This paper compares the demand for the three individual components of aggregate investment demand: (1) demand by businesses for plant and equipment, (2) business inventory investment and (3) residential housing construction. The models tested are largely based on Keynesian theories of business investment demand, with some allowance for residential housing demand being more driven by Keynes’ consumer demand variables. Other possible determinants of investment are also tested, including ”crowd out” effects of government deficits on business investment and demographic effects on the residential construction market. Annual data for the U.S., 1960 – 2000, are tested using two stage least squares regression techniques modified to eliminate heteroskedasticity in the data. The models are estimated in “first differences”, rather than levels of the data to reduce the effects of multicollinearity, non stationarity and autocorrelation. The models explain about 90% of the variance in plant and equipment demand, 85% of the variance in residential housing demand for and 67% of inventory demand. The results indicate that demand for each of these three types of investment goods is driven by different combinations of variables Business investment in plant and equipment appears determined by how much the overall economy is growing (the accelerator effect), the availability of credit (crowd out), the availability of depreciation reserves, the prime interest rate lagged three years, business profits and stock values lagged one year, and the effects of an exchange rate change over the four year period following the change. Inventory investment seems mainly determined by availability of depreciation reserves, crowd out, interest rates, unexpected changes in consumer demand and the accelerator. Residential construction demand seems mainly driven by disposable income, the effect of general growth in the economy on consumer spending (the accelerator), credit availability (crowd out), current year mortgage rates, and prior year consumer wealth levels.
