This study brings together different facets of capital management policies. Specifically, it investigates structural estimation of investment, capital retirement, and exit decision rules of plants using data from the U.S. cement industry. The data available covers 1973-80 and spans the oil-shock period. Because cement is a very fuel intensive product, a marked increase in investment, capital retirement and plant closings in the cement industry were observed during this period. Compared to the existing empirical models of investment, the salient features of this project are that it allows for a more flexible adjustment cost function, endogenous capital retirement and endogenous exit of plants. Methodologically, a stochastic dynamic programming model within an iterative estimation procedure will be solved numerically. A sensitivity analysis will be conducted using the estimates obtained to quantify the responses of investment, capital retirement and plant closings to changes in demand and cost parameters as well as to policies such as the investment tax credit and depreciation allowances.
