28 February 2012

Gasoline Intensity of The Economy

UPDATE 3/1: Please also see the follow up post here.

Following up yesterday's post on US gasoline prices, here is a comparison figure for UK petrol spending as a proportion of GDP over the period 1991-2010 (the period for which data is available) with petrol data from the UK government's DECC and GDP data from ONS (conversion factors courtesy BP).

In this post I introduce a new concept (at least new to me and to Google) -- the gasoline intensity of the economy -- defined as the amount of total spending on gasoline as a proportion of total economic activity. (More precisely, the gasoline expenditure intensity of the economy.) Successful innovation in energy will lead to sustainable reductions in gasoline intensity of the economy. Reduced gasoline intensity means less economic vulnerability to increases in the price of oil resulting from a greater efficiency in the use of energy, both of which are desirable outcomes.

A few points to note:
  • The UK spends about 100% more than the US on gasoline as a proportion of GDP (in round numbers, the US is at about 0.3% and the UK 0.6%, for the 2010 data)
  • Assuming UK petrol consumption is constant in 2011 and 2012, and GDP in 2012 = 2011, then the index above increases to 75 in 2011 and 82 in 2012, or about the same as it was in 2001.
  • Thus, since 2001 US spending on gasoline as a proportion of GDP has increase by about 30%, while UK spending is about the same level
So the bottom line for the US in comparison to the UK is mixed -- The US spends about half as much of GDP on gasoline than does the UK.  At the same time, over the past decade US gasoline intensity has increased by about 30% while UK gasoline intensity is about the same as it was 10 years ago.

It has been frequently pointed out on this blog that gasoline in the UK and Europe is as much as twice the cost of that in the US. No doubt pricing helps to explain the improvement in gasoline intensity in the UK from about 2000 to 2009, and the recent reversal is explained by a combination of economic stagnation (if not contraction) in the UK coupled with increasing oil prices. It is important to observe that increasing oil prices have proportionately less impact in the UK because the relative change in prices will be smaller, as they start with a much larger base. My guess -- and it is only a guess -- is that economic recovery in the UK is likely to be accompanied by a resumption of improvement in the gasoline intensity metric.

The performance of the US with respect to gasoline intensity since 1983 suggests that after about 15 years of improvement, the US has been moving in the wrong direction. This provides some evidence that gasoline prices are not too high, but too low -- by exactly how much we can certainly debate. Those who'd like to argue the other side should address the following (a) whether you accept that decreasing gasoline intensity of economic activity is a desired outcome, and if yes, then (b) if pricing is not an appropriate tool to influence this outcome, what you would recommend instead.


  1. Why use "Percentage" in the graph title when it's actually an index, not a percentage? Better yet, why not make the left hand scale actually be percentages?

  2. -1-Bret

    Thanks ... at the moment I am interested in relative changes. But I can easily gin up a graph with the absolutes, perhaps for a future post. Thanks.

  3. Roger,

    Are you adjusting for excise taxes? If not, I don't think that the comparison is really that meaningful since fuel taxes are much, much higher in the UK.

  4. -3-Marlowe Johnson

    Thanks ... the comparison here is the price at the pump, which includes taxes.

  5. I don't accept that it's a desired outcome. Or at least that I wouldn't put it above a lot of other outcomes, like the economy growing, higher employement rates, etc. I'd gladly trade a point of gasoline intensity for a point of increased employment.

    I will say that I've been mostly working from home for the last few years, so my personal gasoline intensity has gotten quite a bit lower.

  6. It seems that what you're trying to do with these last couple of posts is measure/compare vulnerability to fuel prices. As I've said previously, GDP really isn't a useful metric for these purposes. It makes more sense to compare expenditures as a percentage of disposable income. Taking this approach avoids some of the problems created by differential tax treatments by jurisdiction on fuel that is offset by more or less taxes in other areas. For example, while fuel tax in the UK may be higher, property taxes may (hypothetically) be lower than in the U.S.

    A couple other points. Total vehicle miles travelled (VMT)in the U.S. peaked in 2007 and still hasn't recovered, although I suspect 2012 will surpass it.

    Second, now that the new CAFE standards are kicking in, we will begin to see a gradual convergence of the North American and European vehicle fleets in terms of fuel efficiency. Irrespective of what you think about climate change, this shift can't happen fast enough in my opinion. Oil prices are increasingly being driven by demand from BRIC countries rather than the western world. Thus, the historically benign relationship between the natural business cycle and fuel prices no longer holds.

