth
terms represent the change in profits for domestic firms, which depends on the signs of the
dyi terms and whether productivity spillovers are positive or negative. If the rise in marginal
costs is the dominant effect, domestic output and profits fall, while if the productivity-
spillover effect dominates (and spillovers are positive), domestic output and profits rise.
The fifth term is the direct cost of the subsidy to multinational production.
To help interpret equation (7), we simplify the expression. First, define φi ≡ ri2Bi' ≥
0, which is the strategic effect of own changes in domestic output on domestic profits.
Given outputs are strategic substitutes (by assumption), as a domestic firm raises its output,

36
From equation (8), for the foreign firm to raise output the net change in output for domestic
firms must be negative. We anticipate, then, that the only domestic firms whose output will
rise will be those receiving a substantial positive productivity spillover.
The third term in equation (9) captures the impact of the productivity spillover on
domestic profits. This term is larger, the larger is the increase in output for foreign firm 1.
But larger increases in foreign firm output put also upward pressure on demand for skilled
labor, making it more likely that the second term in (9) will be negative.
Under what conditions will a subsidy to domestic production by a multinational firm
be likely to raise national welfare? We identify four key conditions:
(i) the factors used most intensively in production by the multinational firm are
in elastic supply,
(ii) the domestic firms that compete for resources with the multinational firm
earn low to zero economic profits,
(iii) multinational production generates large positive productivity spillovers for
domestic firms in competing and non-competing industries, and
(iv) the gain in consumer surplus from increased competition in the domestic
market is small.
Condition (i) guarantees that the impact of the subsidy on factor costs for domestic firms
will be small; condition (ii) guarantees that the welfare consequences from shifting
production away from domestic firms and towards foreign firms will be small; and
condition (iii) is necessary for a subsidy to be worthwhile under any circumstances.
Condition (iv) goes beyond the theoretical analysis we have presented, but is an obvious
point. We do not emphasize changes in consumer surplus, since if FDI does happen to raise
domestic market competition then the optimal policy is not a subsidy to multinational firms
but a generalized production subsidy to offset the distortionary consequences of imperfect

37
competition.37 These conditions form the basis for evaluating actual cases of FDI promotion
policies, to which we turn in the next section.
The Promotion of FDI in Practice: Three Cases
General Motors in Brazil.38 In the last decade, Brazil captured the attention of
global auto manufacturers. With Mercosur providing tariff barriers against competition
from outside the southern cone, Brazil having the world’s eighth largest auto market, and
strong projected growth in regional auto demand (only one in nine Brazilians owns a car),
the major auto producers either established or expanded production capacity in Brazil and
Argentina. By the late 1990s, there were a dozen auto manufacturers in Brazil alone. In
1998, VW had 28% of the Brazilian market, Fiat had 24%, GM had 22%, and Ford had
13%, with the rest of the market divided among smaller players.
In the early 1990s, GM began to move aggressively into Brazil, focusing initially on
higher-end products and touting its confidence in the country. By the late 1990s, GM had
an annual production capacity of 500,000 units, up from 170,000 in 1992. VW, Fiat, and
Ford, in contrast, were introducing new models relatively slowly, reflecting the cautious
approach of many durable-goods manufacturers in the wake of Brazil’s continuing struggle
against high inflation and slow growth. GM decided to make Brazil a showcase for its new

37 That is, the market failure in this case is due to imperfect competition, not multinational production.
38 This subsection is based on material from the following sources: Interview with Mustafa Mohaterem, Chief
Economist for General Motors, May 31, 2000. Peter Frisch, “Ford and GM Clash with Brazilian State,” Wall
Street Journal, April 9, 1999, p. A11. Kathleen Kerwin and Joann Muller, “A Close Look at President Rick
Wagoner and His Strategy for GM,” Business Week, February 1, 1999, p. 114. Peter Fritsch and Gregory
White, “Latin Formula: Even Rivals Concede GM Has Deftly Steered Road to Success in Brazil,” Wall Street
Journal, February 25, 1999, p. A1. Haig Simonian, “GM Plans to Develop Car in Brazil,” Financial Times,
June 3, 1997, p. 5. “GM Picks Suppliers for Blue Macaw,” Automotive News Europe, June 23, 1997, p. 8.
Charlotte Craig, “Ford, General Motors May Lose Their Brazilian Incentives,” Detroit Free Press, March 23,
1999, p. 1. John Kolodziejski Ian Katz, Kathleen Kerwin, and Keith Naughton, “A High Stakes Game of
Chicken: Its Ford and GM vs. a Brazilian Governor,” Business Week, May 10, 1999, p. 66.

