Sydney Petrol Prices

by Stubborn Mule on 29 June 2008 · 21 comments

Soaring petrol prices have led to all sorts of calls for action to help reduce prices. The Opposition called for a 5 cents per litre reduction in the excise on petrol, which currently stands at 38.1 cents per litre. (See note below for an explanation of the strike-throughs). the abolition of the double taxation of petrol by eliminating Goods and Services Tax (GST) on petrol excise. Since the excise is currently 38.1 cents per litre, this would save 3.8 cents per litre. One Victorian Liberal MP, Chris Pearce, went further and called for a 10 cent reduction in petrol excise. The Rudd Government initially claimed that there was nothing more that they could do, but then buckled to the pressure and has proposed the introduction of a national FuelWatch scheme aimed at promoting price transparency at the bowser. The Minister for Competition Policy & Consumer Affairs, Chris Bowen, has indicated that this scheme is expected to save around 2 cents per litre. So, what is going on with petrol prices and what are the merits of these proposals?

As a starting point, it is useful to have a look at the breakdown of the price at the pump. The retail price of petrol is basically equal to

Price of Crude Oil + Refining Margin + Excise + Retailer Margin + GST

This is a bit of an oversimplification and bundles together the cost of refining and profit under “refining margin” and the retailer’s costs (such as transport) and profit under “retailer margin”. Given all the media attention, it is hard to miss the fact that the price of crude oil has been sky-rocketinging and simple regression analysis confirms that this is the primary driver of petrol price rises. Refining and retailer margins have been far more stable, opportunistic service station proprietors are not really the root cause of the problem.

The chart below shows a comparison of Sydney average monthly petrol prices with prices from the regression model (for the statistically-minded, the R2 is 96%). The slope of the regression confirms the commonly quoted rule of thumb that a US $1 increase in the price of a barrel of crude oil will result in an increase of 1 cent per litre at the pump (more precisely the model says 0.87 cents, but that’s close enough to 1 cent: it is a crude model after all…no apologies for the pun).

Regression Model of Sydney Petrol Prices (unleaded)*

The average price of unleaded petrol in Sydney in May was 150.8 cents per litre. Yesterday our local service station was charging 163 cents per litre, so prices are already up more than ten cents in a month. Based on Friday’s crude oil price, the model is suggesting that there is another 10 cents to go on top of that. Last week comments from Libya about restricting oil supply helped push the price up crude oil up by US $5 per barrel (up 5 cents at the pump). A week earlier, moves by China to decrease fuel subsidies triggered a US $4 price fall (down 4 cents at the pump). So the realpolitik of oil is triggering larger moves in petrol prices on a daily basis than either the anticipated 2 cent saving of FuelWatch or even the 4 cent GST reduction proposed by the Opposition. These ideas start to look like a drop in the ocean, and expensive drops as that!

According to the most recent Federal budget, Government income from excise on petrol is forecast to be almost $7 billion per year, so dropping GST on cutting petrol excise by 5 cents could cost around $900 million per year in tax revenue for the Commonwealth States. Chris Pearce’s more extravagant proposal would cost the Commonwealth more that $1.5 billion a year in excise revenue! If diesel is included, the figures would be doubled. That is a lot of tax dollars to spend on savings at the pump which could be blown away by a casual remark to the world press by Libya, China or Saudi Arabia.

By contrast, FuelWatch almost seems like a bargain as it is expected to cost taxpayers a mere $20 million. But is that really $20 million well-spent? Complying with the scheme will cost petrol retailers money and, as their margins are already fairly slim, that is a cost that will certainly end up being passed on to drivers. So I am not convinced that the touted savings will materialise, but even if they do, will anyone notice? If oil prices keep rising, so will petrol prices and by much more than 2 cents. If it turns out that we are in the middle of a speculative bubble, oil and petrol prices will collapse at some point and everyone will forget we ever had a problem (for a while at least).

Either way, the best thing that I can find to say about the FuelWatch scheme is that it is the cheapest of the pointless initiatives that our politicians have come up with. If there is $20 million dollars of budget surplus up for grabs, surely there are better ways to spend it!

