Google fuel

It is all about information. Under the government’s new scheme,

PETROL stations will have to reveal their fuel prices up to 24 hours in advance and commuters will be able to find out the prices for their local area by signing up to an email or SMS alert.

Hopefully, it will be on the web too as per my recommendation to the ACCC. A simple integration into Google maps and you can see all the petrol prices on your commute.

This is one area where the ACCC did its homework. Its econometric analysis of the scheme in WA revealed a fall in petrol prices and an elimination of the weekly petrol price cycle — something that was likely to be inefficient for consumers and the industry. Also, petrol stations are already collecting and providing this data to private sources and so the compliance costs are surely minimal. What is terrific here is that a policy was enacted in a part of Australia, evaluated and then rolled out as it was determined it did its job. Now we only have to get politicians interested in pilots and trials as a matter of course and we will be laughing.

17 thoughts on “Google fuel”

  1. The ACCC analysis does seem to indicate there was a 2cpl decline in WA after Fuelwatch came in.

    But I have two problems with that result. First, the analysis looks at relative price margins – this is a measure of profitability. So the ACCC has found that profitability of selling petrol has fallen.

    I tried to replicate that result using relative
    consumer prices – afterall the policy objective here is to reduce consumer prices at the pump. I used the same technique as the ACCC just with relative consumer prices. The coefficient is not statistically significant (I also made the Newey-West adjustment – same as the ACCC).

    So I interpret the two results as a decline in profitibilty but no decline in consumer price. A priori that constitutes a decline in
    economic welfare. I predict that smaller or independent sellers have exited the WA market – I haven’t hunted down the data but that’s what I expect to find.

    Notwithstanding that, the 2cpl is hardly likely to be economically significant. The interaction between the excise and GST is about 4cpl and the government (old and new) showed no interest in unwinding that
    problem – apparently not worth the effort. With 2c being half of that figure, it must be even less worthwhile.


  2. Hi Joshua,

    I quickly read sections of Chapter 15 of the December 2007 ACCC inquiry – from the discussion on page 248-249 it is not clear that the cycle was eliminated – it states
    ” Around the time the 24-hour rule
    was introduced (January 2001) and when Coles entered the Perth market (March 2004), the average price cycle duration extended slightly and became less uniform in duration than at other times. On both
    occasions, the cycles settled down into more stable pattern of more regular weekly cycles after several months. There was no price cycle during September 2005 to December 2005. Since January 2006 Perth
    has moved to roughly 14-day price cycles.
    The move to a 14-day cycle occurred five years after the introduction of FuelWatch. It is difficult to
    believe that this change in duration is simply a result of FuelWatch.”

    Sorry for the long quote – but thought it should be there in full to avoid a misleading impression.

    I have two concerns. First, would really like to see stronger evidence on what happens to the amplitude of the cycle. If it does reduce it and a lot of sales do take place at the trough (they were doing a good business on Manningham Road last night!) these consumers might be worse off.

    Second, the ACCC does caution on the administrative and compliance costs.

    There are other concerns but these need more thought.



  3. Sinclair

    Unfortunately you have not characterised the ACCC work correctly. The ACCC is not comparing profit. The ACCC compares ‘margins’ in that they start with headline retail prices and take off taxes (otherwise changes in state taxes make interstate comparisons meaningless), the fuel quality premiums (state fuel standards changed over time and if this is not removed the comparison is meaningless) and also the Singapore benchmark price from both WA and east coast prices. Because the same benchmark applies to both the east and west coast, this makes no difference (comparing the WA price less the benchmark and the east coast price less THE SAME benchmark means that the comparison is between east and west coast prices).
    So the comparison is between retail prices adjusting for state taxes and regulatory changes that affected retail prices but were (presumably) unrelated to fuel watch. The results are clear and statistically significant.


  4. would you be able to write aomething about why there is a fuel price cycle to begin with? it seems counter inuitive on an econimc sense

    if service stations know that it will be costly on a weekend, shouldn’t they be taking advantage of that and selling their fuel cheaper and increase demand?


  5. Stephen, I hear what you’re saying, but I don’t buy it.
    The ACCC use the variable
    Price margin = (Retail price – lagged Mogas95 price – net taxes – fuel quality premium) Perth
    – (Retail price – lagged Mogas95 price – net taxes – fuel quality premium) Average of eastern capitals
    and find a 2cpl difference that is statistically significant.

    Mr Pat Walker has made the claim
    “In WA, for example, we’ve now managed to turn Perth around from being one of the dearest capital cities for unleaded petrol in the country to the cheapest.”

    When I tested that claim (using the ACCC method) there is no statistical evidence to support it. The ACCC variable does not reflect consumer prices at the pump. It may be interesting for a whole range of other factors, but does not support Mr Walker’s argument. If anyone other than the ACCC made that argument, it is my opinion that the ACCC would look at whether they had engaged in misleading conduct.

    I also made a prediction that independent’s have gone out of business. People have contacted me with anecdotal evidence supporting that prediction. I haven’t yet looked at for any data on this.


