A rejoinder to 'Measuring production efficiency in a not-for

Mensah and Li (1993) find fault with Hayes and Millar's implied assumption of cross-sectionally fixed cost shares and the use of an additional measure of output and an alternative measure of technology, which raises questions regarding model specification. In a rejoinder, Hayes and Millar discuss 2 areas of concern in order to address any previous omissions, needed clarifications, or differences in purposes. Their intended interpretation is that 1. line-item budgeting requires that the cost shares be fixed for the duration of the budget period for a given operating unit, and 2. fixed cost shares may inhibit the discretionary power of operating unit managers to substitute one input for another in an environment of changing relative input prices. The estimation procedure with cross-sectional data simply contemplates that the budgetary units are operating at different points on an underlying production function.