* Data Sources: Sydney Petrol Prices from the Australian Automobile Association, Brent crude oil prices and A$/US$ exchange rates from Bloomberg.

EDIT NOTE: strike-throughs reflect changes made following useful feedback from Michael Michael.


Possibly Related Posts (automatically generated):

{ 18 comments… read them below or add one }

1 Mike Harmon June 29, 2008 at 6:00 pm

Can you tell me who did your layout? I’ve been looking for one kind of like yours. Thank you.

2 Steven June 29, 2008 at 7:50 pm

Pointless initiative?

Labor’s Fuelwatch policy is based on the idea that petrol is price inelastic and can be subject to uncompetitive pricing. The 2c/L saving estimate is based on the level of this uncompetitive pricing as evaluated by the ACCC.

In WA Fuelwatch has been found to boost competition because it improves pricing information in the economy, and communication of that information with consumers. If you accept economics then this should have had some effect on pricing.

Here is some simplistic analysis that supports the idea that the policy has a point. If we assume that we do have a 2c/L average drop in petrol pricing for the cost of implementing Labor’s Fuelwatch of $20 million dollars annually. Then to make it worth while we would need to buy 20 million/0.02, or 1 billion Litres of petrol per year. Now petrol is integral to the economy, so any savings would have a multiplier effect in the economy, but for simplicity we’ll ignore that beneficial effect. Okay, so according to the ABS Australia consumed 16 billion Litres of fuel in 1999. I think it’s safe to assume that fuel consumption has grown to date and that this is a conservative figure. Of this 87% is petrol, and we can take away say another 20% because of WA’s contribution which already has a Fuelwatch scheme.

So about 10 billion [16billion*(0.87-0.2)] L of petrol is consumed annually to be affected by this policy. This is an order of magnitude over the costs of implementation while being excessively conservative.

Pointless I think not.


3 Michael Michael June 29, 2008 at 9:19 pm

Needs citation – as far as I’m aware, it’s not the opposition policy to abolish GST on excise (provide a reference!). Aside, a 2006 Senate inquiry into Petrol Prices chaired by Brandis concluded:
“The Committee is accordingly not persuaded by the evidence received which proposes any change to the system of excise and GST for petrol and does not make any recommendations in this regard.”

Rather the government _suggested_ that might come out of their taxation review:
(not the best source, left-wing bias, jews in media etc.)

Nelson proposed a flat 5c cut in excise in Budget-in-Reply:

I’m happy to be corrected…

4 Joshygeorge June 30, 2008 at 1:28 am

Very good thoughts SMC

5 stubbornmule June 30, 2008 at 11:03 am

@Steven: Thanks for your comment: I was hoping to spur some debate!

The disagreement I have with this return on investment type of argument is that it assumes that the 2 c/L saving represents a net boost to the economy. While this would be the case for a Government initiative that generates productivity gains, FuelWatch will really just result in redistribution of wealth. If an average saving of 2 c/L does eventuate, the saving will come to those drivers who bought their petrol from the more expensive petrol stations and these petrol stations will bear the cost by reducing their margins or losing business to competitors. While some may approve of this redistribution of income away from the more aggressive retailers, this is a moral judgement not an economic one.

As for the multiplier effect, it is very hard to argue that there would be a different effect on the economy if that 2 c/L was available for drivers to spend as opposed to being in the pockets of certain petrol station proprietors to spend.

So, in my view, FuelWatch does not create any demonstrable economic return on the $20 million investment.

@Michael: I stand corrected, thank you! I have updated the post in the approved strike-through blogging style.

@Joshygeorge: Glad you enjoyed the read!