38
global production strategy, based on simple and flexible manufacturing plants, global
sourcing of auto parts, rapid introduction of new models, and a lean dealer network. This
strategy had shown success in Europe and the idea was to develop it further in Brazil, with
the goal of applying it Asia, Eastern Europe, and perhaps ultimately in the United States.
GM’s long-run aim is to have half of its production capacity outside the United
States, compared with only one-fifth in 1990. (As GM has stepped up its Brazilian
operations, it has also expanded auto production in other emerging economies, including
Argentina, China, Poland, and Thailand.) By the late 1990s, other auto makers in Brazil
were following an approach to similar to GM’s. VW, Fiat, and Ford, among others, all
decided to build new plants or add capacity to their existing plants in the country.
In 1997, GM made the Blue Macaw project the centerpiece of its Brazil strategy.
The project revolved around a new auto assembly plant, which would be GM’s third in the
country, with an annual capacity of 150,000 units. The plant would produce a stripped-
down version of the Opel Corsa, a subcompact car, with an under-$10,000 price tag. Much
of the production of the car would be outsourced to suppliers, who would deliver entire
subassemblies of components to GM for final assembly. GM’s plan was to develop this
concept of “modular assembly” in Brazil and then apply it to other production facilities in
Europe and North America. GM chose the state of Rio Grande do Sul as the site for the
$600 million plant, with the idea of completing the facility by the end of 1999. However,
Brazil’s currency crisis in early 1999 and the ensuing recession caused GM’s Brazil sales to
fall by 27% and delayed completion of the project until early 2000.
GM’s plant, which recently opened, employs 1,300 workers and locally-based
suppliers employ another 1,300 workers. The plant houses 20 suppliers, the most important

39
of which are U.S., French, and Japanese companies. GM outsources all components except
powertrains, body welding, body panels, paint, and final assembly.
The advantages of locating in Rio Grande do Sul is that the state is close to the
southern cone’s major markets in southern Brazil, the Buenos Aires region of Argentina,
and Uruguay. The state has a more educated work force compared to the rest of Brazil, but
lower wages that the highly-industrialized, nearby region of Sao Paulo. (Total
compensation for GM workers in Rio Grande do Sul is expected to be $9 per hour,
compared to $13 for auto workers in Sao Paulo.)
In return for agreeing to build a plant in a lightly-industrialized area, GM received a
package of subsidies from the state government of Rio Grande do Sul. These included
promises for direct subsidies to GM to offset costs of building roads, ports, and other
infrastructure related to the plant; temporary exemptions from value-added taxes; and a
waiver of import duties on machinery used in the construction of the plant. While neither
GM nor the state government is willing to give precise figures, the subsidies were reported
to amount to $250 million and the tax breaks appeared to have the potential to equal $1.5
billion over a 15-year period. GM executives maintain that in the absence of these
subsidies, the firm would have located the plant in a more developed part of Brazil.
In May 1999, the government of Rio Grande do Sul, under a newly-elected, more
populist governor, threatened to renege on the promised subsidies to GM. Sharply higher
interest rates in the aftermath of Brazil’s currency crisis had raised borrowing and debt-
service costs in the country and forced state and federal governments to raise taxes and cut
back on spending. Rio Grande do Sul’s governor claimed that his state could no longer
afford the subsidies. GM felt it was too close to completion of the project to pull out.

43
what Intel needs in its global production sites is speed in ramping up production and access
to a dependable, well-educated labor pool.
In Costa Rica’s EPZs, firms enjoy a series of tax breaks as long as they engage in
export production.41 All firms are exempt from (a) import duties on raw materials,
components, and capital goods; (b) export, sales, excise, and municipal taxes; and (c) taxes
on corporate income during the first eight years after an investment is made, with a 50%
exemption applying for the following four years. Foreign investors face no restrictions on
repatriating profits or foreign currency management. Certain export processing zones offer
additional incentives, including longer exemptions from corporate income taxes and
subsidies for employment or employee training. As of 1997, there were 190 companies in
eight industrial parks operating under Costa Rica’s EPZ system.
With its rapid expansion of production capacity, Intel is constantly looking for new
production facilities. In early 1996, Intel decided to build a 400,000 square feet plant, which
would employ 2,000 workers to assemble and test the latest Pentium microprocessors. At
the time, Intel anticipated that by 1999 the plant would process one-quarter to one-third of
the chips that Intel manufactured. For Intel, the required features of a production site (given
moderately large fixed costs and the need to begin production quickly) include political and
economic stability, a sufficient supply of professional and technical operators (preferably in
a non-union environment), ease of importing components and exporting final products, and
minimal lags in obtaining necessary permits and licenses. Costa Rica showered Intel with
attention and information, but did not employ extraordinary measures to attract the firm.
Costa Rica did offer a few concessions to Intel, all of which were extended to other firms as

41 As in many countries, firms in Costa Rica’s EPZs are allowed to sell a fraction on their output on the
domestic market (though few actually do).

