Journal of International Economics 64 (2004) 89 – 112

www.elsevier.com/locate/econbase

FDI and economic growth: the role of local

financial markets
Laura Alfaro a, Areendam Chanda b, Sebnem Kalemli-Ozcan c,*,
Selin Sayek d
a
Harvard Business School, 263 Morgan Hall, Boston, MA 02163, USA
b
Department of Economics, North Carolina State University, Raleigh, NC 27695-8110, USA
c
Department of Economics, University of Houston, Houston, TX 77204-5882, USA
d
International Monetary Fund (IMF), 700 19th Street, NW, Washington, DC 20341, USA
Received 12 March 2001; received in revised form 5 May 2002; accepted 29 January 2003

Abstract

In this paper, we examine the various links among foreign direct investment (FDI), financial
markets, and economic growth. We explore whether countries with better financial systems can
exploit FDI more efficiently. Empirical analysis, using cross-country data between 1975 and 1995,
shows that FDI alone plays an ambiguous role in contributing to economic growth. However,
countries with well-developed financial markets gain significantly from FDI. The results are robust
to different measures of financial market development, the inclusion of other determinants of
economic growth, and consideration of endogeneity.
D 2003 Elsevier B.V. All rights reserved.

Keywords: Foreign direct investment; Capital markets; Credit markets; Economic growth; Spillovers

JEL classification: F23; F36; F43; O16

1. Introduction

In a trade English capital is instantly at the disposal of persons capable of understanding

the new opportunities and making good use of them. In countries where there is little

* Corresponding author. Tel.: +1-713-743-3824; fax: +1-713-743-3798.

E-mail address: [email protected] (S. Kalemli-Ozcan).

0022-1996/$ - see front matter D 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0022-1996(03)00081-3
90 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

Table 1
FDI facts
Value (billion dollars) Annual growth
1982 1990 2001 1986 – 1990 1991 – 1995 1996 – 2000
FDI inflows 59 203 735 24 20 40
FDI inward stock 734 1874 6846 16 9 18
Gross product foreign 594 1423 3495 19 7 13
affiliates
Notes: The data are from UNCTAD (2002), World Investment Report. UNCTAD (2002) defines FDI as an
investment involving a long-term relationship and reflecting a lasting interest and control of a resident entity in one
economy in an enterprise resident in an economy other than that of the foreign direct investor. FDI inflows
comprise capital provided by a foreign direct investor to an FDI enterprise. FDI stock is the value of the share of the
foreign enterprise capital and reserves (including retained profits) attributable to the parent enterprise plus the net
indebtedness of affiliates to the parent enterprise. A parent enterprise is defined as an enterprise that controls assets
of other entities in countries other than its home country, usually by owning a certain equity capital stake (10% or
more of the equity stake). A foreign affiliate is an incorporated or unincorporated enterprise in which an investor,
who is resident in another economy, owns a stake that permits a lasting interest in the management of the enterprise
(an equity stake of 10% for an incorporated enterprise or its equivalent for an unincorporated enterprise).

money to lend enterprising traders are long kept back, because they cannot at once
borrow the capital, without which skill and knowledge are useless (Bagehot, 1873).

The past decade was marked by the increasing role of foreign direct investment (FDI) in
total capital flows (see Table 1). In 1998, FDI accounted for more than half of all private
capital flows to developing countries 1. This change in the composition of capital flows has
been synchronous with a shift in emphasis among policymakers in developing countries to
attract more FDI, especially following the 1980s debt crisis and the recent turmoil in
emerging economies. The rationale for increased efforts to attract more FDI stems from the
belief that FDI has several positive effects which include productivity gains, technology
transfers, the introduction of new processes, managerial skills, and know-how in the
domestic market, employee training, international production networks, and access to
markets 2.
If foreign firms introduce new products or processes to the domestic market, domestic
firms may benefit from accelerated diffusion of new technology. In other situations,
technology diffusion might occur from labor turnover as domestic employees move from
foreign to domestic firms. These benefits, in addition to the direct capital financing it
generates, suggest that FDI can play an important role in modernizing the national
economy and promoting growth 3. Based on these arguments, governments often have
provided special incentives to foreign firms to set up companies in their country.

1
See World Development Report Bank (200a, b)
2
See Caves (1996) for a review of the empirical and theoretical literature on multinational enterprises.
3
See Grossman and Helpman (1991, 1995), Barro and Sala-i-Martin (1995, 1997) for the role of technology
transfers and market integration in growth. In addition to the technology transfer literature, the positive role of
FDI has appeared in the broader capital market integration and development literature. With particular reference
to FDI within the international financial integration, Rowland and Tesar (2003) and Hull and Tesar (2003)
specifically emphasize that multinationals can allow for greater risk diversification.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 91

Yet, curiously, the empirical evidence of these benefits both at the firm level and at the
national level remains ambiguous. For example, examining plant level data in Venezuela,
Aitken and Harrison (1999) find that the net effect of FDI on productivity is quite small—
FDI raises productivity within plants that receive the investment but lowers that of
domestically owned plants—thus seriously putting in doubt the ‘spillover’ theory. At the
macroeconomic level, growth regressions carried out by Borensztein et al. (1998) and
Carkovic and Levine (2003) find little support that FDI has an exogenous positive effect
on economic growth.
While it may seem natural to argue that FDI can convey greater knowledge spillovers, a
country’s capacity to take advantage of these externalities might be limited by local
conditions. In an effort to further examine the effects of FDI on economic growth, our
research takes its cue from the recent emphasis on the role of institutions in the growth
literature. In particular, we emphasize the role of financial institutions and argue that the
lack of development of local financial markets can limit the economy’s ability to take
advantage of potential FDI spillovers.
Schumpeter recognized the importance of well-developed financial intermediaries in
enhancing technological innovation, capital accumulation, and economic growth almost a
century ago. In a nutshell, the argument goes that well-functioning financial markets, by
lowering costs of conducting transactions, ensure capital is allocated to the projects that
yield the highest returns, and therefore, enhances growth rates. Well-known early
protagonists of this view include Goldsmith (1969), McKinnon (1973) and Shaw (1973) 4.
Although most FDI by its very nature relies on capital from abroad, it is important to
recognize that the spillovers for the host economy might crucially depend on the extent of
the development of domestic financial markets. There are different ways in which financial
markets matter. First, it is unlikely that spillovers are restricted to only costless improve-
ments in the organization of the workforce. In particular, to take advantage of the new
knowledge, local firms need to alter everyday activities and, more generally, reorganize
their structure, buy new machines, and hire new managers and skilled labor. Although some
local firms might be able to finance new requirements with internal financing, the greater
the technological-knowledge gap between their current practices and new technologies, the
greater the need for external finance. In most cases, external finance is restricted to domestic
sources. Furthermore, the lack of financial markets also can constrain potential entrepre-
neurs. This is especially true when the arrival of an entirely new technology brings with it
the potential to tap not just domestic markets but export markets. An excellent case in point
is the emergence of the textile export industry in Bangladesh in the early 1980s, following
the establishment of a textile plant by Daewoo in 1979. Of the 130 Bangladeshi workers
who were trained in Korea to become familiar with the technology, 115 eventually left to set
up their own garment export plants 5. It is difficult to imagine that all these workers
managed to finance factories with their own cash. Had loans not been forthcoming to

4
More recent examples include Boyd and Prescott (1986), Greenwood and Jovanovic (1990) and King and
Levine (1993b). Also, Galor and Zeira (1993) show how credit market frictions can limit human capital
accumulation and exacerbate income inequality.
5
To put this in perspective, before Daewoo set up its joint venture, the size of the labor force in the garment
industry in Bangladesh was only 40.
92 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

finance their enterprises and many more export industries that followed, it is unlikely that
garment exports from Bangladesh would have increased from $55 000 in 1980 to $2 billion
in two decades 6.
In addition, the potential of FDI to create backward linkages, in the absence of well-
developed financial markets, is severely impeded. The importance of linkages that
multinationals can create spawned a huge empirical literature following Albert Hirsch-
man’s (1958) seminal book on this topic 7. Even though backward linkages may allow
existing firms, which already produce inputs in the industry, to achieve economies of scale
that may not have existed earlier, it also can encourage the creation of new firms.
An excellent example is the involvement of Suzuki in India. Suzuki entered into a joint
venture with the Government of India in 1981 to manufacture small-sized affordable cars.
Initially, all the car’s parts were imported from Japan. Within 10 years, the plant had
become the center of gravity of scores of ancillary parts manufacturers that did not exist
earlier. Today, these suppliers provide 90% of a car’s parts 8. Without external financing, it
is unlikely that these manufacturers would have emerged. In similar vein, following Intel’s
construction of a semiconductor assembly plant in Costa Rica in 1996, local software
production in Costa Rica increased dramatically. Evidence indicates that the sector
benefited from newly created training programs in higher education institutions that have
become ‘Intel Associates’. However, producers and potential entrepreneurs in the software
sector continuously complain that lack of funds and/or the high cost of available financing
hinder the growth of the sector and its ability to compete in the international arena 9.
The preceding arguments and anecdotes illustrate the significant role financial markets
play in allowing spillovers and linkages associated with FDI to materialize. Furthermore,
to the extent that significant FDI arrives through mergers and acquisitions, it is not just
easy availability of loans but also well-functioning stock markets that matter. Well-
functioning stock markets, by increasing the spectrum of sources of finance for entrepre-
neurs, play an important role in creating linkages between domestic and foreign investors.
To summarize, one can conjecture that the extent of development of financial institutions
may be a decisive factor in determining whether foreign firms operate in isolated enclaves
with no links whatsoever with the domestic economy (beyond hiring labor). Or, whether
they become the catalysts for technology transfers and other benefits that economists have
long argued these firms should be.
Despite this rather obvious role for financial markets, the literature on FDI seems to
have ignored its importance altogether. In fact, the role of not just financial markets but
other factors, such as potential shortages of skills, knowledge, and infrastructure in the
recipient countries, have been neglected in the development literature. Caves (1999) notes
that the four volumes of The Handbook of Development Economics have nothing to say
about the kind of constraints local firms might face to reap such spillovers. It is only
recently that such issues have been addressed. For example, Borensztein et al. (1998),
6
See Easterly (2001) and Rhee and Belot (1990). Bangladesh, however, does not rank very well in terms of
the financial market indicators that we use. At the same time, it is rather well known for micro-credit institutions.
7
For a theoretical treatment on the ability of FDI to create linkages see Rodriguez-Clare (1996).
8
See Parikh (1997), page 138.
9
On Intel in Costa Rica, see Spar (1998), Hanson (2001) and Larrain et al. (2000). On the financing issues,
see Perez (2000).
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 93

using a data set of FDI flows from industrialized countries to 69 developing countries
show that, FDI allows for transferring technology and for higher growth. However, higher
productivity is possible only when the host country has a minimum threshold stock of
human capital. Likewise, Xu (2000), using data on US multinational enterprises (MNEs),
finds that a country needs to reach a minimum human capital threshold in order to benefit
from the technology transfer of US MNEs, and that most LDCs do not meet this threshold.
The World Bank’s (2001) edition of global development finance talks about the
importance of ‘absorptive capacities’ and the success of FDI 10. Absorptive capacities
here include macroeconomic management (as captured by inflation and trade openness),
infrastructure (telephone lines and paved roads), and human capital (share of labor force
with secondary education and percentage of population with access to sanitation).
Financial markets are not mentioned.
Although the empirical evidence on FDI and economic growth is ambiguous, the
interaction between financial markets and growth itself has been studied extensively and
has reached more positive conclusions—namely, that well-developed financial markets
promote economic growth. The theoretical framework has been well established in the
literature, with supporting evidence at the country level reported in the empirical studies
such as those of King and Levine (1993a,b), Beck et al. (2000a,b) and Levine et al. (2000),
suggesting that financial systems are important for productivity growth and development.
In an analysis of the roles of different types of financial institutions, Levine and
Zervos (1998) show that stock markets and banks provide different services, but both
stock market liquidity and banking development positively predict growth, capital
accumulation, and productivity improvements. At the industry level, Rajan and Zingales
(1998) find that the state of financial development reduces the cost of external finance to
firms, thereby promoting growth. Combining industry and country level data, Wurgler
(2000) shows that even if financial development does not lead to higher levels of
investment, it seems to allocate the existing investment better and hence promotes
economic growth.
Fig. 1, which shows data on FDI and financial development, provides motivation for
our work. We use FDI as a share of GDP and a measure of financial development
introduced by Beck et al. (2000a,b) for the period 1975 – 1995. As Fig. 1 suggests, there is
a positive relationship between the two variables. However, it is also apparent that there is
a wide variation in both variables given their interaction with one another. Indeed, if
financial development plays an important role in influencing the effects of FDI on output,
one can expect countries with the same levels of FDI to have very different outcomes in
terms of income levels.
In this paper, we examine whether economies with better-developed financial markets
are able to benefit more from FDI to promote their economic growth. To do this, we use a
battery of financial market variables that exist in the literature and employ them in growth
regressions to study the impact of the interaction of these variables with FDI on economic

10
The discussions demonstrate how some countries with low absorptive capacities, such as Morocco,
Uruguay and Venezuela (the last based on Aitken and Harrison, 1999), failed to reap spillovers; whereas Malaysia
and Taiwan fared well with higher absorptive capacities. See World Bank (2001), page 62.
94 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

Fig. 1. Countries in this plot are the 71 countries for which all accompanying data are available and form the first
sample in Table 2.

growth. We find that, although FDI alone plays an ambiguous role in contributing to
economic growth, having well-developed financial markets alters the results significantly.
This is consistent with the results of Carkovic and Levine (2003) and Hermes and Lensink
(2000). Countries with well-developed financial markets seem to gain significantly more
from FDI. We find that this result holds true even after controlling for a large number of
other variables that have significant influences on economic growth and also after
addressing concerns regarding endogeneity.
The rest of the paper is organized as follows: data are defined in Section 2; empirical
results are discussed in Section 3; and Section 4 concludes.

2. Data

This section describes the data used in the empirical analysis, specifically the measures
of FDI, financial market development, economic growth, and a number of controlling
variables used in growth regressions.
There are several sources for data on FDI. An important source is the International
Monetary Fund (2000) publication ‘‘International Financial Statistics’’ (IFS), which
reports the Balance of Payments statistics on FDI. Net FDI inflows, reported in the IFS,
measure the net inflows of investment to acquire a lasting management interest (10% or
more of voting stock) in an enterprise operating in an economy other than that of the
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 95

investor. It is the sum of equity capital, reinvestment of earnings, other long-term capital,
and short-term capital as shown in the balance of payments. Gross FDI figures reflect the
sum of the absolute value of inflows and outflows accounted in the balance of payments
financial accounts. Our model focuses on the inflows to the economy; therefore, we prefer
using the net inflow measure 11.
It is very difficult to construct accurate and comparable measures of financial services
data for a broad cross-section of countries over several decades. King and Levine
(1993a), Levine and Zervos (1998) and Levine et al. (2000) have constructed several
financial market series, spanning from the stock market to the volume of lending in an
economy. These variables can be classified into two broad categories: those relating to
the banking sector (or loosely, credit markets) and those relating to the stock market (or
equity markets). For the first set, we draw on variables introduced by Levine et al.
(2000), which in turn builds on King and Levine (1993a). The data associated with the
former are available from the World Bank Financial Structure Database 12. Four variables
are included in our work. First, liquid liabilities of the financial system (henceforth,
LLY): equals currency plus demand and interest-bearing liabilities of banks and non-
financial intermediaries divided by GDP. It is the broadest measure of financial
intermediation and includes three types of financial institutions: the central bank, deposit
money banks, and other financial institutions. Hence, LLY provides a measure for the
overall size of the financial sector without distinguishing between different financial
institutions. Second, commercial-central bank assets (henceforth, BTOT): equals the
ratio of commercial bank assets divided by commercial bank plus central bank assets.
BTOT measures the degree to which commercial banks versus the central bank allocate
society’s savings. King and Levine (1993a) and Levine et al. (2000), as well as others,
have used this measure, which provides a relative size indicator, i.e. the importance of
the different financial institutions and sectors relative to each other. Third, private sector
credit (henceforth, PRIVCR): equals the value of credits by financial intermediaries to
the private sector divided by GDP. The two previous measures do not differentiate
between the end users of the claims of financial intermediaries, i.e. whether the claims
are in the public or the private sector. This measure, and the one that follows, focus
solely on the claims on the private sector. Fourth, bank credit (henceforth, BANKCR):
equals the credit by deposit money banks to the private sector as a share of GDP (it does
not include non-BANKCR to the private sector and, therefore, may be less comprehen-
sive than PRIVCR for some countries). The number of countries for which we have
these financial market variables and FDI shares is 71 13.
The stock market data consist of variables introduced in Levine and Zervos (1998).
Stock market liquidity is measured as the value of stock trading relative to the size of the
economy, labeled as ‘value traded’ (henceforth, SVALT). In order to capture the relative

