The real effects of public investment on private investment
One of the main policy questions that emanated from the macro rational expectations literature, and the focus of many empirical studies, is whether expected government policy has an effect on real output (and/or whether unanticipated government policy has a permanent effect on real output). Many of the empirical studies have examined this question directly by testing whether a statistically significant relationship exists between real output or employment and some form of an aggregate demand policy variable. The conflicting results of these studies fail to either validate of refute the neutrality proposition. A shortcoming of these empirical studies arises because government spending is viewed merely as an additive component of aggregate demand and is, therefore, modelled on the demand side only. If, however, the productivity of private investment is correlated with aggregate demand policy, there will exist supply side effects, also. Another approach to determine any effects of public capital on private capital is to specify investment as two separate components - private and public - to determine whether private sector investment is affected (enhanced or diminished) by the public provision of infrastructure. This study uses a two-equation system that was estimated using a full-information maximum-likelihood statistical technique with non-linear parameter restrictions to determine separate effects of past government investment spending and past government deficit spending on private investment spending. The empirical results indicate that there is a positive, statistically significant relationship between private and public capital should be explicityly taken into account when examining aggregate effects of fiscal policy. In addition, combined federal, state and local deficit spending reveals no statistically significant effect on private capital investment spending.
Erenburg, S. J. (1993). The real effects of public investment on private investment. Applied Economics, 25(6), 831–837. doi:10.1080/00036849300000137