TY - JOUR

T1 - Incorporating quality in economic regulatory benchmarking

AU - Heesche, Emil

AU - Asmild, Mette

N1 - Publisher Copyright: © 2022 Elsevier Ltd

PY - 2022

Y1 - 2022

N2 - The Danish water regulator uses, amongst other things, Data Envelopment Analysis to create a pseudo-competitive environment for the water companies. The benchmarking results are used to set an individual revenue cap for each company. The benchmarking model is currently criticised for not including the companies’ supply quality and thereby having an omitted variable bias problem. One problem the regulator has encountered when trying to incorporate supply quality in the benchmarking model is that it tends to increase the revenue caps more than desired. The regulator does, however, not have any prior information about the marginal rates of substitution between the quality variables and costs, which makes it challenging to reduce the supply quality's impact on the revenue caps. In this paper, we analyse the facet structure when incorporating three quality variables into the existing benchmarking model. We argue that it is generally sensible to investigate the facet structure and ensure that it is trustworthy before calculating efficiency scores, to increase the credibility of the results. By using an outlier detection model on the estimated marginal rates of substitution, we use the insights into the facet structure to create weight restrictions between costs and quality. This can help the regulator incorporate quality in the model without allowing the efficiency scores to increase excessively due to the increase in dimensionality. In addition, our proposed method reduces the need to extract the companies’ private information about their marginal rates of substitution, which is costly to verify for both the companies and the regulator. Lastly, we propose to add weight restrictions based on the consumer's willingness to pay for quality to avoid the companies choosing a level of quality that is higher than what the consumers are willing to pay for.

AB - The Danish water regulator uses, amongst other things, Data Envelopment Analysis to create a pseudo-competitive environment for the water companies. The benchmarking results are used to set an individual revenue cap for each company. The benchmarking model is currently criticised for not including the companies’ supply quality and thereby having an omitted variable bias problem. One problem the regulator has encountered when trying to incorporate supply quality in the benchmarking model is that it tends to increase the revenue caps more than desired. The regulator does, however, not have any prior information about the marginal rates of substitution between the quality variables and costs, which makes it challenging to reduce the supply quality's impact on the revenue caps. In this paper, we analyse the facet structure when incorporating three quality variables into the existing benchmarking model. We argue that it is generally sensible to investigate the facet structure and ensure that it is trustworthy before calculating efficiency scores, to increase the credibility of the results. By using an outlier detection model on the estimated marginal rates of substitution, we use the insights into the facet structure to create weight restrictions between costs and quality. This can help the regulator incorporate quality in the model without allowing the efficiency scores to increase excessively due to the increase in dimensionality. In addition, our proposed method reduces the need to extract the companies’ private information about their marginal rates of substitution, which is costly to verify for both the companies and the regulator. Lastly, we propose to add weight restrictions based on the consumer's willingness to pay for quality to avoid the companies choosing a level of quality that is higher than what the consumers are willing to pay for.

KW - Data envelopment analysis

KW - Facet structure

KW - Marginal rates of substitution

KW - Regulation

KW - Weight restrictions

U2 - 10.1016/j.omega.2022.102630

DO - 10.1016/j.omega.2022.102630

M3 - Journal article

AN - SCOPUS:85125278351

VL - 110

JO - Omega: The International Journal of Management Science

JF - Omega: The International Journal of Management Science

SN - 0305-0483

M1 - 102630

ER -