On the role of public capital in production
This thesis examines the role of public capital, in particular, "core infrastructure", in private sector production in the United States. The underlying theme is the importance of the individual infrastructure stocks, in particular highways and streets, water and sewer systems and "other structures". Two different empirical approaches are used to shed light on a number of issues. In the first study in Chapter 3, two cost function models are estimated using data for the total private business sector, one using aggregate infrastructure data and the other using disaggregated infrastructure. The parameter estimates are used to calculate optimal infrastructure stocks (the optimal total infrastructure stock and the optimal individual stocks). The results reveal that, despite the fall in infrastructure investment from 1968-82, none of the infrastructure stocks was undersupplied over the sample period. The estimated output elasticities of the different infrastructure stocks are significantly lower than those obtained in previous research. In the second study in Chapter 4, use is made of recent development sin the productivity literature to construct a measure of manufacturing total factor productivity (TFP) that takes account of varying returns to scale and variable labour and capital utilisation over the cycle. The adjusted TFP measure is used to shed light on the causal relationship between infrastructure (total, core and disaggregated core) and productivity using a selection of autoregressive model-building techniques and causality testing procedures. Contrary to the stated view of many infrastructure researchers there is no evidence of "reverse causality", i.e. productivity causing infrastructure investment. There is, however, evidence that infrastructure has a small but statistically significant positive effect on TCP. Highways and other roads are the most productive types of infrastructure, followed by "other structures". When the TFP data is disaggregated, the findings is that core infrastructure affects some industries more than others, especially those that are capital intensive and have the largest motor vehicle shares.