An Optimization‐based Composite Indicators Approach to Performance Assessment

Composite indicators (CIs) approach has been widely accepted as a useful tool for assessing system performance at macro level. Several recent studies have shown that the weighted product (WP) method, a multiple criteria decision analysis method, may be a good choice in constructing CIs. However, a problem in its application is the subjectivity in determining the weights for sub-indicators. This paper extends the WP method and proposes an optimization-based approach to constructing CIs. The proposed approach requires no prior knowledge of the weights for sub-indicators. The weights used can be generated by solving a series of multiplicative DEA type models that can be transformed into equivalent linear programs. Additional information on the weights can be easily incorporated into the proposed models. A case study on assessing the performance of APEC economies towards sustainable energy development is finally presented to illustrate the use of the approach.