Public Capital, Inequality, and Politics

Christopher Arthur Clarke

This dissertation addresses the effect of public capital stock on the economy. Chapter One replicates the results of the landmark paper by David Aschauer on the impact of public capital on output for the US for 1949 - 1985. He estimated the elasticity of output with respect to public capital of 0.39. We obtained data from his stated sources and used his estimation techniques. We successfully replicate his main results. We also extend his data to the period 1949 – 2015, use several different sources for the input data, use DOLS and VECM estimation to account for various econometric issues, and undertake Granger causality tests. We are again able to generate results that are very close to Aschauer’s. Chapter Two studies the empirical relationship between public capital investments and the complementarity of skilled labor, unskilled labor, and private capital used in the production of private output. Investment in public capital has declined in the US relative to the size of the economy since the early 1970’s. If public investment is more complementary to skilled labor, then the slowdown in spending may cause a decline in the skilled wage premium. Using aggregate US time series from 1964-2014, we employee a Monte Carlo Markov Chain Metropolis-Hastings algorithm with Gibbs Sampling to estimate the elasticities of substitution of public capital. We find that public capital complements skilled labor more than unskilled labor. Had public investment maintained the post-war and pre-1970’s growth rate, the college wage premium would be 14% higher than it was in 2014. Chapter Three asks whether politicians allocate public capital to reward political loyalty or to persuade ideologically hesitant voters? I extend the theoretical model of Dixit and Londregan (1996) to include public capital. It predicts greater public capital allocation in political minority districts when the voting distance is small relative to the majority jurisdiction. When the voting gap increases, the relative allocation to minority-leaning districts decreases. Using unique data of census-level infrastructure and presidential elections, the empirical tests find the opposite result. The findings are robust at the county-level, in rural tracts, and in “safe” states.

Public Capital, Inequality, and Politics

Files and links (1)

Abstract

Metrics

Details