Restricting weights in value efficiency analysis

In this paper, we consider the problem of incorporating additional preference information into value efficiency analysis by using the 'price' information of inputs and outputs. This is done to improve the accuracy of the estimation of the value efficiency scores, Value efficiency developed by Halme et al. (1999) is an efficiency concept, which takes into account the decision maker's preferences. Value efficiency analysis is based on the assumption that an implicitly known value function reaches its maximum at the most preferred solution on the efficient frontier. The most preferred solution is an input-output vector preferred to all other possible input-output vectors. The ultimate goal is to measure a need to improve (radially) the values of inputs and/or outputs to make them equally preferred to the most preferred solution. Because we do not know the value function, we approximate the indifference curves of all possible value functions satisfying certain assumptions by their tangents at the most preferred solution. The resulting cone of the tangents consists of points less or equally preferred to the most preferred solution. However, in addition to the most preferred solution information about the 'prices' of inputs and outputs may be available as well. We show how this information can be incorporated into the analysis and illustrate the approach by an example on the performance of municipal dental units in Finland. (C) 2000 Elsevier Science B.V. All rights reserved.