explanatory power, or that are highly correlated with other variables that have explanatory power. This approach may be well suited to the assessment of the disputed points, insofar as much of the debate between Ofwat and the Disputing Companies revolves around the degree of correlation between potential explanatory variables, and which set of variables can best predict costs given the limited number of observations. Depending on the robustness of the results and the extent to which they differ from Ofwat’s, we may use the results to inform our assessment of the claims submitted by the Disputing Companies and our decisions on whether to revise allowances. 44.
Second, several issues relating to cost adjustment claims are relevant to the modelling . These include Southern’s coastal population, regional wages and energy cost adjustment claims, alongside South East’s economies of scale in wastewater treatment services claim and Anglian’s leakage cost adjustment claim. For these we plan to assess a) if the claim is valid and if so then b) the extent to which they can be resolved via edits to the models in conjunction with the results of our work on econometric benchmarking set out in paragraph 43. If that is not possible, but we find there is still merit in the claim, we will evaluate these as they have been put forward by the Disputing Companies:
as individual cost adjustments. 45. Third, South East has asked us to acknowledge the limited reliability of these models and reflect this in our decisions . Where relevant, we will consider the reliability of models, for example in our consideration of the catch-up efficiency challenge, frontier shift and/or Performance Commitment Levels. 46. Finally, Wessex asked us to set aside the results of the models and set its base costs based on the ‘bottom-up’ evidence submitted in the Wessex business plan. We plan to deprioritise this request as the use of econometric benchmarking is a well-established practice.
We acknowledge that all econometric models are imperfect, and that it is not possible to establish with certainty that they incorporate every single determinant of costs. However, these models contain important information about the relative performance of companies and are the most important means by which Ofwat and we can mitigate the asymmetry of information that exists between regulators and the companies. We also note that Ofwat’s process leaves room for companies to submit focused cost adjustment claims to capture the impact of specific factors not included in the models.
Against that backdrop, setting aside the entirety of the modelling results and using Wessex’s evidence does not appear to be consistent with our duty to protect the interest of consumers. 13