44
well. These included waiving a 1% tax on assets (extended to all firms in EPZs), increasing
the number of foreign air carriers allowed to fly into Costa Rica, lowering energy prices for
large buyers of electricity, and expanding training in electronics and English in several of
Costa Rica’s technical high schools. Intel executives stated that they chose Costa Rica
based on the country’s long history of stability, open trade and investment regime, relatively
high-quality primary and secondary educational systems, and recent success in attracting
other multinational firms in electronics.
Evaluation of FDI Promotion Cases
In this subsection we use the insights from the theoretical model developed earlier to
examine whether on welfare grounds (a) Brazil was justified in offering subsidies to GM
and Ford, and (b) Costa Rica would have been justified in subsidizing Intel. We focus on
the issue of subsidies and not tax breaks, as tax breaks on income to foreign capital tend to
move a country towards to the residence principle and so more efficient taxation (Costa
Rica, in particular, appears to have been close to having zero taxes on direct foreign capital).
Since state governments are the entities offering subsidies in Brazil, we consider whether
subsidies to FDI would be likely to raise state welfare. A full treatment of this issue requires
a much more complete analysis than we offer here. Our goals are simply to identify the key
questions behind whether subsidizing FDI was likely to raise welfare and, based on casual
evidence, to identify some possible answers to these questions.
The model we developed earlier appears to be applicable to both the Brazilian and
Costa Rican contexts. In each foreign firms are technologically advanced relative to
domestic firms, operate in markets which appear to be imperfectly competitive at a regional,

45
national, or global level, and produce primarily for export (to the world market in Intel’s
case and to the broader Mercosur market in GM’s case and Ford’s case).
(i) Is FDI likely to raise production costs for domestic firms (i.e., is the
multinational firm relatively intensive in the use of inelastically-supplied factors)?
To the extent that the scale of new production by a multinational firm is large, the
prices of inelastically-supplied factors used in production may rise. As we showed earlier, a
subsidy to FDI in this context, all else equal, would lower national welfare. Higher incomes
(resulting from the subsidy) to factors employed in domestic firms would simply offset
higher production costs to domestic firms, and higher incomes to factors employed in the
multinational would be insufficient to cover the direct cost of the subsidy.
Production of both autos and semiconductors appear to be relatively intensive in the
use of skilled labor. Intel prefers workers with at least a high-school education, a level of
schooling well above the average in Costa Rica, and GM’s (and presumably Ford’s, also)
hourly wage of $9 is well above the Brazilian average. We would expect skilled labor to be
relatively scarce in both Costa Rica, Rio Grande do Sul, and Bahia, such that the arrival of a
multinational firm would put upward pressure on the local relative price of the factor.
(ii) Are there domestic firms which compete directly with the multinational firm?
To the extent that a foreign firm faces no domestic rivals, then any increase in its
production in the host economy is unlikely to directly lower the profitability of domestic
firms. In this case, a subsidy to FDI has no direct consequences for competition in the
industry which are relevant for host economy welfare. (The absence of domestic rivals,
however, also means the potential absence of firms which could benefit from productivity
spillovers.) In Brazil, foreign firms overwhelmingly dominate production of autos and also

46
appear to dominate production of auto parts. In Costa Rica, Intel faces no domestic
competitors and would appear to have few domestic suppliers.
(iii) Are there domestic firms which are likely to benefit from productivity spillovers
or forward or backward linkage effects from FDI?
If no domestic agents benefit from productivity spillovers from FDI, then a subsidy
to FDI will lower national welfare (assuming the absence of other distortions in the
economy). This issue is obviously difficult to evaluate in full. However, that GM, Ford,
and Intel face no domestic rivals of any significance and rely primarily on foreign firms for
parts and components suggests that there are few candidate domestic firms to benefit from
productivity spillovers in either Brazil or Costa Rica (to the extent that such spillovers even
exist). Of course, domestic firms in disparate industries may learn basic management skills
simply from observing how multinationals like GM, Ford, and Intel structure their supply
chains, introduce new products, treat their workers, etc. But such learning effects would
have to be substantial to justify the subsidies Brazil gave to GM and Ford. Empirical
literature to date shows little evidence of such effects.
(iv.) Is the multinational likely to repatriate most profits to its home country?
To the extent that a foreign firm does not share rents with host-country suppliers or
employees, all returns to FDI accrue to shareholders of the multinational, who are likely to
be located abroad. This issue is also difficult to evaluate, but given the global reach of firms
likely GM, Ford, and Intel (and their clear preference for non-union environments) there is
no obvious reason to expect that these firms are likely to share profits with their host-
country workers. Again, the magnitude of such profit sharing would have to be large to
justify the subsidies Brazil gave to Ford and GM.

