the patient’s health, the outcome of which is difficult to measure accurately. However, output, being a crucial
variable for the estimation of cost, needs to be measured. Breyer (1987) argues that given the ambiguous nature
of the concept of output in health sector, it is essential to identify the level of easily observable intermediate
products and use them as proxies for output. These proxies for output may be in terms of either total number of
patients admitted or the total number of bed days respectively.
In the literature, the unit case approach has been used more frequently than the patient day approach because of
two major reasons. First, it is more closely related to the true output, that is, health improvement than that of
the patient days. And second, under the patient day approach, length of stay can be easily influenced by the
hospital rendering the patient days an endogenous output variable. This violates an important property of cost
functions. Adoption of the unit case approach overcomes this problem. However, the unit case approach
assumes away the differences in case mix, length of stay and severity of the ailment, the factors which have an
important bearing on the cost incurred for cure. The unit case approach overcomes the problem of the length of
stay becoming endogenous variable. If the study does not cover a cross-section of hospitals, length of stay does
not become an endogenous variable. Thus, days rather than the number of patients become important, and the
patient day approach becomes the better alternative. However, this approach is not devoid of limitations.
Following this approach would not help to account the out patient services. The only possibility to overcome
this problem is to ascertain the number of outpatient visits. The cost of a day of inpatient also may vary
substantially with respect to the severity and complexity of the ailment. For example, the cost of an inpatient
day for Cardiac Surgery could be higher than that for many other ailments.
Cost classification for the health sector: The classification of cost into fixed and variable will not capture all
the cost heads for the health services. It calls for analysing cost in a broader perspective. A widely accepted
way of classification of cost in the health sector is to categorise cost into capital costs and operating costs (De
Ferranti, 1985). Operating costs includes those costs directly attributable in the production process, connoting
the variable cost. There are certain costs which change with the changes in output, but cannot be attached or
identified to any single output, as they are spread in the different products (services) produced in a multi-
product (service) system. These costs are called common costs. Panzar and Willig (1981) has classified them as