1. Introduction
The Swedish health care system is to a large extent publicly managed with most health care
facilities owned and operated by 21 local, independent and politically governed county
councils. During recent years there has, however, been a movement were the local county
councils have opted to allow more of the production to be performed in production facilities
owned by others (i.e. private firms, cooperatives etc.) than the local county council. Also,
studies from other sectors in the economy have shown that public providers subject to
competition are more efficient than public providers with a local monopoly (e.g. Millward and
Parker, 1983; Sandstrom and Bergstrom, 2005).
There are some previous studies of performance in the Swedish health care sector. Fare et al.
(1995) and Fare et al. (1996) estimated Malmqvist productivity indices for Swedish hospitals,
and found that productivity had declined during the period 1980 to 1991. Jonsson (1994)
compare productivity measures in a county council who implemented a purchaser/provider
split to fourteen county councils who did not implement such a system. The results indicated
that productivity had increased more in the county council where the purchaser/provider split.
Finally, Gerdtham et al (1999) had the same main purpose as Jonsson (1994), but use
empirical methods closely related to this paper, of measuring technical efficiency. In their
paper, Gerdtham et al (1999) uses the approach of Borden (1988) and Bedard and Wen (1990)
in their studies of how different reimbursement systems affected efficiency in US health care
systems. The analysis is performed in two steps; first, economic efficiency is estimated using
data envelopment analysis (DEA), and second, the resulting efficiency scores are used as the
dependent variable in regression models trying to explain the efficiency score of different
production facilities or different county councils. The results from Gerdtham et al. (1999) are
mixed, but when pooling the data the results show that the introduction of purchaser/provider
splits increased economic efficiency in Swedish county councils. It should also be noted that
Gertham et al. (1999) includes the share of private physician visits in out-patient special care
in their estimations. The results indicated that the share of private physician visits did not
affect economic efficiency in the county councils, while having a non-socialist local
government, the proportions of elderly persons and the number of hospital beds all had a
positive impact on economic efficiency. However, during the last decade there has been an
increase in the amount of health care provided by others than the local county councils, which