Thus, there is some evidence that investment by
local competitors may force the affiliates to higher
imports of the kind of technology that is embodied in
capital goods. The costs posed by the host countries’
performance and technology transfer requirements do
not seem to discourage imports of capital equipment
- if anything, there is a slight positive effect of
requirements, that perhaps reflects the preference to
import more embodied technology when the costs for
other transfer modes are high. Differences in the level
of education do not have any determinate effects on
the imports of capital equipment. As discussed above,
the reason may be that variations in the level of education
are related to two opposite effects: the negative
relation between education and technology transfer
costs may be balanced by a positive relation between
education and the availability of locally produced capital
equipment. Yet, the weak fit of the model suggests
that we have omitted some important determinants of
affiliates’ capital equipment imports, or that our
proxies are weak measures of the affiliates’ technology
imports, because of transfer pricing or other data
errors. In either case, it is clear that more detailed
analyses of intra-MNC trade in machinery and equipment
are called for.