11
A limitation of this definition is that it may overestimate the amount of ‘new capital’ in the economy, since
it might simply involve a multinational enterprise buying out a local manufacturer. Carkovic and Levine (2003)
use gross FDI flows instead. However, it is not clear to us that outward foreign investment should generate
technological spillovers within the source economy.
12
The URL for the database is http://www.worldbank.org/research/projects/finstructure/database.htm. We are
grateful to the referee for directing us to this website.
13
In keeping with the literature, we use the logarithm of the financial sector variables.
96 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

size of the stock market, we use the average value of listed domestic shares on domestic
exchanges in a year as a share of the size of the economy (the GDP). This series is labeled
‘capitalization’ (henceforth, SCAPT). The stock market data series are also available from
the World Bank Financial Structure Database. The restrictiveness of the availability of
stock market measures, accompanied by those of FDI data, limits the sample size to
approximately 50 and also the length of the period to 1980 –1995. The countries included
in the various regressions are listed in Appendix A 14.
Growth rate of output is measured as the growth of real per capita GDP in constant
dollars, and the data are obtained from World Development Indicators (WDI) (World
Bank, 2000a,b). Gross domestic investment data come from World Bank (2000a,b), which
consists of outlays on additions to the fixed assets of the economy plus net changes in the
level of inventories. Inflation, measured as the percentage change in the GDP deflator, is
used as a proxy for macroeconomic stability. The data are from World Bank (2000a,b).
The institutional stability and quality in the economies are proxied by using data from the
International Country Risk Guide (ICRG), a monthly publication of Political Risk Services
that reports data on the risk of expropriation, level of corruption, the rule of law, and the
bureaucratic quality in an economy. A detailed description of all the data is included in
Appendix A.
To capture openness to international trade, we use the ratio of the sum of exports plus
imports to total output (GDP). Human capital is measured as the ‘average years of
secondary schooling’, obtained from Barro and Lee (1996) series 15. The government
consumption data come from the World Bank (2000a,b) and is the ratio of central
government expenditures to GDP. Finally, the population growth data are also obtained
from World Bank (2000a,b).

3. Empirical analysis

The first data set, relating to the ‘credit market indicators’ includes 20 OECD countries
and 51 non-OECD countries. The second data set, concentrating on ‘equity market
indicators’ consists of 20 OECD countries and 29 non-OECD countries 16.
Table 2 presents descriptive statistics for investment, growth, and financial develop-
ment data. There is considerable variation in the share of FDI in GDP across countries,
ranging from
0.15% in Sierra Leone (1975 – 1995) to 10% in Singapore (1980 – 1995).
GDP growth also shows variation, ranging from
4% for Guyana to 7% for Korea (both
for 1975– 1995). The financial development variables also range extensively; capitaliza-
tion of the stock market ranges from 1% for Uruguay to 126% for South Africa; SVALT
ranges from close to 0% for Uruguay to 130% for Switzerland. Finally, the liquidity

14
For value traded, we have data on 53 countries; for Capitalization we have data on 49 countries. The four
countries for which we do not have data for the latter variable are Costa Rica, Ireland, Honduras and Panama.
15
We used the updated data available at http://www.cid.harvard.edu/ciddata/ciddata.html.
16
Here OECD countries refer to those that were ‘early’ members and therefore exclude newer members, such
as Mexico and Korea among others. For value traded, we also have Ireland in the sample, increasing the number
of OECD countries to 21.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 97

Table 2
Descriptive statistics
Mean Standard deviation Minimum Maximum
Sample 1: 71 Countries (1975 – 1995)
Growth 0.01 0.02
0.04 0.07
FDI/GDP 0.01 0.008
0.001 0.041
Investment/GDP 0.23 0.06 0.11 0.41
PRIVCR 0.44 0.34 0.03 1.64
BANKCR 0.33 0.24 0.03 1.37
BTOT 0.77 0.19 0.27 0.99
LLY 0.48 0.28 0.16 1.61

Sample 2: 49 countries (1980 – 1995)

Growth 0.02 0.02
0.02 0.07
FDI/GDP 0.012 0.015 0.00 0.10
Investment/GDP 0.23 0.05 0.12 0.39
SVALT 0.11 0.21 0.00 1.30
SCAPT 0.27 0.30 0.01 1.26

measures (M2/GDP) ranges from 16% for Argentina to 161% for Japan. The PRIVCR
variable ranges from 3% for Ghana to 164% for Switzerland. Ghana and Switzerland also
form the two ends of the spectrum for the BANKCR variable. Ghana also has the lowest
value for the share of BTOTs; Austria records the highest.

3.1. Growth and FDI: financial markets as a channel

The purpose of our empirical analysis is to examine the financial markets channel
through which FDI may be beneficial for growth. In an influential paper, Mankiw et al.
(1992) (MRW) derive an empirical specification based on the assumption that countries
are unlikely to be at their steady states and, therefore, transitional dynamics should be
more important. We employ a specification similar to theirs 17. As a starting exercise, we
look at the direct effect of FDI on economic growth and estimate the following equation by
OLS:

GROWTHi ¼ b0 þ b1 logðINITIAL GDPi Þ þ b2 FDIi þ b3 CONTROLSi þ mi ð1Þ

Table 3 presents results based on regressions for the two samples that we have (the
larger sample of 71 countries for which we have data on all four credit market variables
and the smaller sample of 49 countries for which we have data on both equity market
variables). Columns (1) and (3) show results for a selection of control variables that
include initial income, human capital, population growth, government consumption, and a
sub-Saharan Africa dummy variable. For the sample of 71 countries, it is clear that FDI is
not significant at all, whereas in the smaller sample it clearly is. The results could be

17
Further, to ensure comparability, we include a number of controls that are present in Beck et al. (2000a,b)
and Levine et al. (2000); Carkovic and Levine (2003).
98 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

Table 3
Growth and FDI. Dependent variable—average annual per capita growth rate
(1) (2) (3) (4)
Period 1975 – 1995 1975 – 1995 1980 – 1995 1980 – 1995
Observations 71 71 49 49
log (initial GDP)
0.009
0.011
0.007
0.016
(
2.55) (
3.87) (
2.80) (
3.51)
FDI/GDP 0.16
0.076 0.347 0.063
(0.48) (
0.25) (2.31) (0.27)
Schooling 0.014 0.011
0.006 0.0001
(3.23) (2.62) (
1.41) (0.02)
Population growth
0.805
0.192
0.948
0.265
(
2.51) (
0.61) (
3.59) (
0.91)
Government consumption 0.0001
0.0003 0.008
0.003
(0.02) (
0.07) (0.98) (
0.35)
Sub-Saharan Africa dummy
0.007
0.017
0.021
0.021
(
1.15) (
2.63) (
4.78) (
3.80)
Institutional quality – 0.005 – 0.011
(2.62) (2.82)
Black market premium –
0.006 – 0.007
(
1.68) (2.00)
Inflation –
0.018 –
0.003
(
1.86) (
0.25)
Trade volume – 0.000005 – 0.008
(0.000) (1.25)
R2 0.37 0.59 0.34 0.60
Notes: All regressions have a constant term. t-values are in parentheses. The first two columns refer to the
sample of countries for which we have data on bank credit (BANKCR), commercial bank assets as a ratio of
total bank assets (BTOT), private sector credit (PRIVCR), and liquid liabilities (LLY). The second two columns
refer to the sample of countries for which we have data on stock market capitalization (SCAPT) and stock
market value traded (SVALT). The schooling variable is the log of (1 + average years of secondary schooling) for
the period of the regression. Population growth is the average growth rate for the period. Government
consumption is log(average share of government spending/GDP) over the period. Institutional quality is
measured by the average risk of expropriations. The black market premium is log (1 + average BMP) and
inflation is log (1 + average inflation rate) for the period. Trade volume is log (average of exports + imports as a
share of GDP) for the period.

driven by the composition of the two samples; approximately 28% of the first sample
(column (1)) and 41% of the second sample (column (3)) consist of developed countries.
In columns (2) and (4), we have an expanded set of control variables that include the
black market premium, institutional quality (captured by the ICRG measure called ‘risk of
expropriation’), rate of inflation, and trade volume. The FDI share no longer is significant
in either of the samples. This nicely summarizes the problem that exists in the literature:
whereas on theoretical grounds there is a strong basis for expecting FDI to have a positive
role in growth, the empirical evidence is fragile, to say the least 18. This ambiguous effect
of FDI is what forms part of the motivation for this research.

