Scale effects in growth: A role for institutions

Reliance on a Dixit–Stiglitz production function leads the growth literature naïvely to associate economic scale with the size of a country's population. I develop an alternative approach in which market size is endogenous, reflecting a trade-off between the gains to exploiting non-rival skills and market transaction costs. Transaction costs reflect a country's institutional framework. The model supports scale effects in transitional growth rates and steady state income levels, suggesting scale effects may be important for developing countries. In this framework effective market size depends on a country's institutions. It is independent of population size and other macroeconomic variables.