3.2 Panel Studies with Heterogeneous Effects
The six papers reviewed in this section are based on the premise that the relative im-
pact of FDI determinants should be allowed to vary across regions, and the specification
therefore includes interactions between FDI determinants and the regional dummy vari-
ables like in (2).4 The results suggest that agglomeration effects, growth and openness
are equally important in all regions, whereas the return on investment, infrastructure,
political instability and fiscal incentives, among others, have a heterogeneous impact on
FDI across regions. We also find that region-specific factors should be taken into account.
A number of studies analyse if countries in one region are different from other devel-
oping countries. Asiedu (2002) and Kolstad and Villanger (2008) find that openness has
an equal impact on FDI irrespective of regional location. Asiedu (2002) also finds that
the provision of infrastructure and the return on investment have a larger impact on FDI
to African than non-African countries. The latter result is confirmed by Razafimahefa
and Hamori (2005). Kolstad and Villanger (2008) find that FDI to Latin America is
particularly sensitive to political instability, while the absence of regulation appears to
have been a particularly beneficial factor.
Another set of studies compare determinants of FDI in several regions. Asiedu and
Lien (2004) find that the impact of capital controls on FDI varies by region: capital
controls have no effect on FDI to African countries but affect FDI to East Asia and
Latin America adversely. Chen (1998) finds that agglomeration, growth and government
expenditures are equally important in Latin America and South East Asia, whereas fiscal
incentives and growth of export have a heterogeneous impact on FDI in the two regions.
In his comparison of Asia and Latin America, Nasser (2007) finds a great degree of
heterogeneity between the two regions. While agglomeration effects are equally important
in the two regions, infrastructure (telephone lines) and political instability (revolutions
and assassinations) have significant but different effects in the two regions. A large number
of factors (GDP, inflation, current account, schooling and political rights) only turn out
significant in one of the regions.
3.3 Regional Studies
Table 3 and Table 4 review 21 regional studies of FDI that base their FDI specification
on (1) for the group of either African, Asian or Latin American countries.5 To ease inter-
pretation and comparison, the FDI determinants have been divided into return (market
4Chen (1998) also uses dummy variables to compare FDI in Latin American and South East Asian
countries. However, this paper is excluded since the dependent variable is FDI in per capita terms which
invalidates comparisons with the other papers in the review.
5Kandiero and Chitiga (2003), Quazi (2007b), Chen (1998), Trevino et al. (2002a, 2002b), Vogiatzou-
glou (2007) and Trevino and Mixon (2004) are excluded from the review since they use absolute FDI or
FDI in per capita terms as their dependent variable.