47
A preliminary evaluation of the three cases of FDI promotion suggests that the case
for subsidies to GM and Ford in Brazil was weak and that subsidies to Intel in Costa Rica
would have been difficult to justify. In both Brazil and Costa Rica, subsidy-induced
increases in production would appear to likely to put upward pressure on the relative wage
of skilled workers, reducing the profitability of domestic firms. While the foreign firms in
question would appear to face few domestic rivals, limiting the direct consequences of FDI
for domestic profitability, there would also appear to be a few firms in the same lines of
business which would benefit from spillovers.
A particularly troubling feature of FDI subsidies in Brazil is that they appeared to
result from competition between Brazilian states for the right to host foreign plants. Both
GM and Ford appeared to have already concluded that they needed to increase production
capacity in Brazil. The subsidies, then, may have had little effect on whether the auto
companies invested in the country and only impacted where they located their facilities
within Brazil. The end result of this inter-state fiscal competition for FDI may be limited to
extra burdens for Brazilian taxpayers and excess capacity in the regional auto industry.
V. Concluding Remarks
In this paper we have examined whether it makes sense for countries to promote
FDI. Our work falls short of an in depth analysis. The goal has been, instead, to identify the
key issues which determine whether FDI promotion policies are justified and then to
examine relevant academic literature to see whether there is evidence that such conditions

48
appear to hold in practice. Our focus has been, in particular, on whether spillovers related to
multinational production justify FDI promotion.
For small open economies, efficient taxation may or may not require lower taxes on
capital income to foreign residents than on capital income to domestic residents. Absent
market failure, there is no reason to favor FDI over foreign portfolio investment. In
practice, countries appear to tax income from foreign capital at positive rates (if rates that
are lower than those for domestic capital) and to subject different forms of foreign
investment to unequal tax treatment.
FDI appears to be quite sensitive to host country characteristics. Higher taxes deter
foreign investment and a more educated work force and larger consumer and industrial
markets attract FDI. There is also some evidence that multinationals tend to agglomerate in
manner consistent with location-specific externalities.
There is weak evidence that FDI generates positive spillovers for host economies. A
oft-mentioned possibility is that FDI raises the productivity of domestic agents. While
multinationals are attracted to high-productivity countries, and to high-productivity
industries within these countries, there is little evidence at the plant level that FDI raises the
productivity of domestic enterprises. Indeed, it appears that plants in industries with a larger
multinational presence enjoy lower rates of productivity growth. Empirical research thus
provides little support for the idea that promoting FDI is warranted on welfare grounds.
Using a simple theoretical model, we derived conditions under which subsidies to
FDI would be more likely to raise host-country welfare. Subsidies to FDI are likely to be
warranted where multinationals are intensive in the use of elastically-supplied factors, the
arrival of multinationals to a market does not lower the market share of domestic firms, and

50
may benefit specific constituencies, from whom politicians derive support, or may fit into
political strategies of empire-building. Whatever the explanation, countries are likely to be
better served by being cautious about promoting FDI, until we see strong empirical evidence
that the social rate of return on FDI exceeds the private rate of return.

53






∆
−
γ∆
γφ
+ργ−ργφ+β=


 ∑ ∑ ∑
∑
∑
≠
≠
≠i 1i 1
1
i
i1
1i
ii
1
1i
iiiiii
Y)D(signds
dWsign (A12)
To interpret equation (A12), consider each term inside the brackets on the right. The first
term is the direct effect of spillovers on domestic profits. Positive spillovers increase
revenues and profits directly, as captured by βi (see main text and preceding appendix for
definitions). The second term is the strategic effect of domestic spillovers on profits. Even
if higher factor costs make a domestic firm lower its output, positive spillovers make the
firm more aggressive relative to its foreign rival. The firm lowers its output by less than it
would have in the absence of spillovers. Spillovers thus moderate the loss in market share a
firm experiences from an increase in factor costs (but only for firms that do not compete
directly against foreign firm 1). (Of course, the impact of the spillover may be so large that
the firm raises its output, even in the face of rising factor costs. Some domestic firms,
however, must lower their output (see equation (8).) This effect is the product of three
terms: φi, the impact of the foreign rival firm’s output response to a change in the domestic
firm’s output on the domestic firm’s profits; γi, the marginal responsiveness of domestic
firm profits to changes in domestic firm output; and ρi, the responsiveness of domestic
firm’s output to productivity spillovers from FDI.
The third term on the right in (A12) captures the loss in domestic profits due to rising factor
costs. This term depends on the interaction between two effects: the increase in demand for
labor at unchanged factor prices, which is the sum of labor demand from foreign firm 1 (D1)
and the extra labor demand from domestic firms induced by productivity spillovers (γiρi,
which is the output response of a domestic firm to the spillover); and the strategic effect of
higher factor costs on domestic firm profits (the ratio term), which results in domestic firms
lowering output and ceding market share to the foreign rival firm. The fourth term on the
right captures the cost of the subsidy, which intuitively is proportional to the level of
production by the multinational firm.