  7. -5-Matt

    Thanks .... actually, the causality may work the opposite way, e.g.:


  8. -6-Marlowe Johnson

    Thanks ... a quick response:

    "It seems that what you're trying to do with these last couple of posts is measure/compare vulnerability to fuel prices."

    No. I am looking at the role of gasoline expenditures in the overall economy. Conventional metrics such as you describe certainly are useful, but not for what I am looking at here. Thanks!

  9. Those who'd like to argue the other side should address the following (a) whether you accept that decreasing gasoline intensity of economic activity is a desired outcome, and if yes, then (b) if pricing is not an appropriate tool to influence this outcome, what you would recommend instead.

    If I understand the question correctly, then a) not especially, and b) n.a.

    I never woke up in the morning thinking 'if only decreasing gasoline intensity of economic activity would happen...' It would be nice if money grew on trees and hot coeds dug middle aged men with receding hairlines, but I don't consider either to be goals.

    I would see a great benefit from new technology that allowed decreasing gasoline intensity of economic activity IF it came about without increased cost. But I see no benefit to taking my mind off the pain in my thumb by dropping a bowling ball on my toe.

    Increased gas taxes would come straight out of my wallet, and would limit my mobility and directly lessen my standard of living. The benefit would be the pleasure of seeing an up-tick on a graph. Thanks, but no thanks.

  10. A) yes decreasing gasoline intensity overall is a good thing IMO BUT it also depends on how it is achieved

    for example considering the make up of the US economy and how those changes have occurred like the financialization phenomenon isn't a good thing

    B) as far as using pricing as a tool I would argue that using it is fine as long as it is coupled with policy which allows for reasonable alternatives rather than subsidizing more failure or dead ends which is what the US and many other nations are attempting

    for example why continue to subsidize auto's and roads while at the same time discouraging other modes of surface travel which are far more efficient, proven, and require no new technology?

    for some other views on this same topic stuart staniford has done some similar work lately:



  11. -9-Mark B.

    Thanks, but you'll want to check your math. Let's say that your income goes up by 10% and the cost of gasoline goes up by 5%. Is that preferable to the reverse (income 5% and gasoline 10%). That is what is meant by gasoline expenditure intensity.

  12. The UK spends about 100% more than the US on gasoline as a proportion of GDP (in round numbers, the US is at about 0.3% and the UK 0.6%, for the 2010 data)

    Actually that is not true. They spend almost the same. The problem with that claim is the pump price is a combination of the price of gasoline and the price of taxation.

    To use your methodology on the new car market the average price of a rear view mirror is around $20,000 because the sticker price on the new car includes that mirror.


  13. You could actually turn this metric into something like the Economist's Big Mac index for inflation. As there are longer records of gas prices, it could be extended back further.

  14. -12-AAA

    Thanks for your comment, but your logic fails. I can go to an auto parts store and buy a rear-view mirror. I cannot go to the gas station and buy untaxed gasoline.

    The price paid for gasoline includes taxes.

  15. Pricing appears to be a symptom of the following “tools”: saber-rattling about the mid-East generally (and Iran in particular) leading to price speculation; government inertia in granting permits for off shore drilling or for pipelines (with consequent effect of refinery closures); any government influence in spurring the presently rising gasoline exports; and, quantitative easing of the money supply to make imported oil more uneconomical. Recession, unemployment, and over-regulation of the environment and business credit market also contribute to decreasing gasoline intensity of economic activity.

    In a CNBC interview about gasoline at $4 per gallon as the June, 2008 national average, then Presidential candidate Obama responded, “I think that I would have preferred a gradual adjustment. The fact that this is such a shock to American pocketbooks is not a good thing. But if we take some steps right now to help people make the adjustment, first of all by putting more money into their pockets, but also by encouraging the market to adapt to these new circumstances more quickly, particularly US automakers, then I think ultimately, we can come out of this stronger and have a more efficient energy policy than we do right now.” During a declining economy and declining demand for gasoline in late 2008, then future Energy Secretary Chu said, “Somehow we have to figure out how to boost the price of gasoline to the levels in Europe.” (otherwise, efforts to promote energy alternatives and more efficient automobiles would be competitively undermined). Thus, despite new discoveries of and innovations in production of oil, it appears the “tools” have been applied and the “gradual adjustment” in price is at hand.

  16. But Roger how is your approach useful when it doesn't factor in the role that excise taxes play in the overall tax burden of the jurisdiction you're looking at?

    Put slightly differently, what does a fuel expenditure to GDP ratio tell you that a fuel expenditure to disposable income does not?

