Modeling the Role of Infrastructure for Regional Development

Abstract

Simplistic treatment of infrastructure during regional modeling often produces erroneous results and potentially wrong policy prescriptions. This paper attempts to address this issue and proposes a non-linear formulation for a productivity index in which infrastructure and economic mass (to reflect degree of agglomeration economies) of a region enter as interacting variables. The formulation is numerically tested to examine the effect of infrastructure investment at different level of agglomeration. Finally, the productivity index is used as a part of Constant Elasticity of Substitution (CES) production function, and the model is validated through estimation and calibration techniques using time-series data of Japan's regions. The validated formulation for productivity index, which realistically encompass the role of infrastructure, can potentially be utilized to assess the dynamic impacts of infrastructure in a multiregional context. The result is theoretically robust and practically useful for policy making particularly in developing Asian countries.