18
We repeated these regressions by adding the financial market variables as well. Although these variables
were significant and positive, they did not alter the insignificance of FDI.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 99

The regressions in Table 4 examine the role of FDI on growth through financial
markets. We interact FDI with financial markets and use this as a regressor to test for the
significance of financial markets in enhancing the positive externalities associated with
FDI flows. To ensure that the interaction term does not proxy for FDI or the level of
development of financial markets, both of the latter variables were included in the
regression independently. Thus, we run the following regression:

GROWTHi ¼ bV
0 þ bV
1 FDIi þ bV
2 ðFDIi  FINANCEi Þ þ bV
3 FINANCEi

þ bV
4 CONTROLSi þ mi ð2Þ

As shown in Table 4, the interaction term turns out to be positive and significant in all
columns. Each regression uses a different indicator for financial market development and
hence, samples may differ from one regression to another. Column (1) uses BTOT, column
(2) uses BANKCR, column (3) uses LLY, column (4) uses PRIVCR, column (5) uses
SCAPT and column (6) uses SVALT 19. The main result is that the interaction term is
significant at the 10% level for the entire range of financial sector variables used.
Moreover, the interactions with LLY, PRIVCR and BANKCR are significant at the 1%
level. On the other hand, financial market indicators by themselves are insignificant and
even negative for the non-stock market variables 20. This may in part be due to the
interaction term capturing an important allocation function that the financial sector
performs—having a well-developed financial sector is a means to an end and not an
end in itself. Interestingly, the coefficient of FDI displays considerable variation even
within the same sample of countries as the financial sector variable changes—clearly
making the case for looking at the range of financial sector variables rather than a few.
Table 4 also reports (a) the joint significance test of financial markets with the interaction
term and (b) the joint significance test of FDI with the interaction term. For most financial
market variables, the tests confirm the importance of both financial markets and FDI. The
hypothesis that the coefficients of both FDI and the interaction between FDI and financial
markets are zero cannot be rejected outright at the 10% level only in the case of BTOT and
SVALT. Not surprisingly the coefficients of the interaction terms in these two regressions
also report the lowest t-statistics compared with the counterparts in the other columns. The
hypothesis that the coefficients of both financial markets and the interaction between FDI
and financial markets are zero is rejected in all regressions.
To get an estimate of how important the financial sector has been in enhancing the growth
effects of FDI, one can ask the hypothetical question of how much a one standard deviation
increase in the financial development variable would enhance the growth rate of a country
receiving the mean level of FDI in the sample 21. If we use the PRIVCR variable (i.e. column
(4)), it turns out that having better financial markets would have allowed countries to

19
See the data section for detailed definitions.
20
The literature that tests the effects of financial development on growth has not considered FDI and its
interaction term with financial markets, thus limiting comparisons.
21
The mean value for FDI is 1.003% in the 71-country sample. Note that the financial development variable
here is the log of the financial market indicator.
 100
Table 4
Growth and FDI: the role of financial markets. Dependent variable—average annual real per capita growth rate
(1) BTOT (2) BANKCR (3) LLY (4) PRIVCR (5) SCAPT (6) SVALT
Period 1975 – 1995 1975 – 1995 1975 – 1995 1975 – 1995 1980 – 1995 1980 – 1995
Observations 71 71 71 71 49 53
log (initial GDP)
0.013
0.012
0.01
0.012
0.017
0.017
(
4.00) (
3.81) (
3.18) (
3.76) (
3.60) (
4.22)

L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

FDI/GDP 0.154 0.917 0.504 0.588 0.121 0.341
(0.45) (2.01) (1.67) (1.56) (0.68) (1.83)
(FDI/GDP)  financial markets 0.899 0.893 1.169 0.777 0.335 0.169
(1.91) (2.85) (3.08) (2.68) (2.61) (1.89)
Financial markets
0.0003
0.004
0.004
0.002 0.00007 0.0005
(
0.00) (
1.00) (
0.77) (
0.55) (0.03) (0.26)
Schooling 0.012 0.014 0.008 0.009 0.008 0.002
(2.85) (2.49) (1.92) (2.15) (0.15) (0.51)
Population growth
0.361
0.149 0.078
0.146
0.561
0.581
(
1.24) (
0.57) (0.29) (
0.56) (
1.70) (
1.80)
Government consumption 0.002 0.001
0.002 0.001
0.001 0.0004
(0.48) (0.27) (
0.37) (0.22) (
0.15) (0.06)
Sub-Saharan Africa dummy
0.016
0.02
0.021
0.02
0.025
0.023
(
2.42) (
3.14) (
3.25) (
3.08) (
5.08) (
4.83)
Institutional quality 0.004 0.004 0.005 0.004 0.008 0.009
(2.32) (2.44) (2.92) (2.45) (2.32) (2.64)
Black market premium
0.005
0.007
0.009
0.008 0.006 0.008
(
0.88) (
1.80) (
2.24) (
1.72) (2.15) (2.81)
Inflation
0.016
0.014
0.011
0.013
0.003
0.003
(
1.54) (
1.36) (
1.11) (
1.15) (
0.33) (
0.39)
Trade volume 0.0002 0.006
0.0002 0.001 0.008 0.085
(0.06) (0.12) (
0.06) (0.20) (1.27) (1.56)
R2 0.62 0.64 0.66 0.64 0.67 0.68
F-statistic for financial markets (Prob>F) 2.35 4.31 6.31 3.94 3.67 3.17
(0.10) (0.018) (0.003) (0.024) (0.035) (0.052)
F-statistic for FDI (Prob>F) 2.29 4.37 4.82 3.88 4.08 2.32
(0.11) (0.017) (0.011) (0.026) (0.025) (0.11)
Notes: All regressions have a constant term. Heteroscedastic consistent t-values are in parentheses. The financial market variable changes with each column. The financial market variables are all
logarithms of the actual values. See notes to Table 3 for the definitions of remaining variables. The F-statistics test the joint significance of coefficients. The F-statistic for financial markets tests
the null hypothesis that the coefficient for financial markets and the interaction terms are jointly zero. The F-statistic for FDI tests that the coefficient for FDI and the interaction term are jointly
zero. The numbers in parentheses below the test statistics indicate the P-values.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 101