  6. If the price cycle is eliminated and petrol prices fall on average but most consumers (I have seen figures quoted of approx 70%) purchase petrol at the low point of the cycle, doesn’t this mean that consumers could be worse off as they may pay a higher price for petrol?

    For example, pre-change, assume the price cycle has a low (or Tuesday) price of 140cpl and the average price is 144cpl. After the change, assume there is variation of only 1cpl and the average price is 142cpl, then at least 70% of consumers are worse off because they are buying in the range of 141-143cpl rather than 140cpl.


  7. I was contacted a few days ago by an organisation called Informed Sources who had undertaken a detail analysis of Fuelwatch. They are happy for me to publicise their results. They say the biggest impact on petrol prices in Perth was the introduction of supermarket competition.

    In early 2004 a significant and lasting benefit occurred with the arrival of the second supermarket brand. The Perth market stood to benefit most from the introduction of supermarket petrol competition – it was the last capital city to abolish weekend rostered trading and a capital with inherent competition challenges due to its remoteness. This can be seen in the large fall in average prices of 2.5 cents per litre compared with other capital cities – Adelaide enjoying a 0.9 cents per litre drop and Brisbane a 0.5 cents per litre improvement. Sydney and Melbourne delivered negligible improvement from the introduction of supermarket competition.

    The ACCC have a serious problem: I can’t replicate their empirical results using consumer prices, while another source says any price differential exists due to supermarket competition. The ACCC caveat to the research itself is worth reading too. Read the whole thing, pay particular attention to

    Of potentially greater concern is the possibility that something else entirely has driven the improvement in the relative price margin.

    Bottom line, the ACCC doesn’t really know what happened and the person doing the analysis isn’t really sticking his neck out too far.


  8. Are we seriously calling three pages in Appendix S of the ACCC report “homework” or “evaluation”??

    Where are the full details of the ACCC’s econometric analysis? There are a couple of tables buried in the last appendix but that’s about it. Hardly a rigorous analysis. Will they allow independent researchers to inspect their entire body of work and evaluate it?

    Also, there is a huge difference between the following three statements:

    (a)”Fuelwatch led to lower prices, full stop” and
    (b) “Fuelwatch led to lower prices than in other states”
    (c) “Allowing for all other possible factors (i.e. ceteris paribus), fuel price margins in Perth would have been 2 cents higher, if not for Fuelwatch”

    Clearly (a) and (b) are not true, as you can check by just eyeballing the data. And (c) is open to debate; just because the ACCC makes a three page empirical claim does not make it so.

    And yet the political message that is being sold to the public is (a). Don’t believe me? The latest press release from Assistant Treasurer Chris Bowen says:

    “The econometric analysis undertaken by the ACCC indicates that not only has the introduction of FuelWatch in WA seen petrol prices around 1.9 cents per litre less on average, on the low points of the current weekly cycle, like Tuesdays, motorists were still paying less in Western Australia under FuelWatch, an average about one cent per litre less.”

    Would you say that is a fair representation of the ACCC’s findings?


  9. Well, this is certainly a passionate issue.

    On one point, Sarah, most people do not purchase at the low part of the cycle. It is more like 30% at best. Indeed, if that were not so, then there would be no cycle.


  10. It is not a passionate issue if you stick to the facts.

    So let’s stick to the facts, shall we? The fact is that the evidence presented by the ACCC does not support the claims being made by Labor.

    Bowen has claimed that “The ACCC did a very detailed analysis” and they presented “substantial evidence.”
    The reality is that the ACCC analysis is three pages long and would not even pass as a report from a first year university student.

    Bowen infers that Fuelwatch turned the weekly Perth price cycle into a fortnightly one. But the reality is that the weekly cycle still persisted in Perth well after Fuelwatch was introduced. It only changed to a fortnightly one in January 2006 – some five years after Fuelwatch was introduced. Of course, Bowen would know this if he had read the ACCC report.

    Bowen also infers that we are the only country in the world with regular petrol price cycles.
    But again, as the ACCC report shows, cycles exist in many other countries eg Canada, US, Europe. Bowen just has to read the ACCC report (p162) and he would know this!

    Notice also that Labor has conveniently moved the goalposts. We’re now being told that the intra-day cycle — not high fuel prices themselves — are the world’s greatest evil.

    But where did Labor in its election campaign rhetoric promise to abolish the intra-day cycle? No — they always said that prices would be lower, full stop.

    Less cyclical or flatter is not the same as lower.

    Fuelwatch will certainly eliminate the intra-day cycle – it bans it by law. But that was never Labor’s major promise. In my book, that is known as getting conned.


  11. The reason why Perth Prices are lower (and should be lower) than those on the east coast is that Perth is closer to the Singapore market where the U95 base quote is established. The 2 Acpl difference established by this ‘analysis’ is simply a pretty good estimate of this relative freight differential that should exist between these cities under the well established principle of import parity pricing.

    It is indeed disappointing how political this issue has become and how we as taxpayers are funding this ever increasing bureaucracy monitoring petrol pricing. Far better off to spend the money funding alternative energy research.


Comments are closed.

%d bloggers like this: