Production data concerning inputs and outputs in health care production for the Swedish local
county councils have been collected from the official records published by the Swedish
Association of Local Authorities and Regions. Data have been collected annually for the years
2001 until 2004, and all data expressed in monetary terms have been discounted using the
producer price index. Due to missing data, two local county councils (Uppsala and Gotland)
had to be excluded, leaving 19 local county councils to be studied. Also, in order to increase
the degrees of freedom in the Tobit regression models, the data have been pooled creating a
dataset consisting of 76 observations.
In this paper it is assumed that health care is produced using three types of inputs; labour (L),
capital (K), and health care from other sources (I). Labour is measured as total labour cost in
the different municipalities in million Swedish kronor (SEK) and capital is measured as the
total reported capital stock including buildings, machines and other inventories in million
SEK. If we were to only include the amount of labour and capital in each local county council
in the empirical analysis this amounts to making the implicit assumption that all health care is
provided by the local county councils. However, some local county councils buy part of their
health care from private firms or other county councils. In order to account for this, the
amount of health care bought from other sources (private firms, cooperatives etc. or other
county councils) is also included in the estimations of economic efficiency.
Turning to the measurement of output, the conceptual output of interest for the researcher is
the change in the population health status due to health care production in the local county
council. This is difficult to measure, and as most previous studies we are forced to use a
number of intermediate output measures instead. In this paper output will be measured by the
number of physician visits (PVISIT) in thousands in the local county council, the number of
visits to district nurses in thousands (NVISIT), the number of short-term hospital beds
available (BEDS), and the amount of health care sold to other producers of health care
(SALES) in million SEK. Descriptive statistics for these inputs and outputs are presented in
Table 1.
Table 1 about here.
In order to estimate the economic efficiency scores of the local county councils, the following
linear programming problem is solved;
6