experience an annual growth rate increase of 0.60% points during the 20-year-period, where
the net effect being measured is (b2meanFDIirlog(PRIVCR))+b3rlog(PRIVCR) 22.
An alternative way to see how countries performed is to simply use the estimated
coefficients for the sample of countries and calculate the net effect of FDI on growth for
each country. It turns out that most countries actually had a negative effect from FDI. The
net effect of FDI on growth is equal to b1FDIi+(b2FDIilog(FINANCEi)) 23. Table 5a
lists the distribution of the sample in terms of number of countries that benefited and
number of countries that actually experienced negative growth because of FDI. As can be
observed, there is considerable variation depending on which financial market variable we
look at. The stock market variables are particularly disturbing since they suggest that most
countries experienced a negative effect due to FDI. Of course this might partly be due to
the fact that most countries’ stock markets are even less developed compared with banks
and thereby exaggerating the problem. However, irrespective of which financial market
variable we use, there remains the concern that an unusually large number of countries
seem to experience negative effects. One explanation could be that we have forced a linear
relationship on what is essentially a non-linear interaction between FDI and financial
markets 24. Other than this problem, the results confirm our conjecture that insufficiently
developed financial institutions can choke the positive effects of FDI.
Table 5b reports the results of the significance tests of linear combinations of
coefficients at different levels of financial development. The null hypothesis is that
b1+(b2FINANCE)=0 at different levels of ‘FINANCE’. Therefore, here we report the
significance of FDI for different values of the financial market variables. As a crude guide,
we present the results at the minimum, mean and maximum values for each of the six
financial market variables. As expected, at the lowest levels of financial development, FDI
registers strong negative effects. This reconfirms the results of Table 5a. From Table 5b, it
is also apparent that even countries with levels of financial development equal to the
sample average did not derive significant positive effects from FDI. In fact, though not
significant, the effect of FDI at the average level of financial development also remains
negative for most of the variables. It is only at the maximum level of financial
development that the effects of FDI seem to be positive and significant. However, here
too the effects are not strong for at least two financial market variables: BTOT and SCAPT.
The results for BTOT are in keeping with the failure of the F-test for joint significance of
variables involving FDI in Table 4. The findings for SCAPT reinforce the results for the
same variable in Table 5a.
The strong positive correlation between the domestic investment ratio and the growth
rate of an economy is one of the few consistent results to have emerged from the multitude
of cross-country growth regressions that have appeared in the past decade. One could

22
Here mean FDI is 1.003% as mentioned in the earlier footnote. The standard deviation of log (PRIVCR) is
equal to 0.78.
23
Again, note that the financial market variable is a logarithm of the actual indicator and hence is negative for
any country with a value less than 1 (i.e. less than 100% of GDP). Therefore, even if the estimated coefficients are
positive, the net effect may still be negative if log(FINANCEi) is sufficiently negative.
24
Borensztein et al. (1998) suggest a similar possibility for the interaction with human capital. Such non-
linearities seem to provide support to theories of ‘poverty traps’ (see Galor, 1996).
 102
L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112
Table 5
(1) BTOT (2) BANKCR (3) LLY (4) PRIVCR (5) SCAPT (6) SVALT
(a) Net effects of FDI
Period 1975 – 1995 1975 – 1995 1975 – 1995 1975 – 1995 1980 – 1995 1980 – 1995
Observations 71 71 71 71 49 53
Number of countries that 34 29 19 27 5 13
had a net positive effect
Number of countries that 37 42 52 44 44 40
had a net negative effect
Maximum 0.6% Malaysia 1.4% Malaysia 2.1% Malta 1.3% Malaysia 1.9% Singapore 2.0% Singapore
Minimum
4.0% Guyana
2.1% Guyana
2.4% Papua New Guinea
1.7% Papua New Guinea
2.0% Egypt
1.5% Costa Rica

(b) Significance of FDI at different levels of financial development

Period 1975 – 1995 1975 – 1995 1975 – 1995 1975 – 1995 1980 – 1995 1980 – 1995
Observations 71 71 71 71 49 53
Minimum
2.12 1.35
2.98
2.78
2.23
1.55
Mean
0.43 1.65
1.99
1.11
1.54
0.86
Maximum 0.43 1.96 2.41 1.97 1.14 1.94
Note: See notes to Table 3 for the definitions of variables. The columns report the t-statistics for the Null Hypothesis that b1+(b2  Finance = 0) at different values
(minimum, mean and maximum) for each financial market variable where b1 and b2 are the coefficients of FDI and the interaction term, respectively.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 103

argue that the reason FDI appears significant in the above analysis is because the domestic
investment ratio was not controlled for. Therefore, for further robustness checks, we add
domestic investment to the list of independent variables, and the results are reported in
Table 6. Including domestic investment leads to a couple of interesting results. First, the
significance of the interaction term increases, particularly for BANKCR, BTOT, PRIVCR
and LLY. Only for stock market capitalization does the coefficient become less significant.
Second, the t-ratio of the FDI term also increases across all the columns, though still not
always significant. This suggests that FDI may have positive effects over and above its
direct role in capital accumulation. In particular, the so-called ‘positive externality’ effect
may be what is reflected here, though one would need more convincing results to come to
a firm conclusion. As expected, domestic investment enters significantly in all the
regressions 25. A final issue of robustness concerns the interaction between FDI and
human capital since this was shown to have a significant positive effect on economic
growth in earlier research 26. Column (7) reports the results for this regression. While FDI
and schooling both register significant effects, the interaction between the two does not. To
the contrary of previous findings, the interaction term is negative. However, we are using a
different human capital variable for a slightly different time period and, therefore, our
result may not be completely comparable with previous findings. The interaction between
FDI and financial markets (PRIVCR) remains robust.

3.2. Endogeneity issues

So far there has been no discussion of the endogeneity problem. Theoretically it is

plausible, and also very likely, that both the magnitude of FDI and the efficiency of
financial markets increase with higher growth rates. This would lead to an overstatement
of the effects of each of the two variables and their interaction on growth, since we use the
average values of these variables to estimate the coefficients in Table 4. Thus, we need to
construct instruments for both. For financial markets, we use variables that are not subject
to reverse causality, such as origins of a country’s legal systems and creditor rights. La
Porta et al. (1998) examine the laws governing investor protection, the enforcement of
these laws, and the extent of concentration of ownership of shares in firms across countries
(more popularly known as LLSV variables). They find that countries with different legal
histories offer different types of legal protection to their investors. Most countries’ legal
rules, either through colonialism, conquest, or outright borrowing, can be traced to one of
four distinct European legal systems: English common-law, French civil law, German civil
law, and Scandinavian civil law. They show that countries whose legal rules originate in
the common law tradition offer the greatest protection to investors. As far as law

25
In initial stages of our research, we found that the introduction of domestic investment made FDI
insignificant. We further found that this could be explained by the fact that both types of investment were highly
correlated, and FDI seemed to be a significant determinant of domestic investment. However, with the current
expanded sample, as noted above, we find that FDI can have significant positive effects on growth even when
controlling for domestic investment. Furthermore, there is little evidence that the two types of investment are
correlated any longer.
26
See Borensztein et al. (1998) and Xu (2000).
 104
Table 6
Growth and FDI—robustness: domestic investment and human capital. Dependent variable—average annual per capita growth rate
(1) BTOT (2) BANKCR (3) LLY (4) PRIVCR (5) SCAPT (6) SVALT (7) Schooling and PRIVCR
Observations 71 71 71 71 49 53 71
log (initial GDP)
0.011
0.01
0.009
0.01
0.017
0.017
0.01
(
4.15) (
3.55) (
3.10) (
3.42) (
4.36) (
4.87) (
3.40)

L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

Investment/GDP 0.119 0.096 0.069 0.096 0.143 0.128 0.99
(4.18) (3.35) (1.86) (3.32) (3.17) (3.11) (3.54)
FDI/GDP 0.311 1.066 0.501 0.672 0.194 0.352 1.59
(1.13) (2.70) (1.67) (2.04) (1.26) (2.11) (2.28)
(FDI/GDP)  financial markets 1.684 1.059 1.158 0.912 0.241 0.161 1.167
(3.74) (3.49) (3.22) (3.19) (1.86) (1.94) (3.75)
Financial markets
0.013
0.007
0.007
0.007
0.001
0.0009
0.009
(
1.24) (
1.96) (
1.55) (
1.72) (
0.51) (
0.52) (
2.24)
(FDI/GDP)  schooling – – – – – –
0.429
(
1.31)
Schooling 0.01 0.007 0.006 0.007 0.008 0.008 0.012
(2.40) (1.95) (1.69) (1.71) (1.22) (1.43) (1.93)
Population growth
0.454
0.262
0.045
0.237
0.673
0.681
0.304
(
1.71) (
0.98) (
0.14) (
0.87) (
2.06) (
2.23) (
1.16)
Government consumption 0.002 0.0002
0.002 0.0002 0.004 0.004
0.0006
(0.47) (0.04) (
0.46) (0.04) (0.59) (0.58) (
0.11)
Sub-Saharan Africa dummy
0.014
0.018
0.019
0.018
0.019
0.020
0.015
(
2.04) (
2.81) (
2.94) (
2.80) (
2.82) (
3.57) (
2.27)
Institutional quality 0.004 0.004 0.005 0.005 0.008 0.008 0.004
(2.49) (2.65) (3.07) (2.73) (2.61) (2.89) (2.64)
Black market premium
0.007
0.008
0.009
0.009 0.012 0.012
0.008
(
2.08) (
2.63) (
2.24) (
2.65) (2.73) (3.33) (
2.39)
Inflation
0.009
0.011
0.012
0.011
0.003
0.003
0.014
(
1.30) (
1.37) (
1.45) (
1.33) (
0.29) (
0.37) (
1.60)
Trade volume
0.0007
0.0007
0.001
0.0006 0.002 0.004
0.001
(
0.17) (
0.15) (
0.31) (
0.13) (0.46) (0.089) (
0.35)
R2 0.70 0.70 0.68 0.69 0.75 0.75 0.70
Notes: All regressions have a constant term. t-values are in parentheses. See notes to Table 3 for definitions of the variables.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 105