54
References
Aitken, B., G. Hanson, and A. Harrison. 1997. Spillovers, Foreign Investment, and Export
Behavior. Journal of International Economics 43: 103-132.
Aitken, B. and A. Harrison. 1999. Do Domestic Firms Benefit from Foreign Investment?
Evidence from Venezuela. American Economic Review 89: 605-618.
Aitken, B., A. Harrison, and R. Lipsey. 1996. Wages and Foreign Ownership: A
Comparative Study of Mexico, Venezuela, and the United States. Journal of International
Economics 40: 345-371.
Avi-Yonah, R.S. 1999. Globalization, Tax Competition, and the Fiscal Crisis of the
Welfare State. Mimeo, Harvard University.
Belderbos, R. and L. Sleuwaegen. 1996. Japanese Firms and the Decision to Invest
Abroad. Review of Economics and Statistics 78: 214-220.
Behrman, J. and H. Wallender. 1976. Transfer of Manufacturing Technology within
Multinational Enterprises. Cambridge, MA: Ballinger.
Bhagwati, J., R. Brecher, E. Dinopoulos, and T.N. Srinivasan. 1987. Quid Pro Quo Foreign
Investment and Welfare: A Political-Economy Approach. Journal of Development
Economics 27: 127-138.
Black, D.A. and W.H. Hoyt. 1989. Bidding for Firms. American-Economic-Review 79:
1249-1256.
Blomstrom, M., 1986. Foreign Investment and Productive Efficiency: The Case of Mexico.
Journal of Industrial Economics 35: 97-110.
Blomstrom, M. and A. Kokko, 1998. Multinational Corporations and Spillovers. Journal of
Economic Surveys 12: 247-277.
Blomstrom, M. and H. Persson. 1983. Foreign Investment and Spillover Efficiency in an
Underdeveloped Economy: Evidence from Mexican Manufacturing Industry. World
Development 11: 493-501.
Blomstrom, M. and F. Sjoholm. 1999. Technology Transfer and Spillovers: Does Local
Participation with Multinationals Matter? European Economic Review 43: 915-923.
Blomstrom, M. and E.W. Wolff. 1989. Multinational Firms and Productivity Convergence
in Mexico. Mimeo, New York University.
Blonigen, B. and D. Figlio. 1998. Voting for Protection: Does Direct Foreign Investment
Influence Legislator Behavior? American Economic Review 88: 1002-1014.

55
Blonigen, B. and Y. Ohno. 1998. Endogenous Protection, Foreign Direct Investment and
Protection-Building Trade. Journal of International Economics 205-228.
Bond, E. and L. Samuelson. 1986. Tax Holidays as Signals. American Economic Review
76: 820-826.
Borensztein, E., J. De Gregorio, and J-W. Lee. 1998. How Does Foreign Direct Investment
Affect Economic Growth? Journal of International Economics 45: 115-135.
Brainard, L. 1997. An Empirical Assessment of the Proximity-Concentration Tradeoff
between Multinational Sales and Trade. American Economic Review 87: 520-544.
Buckley, P. and M. Casson. 1981. The Optimal Timing of Foreign Direct Investment.
Economic Journal 91: 75-87.
Caves, R.E., 1974. Multinational Firms, Competition, and Productivity in Host-Country
Markets. Economica 41: 176-193.
Caves, R.E. 1995. Multinational Enterprise and Economic Analysis, 2nd Edition.
Cambridge: Cambridge University Press.
Caves, R.E. 1999. Spillovers from Multinationals in Developing Countries: The
Mechanisms at Work. Mimeo, Harvard University.
Chen, E.K.Y. 1983. Multinational Corporations, Technology, and Employment. London:
Macmillan.
Chung, W. 1999. Up and Downstream Foreign Direct Investment and U.S. Manufacturing
Productivity. Mimeo, NYU.
Davies, H., 1977, Technological Transfer Through Commercial Transactions. Journal of
Industrial Economics 26: 165-171.
Dixit, A. and G. Grossman. 1986. Targeted Export Promotion with Several Oligopolistic
Industries. Journal of International Economics 21: 233-249.
Dunning, J.H. 1981. International Production and the Multinational Enterprise. London:
George, Allen, and Unwin.
Dunning, J.H. 1993. The Globalization of Business. London: Rutledge.
Feldstein, M., J. Hines, and R.G. Hubbard. 1995. The Effects of Taxation on Multinational
Corporations. University of Chicago Press.

