the above institutional and cultural variables. On the assumption that similarity boosts investments
abroad, the expected signs of the cultural, institutional an dummy variables are positive. The time
dummies Dt are meant to capture a variety of macroeconomic and FDI policy factors that affect
aggregate investments abroad of each of the four countries and the foreign investments in each
partner country. Possible omitted variables that vary through time but affect our four European
countries and their foreign partners are subsumed in these dummies.
Like Girma and Yu (2002), we choose not to use country-specific fixed effects in the
empirical model because this would make impossible to identify the impact of time-invariant
variables as distance, religion and governance, which are essential for this study. These same
variables, on the other hand, are specific fixed-effect dummies that capture potentially distinct
effects on the level of FDI.
The gravity model is then further augmented to include our variables of interest, the
immigrant stock from each country of origin (IMMIGRANTS) and, in the case of Italy, the stock of
emigrants living in each foreign country (EMIGRANTS). Following the theory of networks, these
variables are expected to have positive effects on the four economies’ bilateral FDI.
Because of the complexity of investment operations abroad, the networks of skilled
immigrants are expected to have a higher impact than that of the low-skilled immigrants, both for
the outward and the inward FDI. In the notation of the previous Section, this is consistent with the
assumption that the impact of skilled networks on international informal barriers is higher than that
of the networks of less skilled migrants, πS*'<πU*' < 0. Hence, in a subsequent specification of the
model, the stocks of immigrants are split into the SKILLED_IMMI and LOWSKILLED_IMMI
subsets.
As recalled previously, the biggest share of world FDI remains within the group of
developed countries. A closely related question is whether the networks’ ties can have an extra
value in influencing the international investments going to the developing countries. A way of
measuring this effect is by interacting the immigration and emigration variables with the OECD and
non-OECD dummies, where the latter represent the group of less developed economies. Given the
presumption that the information, matching and referral services that migrants provide may be more
valuable when referred to the less developed economies, we expect the sign of the coefficients of
the non-OECD subset to be higher. In particular, we expect this stronger effect to hold for the
outward FDI equations of our four countries. Prices in the developed economies generally convey
more information than in the poor economies. In the notation of the theoretical model, πt < πt* .
Hence, the transnational ties of networks are more important for firms doing foreign investments
into the developing economies.
10