The crowd out effects of the government deficit is tested by adding it to consumption and investment models which control extensively for other factors. Effects are calculated for recession and non-recession periods, and compared to models with average crowd out, and models without crowd out. Test results indicate 1) deficits crowd out private consumption and investment, are statistically significant, and add substantially to explained variance. They also predict “IS” curve coefficients better than no crowd out models. Crowd out was found to have roughly equal effects in recessions and non-recession periods. Government spending deficits were found to result in complete crowd out, tax cut deficits resulted in more than complete crowd out. Both findings consistent with crowd out theory. Increases in M2, especially it’s saving (Non-M1) components, in the three years prior to a deficit, were found to offset the crowd out effects of government spending deficits, but not the effects of tax cut deficits. M1 increases were not found effective in eliminating crowd out effects. Foreign borrowing can be used to supplement domestic saving, reducing the possibility of crowd out when deficits occur. This paper uses a one variable definition of the deficit (T-G). This implies marginal effects of spending and tax deficits are the same. In working Paper 1104 of this series, these two kinds of deficits are tested separately to determine if they have separate effects.