56
Fernandez-Arias, E. and R. Hausmann. 2000. If FDI a Safer Form of Financing? Mimeo,
Inter-American Development Bank.
Figlio, D.N. and B.A. Blonigen. 1999. The Effects of Direct Foreign Investment on Local
Communities. Mimeo, University of Florida.
Gao, T. 1999. Economic Geography and the Department of Vertical Multinational
Production. Journal of International Economics 48.
Gerschenberg, I. 1987. The Training and Spread of Managerial Know-How: A
Comparative Analysis of Multinational and Other Firms in Kenya. World Development 15:
931-939.
Glass, A, and K. Saggi. 1999a. FDI Policies under Shared Factor Markets. Journal of
International Economics 49: 309-332.
Glass, A. and K. Saggi. 1999b. Multinational Firms and Technology Transfer. Mimeo,
Ohio State University.
Globerman, S. 1979. Foreign Direct Investment and "Spillover" Efficiency Benefits in
Canadian Manufacturing Industries. Canadian Journal of Economics 12: 42-56.
Gordon, R. 1986. Taxation of Investment and Savings in a World Economy. American
Economic Review 76: 1086-1102.
Gordon, R. 1992. Can Capital Income Taxes Survive in Open Economies? Journal of
Finance 47: 1159-1180.
Gordon, R. and L. Bovenberg. 1996. Why Is Capital so Mobile Internationally? Possible
Explanations and Implications for Capital Income Taxation. American Economic Review
86: 1057-1075.
Gordon, R. and J. MacKie-Mason. 1995. Why Is There Corporate Taxation in a Small
Open Economy? The Role of Transfer Pricing and Income Shifting. In M. Feldstein, J.R.
Hines, Jr., and R.G. Hubbard, eds., The Effects of Taxation on Multinational Corporations,
Chicago: University of Chicago Press.
Gorg, J. and E. Strobl. 2000. Multinational Companies and the Development of Firm Start-
Up Size: Evidence from Quantile Regressions for Ireland. Centre for Research on
Globalization and Labour Markets Research Paper No. 11.
Haddad, M. and A. Harrison. 1993. Are There Positive Spillovers from Direct Foreign
Investment? Journal of Development Economics 42: 51-74.
Hanson, G.H. 2000a. Scale Economies and the Geographic Concentration of Industry.
Journal of Economic Geography, forthcoming

57
Hanson, G.H. 2000b. Firms, Workers, and the Geographic Concentration of Economic
Activity. In Gordon L. Clark, Meric S. Gertler, and Maryann Feldman, eds., Handbook of
Economic Geography, Oxford University Press, forthcoming.
Hausmann, R. and E. Fernandez-Arias. 2000. Foreign Direct Investment: Good
Cholesterol? Mimeo, Inter-American Development Bank.
Head, K., J. Ries, and D. Swenson. 1995. Agglomeration Benefits and Location Choice:
Evidence from Japanese Manufacturing Investment in the United States. Journal of
International Economics 38: 223-247.
Head, K., J. Ries, and D. Swenson. 1998. Attracting Foreign Manufacturing: Investment
Promotion and Agglomeration. Regional Science and Urban Economics 29: 197-218.
Helpman, E. 1984. A Simple Theory of Trade with Multinational Corporations. Journal of
Political Economy 92: 451-471.
Helpman, E. and P. Krugman. 1985. Market Structure and Foreign Trade. Cambridge,
MA: MIT Press.
Hines, J. 1996. Altered States: Taxes and the Location of Foreign Direct Investment in
America. American Economic Review 86: 1076-1094.
Hines, J. 1997. Tax Policy and the Activities of Multinational Corporations. In Alan J.
Auerbach, ed., Fiscal Policy: Lessons from Economic Research, MIT Press, pp. 401-445.
Janeba, E. 1996. Foreign Direct Investment under Oligopoly: Profit Shifting or Profit
Capturing? Journal of Public Economics 60: 423-445.
Janeba, E. 1998. Tax Competition in Imperfectly Competitive Markets. Journal of
International Economics 44: 135-154.
Kokko, A. 1994. Technology, Market Characteristics, and Spillovers. Journal of
Development Economics 43: 279-293.
Lall, S. 1980. Vertical Interfirm Linkages in LDCs: An Empirical Study. Oxford Bulletin
of Economics and Statistics 42: 203-226.
Lipsey, R. 1995. Internationalized Production in World Output. NBER Working Paper
No. 5385.
Lipsey, R. 1998. Internationalized Production in Developed and Developing Countries and
in Industry Sectors. NBER Working Paper No. 6405.

58
Lipsey, R. 1999. The Role of Foreign Direct Investment in International Capital Flows.
NBER Working Paper No. 7094.
Madani, D. 1999. A Review of the Role and Impact of Export Processing Zones. Mimeo,
The World Bank.
Mansfield, E. and A. Romeo. 1980. Technology Transfer to Overseas Subsidiaries by US
Based Firms. Quarterly Journal of Economics 95: 737-750.
Markusen, J.R. 1984. Multinationals, Multi-Plant Economies, and the Gains from Trade.
Journal of International Economics 16: 205-226.
Markusen, J. 1995. The Boundaries of Multinational Enterprises and the Theory of
International Trade. Journal of Economic Perspectives 9: 169-189.
Markusen, J. 1998. Multinational Firms, Location, and Trade. The World Economy 21:
733-756.
Markusen, J.R. and K. Maskus. 1999a. Multinational Firms: Reconciling Theory and
Evidence. NBER Working Paper No. 7163.
Markusen, J.R. and K. Maskus. 1999b. Discriminating among Alternative Theories of the
Multinational Enterprise. NBER Working Paper No. 7164.
Markusen, J. and A. Venables. 1998. Multinational Firms and the New Trade Theory.
Journal of International Economics 46: 183-203.
Markusen, J. and A. Venables. 1999a. The Theory of Endowment, Intra-Industry and
Multinational Trade. Mimeo, University of Colorado and London School of Economics.
Markusen, J. and A. Venables. 1999b. Foreign Direct Investment as a Catalyst for
Industrial Development. European Economic Review 43: 335-356.
Matouschek, N. 1999. Foreign Direct Investment and Spillovers through Backward
Linkages. CEPR Discussion Paper No. 2283.
Motta, M., A. Fofur, and T. Ronde. 1999. Foreign Direct Investment and Spillovers
through Workers' Mobility. CEPR Discussion Paper No. 2194.
Price Waterhouse. Various years. Corporate Taxes: A Worldwide Summary. New York:
Price Waterhouse.
Razin, A. and E. Sadka. 1991. International Tax Competition and Gains from Tax
Harmonization. Economics Letters 37: 69-76.
Razin, A., E. Sadka, and C.W. Yuen. 1998. A Pecking Order of Capital Flows and

