This paper addresses the well-known paradoxes of high rates of protection, underinvestment in agricultural research, and relatively high productivity that characterize developed country agriculture, while developing country agriculture is typically characterized by taxation of the sector, research underinvestment, and low sectoral productivity. The paper tests the proposition emergiog from political economy theory that productive policies (e.g., research) and redistributive policies (e.g., subsidies) can be viewed as complementary in that the latter compensate producers who lose from the price-reducing effects of the former. The economic relationships between agricultural research expenditure, total policy transfers, sector productivity, and other variables are examined for a sample of developed and developing countries. The results confirm the complementarity hypothesis and show that increased relative rates of research expenditure are associated with higher agricultural productivity, higher country incomes, and higher rates of agricultural protection found in developed countries. The reverse is shown to occur in lowincome countries. The results suggest that both policy and trade reforms in developed countries and increased agricultural research allocation and sector productivity in developing countries may be harder to accomplish than previously thought due to the complementarity phenomenon.