Intra-sectoral externalities, inter-sectoral externalities and international technological specialization.

Summary This thesis proposes a theoretical and empirical reflection on the implication of implicit knowledge exchanges (or externalities) on the technological performance of countries. Two categories of externalities are taken into account: "similarity" externalities and "difference" externalities. The theoretical contribution would be to construct a model in which externalities of difference would act simultaneously with externalities of similarity. The objective is to show how knowledge externalities manage to account for several phenomena linked in one way or another to international specialization in a knowledge economy. In particular, the proposed evolutionary model provides a theoretical explanation for the diversification of territories that is largely neglected in the literature. Empirically, our main work is in line with the geography of innovation. In addition to highlighting the impact of externalities on the capacity for innovation, our estimates confirm the capacity of sectors without comparative advantage (SAC) to generate positive intersectoral externalities. Thus, diversification in sectors without a comparative advantage (SAC) is transformed from a "harmful" and "involuntary" phenomenon into a "beneficial" and "intentional" phenomenon likely to encourage proactive policies for the capture of externalities. In general, by distinguishing externalities according to technological and geographic dimensions, our results allow us to draw the contours of an efficient innovation policy that takes into account technological complementarities and geographic environment. Thus, our thesis proposes an overall vision of the "externalities" phenomenon. It highlights the explanatory potential of externalities in understanding the dynamics of international specialization. Similarly, it allows us to understand why certain types of externalities have a stronger impact on innovation capacity than others.