50
Table 1:
Corporate Tax Rates for G-24 and Other Countries, 1990 and 1998
1990 1990 1998 1998
Min. Rate Max. Rate Min. Rate Max. Rate
GROUP OF 24
CONGO 49.0 49.0 45.0 45.0
EGYPT 34.0 42.0 34.0 42.0
GABON 40.0 45.0 40.0 45.0
GHANA 8.0 35.0
NIGERIA 20.0 40.0 20.0 30.0
INDIA 50.0 64.8 40.0 43.0
IRAN 12.0 75.0 15.0 57.0
PAKISTAN 45.0 66.0
PHILIPPINES 35.0 35.0 34.0 34.0
SRI LANKA 35.0 35.0
ARGENTINA 33.0 33.0 33.0 33.0
BRAZIL 6.3 41.5 15.0 25.0
COLOMBIA 30.0 30.0 35.0 35.0
GUATEMALA 12.0 34.0 30.0 30.0
MEXICO 36.0 36.0 17.0 34.0
PERU 35.0 35.0 30.0 30.0
TRINIDAD &
TOBAGO
15.0 40.0 15.0 35.0
VENEZUELA 15.0 50.0
OTHER COUNTRIES (REGIONAL AVERAGES)
AFRICA 36.2 46.5 25.8 35.6
EAST ASIA 24.1 39.8 20.0 42.1
EASTERN EUROPE 0.0 40.0 33.4 35.0
LATIN AMERICA 19.2 31.7 24.5 28.5
MIDDLE EAST 12.7 39.3 11.3 34.4
NORTH AMERICA 27.5 47.3 26.6 48.3
OCEANIA 41.0 41.0 34.5 34.5
SOUTH EAST ASIA 26.0 39.8 22.6 29.2
WESTERN EUROPE 37.2 43.1 33.5 37.2
This table shows minimum and maximum corporate income tax rates for selected countries. See text
for details. Data is more detailed for some countries than others. Approximations are made in certain
cases. Source: PriceWaterhouse, Corporate Taxes: A Worldwide Summary 1990.

52
Table Notes:
1. Maximum of 20 years if is a project for medium-size and economical housing whose whole units are leased vacant for dwelling.
2. In July 1989, a new investment law was issued and it offered a projects’ profits to be exempt from tax on industrial and commercial profits and from
corporate tax.
3. More restrictions apply on domestic corporations than on foreign corporations.
4. Excise duty paid on export manufactures is refundable.
5. Refund of import duty.
6. Excise and sales and service tax exemptions are granted to exporters of manufactured goods.
7. The government may grant exemptions from duties and taxes if the enterprise is classified as either basic, necessary or useful, or is to be located outside the
municipality of Guatemala.
8. See 7.
9. Exemptions from income taxes and import duties (up to 100% for a maximum ten-year period) may also be granted to industries originally located (or, if
existing, transferred) outside of the Department of Guatemala, site of the capital city.
10. See 9.
11. New industrial undertaking in Free Trade Zone or a 100% export-oriented undertaking is entirely exempt from income tax, subject to certain conditions.
12. Companies exporting goods manufactured in Pakistan can claim a rebate in tax of 50% of the tax attributable to such export sales. However, in respect of
certain specified goods, the rebate of tax is available at 25% or 75%.
13. Income tax holiday giving full exemption from corporate income tax for six years for pioneer firms and four years for non-pioneer firms from date of
commercial operation; expanding firms are given three years.

53
Table 3: Tax Concessions for Inward FDI in Comparison Countries, 1990
(I) (II) (III) (IV)
Country Corporate Income
Tax Exemption
Period Sectors Value-Added
Tax Exemption
Items Import Duty
Exemption
Items EPZ
Provision
Available to Domestic
Corporations?
(I) (II) (III) (IV)
Chile X1 P X2 E X
Ireland X All3
Japan
South
Africa
Thailand X 3-8 years P X4 E, R,
K
X X X
X = country offered concession in indicated year, * = taxed at lower rate, E = exported goods/exporting, K = capital, R = raw materials, A= agriculture, P =
priority companies/industries, M= manufacturing. Data is more detailed for some countries than others. Approximations are made in certain cases. Source:
PriceWaterhouse, Corporate Taxes: A Worldwide Summary 1990.
Table Notes:
1. Tax benefits and other incentives for companies operating in northernmost and southernmost parts of the country. Tax benefits to forestry companies also.
2. Reimbursement of taxes paid.
3. Companies that commenced operation within the Shannon Free Airport before January 1, 1981 can obtain full exemption from corporation tax until April 5,
1990 if the income is derived from the carrying on of certain activities, including exporting goods and a wide range of services.
4. Either exemption or reduction.

54
Table 4: Tax Concessions for Inward FDI in G-24 Countries, 1998
(I) (II) (III) (IV)
Country Corporate Income
Tax Exemption
Period Sectors Value-Added
Tax Exemption
Items Import Duty
Exemption
Items EPZ
Provision
Available to Domestic
Corporations?
(I) (II) (III) (IV)
Cote
d’Ivoire
5-8 years All
Egypt
X
X 5-20
years
All1
X All
X
Gabon X 0-10
years
All X2
Nigeria X 0-5 years P, E, A,
M
X All3 X E, R X X X X X
Congo X4 0-10
years
A, P, M
Argentina (tax credit bonds) X E X E, R X X
Brazil X E, P X E, P X5
Guatemala6 X E X K, E,
R
X X X X
Mexico (tax credits) X E
Peru X All7 X All8 X X
India X 5 years All9 X E X X X
Philippines X 3-6 years All10 X All11 X All X X X X X
Sri Lanka X P X P X X
X = country offered concession in indicated year, * = taxed at lower rate, E = exported goods/exporting, K = capital, R = raw materials, A= agriculture, P =
priority companies/industries, M= manufacturing. Data is more detailed for some countries than others. Approximations are made in certain cases. Source:
PriceWaterhouse, Corporate Taxes: A Worldwide Summary 1998.

55
Table Notes:
1. An investment and guarantee law effective as of May 11, 1997 offers the profits of a project formed under it to be exempt from tax on industrial and
commercial profits and from corporate tax.
2. More restrictions apply on domestic corporations than on foreign corporations.
3. 1998 budget abolishes payment of excise duties.
4. Tax priority status giving exemption (or a reduction ) from various taxes and custom duties for up to ten years can be obtained by notification of agreement.
5. Excise and sales and service tax exemptions are granted to exporters of manufactured goods.
6. In general, exemption from payment of import duties on machinery and equipment and on raw and packaging materials and from income tax is available for
those corporations classified as exporting companies. These exemptions also apply to Free Trade Zones.
7. Industrial entities established in the jungle, frontier zones and free zones are exempt from income tax.
8. Exemption from VAT is provided for industrial entities established in the jungle and frontier zones.
9. New industrial undertaking satisfying certain conditions established in a Free Trade Zone, software technology park, or electronic hardware technology park
or a 100% export-oriented undertaking is entirely exempt from income tax.
10. Income tax holiday giving full exemption from corporate income tax for six years for pioneer firms and those locating in less-developed area and four years
for non-pioneer firms from the date of commercial operation; expanding export-oriented firms are given three years.
11. Local purchases of goods and services from VAT-registered entities are either VAT exempt or zero rated.

56
Table 5: Tax Concessions for Inward FDI in Comparison Countries, 1998
(I) (II) (III) (IV)
Country Corporate Income
Tax Exemption
Period Sectors Value-Added
Tax Exemption
Items Import Duty
Exemption
Items EPZ
Provision
Available to Domestic
Corporations?
(I) (II) (III) (IV)
Chile X1 X2 E X
Ireland *3
Japan (tax credits)4 X
South
Africa
X5 X
Thailand X 3-8 years P X6 K, R,
E
X = country offered concession in indicated year, * = taxed at lower rate, E = exported goods/exporting, K = capital, R = raw materials, A= agriculture, P =
priority companies/industries, M= manufacturing. Source: PriceWaterhouse, Corporate Taxes: A Worldwide Summary 1998. Note: Data is more detailed for
some countries than others. Approximations are made in certain cases.
Table Notes.
1. Tax benefits and other incentives for companies operating in northernmost and southernmost parts of the country. Tax benefits to forestry companies also.
2. Reimbursement of taxes paid.
3. Reduced rate of corporation tax of 10% of profits from manufacturing operations arising between January 1, 1981 and December 31, 2010, regardless of
whether the goods are exported. Definition of manufacturing operation is rather lenient.
4. A corporation tax credit of 3.5% or 7% of the adjusted acquisition cost (25% to 100%) of certain designated machinery and equipment that save energy or
7% of the acquisition cost of certain designated machinery and equipment containing electronic computer systems acquired by designated small or medium-
size corporations is available. The credit is limited to 20% of the corporation tax otherwise payable.
5. Tax holiday granted at discretion to an enterprise with qualifying assets in excess of R3 million, incorporated on or after October 1, 1996, for the sole object
of carrying out a qualifying project.
6. Either exemption or reduction.