Emissions, market power and electricity

The recent Frontier Economics proposal to treat emissions differently in electricity has been framed as being about whether price signals flow through to consumers or not. But there were no claims that the Frontier proposal might be favourable in encouraging low emissions generation in electricity. I thought that the CPRS and Frontier’s scheme would be equivalent in that regard.

According to a new paper released today by Stephen Holland there may be a difference. The abstract is over the fold. But the bottom line is this: if generators have significant market power in wholesale electricity markets, then an emissions intensity standard (similar to the Frontier proposal) dominates an emissions tax (equivalent in the  model to a CPRS scheme).

The intuition of these results is relatively straightforward. The weakness of an intensity standard is that it relies more on substitution effects than output effects to reduce emissions. With incomplete regulation or market power, this disadvantage may be helpful since leakage  may offset gains rom reducing output and since market power already inefficiently reduces  output.

The reason is that with market power, output is restricted and carbon prices have less of an impact on it. Now, I didn’t see this in the Frontier report but they are the experts on electricity markets and, in light of this paper, I’d be happy to admit that their scheme may be superior in driving the optimal mix of emissions in electricity if they would provide clear statements and evidence about the existence of significant market power held by electricity generators in Australia. (The issue on price signals remains live, however).

[DDET Read the abstract]

This paper investigates whether an emissions tax (equivalent to an emissions cap) is the  best policy in the presence of incomplete regulation (leakage) or market power by analyzing  an intensity standard regulating emissions per unit of output. With no other market failures, an intensity standard is indeed inferior, although combining it with a consumption tax  eliminates this inferiority. For incomplete regulation, I show that under certain conditions an  intensity standard can dominate any emissions tax (including the optimal emissions tax). This  dominance persists even with the addition of a consumption tax, which ameliorates output distortions and can sometimes help the intensity standard attain the first best (when an emissions  tax/consumption tax combination cannot). Comparing intensity standards to output-based  updating shows that the latter dominates because of its additional flexibility. Finally, I show  that with market power an intensity standard can dominate the optimal emissions tax.


3 thoughts on “Emissions, market power and electricity”

  1. Frontier did not assume market power or rely on it for their results.  They stated (P8):
    “A model of an efficient and competitive electricity market is used to examine the effects of various CPRS schemes, where costs, prices and generator returns are
    determined on an optimal least-cost basis” [my emphasis]
    I interpret the Holland paper as arguing that: under market power electricity prices are inflated, under an intensity standard they are subsidised, so the intensity standard can be adjusted so that these two effects cancel out and so prices are just right.
    This sounds like a great new way to regulate oligopolies.  Don’t bother with anti-trust legislation.  Simply subsidies the producers sufficiently so that they bring prices down to competitive levels.  In other words, so that the rent they obtain comes from the government rather than the consumer.
    Have I got this right?


  2. If you look at the <a href=”http://www.aemo.com.au/data/GRAPH_30NSW1.html”>price/demand data provided by AEMO</a> it looks like the generators have significant market power during peak demand periods, and not during the rest of the day.


  3. Market power in the Australian electricity sector is complicated.  Electricity being non-storable it doesn’t behave quite the same way as other commodity markets (similarly the price distribution is highly asymmetric, average price is $40, price cap is $10,000).
    Simple version: 80%+ of the time there is no market power and generators don’t play bidding games.  Overnight and during low demand periods of the year (spring/autumn) generators basically have to bid at cost since there is surplus capacity (e.g. Vic has 10,000 MW capacity, overnight demand is around 4000-5000 MW).  Overnight the coal stations are close to losing money because they don’t want to shut-down and restart.
    Side issue. This is a big problem with modelling, most coal stations have to run at 40-50% capacity all the time since they have to keep everything warm.  This means that their behavior isn’t flexible or responsive to small price signals.  Just looked at his paper, its standard economist assumption land, not an actual market simulation.
    However come hot summer days the story changes and certain generators can arbitrarily set prices in particular states.  This happened in SA in the 2008 summer and previously in NSW in the 2007 winter.  Most of the action in the electricity market occurs on around 10-50 days a year.  When market gaming is getting exercised prices will go above $1000 /MWh.
    A carbon price of $20 or even $50 won’t effect this gaming.  Losing a bit of volume (say 25%) on 10 days a year to gas peakers doesn’t really reduce emissions.


Comments are closed.