Development of ESG Performance and Sustainability Indicators: A Data Envelopment Analysis Approach with Overlapping Criteria

In developing a firm’s environmental, social, and governance (ESG) performance indicator, we note that some criteria belong to more than one dimension. For example, “green energy” can be classified as environmental or social. We call such criteria “overlapping criteria.” The main issue is assigning appropriate weighting scores to overlapping and non-overlapping criteria. Another concern is that if there is a negative correlation between criteria, it is inappropriate to construct a composite indicator (CI). To resolve these shortcomings, we propose a two-level approach. The first level involves using data envelopment analysis (DEA) to find three-dimensional CIs. Each pillar of DEA includes the corresponding overlapping and non-overlapping criteria. The second level involves finding the aggregate score of the dimensional CIs by simple weighted average or DEA, depending on the dimension’s negative correlation. Due to the nature of DEA, some firms can use a few criteria or dimensions to achieve the best ESG performance, which deviates from the ESG development purpose. To resolve this issue, we propose to impose a set of sustainability constraints so that firms must consider all the criteria in a balanced manner and maintain flexibility to obtain a set of optimal DEA weighting scores.