Table 3 presents some of the estimations for
imports of capital equipment by US MOFAs. The
dependent variables in equations (7) and (8) are based
on the directly observable data for the affiliates’
imports of capital equipment (CAPIMP), but has been
adjusted for the industry mix of the affiliates operating
in the host country in equation (9) (CAPDZF), and also
for the host country’s development level in equation
(10) (CAPDEV). The results are weaker than those for
the license measures - most notably, R is significantly
lower - but the positive coefficient of local
investment remains significant at the 10% level for all
variants of the equation. The coefficient of PREQS is
positive in all estimations, although never significant.
The signs of the coefficients of TREQS and the education
proxies vary, but none of them is significantly different
from zero. Furthermore, there are only small
differences between equation (7) and equations (9)
and (IO), which suggests that neither the industry distribution
of affiliates nor the development level of the
host country have any unambiguous impact on the
affiliates’ imports of capital equipment.