6 Chatterick June 30, 2008 at 11:51 am

I think your argument is spot on. There has been no demosntratable long term saving on fuel in WA. While price fluctuations may have been minimised the price of fuel is not substantially cheaper. At the end of the day Fuel Watch is a layer of bureaucracy that doesn’t currently exist which will ultimately be paid for by the consumer. It targets the retail price margin which is one of the smallest parts of the petrol pricing equation. Structural reform of fuel standards and introducing competition at the local refinery level will do more to bring the price of fuel down than Fuel Watch will. At the end of the day if oil prices keep rising petrol prices will too. If the government feels so concerned about it then they should give up some of their margin……

7 Michael Michael June 30, 2008 at 7:23 pm

Ta – yup, I’m a badboy. But that other correction about Government vs. State revenue is a damn good get that I’m saddened to have missed.

8 dan June 30, 2008 at 10:01 pm

all very interesting food for thought. a more interesting question i think (having today put 60 odd litres of unleaded into my mazda 626 and for the first time broken into a three digit price range) is at what point the petrol price will begin to have a significant measurable impact on our reliance on cars.

in all the debate about using pricing messages to change public behaviour (carbon taxes being the most common) i don’t think any government would have had the courage to raise the price of petrol from 70c a litre to $1.60 … yet this is what has happened.

the traffic is still jammed.

so the question is: can you work out an algorithm/formula that will predict traffic decline as petrol price rises. so say at $7 per litre, you might be able to do the parramatta-sydney commute at 8am in 40mins, whereas at $10 per litre it’ll only take half an hour.

obviously some roads don’t count. i suspect the line of cars going up to kings cross from william street on a saturday night will always be the slowest lane of traffic in the world. and there will never be any congestion in the cross city tunnel.

but you get the general idea.

no doubt someone in a bunker somewhere in the rta control room (you can have peak here: is trying to work out this very question.


9 Mark L July 1, 2008 at 1:12 pm

Another nice piece, Sean.

However, I don’t agree with what I understand to be the basic rhetorical position, which seems to be that FuelWatch is pointless because the impact on retailer margins will be much smaller than the variability coming from oil price movements. Just because one component of a quantity is varying widely, does not mean it is (completely) pointless to make a narrower improvement to another more stable component. Yes, the presence of the widely variable component reduces the utility of the improvement, but an improvement is still an improvement. If you think petrol should be cheaper, then 2c/L is 2c/L better than before.

I think Steven’s argument is reasonable (if you accept the projection of a 2c/L reduction, and you think petrol should be cheaper — which I don’t), although he did ignore your point about compliance costs for retailers (unless these were already factored into the 2c/L calculation). Your response that this only represents a redistribution is an argument that can also be applied to the $20 million cost that you are complaining about: this Treasury expense will be used to pay lawyers, clerical staff and other government service providers and thus stimulate its own economic activity. If you want to explore this properly, you will need to talk about the economic distortion, not the headline cost.

Personally I think we should be increasing the excise. It represents the many negative externalities that car use in our society creates: carbon emissions, use of space for (larger) roads; materials, construction and maintenance of (larger) roads; particulate pollution; deaths and injuries from road accidents, and so on, all of which scale with mileage and hence fuel use. And our understanding of all these factors is moving in the direction of increased disutility. Hence the charges should be increased.


10 stubbornmule July 1, 2008 at 2:19 pm

@dan You raise some interesting questions here. I’ve been thinking about a chapter in Predictably Irrational which argues that behaviour changes more in response to sudden price changes than in response to gradual inflation (possible topic for a future Stubborn Mule piece…). Of course some changes in behaviour will themselves be gradual. If fuel costs go up, you may not go our and buy a more fuel-efficient car the next day, but perhaps you will change your decision next time you need to but a new car. There is some evidence that behaviour is changing in the US and the Honda Civic is now the best-selling car over there. While you may not be seeing much difference on the roads yet, changes may well be afoot. Anecdotally, my train does seem a little more crowded these days.

@Mark I think that my counter to Steven about wealth re-distribution does hold as he was arguing on the basis of a straight-forward return on investment for the $20 million. I do agree, however, that the redistribution idea can be extended further to factor in the destination of the $20 million. Then, while Steven argues that savings over $20 million represent net benefit from the scheme, the redistribution view sees the scheme as neutral to first order with some possible second order multiplier effects. However, it is then possible to argue that not all economic activity is “productive” in the same way, otherwise Goverments could simply hire as many people as possible to stack and unstack papers all day and point to the economic activity that they have generated.

