Can ‘benchmarking’ help network pricing?


It is almost certainly a sign of real sickness when the first thing I write about after crawling out of bed with the flu is electricity network pricing. But on Friday the AEMC released draft rules to improve the regulation of network pricing. The draft determination is here. One element that is given a lot of play in the press release is the potential use of ‘benchmarking’ for network companies. According to the press release:

The draft determination would:

  • allow the Australian Energy Regulator (AER) to benchmark network businesses against each other;
  • require the AER to publish annual benchmarking reports on network business performance;
  • allow the AER to take past efficiency into account when determining future spending allowances for network businesses;

Most regulatory economists will say ‘about time’. Benchmarking network firms’ performance (either intra-state between distribution networks, or interstate between transmission networks) has been a long time coming. Indeed, my understanding of why we have five separated distribution networks in Victoria (which goes back to the 1990s) was to facilitate benchmarking or ‘yardstick competition’ that never actually occurred.

The AEMC is taking a ‘softly, softly’ approach to benchmarking. It notes that the process for benchmarking is well understood and that the AER should be able to use benchmarking when determining appropriate operational expenditure and capital expenditure allowances for network companies. It leaves the ‘how’ largely to the AER. However, the AEMC does want the AER to undertake annual benchmarking reports and suggests changes to the National Electricity Law may be needed to allow the AER to gather appropriate information.

These are all good first steps – but come to a total of about five pages in a 200+ page document! So the real challenge will be for the AER. Can it grasp the lead given by the AEMC and start us on the path to full benchmarking?

To be fully implemented, benchmarking requires (at least) five things. First, it requires minimum service standards to be set for the network providers (e.g. reliability of supply). In other words, someone has to make explicit the level of standard that consumers can expect from their electricity networks. And there must be significant penalties in place for network providers who fail to meet those minimum standards.

Second it requires proper econometric analysis of differences in networks’ costs and performance. Network providers will always argue that they are ‘different’ and ‘can’t be compared to others’ because of ‘unique characteristics’ of their network. The networks have a strong incentive to overstate their ‘unique’ and ‘costly’ differences. However, real (exogenous) differences need to be identified and measured to allow benchmarking to work.

Third, benchmarking can only work if the network owner is given the latitude to determine its approach to investment. The network owner has to be given the latitude to try different approaches (e.g. rejig prices to encourage particular generation locations, use demand-side management rather than network augmentation, and so on). The idea of benchmarking is to encourage efficiency and innovation. And of course the network owner should be allowed to keep some of the surplus created by efficiency and innovation. Which brings us to …

Fourth, the network companies have to be allowed to lose money if they are inefficient. While there needs to be protections to ‘keep the lights on’, if a network provider is unable to make a profit due to inefficiency then it must be allowed to lose money and, if required, be on-sold to a better operator. This holds for public as well as private network companies.

Finally, benchmarking requires that the  AER can use the benchmarking data according to well-defined rules to compensate the network owners for the provision of network services. Put simply, the AER requires ‘real teeth’ to use the data and to ensure that inefficient network operators do not get to pass inefficiencies on to prices.

So overall, the AEMC benchmarking proposals are a good first step. But they are no more than that and there is a long way to go to achieve the benefits that full benchmarking can offer.

It looks like there will also be a Senate select committee to look at electricity prices. I hope that the committee will grab the lead presented by the AEMC – and push it a lot further. Because of its historic state-based structure, the Australian electricity system is a model case for benchmarking networks and using yardstick competition.

One Response to "Can ‘benchmarking’ help network pricing?"
  1. Again, much to agree with here. And a testament to my “sickness” that I comment on this on a Sunday night!

    Is the AEMC due for credit here? Hard to see really, the existing National Electricity Rules says that in setting expenditure allowances, the AER “MUST” benchmark opex and capex. How much clearer could it be?

%d bloggers like this: