PhD defence: Benchmarking Economic Efficiency and Supply Quality in Regulation of Water Utilities

Emil Heesche portrait

PhD defence

Emil Heesche

Title of thesis

Benchmarking Economic Efficiency and Supply Quality in Regulation of Water Utilities


Utility regulators often use state-of-the-art productivity methods to ensure low consumer prices and high product quality in a political environment where theory and practice are not always aligned. This thesis aims to help regulators navigate between the theory and the practical problems they encounter when using benchmarking models, keeping the various political interests in mind. More specifically, the thesis examines the Danish water regulator's current benchmarking model to help develop the model to better consider that the water companies operate under different environmental conditions and with varying water quality levels.

The thesis presents four papers that discuss and propose solutions to several issues. 

The first paper shows how political considerations have resulted in untraditional model choices in the Danish water regulation and how these influence the properties of the model. We show how correlated outputs in a Data Envelopment Analysis (DEA) affect the results and argue that the regulator's benchmarking model yields higher efficiency scores than necessary because the regulator is too cautious. In addition, we propose to use DEA multipliers in a second stage analysis to examine how environmental variables influence the results. 

The second paper develops a method to incorporate quality variables into an economic benchmarking model. We show how a standard DEA model estimates unrealistic trade-offs between quality and costs, resulting in too high efficiency scores. We propose to use an outlier detection method to identify the most extreme trade-offs and use weight restrictions to remove these. Hereafter, we let the companies use their private information about their trade-offs to argue if some of the identified unrealistic trade-offs, in reality, are realistic. 

The third paper examines the incentives to have low costs and high quality in a standard RPI-X regulation with or without quality variables included in the associated benchmarking model. We use the Danish water regulation to show that the water companies have incentives to reduce both costs and quality in the benchmarking model, both with and without quality considered. However, the model with quality does improve the incentives to increase quality, although not sufficient. 

The last paper examines how aggregation uncertainty for inputs and outputs influences the DEA results. Regulators, researchers and other practitioners often use aggregations to reduce the dimensionality of a model. However, by doing so, they should be aware that they make implicit assumptions about the underlying estimated DEA technology set. In addition, uncertainty in the aggregation scheme can greatly influence the estimated efficiency scores. We use the Danish wastewater regulation to show that an aggregation uncertainty of 20 % can be severe for individual companies, while the average company is unaffected in this specific regulation. 

The model development and empirical results are based on the Danish drinking water and wastewater sectors. Still, they should easily be converted to most European utility regulators within water distribution, wastewater distribution, water treatment, wastewater treatment, electricity networks, gas networks, and other practitioners who encounter some of the same issues.


Professor Mette Asmild, Department of Food and Resource Economics (IFRO), University of Copenhagen

Assessment Committee


  • Professor Jens Leth Hougaard, Department of Food and Resource Economics, University of Copenhagen
  • Professor Niels Christian Petersen, University of Southern Denmark (SDU)
  • Professor Joakim Agrell, Université catholique de Louvain

The defence is open to all.


If you are interested in a full copy of the thesis, please contact the PhD student or the PhD Secretary.



No. The doors close when the defence starts and will not be opened again until the defence is finished.
You cannot leave early either unless there is an emergency.



It is not allowed to take pictures or record the defence without prior agreement with the PhD student and supervisors.



There is usually a reception after a defence. We kindly ask you to contact the PhD student if you want to know the location of the reception.