How large are spatial barriers to transferring knowledge? We analyze the international operations of multinational firms to answer this fundamental question. In our model firms can transfer bits of knowledge to their foreign affiliates in either embodied (traded intermediates) or disembodied form (direct communication). Knowledge transfer costs interact with the knowledge intensity of production to determine the geographic structure of multinationals' input sourcing as well as its competitiveness in foreign markets. The model shows how data on observable trade costs and features of multinationals'global operations reveal the size and nature of knowledge transfer costs. Our empirical analysis confirms the model's predictions using firm-level data, quantifies the aggregate implications of the model for the structure of multinationals' operations, and demonstrates that transfer costs shape the knowledge content of intra-firm trade flows.

We develop a quantifiable general equilibrium model of trade and multinational production (MP) in which countries can specialize in innovation or production. Home market effects or comparative advantage leads some countries to specialize in innovation and relegate manufacturing operations to other countries via outward MP. Counterfactual analysis reveals that the reduction in the cost of MP or the integration of China into the world economy may hurt countries that are driven to specialize in production, although these losses tend to be very small. Contrary to popular fears, production workers gain even in countries that further specialize in innovation.