  17. -16-Marlowe Johnson

    Thanks, but I am not sure I understand what you are asking. So please excuse if the following is not responsive.

    1. Do we spend $$ on gasoline as part of our economic activity? YES
    2. Would we be richer if we could accomplish the same level of economic activity while spending less on gasoline? YES
    3. Do people generally want to be economically richer? YES
    4. Thus, it is preferable to see the gasoline expenditure intensity of economic activity decline. QED. ;-)

  18. Roger,

    I can't track where you got your stats from re petrol consumption, but are you aggregating petrol and diesel? While commercial vehicles have always been diesel, diesel was a small fraction of the car market in 1991 and is now over 50% (new vehicle sales).

    I assume you're using road transport fuel as a whole? DECC seem to provide all the stats, and diesel has been out-growing petrol over the period in question. Petrol on its own would be unrepresentative for any purpose.

    I only ask because the post specifically says petrol.

  19. -18-Roddy

    Great question. The data comes from DUKES Table 3.1.2:

    In both the US and UK cases I have focused on gasoline (Motor Spirit) and have ignored diesel, jet fuel, other oils and gases and non-energy uses. But you are correct that a focus on "vehicle transport fuel" makes more sense.

    In the UK gasoline as a proportion of petroleum has varied from 0.21 to 0.30. Ultimately, it may be that a better metric here is "petroleum expenditure intensity" but that would require final prices for a much larger basket of oil-based products.

    I have just now calculated a "transport fuel expenditure intensity" by combining MS + DERV and it does change the curve (though the 2001 on decline is still present). I need to track down US diesel data and I will then provide an update. Excellent advice, Thanks!

  20. Wouldn't be better to remove taxes from the data? Since it would be difficult to come up with a pretax price, why not use gallons or liters/unit GDP?

  21. -20-Mike McHenry

    Thanks, I don't see any point in removing taxes, as taxes are part of the price (and indeed, as I've hypothesized, an important causal factor in the trend in intensity).

  22. Roger I still don't understand what information your approach provides that makes it preferable to others.

    if you don't adjust for tax effects then you're getting a distorted view of the impact that fuel expenditures play in an economy. Consider the following hypothetical example. In the U.S., fuel prices are half of what they are in the U.K. because fuel taxes are substantially lower. This results in the U.S. government receiving $100 billion less in excise tax revenue compared to the UK. However, property taxes are substantially higher in the U.S. (to pay for road maintenance, etc.) resulting in an additional $100 billion annually compared to the UK.

    If you only look at the expenditures as a proportion of GDP you get the mistaken impression that on-road transportation is much cheaper in the U.S. than it is in the U.K., when in fact it would be more expensive for U.S. property owners relative to their UK counterparts, since they would in effect be subsidizing motorists who don't pay property tax.

  23. -22-Marlowe Johnson

    Thanks, but I don't understand your comment.

    Imagine if gasoline was free, except for taxes, and the US charged $1.00 per gallon and the UK $2.00 per gallon. Consumers would still be paying for gasoline and the total of that consumption would be some fraction of GDP.

    Fraction of GDP is a very common economic metric. Thanks!

  24. -18-Roddy

    After some investigation, I don't see a simple way to calculate the transport intensity for the US (UK data is more amenable). But may not matter as gasoline is used in the US much more than diesel. I will continue looking. Thx.

  25. Roger 21

    In other energy intensity comparisons you've used quads/GDP. Another note: A substantial part of UK "petrol" is diesel. My recollection is that its taxed differently than gasoline. I don't know if that matters.

  26. -25-Mike McHenry

    Thanks ... yes, diesel does matter in the UK (having grown as a proportion of transport fuel). Plz see exchange in 18, 19 above ... Thx!

  27. Roger - sorry to add a complication! I saw a scatter-plot in the DECC data showing diesel as being below 50% a while ago as % of road transport fuel, and now significantly over 50%, so us Brits have been steadily increasing our gasoline intensity, showing our brilliance, by duplicitously switching to diesel. As has the rest of EU I imagine.

    I don't know the US diesel picture at all, except that no cars are? So it may make much less difference.

    I've been wondering about Road Transport Fuel price elasticity - anecdotally we're driving fewer miles (and I find the traffic less) as a combo of price and recession, but I don't know where to look for RTF by litre sales regressed against pump price.

    Slightly OT, but the regressive tax nature of higher energy prices, a necessary consequence of FiTs and the like, which is coming through in utility bills a lot also bothers me, I wondered who had done a decent study of that for UK, I've seen newspaper articles and political venting only.