As to your points about negative externalities, that’s where things really start getting interesting and it’s a subject that I plan to discuss here at the Mule in the future. In the meantime, it is very interesting to note that the response from China to soaring oil prices is to push up domestic prices by cutting subsidies, while our own Government’s response is to try to lower price. As Dan says, it doesn’t leave one with very much confidence in the Government’s commitment to economic approaches to tackling climate change, whether through trading schemes or carbon taxes.

11 Mark L July 1, 2008 at 9:23 pm

I agree that taking Steven’s argument as being one about return-on-investment leads to problems, but I’m not sure that’s how he meant it. The other way to take it is simply as a comparison between the options of reducing fuel excise (where each dollar of treasury expense generates one dollar of petrol cost saving to the consumer) and implementing FuelWatch (where an order of magnitude lower treasury expense generates the same outcome by redistributing wealth from retailers). If your primary goal is to cut petrol prices, subject to government budget constraints, then the latter option is clearly better.

12 Mark L July 1, 2008 at 9:27 pm

Well, assuming it works as predicted. I have some doubts about the 2c/L estimate.

13 dan July 2, 2008 at 10:55 am

@stubborn mule “I’ve been thinking about a chapter in Predictably Irrational which argues that behaviour changes more in response to sudden price changes than in response to gradual inflation (possible topic for a future Stubborn Mule piece…).” …

You mean the frog will really jump straight out of the boiling pot. But for the rest of us, as the price of petrol slowly creeps up – we’re making a nice Zuppa Di Rane

14 stubbornmule July 2, 2008 at 11:20 am

@dan Ahh, the old slow boiled frog analogy! My mother used exactly the same reference the other night. The look on my father’s face suggested he wasn’t impressed: that’s the physiologist seeing through the myth!

15 stubbornmule July 4, 2008 at 2:48 pm

There’s an interesting post over on the Possum Pollytics blog which discusses the analysis of FuelWatch carried out by and economics professor at La Trobe University. He concludes that the impact of FuelWatch on retail price margins was not a clear decrease. He concludes that a 95% confidence band ranges from a decrease of 1.0 cents/litre to an increase of 0.4 cents/litre.

16 Steve Reynolds July 15, 2008 at 10:58 pm

Nice to see you syndicated!

Aren’t we missing something in this whole debate (not on your blog, but nationally)? Even the most hopeful outcome of Rudd’s FuelWatch – a 2c per litre reduction – is utterly negligible, isn’t it? On average, Australian households consume 150 litres of petrol per month, so that amounts to a monthly saving of around $3. Fantastic, maybe I can buy that large coffee I’ve been saving up for. Under Nelson’s even more ludicrously populist proposal the saving is about $7 per month. If I drive to work, I spend more than that on tolls EVERY MORNING.

We live in a market economy. We rely on a commodity which is sourced in unstable nation states, subject to global market speculation, and on-sold by oligopoly retailers. We have no control – which is precisely why fuel is the product of choice for any government wanting to collect a reliable indirect tax via excise. We’ll pay whatever is asked.

Or we can use less of it. Which isn’t a bad idea anyway. Wouldn’t it be strange if capitalism included its own self-correcting mechanism that helped save the planet?

17 stubbornmule July 16, 2008 at 12:13 pm

@Steve: You made a good point: I don’t need FuelWatch as long as I can give up one coffee a month!

Speaking of self-correcting mechanisms, I heard Barnaby Joyce on the radio saying that this global credit crunch will slow down the world economy to much that our emissions problems may sort themselves out. Now that is optimisim!

Also I see that the Rudd Government is now planning to cut excise after all to offset price increases due to emissions trading.

18 stubbornmule July 31, 2008 at 7:40 pm

Here is the link to the syndicated version of the article on the Oil Drum.

Leave a Comment

{ 3 trackbacks }

Previous post:

Next post: