Recovering the sunk costs of R&D: the moulds industry case

Sunk costs for R&D are an important determinant of the level of innovation in the economy. In this paper I recover them using a Markov equilibrium framework. The contribution is twofold. First, a model of industry dynamics which accounts for selection into R&D, capital accumulation and entry/exit is proposed. The industry state is summarized by an aggregate state with the advantage that it avoids the "curse of dimensionality". Second, the estimated sunk costs of R&D for the Portuguese moulds industry are shown to be important (3.4 million Euros). They become particularly relevant since the industry is mostly populated by small firms. Institutional changes in the early 1990s generated an increase in demand from European car makers and created the incentives for firms to pay the costs of investment. Trade-induced innovation reinforced the selection effect by which international trade leads to productivity growth. Finally, using the estimated parameters, simulations evaluate the effects of changes in market size, sunk costs and entry costs.