enforcement is concerned, German civil law and Scandinavian civil law countries emerge
superior. The French civil law countries offer both the weakest legal protection, and the
worst enforcement. These legal origin variables have been increasingly adopted as
exogenous determinants of institutional quality in the economic growth literature. In
particular, given their usefulness in predicting various indicators of investor rights and
protection, they have been used as instrumental variables for financial market development
in La Porta et al. (1997), Beck et al. (2000a,b) and Levine et al. (2000).
In addition to using the legal origin variables, we also use a measure of creditor rights (a
LLSV variable) as instrument for financial development. We will focus primarily on the
legal origin variables for two reasons: first, compared with creditor rights, these are less
controversial in terms of exogeneity—they are functions of colonization and occupation
usually before the second half of the 20th century. Second, as noted by La Porta et al.
(1997), the chain of links begins with legal origins, which, in turn influence the
shareholder and creditor rights, providing a basis for financial development. Shareholder
rights, creditor rights, and enforcement of legal – political rights variables also have been
used to instrument capital market integration in Kalemli-Ozcan et al. (2003). They provide
strong empirical evidence for an important mechanism through which a developed and
reliable financial system, backed by a legal environment that protects investor rights,
enhances specialization in industrial production. Hence, in addition to the legal origin
variables, their study provides a basis for additional instruments for financial development
such as creditor rights.
Table 7 reports the results of the IV regressions using legal origin variables and creditor
rights as instruments for some of the financial sector variables. In columns (1) – (3), the
financial sector variables PRIVCR, BANKCR and SCAPT are instrumented by the
English and Scandinavian legal origin dummy variables. In column (4), the French legal
origin variable is added to the list of instruments for SCAPT. In column (5), the creditor
rights variable also is added to the list of instruments (note that it significantly reduces the
sample size). All columns show that the interaction term is still positive and significant and
results are very similar to the OLS results in column (1). All of the columns also report the
test statistic for no overidentifying restrictions to confirm the validity of the instruments 27.
Among the few consistently significant determinants of FDI are real exchange rates and
lagged FDI. Real exchange rates, either through altering relative costs or relative wealth,
impact the foreign investment decisions of multinational firms. In a model with imperfect
capital markets, Froot and Stein (1991) link FDI decisions with real exchange rate
variations where, for example, a depreciation of the domestic currency increases the
relative wealth of foreign firms, which leads them to increase their investment abroad.
Similarly, Blonigen (1997), assuming imperfections in the goods market, shows that the
real exchange rate influences the relative wealth of firms, thereby generating foreign
investment flows. In the empirical literature, Klein and Rosengren (1994) find supporting

27
We experimented with using at least three legal origin variables as instruments for each of the financial
market variables. It was only in the case of SCAPT that the null hypothesis of no overidentifying restrictions was
not rejected. Further, the sample correlation between English and the French legal origins was approximately

0.8 making it difficult to enter both simultaneously as instruments. In addition the first stage regressions where
we use French and English dummies in the same regression with another dummy do not provide a good fit.
 106
Table 7
Growth and FDI: the role of financial markets—endogeneity (IV). Dependent variable—average annual per capita growth rate
(1) (2) (3) (4) (5) (6) (7)
Period 1975 – 1995 1975 – 1995 1980 – 1995 1980 – 1995 1980 – 1995 1980 – 1995 1980 – 1995
Observations 73 73 50 50 36 48 32

L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

log (initial GDP)
0.01
0.013
0.011
0.012
0.013
0.01
0.006
(
2.58) (
2.15) (
1.90) (
2.16) (
2.57) (
2.17) (
0.82)
FDI/GDP 2.75 1.585 0.213 0.148
0.178 0.243 1.525
(1.92) (1.60) (0.89) (0.62) (
0.75) (0.79) (1.84)
(FDI/GDP)  financial markets 2.51 (2.04) 1.918 0.552 0.514 0.441 0.68 1.221
(1.85) (2.47) (2.41) (1.77) (1.69) (1.89)
Financial markets
0.014
0.009
0.0009 0.002 0.011
0.003 0.001
(
0.92) (
0.50) (
0.09) (0.24) (1.67) (
0.37) (0.13)
Schooling 0.014 0.012
0.007
0.002
0.007 0.001
0.016
(2.66) (1.99) (
0.08) (
0.28) (
0.71) (0.10) (
1.46)
Population Growth
0.225
0.228
1.108
1.28
1.43
1.00
1.50
(
0.85) (
0.93) (
1.55) (
1.85) (
1.83) (
1.69) (
2.06)
Government consumption 0.009U 0.007
0.001
0.001 0.004
0.003 0.006
(0.85) (0.73) (
0.17) (
0.10) (0.35) (
0.31) (0.52)
Sub-Saharan Africa dummy
0.021
0.019
0.02
0.021
0.031
0.02
0.025
(
2.51) (
2.41) (
1.88) (
1.98) (
4.02) (
2.13) (
1.48)
Black market premium
0.012
0.013 0.001 0.002 0.01 0.0004 0.005
(
2.88) (
2.63) (0.26) (0.51) (1.30) (0.09) (0.49)
Inflation
0.011
0.009
0.020
0.014 0.003
0.025 0.047
(
0.60) (
0.55) (
0.80) (
0.57) (0.15) (
1.15) (1.20)
Trade volume 0.002 0.003 0.007 0.007 0.011 0.007
0.005
(0.53) (0.66) (1.18) (1.19) (1.80) (1.16) (
0.76)
OIR test (Prob>v2) 0.175 0.028 0.311 0.291 7.22 3.477 1.42
(0.915) (0.989) (0.855) (0.571) (0.30) (0.481) (0.83)
Notes: All regressions have a constant term. t-values are in parentheses. See notes to Table 1 for the definitions of other variables. In columns (1) – (3), the financial sector variables PRIVCR,
BANKCR and SCAPT are instrumented by the English and Scandinavian legal origin dummy variables. In column (4), the French legal origin variable is added to the list of instruments for
SCAPT. In column (5), the creditor rights variable is also added to the list of instruments. Columns (6) and (7) control for both the endogeneity problem in FDI and in financial market indicators
by instrumenting FDI with one-period lagged FDI (FDI in 1979) and real exchange rate levels, respectively, and financial markets with the LLSV variables used in column (4). The OIR Test
reports the v2-test statistic for overidentifying restrictions. The null hypothesis is that there are no overidentifying restrictions. The terms in parentheses represent the P-value.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 107

evidence that the real exchange rate is a significant determinant of FDI. Along these lines,
real exchange rate is used as an instrument for FDI in the following analysis, where the
real effective exchange rate is calculated as the ratio of the local price index to the US price
index converted to the local currency. Likewise, following the evidence provided by
Wheeler and Mody (1992) that FDI is self-reinforcing, i.e. existing stock of foreign
investment is a significant determinant of current investment decisions, lagged FDI is used
as an additional instrument for FDI in the following analysis. This result is further
reinforced in several country level studies in the literature 28. Columns (6) and (7) control
for both the endogeneity of FDI and financial market indicators by instrumenting FDI with
one-period lagged FDI and real exchange rate levels, respectively, and financial markets
with the legal origin variables used in column (4). The results continue to support the
finding that FDI promotes growth through financial markets. The coefficients, however,
increase considerably in values compared with the earlier OLS results in Table 5.
Instrumental variable estimation here corrects for classical measurement error, which
biases the OLS coefficients to zero. The higher values of the coefficients also alter the
balance between countries that lose and those that benefit from FDI. For example, if we
repeat the earlier exercise of figuring out how many countries in the sample benefited from
FDI but now use the coefficients from column (7) for our original sample of 49 countries,
we find that as many as 20 countries benefited. This is a much higher figure compared
with our earlier finding that only five benefited.

