Investing in Public Infrastructure: Roads or Schools?

Abstract

This paper analyzes public investment scaling-ups in economic and social infras-tructure—say roads and schools—through the lens of a general equilibrium model. In the context of developing economies, investing in schools (relative to roads) is characterized by much larger long-run returns, but also by a much more pronounced intertemporal substitution of labor and crowding-out of private investment. There-fore, the public investment composition has profound repercussions on government debt sustainability, and is characterized by a trade-o with important welfare im-plications. A myopic government would not invest in social infrastructure at all. The model predicts an horizon of at least thirty years for political leaders to start investing in schools and twice-as-long an horizon for the size of expenditures to be comparable to the socially-optimal level. A big-push in investments can mitigate concerns arising from myopia, but at the cost of worsening fiscal concerns, although in this case the relative disadvantage of investment in social infrastructure is con-siderably reduced.