Public Infrastructure and Productivity: Updating the Canadian Case

Abstract

This report revisits the public capital hypothesis in Canada using a classical model with updated estimation techniques. Provincial panel data spanning 1998-2020 are leveraged to determine the effect of public infrastructure on business sector output and productivity. Several specifications (levels and first-differences) of the Cobb-Douglas production function are estimated using OLS and FGLS techniques that account for heteroskedasticity, autocorrelation, cross-sectional dependence, and province-specific effects. I find output and productivity elasticities for aggregated public infrastructure to be predominantly insignificant from zero. A detailed disaggregation of public infrastructure by asset type and function also yields largely insignificant results.

Download full PDF report