Debunking the myths about recent regulatory appeals

Garth Crawford, Executive Director, Economic Regulation

Twitter: @garth_crawford

For the past week, many in the energy sector have been carefully absorbing and analysing a lengthy ruling by the Competition Tribunal on recent determinations by the Australian Energy Regulator (AER). In Get set for bill shock: inside the legal fiasco pushing up electricity prices (3 March 2016), Sydney Morning Herald Senior Writer, Jessica Irvine, claims that the appeals process represents a ‘fiasco’ that will cost consumers.

Much of the piece appears to be based on a mistaken assumption that the use of economic benchmarking has been broadly rejected by the Tribunal, and the AER’s discretion has now been restricted to merely passing on proposed or actual costs.

In fact, the Competition Tribunal has reaffirmed the AER’s ability to use robust economic benchmarking evidence, while identifying legitimate problems in some aspects of the modelling and application. Clearly, the public does not benefit when regulatory decisions are based on flawed evidence.

As a self confessed ‘professional sceptic’ about statistical evidence, red flags about the original decisions should have been flying high as the correspondent read through the judgement.

Given the number of factual errors in the article, we address below each of the false claims that could be misleading to readers.


Get set for bill shock: inside the legal fiasco pushing up electricity prices

Everyone knows prices are set by the relative forces of supply and demand, right?


Most people would be shocked to know the extent to which many every day prices in our economy are regulated.

Take electricity prices, and, in particular, the price paid for the transmission of electricity across the poles and wires networks, which makes up about half of household electricity bills.

As publicly-owned natural monopolies, much like Telstra’s copper wire network used to be, the price electricity networks can charge has long been set by an independent regulator.

It used to be the Independent Pricing and Regulatory Tribunal (IPART) in NSW, and today it is the national Australian Energy Regulator (AER).

Every five years, much like Stalin used to do, the regulator comes up with a determination on how much revenue the networks should be allowed to recoup from customers to cover their efficient costs, plus a reasonable profit.

Note the emphasis on “efficient” there.

Devoid of any competition, and left to their own devices, networks could in theory charge as much as they wanted for delivering their service.

They could be fat, lazy and over-invest in infrastructure, so-called “gold plating” of networks.

Consumers would end up paying highly inflated prices and the economy as a whole would suffer.

State governments might find themselves complicit, more inclined to turn a blind eye to the excessive costs imposed on households so they could keep collecting supercharged dividends to fatten their budgets.

In fact, that’s a pretty good description of exactly what has happened over the past decade, when electricity prices doubled.

In response, the national electricity regulator was given increased powers to take a more nuanced approach to setting prices, which took into account how the efficiency of Australia’s electricity networks compared to a benchmark of international networks.


The recent rule changes specifically direct the regulator to examine Australian benchmarks, not international benchmarks. Neither the regulator, nor the consultants that undertook the benchmarking, claim to be making international comparisons of efficiency.

The comparison was not a pretty one. Australian networks, and NSW networks in particular, emerged as some of the most bloated and least efficient in the world.


No part of the AER’s industry-wide benchmarking report made international comparisons.

In fact, the only international information used in the benchmarking comparison was added by the AER’s consultants to help their econometric models work, because they initially found there was no way to reliably estimate even a simple cost function without it.[1]

The electricity regulator responded last year, slashing the amount the NSW networks would be allowed to recoup from customers by billions of dollars over five years. This would have delivered households a $100 to $300 annual saving, depending on their network.

The AER, and the review body, are tasked with promoting the long-term interests of consumers. This is not synonymous with the delivery of short-term price falls. Short-term price reductions that are not reflective of long-term changes in underlying costs are not genuine ‘savings’.

Rather, reductions that are not evidence-based put consumers at risk of an unstable cycle of under-investment, required catch-up investment, driving price volatility and risking poor service outcomes. Independent reviews have pointed to this type of cycle affecting NSW consumers in the past.

Retaliation from the networks was swift. Amid wild threats of massive job losses and increased bushfire risks, they lawyered up, taking the regulator to the Australian Competition Tribunal seeking to have it overturn the decision.

In the three years prior to the AER’s final decision, NSW electricity distribution networks had already collectively reduced their workforce by more than 25 per cent, through company-led savings initiatives. Both network firms and the regulator are under a legal obligation to ensure that decisions permit the safe and reliable delivery of electricity.

Rather than a “wild threat”, the issue of managing bushfire risks was clearly relevant for the regulatory determination. In fact, the AER received a submission from fire authorities raising a range of concerns with the proposed cuts in the AER’s draft determination.

A legal fee feeding frenzy ensued. Pretty soon, the networks had engaged all the top silks in Sydney, forcing their opponents, the Public Interest Advocacy Centre, to fly barristers in from Melbourne to mount the case that consumers should pay even less than the regulator had determined.

More than 40 barristers assembled in a Phillip Street court for three weeks of hearings last year.

Together they threw thousands of pages of submissions and evidence at the tribunals’ three members: a judge and two economists.

No doubt the tribunal members assessed the evidence before them carefully, and to the best of their ability.

But it is it any wonder that, under such an onslaught, they ultimately ruled last Friday in favour of many of the network’s grievances?

The Tribunal ruled in favour of the AER, or made no determination either way, on seven out of 12 issues raised.

On the remaining issues, the Tribunal found that energy consumers’ long-term interests would be best promoted by the AER reconsidering its decision.

The effect of their 300-plus page judgment is that the regulator must go back to the drawing board on the way it determines allowable revenue. Instead of benchmarking networks against an international benchmark, revenue must be determined on a “bottom up” assessment of costs.


The AER was not proposing to benchmark Australian networks against any international benchmark and this was never questioned in the appeals. 

The actual effect of the judgement is that the regulator has to take into account the information it has already received, and must apply the existing Electricity Law and Rules in a way that best promotes the interests of consumers.

This will require:

  • an assessment of benchmarking evidence that does not rely on flawed data or weak models; and
  • avoiding arbitrary ‘after modelling’ adjustments to the results.

The AER’s use of benchmarking was criticised by both networks and consumer groups, such as the Public Interest Advocacy Centre, in the appeal.

The Tribunal found that the benchmark modelling the regulator relied upon:

  • included flawed data, despite clear and repeated warnings being given;[2]
  • used dummy variables in way that is not statistically valid to seek to account for incomparable data;[3]
  • failed to adequately account for distortions that were created when international data was added to help the econometric model to function at all;[4]
  • appeared to use the international data to populate the modelling merely because it was available, rather than because it was comparable;[5] and
  • failed to adopt a ‘reasonableness’ check to the results of econometric modelling.[6]

These weaknesses in the modelling used led the Tribunal to conclude:

Where, as here, the application of a new untested benchmarking model is applied to arrive at a total opex figure, sensible administration suggests that the regulator responsible for its application would apply some form of quantitative “reasonableness check” bottom up analysis to at least some, if not all, of the opex components. That is, however, not the case here.[7]

The decision does not mean that revenue must be determined only on a ‘bottom up’ assessment of costs. In fact, the Tribunal clearly directs the AER to use a broader range of modelling and benchmarking, and to have regard to reasonable forecasts of the likely costs of an efficient network provider. 

Whereas the regulator had pointed the finger at enterprise bargaining contracts signed by the networks and unions as too generous, the tribunal has ruled all those costs as given.


The Tribunal has made no such ruling. The Tribunal indicated that modelling used to set operating costs cannot simply ignore the existence of current enterprise bargaining agreements.[8] In fact, the Tribunal specifically stated that any identified and evidenced inefficiencies that the AER finds in its future determinations can be reflected in lower cost benchmarks, to the benefit of customers.[9]

So too, where the regulator had deemed contracts signed by the networks for maintenance of vegetation around poles and wires as excessively priced, the tribunal says the costs are allowable in full.

Where the regulator had tried to apply “incentive regulation”, designed to squeeze the networks to curtail costs, the tribunal’s decision essentially green lights a “cost of service” methodology, in which all costs incurred, however inefficient, can be passed on to customers.


A ‘cost of service’ methodology as described would be against the law today.  The National Electricity Law and Rules require the AER to reject a proposed operating cost forecast that does not reflect the efficient costs of a prudent operator meeting expected demand (See National Electricity Rules Clause 6.5.6 (c)). A regulator that passed all actual costs on to customers would be in breach of the Rules.

What the Tribunal actually said was don’t rely overly on flawed benchmarking models and that:

…in circumstances where benchmarking in Australia is in its infancy, sensible administration dictates that the AER should not have cast aside its previous practice of conducting bottom-up reviews in favour of the emphasis it placed on benchmarking.[10]

The bottom line is that you will end up paying more for electricity. It’s not clear exactly when the bill shock will hit. The tribunal’s judgment is so complex, it will likely take the regulator at least a year to come up with another price determination. We are already two years into the five year pricing period, so any increase in costs will have to be recovered by large increases in electricity prices, possibly as soon as July 1, but more likely heavily back loaded in years to come.

Networks share concerns on the potential for the reversal of any short-term cuts flowing from AER reconsidering its decisions, to result in unnecessary volatility in network charges.

This reinforces the importance of creating the right incentives for the best possible determinations in the first instance, which is a policy objective of having merits review avenues. It should also be a caution in future regulatory rulings to avoid ‘front-loading’ cuts, or increases, where the evidence to support these is not robust.

Worse, there is nothing to say the networks won’t simply haul the regulator back to the tribunal for an even more favourable decision. 

An extraordinary situation now exists where the national electricity regulator is effectively subordinate to a three-member tribunal.

Both the Australian Competition Tribunal and Australian Energy Regulator are both constituted by three members.

The Competition Tribunal is the recognised appeal body on access, regulation and competition matters across the entire Australian economy, under the Competition and Consumer Act. Its membership includes ex-Commissioners of the AER and ACCC and Federal Court judges.

And the best bit? Because the networks are still government-owned, taxpayers have been picking up the legal bill for this fiasco.

And we call this capitalism?

[1] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [157]

[2] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [254, 335]

[3] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [295]

[4] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [295]

[5] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [262]

[6] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [408]

[7] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [408]

[8] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT [436]

[9] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT[436]

[10] Applications by PIAC (Ausgrid) and Ausgrid [2016] ACompT[389]