4. Conclusion

Following the debt crisis in the 1980s and the recent turmoil in emerging markets in
the late 1990s, developing countries have changed their attitude towards FDI because it
is believed that FDI can contribute to the development efforts of a country. In general,
a multinational firm’s decision to extend production to another country is driven by
lower costs and higher efficiency considerations. From the host country’s perspective
though, the benefits of FDI are not restricted to improved use of its resources, but also
stem from the introduction of new processes to the domestic market, learning-by-
observing, networks, training of the labor force, and other spillovers and externalities.
Due to the ‘growth-development’ benefits FDI seems to convey, different countries and
regions have pursued active policies to attract FDI. Most countries, including both
developed and emerging nations, have established investment agencies, and have
policies that include both fiscal and financial incentives to attract FDI as well as
others that seek to improve the local regulatory environment and the cost of doing
business.
Even though such policies can be very effective in attracting foreign investment, local
conditions can limit the potential benefits FDI can provide to the host country by not
generating benefits that go beyond the ‘capital’ FDI brings and the wages it generates. In

28
Markusen and Maskus (1999) use different FDI determinants, such as lagged FDI, to discriminate among
alternative FDI theories. Borensztein et al. (1998) used these variables as instruments for FDI in their work.
108 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

this paper, we focused, in particular, on the role of local financial markets and the link
between FDI and growth. We believe that the lack of development of local financial
markets, in particular, can adversely limit an economy’s ability to take advantage of such
potential FDI benefits. Whereas bad financial markets may mean that a country is not in a
position to cope with unregulated short-term capital flows, our work suggests that the full
benefits of long-term stable flows also may not be realized in the absence of well-
functioning financial markets.
Our empirical evidence suggests that FDI plays an important role in contributing to
economic growth. However, the level of development of local financial markets is crucial
for these positive effects to be realized, and to the best of our knowledge this has not been
shown before. We also provide evidence that the link between FDI and growth is causal,
where FDI promotes growth through financial markets. The result of this paper suggests
that countries should weigh the cost of policies aimed at attracting FDI versus those that
seek to improve local conditions. These two policies need not be incompatible. Better local
conditions not only attract foreign companies but also allow host economies to maximize
the benefits of foreign investments.

Acknowledgements

The authors thank the referee, Eduardo Borensztein, Michael Devereux, Eduardo
Fernandez-Arias, Alex Hoffmaister, Tim Kehoe, Ross Levine, Edmundo Murrugarra, Julio
Rotemberg, Bent Sorensen, Beata K. Smarzynska and seminar participants at Bentley
College, Brandeis University, Harvard Business School, University of Houston, University
of North Carolina-Chapel Hill, Union College, World Bank; the 2002 ‘FDI Race: Who
gets the Price? Is it Worth the Effort?’ Conference of Inter-American Development Bank
and World Bank, Washington, DC, and the 2001 SED Conference in Stockholm for
valuable comments. The views expressed in this paper are those of the authors and do not
necessarily represent those of the IMF.

Appendix A

A1. Countries in the samples

1. Sample of 71 countries for which data on credit markets are available (BANKCR,
BTOT, PRIVCR, LLY).
2. Sample of 49 countries for which data on SCAPT and SVALT are available.
3. Sample of 53 countries for which SVALT was available but SCAPT was not: Sample of
49 plus Costa Rica, Honduras, Ireland, and Panama.

List: Algeria (1), Argentina (1, 2), Australia (1, 2), Austria (1, 2), Bangladesh (2),
Belgium (1, 2), Bolivia (1), Brazil (1, 2), Cameroon (1), Canada (1, 2), Chile (1, 2),
Colombia (1, 2), Congo (1), Costa Rica (1), Cyprus (1, 2), Denmark (1, 2), Dominican
Republic (1), Ecuador (1), Egypt (1, 2), El Salvador (1), Finland (1, 2), France (1, 2),
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 109

Gambia (1), Germany (1, 2), Ghana (1, 2), Greece (1, 2), Guatemala (1), Guyana (1), Haiti
(1), Honduras (1), India (1, 2), Indonesia (1, 2), Iran (1), Ireland (1), Israel (1, 2), Italy (1,
2), Jamaica (1, 2), Japan (1, 2), Jordan (2), Kenya (1, 2), Korea (1, 2), Malta (1),
Malawi(1), Malaysia (1, 2), Mexico (1, 2), Netherlands (1, 2), New Zealand (1, 2),
Nicaragua (1), Niger (1), Norway (1, 2), Pakistan (1, 2), Panama (1), Papua New Guinea
(1), Paraguay (1), Peru (1, 2), Philippines (1, 2), Portugal (1, 2), Senegal (1), Sierra Leone
(1), Singapore (2), South Africa (1, 2), Spain (1, 2), Sri Lanka (1, 2), Sudan (1), Sweden
(1, 2), Switzerland (1, 2), Syria (1), Thailand (1, 2), Togo (1), Trinidad Tobago (1, 2),
Turkey (2), United Kingdom (1, 2), United States (1, 2), Uruguay (1, 2), Venezuela (1, 2),
Zimbabwe (1, 2).

A2. Data sources and descriptions

Foreign direct investment: The net FDI inflows measure the net inflows of investment
to acquire a lasting management interest (10% or more of voting stock) in an enterprise
operating in an economy other than that of the investor. It is the sum of equity capital,
reinvestment of earnings, other long-term capital, and short-term capital as shown in the
balance of payments. Source: IMF ‘‘International Financial Statistics’’.
Output levels and growth: Output level and growth data is the growth of real per capita
GDP, constant dollars. Source: WDI, World Bank (2000a,b).
Value traded: Value of stock trading relative to the size of the economy. Source: World
Bank Financial Structure Database. (http://www.worldbank.org/research/projects/finstruc-
ture /database.htm).
Capitalization: Captures the size of the stock market, measures the average value of
listed domestic shares on domestic exchanges in a year as a share of the size of the
economy (the GDP). Source: World Bank Financial Structure Database.
Liquidity (LLY): Liquid liabilities of the financial system (currency plus demand and
interest bearing liabilities of the financial intermediaries and non-blank financial interme-
diaries) divided by GDP. Source: World Bank Financial Structure Database.
Private sector credit (PRIVCR): The value of credits by financial intermediaries to the
private sector divided by GDP. It excludes credits issued by central and development
banks. Furthermore, it excludes credit to the public sector and cross claims of one group of
intermediaries on another. Source: World Bank Financial Structure Database.
Bank credit (BANKCR): Credit by deposit money banks to the private sector as a share
of GDP. Source: World Bank Financial Structure Database.
Commercial-central bank (BTOT): Ratio of commercial bank domestic assets divided
by central bank plus commercial bank domestic assets. Source: World Bank Financial
Structure Database.
Creditor rights: An index aggregating different creditor rights. The index is formed by
adding 1 when: (1) the country imposes restrictions, such as creditor’s consent or
minimum dividends to file for reorganization; (2) secured creditors are able to gain
possessions of their security once the reorganization petition has been approved (no
automatic stay); (3) secured creditors are ranked first in the distribution of the proceed that
result form the disposition of the assets of a bankrupt firm; and (4) the debtor does not
110 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

retain the administration of its property pending the resolution of the reorganization. The
index ranges from 0 to 4. Source: La Porta et al. (1997, 1998).
Domestic investment: ‘Gross domestic investment’ measuring the outlays on additions
to the fixed assets of the economy plus net changes in the level of inventories. Source:
World Bank (2000a,b).
Inflation: Percentage changes in the GDP deflator. Source: World Bank (2000a,b).
Government consumption: Total expenditure of the central government as a share of
GDP. It includes both current and capital (development) expenditures and excludes
lending minus repayments. Sources: World Bank (2000a,b).
Trade volume: Exports plus imports as a share of GDP. Source: World Bank (2000a,b).
Schooling: Human capital measured as the average years of secondary schooling in
total population. Source: Barro and Lee (1996). Updated version downloadable from:
http://www.cid.harvard.edu/ciddata/ciddata.html.
Bureaucratic quality: The institutional strength of the economy. High levels of quality
imply that the bureaucracy has the strength and expertise to govern without drastic
changes in policy, or interruption to public services. Source: ICRG.
Risk of expropriation: The probability that the government may expropriate private
property. Source: ICRG.
Black market premium: It is calculated as the premium in the parallel exchange market
relative to the official market (i.e. the formula is (parallel exchange rate/official exchange
rate
1)100). The values for industrial countries are added as zero. Source: World Bank.
(http://www.worldbank.org/research/growth/GDNdata.htm).
Real effective exchange rate: Calculated as the ratio of local price index to the
multiplication of the US price index and the official exchange rate. Source: World Bank.
(http://www.worldbank.org/research/growth/GDNdata.htm).

