These last two figures (and the stochastic dominance tests reported in Appendix B)
confirm the findings of the parametric analysis on the direction of trade. In fact, after
controlling for firm characteristics, the density distribution of productivity associated to
trade flows that have the same origin and destination always lies on the right of the density
distribution of productivity associated to other flows (Fig.4). While in the case of trade
flows with the “North” there seem to be no diversification between the density
distributions (Fig.5).
Conclusions
Indian exporters and importers have higher productivity than firms that are not engaged in
these practices. This replicates similar findings for a number of other countries.
Specifically, we find that such positive correlation with performance is stronger and more
significant when the share of inputs imported or the share of output exported are
introduced substantiating the idea that markets are segmented and it is necessary to
specialize both as a “buyer” and as a “seller”. Therefore the more a firm is oriented
towards foreign markets, the more advantages it can reap given also the fact that there are
fixed costs in entering foreign market that need to be compensated. Moreover the most
productive firms are those that choose to have both backward and forward linkages with
foreign counterparts. In particular, the higher is the magnitude of the combined flows that
involve the same region the higher is the efficiency advantage. This combined effect of
imports and exports within trade networks on firm performance has not being documented
before and it deserves further investigation.
Besides, further analysis on the relationship between import and exports can have
important trade policy implications. India has moved from an import-substitution
industrial policy to more liberalized import and export policies. However the country
liberalization process is not yet completed and for this it can be relevant to focus on the
import content of export at the firm level.
Finally from the analysis on the origin and destination of trade it has emerged that firms
that are exporter-to or importer-from North America and Western Europe do not
necessarily have a productivity advantage with respect to the others. This is quite
surprising if we interpret the results on the light of previous studies. However this could
correspond to the involvement of Indian firms in production networks with Northern
31