This study introduces an econometric model to explain the determinants of expenditures in capital maintenance and repair and, indirectly, to question the assumption implicit in most investment studies that such expenditures do not "matter" in the process of capital accumulation. The model is estimated by a consistent-systems technique, with data pertaining to the rolling stock of class-I line-haul railways in the United States from 1944-70. The empirical results indicate that maintenance expenditures are determined by gross additions to and retirements from the rolling stock as well as by the cost of funds and the rate of utilization. Moreover, in light of the uncovered rade-offs between maintenance expenditures and gross investment, the paper concludes that in estimating investment models, a proxy for maintenance expenditures should be included among the independent variables.

George C. Bitros, Harry H. Kelejian, On the Variability of the Replacement Investment Capital Stock Ratio: Some Evidence from Capital Scrappage, The Review of Economics and Statistics, Vol. 56, No. 3 (Aug., 1974), pp. 270-278