References

Aitken, B.J., Harrison, A., 1999. Do domestic firms benefit from direct foreign investment? Evidence from
Venezuela. American Economic Review 89, 605 – 618.
Bagehot, W., 1873. Lombard Street. Irwin, Homewood, IL.
Barro, R., Sala-i-Martin, X., 1995. Economic Growth. McGraw-Hill, New York, NY.
Barro, R., Lee, J.W., 1996. International measures of schooling years and schooling quality. American Economic
Review 86, 218 – 223.
Barro, R., Sala-i-Martin, X., 1997. Technology diffusion, convergence and growth. Journal of Economic
Growth 2, 1 – 26.
Beck, T., Demirguc-Kunt, A., Levine, R., 2000a. A new database on financial development and structure. World
Bank Economic Review 14 (3), 597 – 605.
Beck, T., Levine, R., Loayza, N., 2000b. Finance and the sources of growth. Journal of Financial Economics 58,
261 – 300.
Blonigen, B.A., 1997. Firm-specific assets and the link between exchange rates and foreign direct investment.
American Economic Review 87 (3), 447 – 465.
Borensztein, E., De Gregorio, J., Lee, J.-W., 1998. How does foreign direct investment affect economic growth.
Journal of International Economics 45, 115 – 135.
Boyd, J.H., Prescott, E.C., 1986. Financial intermediary coalitions. Journal of Economic Theory 38 (2),
211 – 232.
 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112 111

Carkovic, M., Levine, R., 2003. Does Foreign Direct Investment Accelerate Economic Growth? University of
Minnesota, Working Paper.
Caves, R., 1996. Multinational Enterprise and Economic Analysis. Cambridge University Press, Cambridge, UK.
Caves, R., 1999. Spillovers from multinationals in developing countries: some mechanisms at work. Manuscript
prepared for the William Davidson Conference on The Impact of Foreign Investment on Emerging Markets,
University of Michigan, Ann Arbor, 18 – 19 July.
Easterly, W., 2001. The Elusive Quest for Growth. MIT Press, Cambridge, MA.
Froot, K., Stein, J., 1991. Exchange rates and foreign direct investment: an imperfect capital market approach.
Quarterly Journal of Economics 106, 1191 – 1217.
Galor, O., 1996. Convergence? Inferences from theoretical models. Economic Journal 106, 1056 – 1069.
Galor, O., Zeira, J., 1993. Income distribution and macroeconomics. Review of Economic Studies 60, 35 – 52.
Goldsmith, R.W., 1969. Financial Structure and Development. Yale University Press, New Haven, CT.
Greenwood, J., Jovanovic, B., 1990. Financial development, growth and the distribution of income. Journal of
Political Economy 98 (5, Part 1), 1076 – 1107.
Grossman, G., Helpman, E., 1991. Innovation and Growth in the Global Economy. MIT Press, Cambridge.
Grossman, H., Helpman, E., 1995. Technology and trade. In: Grossman, G.M., Rogoff, K. (Eds.). Handbook of
International Economics, vol. 3 North-Holland, New York.
Hanson, G.H., 2001. Should Countries Promote Foreign Direct Investment? United Nations Conference on Trade
and Development, G-24 Discussion Paper 9.
Hirschman, A.O., 1958. The Strategy of Economic Development. Yale University Press, New Haven.
Hermes, N., Lensink, R., 2000, Foreign direct investment, financial development and economic growth, SOM
Research Report, 00027, Groningen, University of Groningen, 2000.
Hull, L., Tesar, L., 2003. Risk, Specialization and the Composition of International Capital Flows,
Working Paper.
International Monetary Fund, 2000. International Financial Statistics on CD-ROM. Washington, DC.
Kalemli-Ozcan, S, Sorensen, B.E., Yosha, O., 2003. Risk sharing and industrial specialization: regional and
international evidence, American Economic Review, in press.
King, R., Levine, R., 1993a. Finance and growth: schumpeter might be right. Quarterly Journal of Economics
108, 717 – 738.
King, R., Levine, R., 1993b. Finance, entrepreneurship and growth: theory and evidence. Journal of Monetary
Economics 32, 513 – 542.
Klein, M., Rosengren, E., 1994. The real exchange rate and foreign direct investment in the United States:
relative wealth vs. relative wage effects. Journal of International Economics 36, 373 – 389.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1997. Legal determinants of external finance. Journal
of Finance 52, 1131 – 1150.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and finance. Journal of Political Economy
106, 1113 – 1155.
Larrain, F., Lopez-Calva, L., Rodriguez-Clare, A., 2000. Intel: a case study of foreign direct investment in Central
America. Center for International Development, Harvard University Working Paper No. 58.
Levine, R., Zervos, S., 1998. Stock markets, banks and economic growth. American Economic Review 88,
537 – 558.
Levine, R., Loayza, N., Beck, T., 2000. Financial intermediation and growth: causality and causes. Journal of
Monetary Economics 46 (1), 31 – 77.
Mankiw, G.N., Romer, D., Weil, D.N., 1992. A contribution to the empirics of economic growth. Quarterly
Journal of Economics 107 (2), 407 – 437.
Markusen, J., Maskus, K., 1999. Discrimination among Alternative Theories of FDI, NBER Working
Paper 7164.
McKinnon, R.I., 1973. Money and Capital in Economic Development. Brookings Institute, Washington.
Parikh, K.S. (Ed.), 1997. India Development Report. Oxford University Press, Delhi.
Perez, C., 2000. Empresa de Technologı́a con cautela en 2002, El Financiero. January, 7 – 13.
Political Risk Services, International Country Risk Guide (ICRG), New York, various issues.
Rajan, R.J., Zingales, L., 1998. Financial dependence and growth. American Economic Review 88 (3),
559 – 586.
112 L. Alfaro et al. / Journal of International Economics 64 (2004) 89–112

Rhee, Y.W., Belot, T., 1990. Export Catalysts in Low Income Countries: A Review of Eleven Success Stories,
World Bank Discussion Paper 72.
Rodriguez-Clare, A., 1996. Multinationals, linkages, and economic development. American Economic Review
86 (4), 852 – 873.
Rowland, P., Tesar, L., 2003. Multinationals and the Gains from International Diversification, Working Paper,
University of Michigan.
Shaw, E.S., 1973. Financial Deepening and Economic Development. Oxford University Press, New York.
Spar, D., 1998. Attracting High Technology Investment: Intel’s Costa Rica Plant, World Bank Occasional
Paper 11.
UNCTAD, 2002. World Investment Report. United Nations, New York.
Wheeler, D., Mody, A., 1992. International investment location decisions: the case of US firms. Journal of
International Economics 33, 57 – 76.
World Bank, 2000a. World Bank Development Report. The World Bank, Washington, DC.
World Bank, 2000b. World Development Indicators on CD-ROM. The World Bank, Washington, DC.
World Bank, 2001. Global Development Finance Report. The World Bank, Washington, DC.
Wurgler, J., 2000. Financial Markets and the allocation of capital. Journal of Financial Economics 58, 187 – 214.
Xu, B., 2000. Multinational enterprises, technology diffusion, and host country productivity growth. Journal of
Development Economics 62, 477 – 493.