26 February 2012

How Economically Significant are US Gasoline Prices?

The figure above shows US spending on gasoline as a proportion of GDP for the period 1983 to 2010, with 1983 set to 100. The data comes from the US EIA (gasoline consumption and prices) and the White House (GDP). The figure shows that in 2010 the total economic cost of gasoline as a proportion of GDP was about half what it was in 1983.

Assuming that gasoline consumption is constant in 2011 ans 2012 and prices go to $4.00 per gallon, with 2.5% annual GDP growth then the index increases in 2012 to 66, or just about the same level as in the lead-up to the 1988 elections.

While everyone likes lower priced energy, the current price of gasoline does not appear particularly exceptional in recent US economic context. $4/gallon gasoline is not what it used to be.


  1. gasoline consumption has fallen since then though, and strip out the hedonic adjustments to GDP to get a better picture

    US GDP like many other econ numbers published by the govt are heavily massaged to make things far more benign than they are

  2. I wonder if the question is, how many tens of billions of $ of spending power are taken from consumers by an increase of, say, $1 per gallon? What is the economic impact of that?

  3. -3-Tom G

    Assuming 2010 gasoline consumption, an increase from $3.56 (the 2010 value) to $4.56 results in additional spending on gasoline of about $18 billion.

  4. Roger,

    You'll get a much different (and more useful) picture if you use disposable income rather than GDP.

  5. -4-Marlowe Johnson

    Mark Perry has that here:


  6. Received from Dan by email:


    blogspot claims that it could not verify my bona fides when I tried to leave a comment.

    But those plots at Mark Perry's place include gasoline and a whole lot more.

    "The reduction in real spending on energy over the last six years, and the ongoing reduction in spending on energy as a share of total consumption since 1995, could reflect increasing energy efficiency (appliances, cars, new homes, etc.), reduced driving, and lower natural gas costs. The downward trend in both series would also suggest that higher gasoline prices in 2012 would have less of an impact on consumers than in past years."

    The numbers presented here and at Perry's are massive economy-wide averages. The usual complaint is that increases in gasoline costs, and energy costs in general, will impact lower-income households more heavily. Are there data laying around to support, or not, this objection to using large-scale averages.

    For example, do data show that for counties, say, for which the per-capita income is among the lowest, in states that are basically rural and so might require more fuel for transportation, indicate that the gasoline ( and / or energy ) cost fraction of income is significantly different from the USA-wide averages?


  7. thanks for the pointer Roger. A couple thoughts.

    First, Perry's looking at total energy spending, not gasoline alone. If you look at gasoline (as Levi does here), you'll see that prices in real dollars are high by historical standards.

    Second, using coarse measures like GDP hide important details. I'd suggest that a more useful metric for evaluating the impact of rising gasoline prices would look at the spending levels across the income distribution. Using GDP hides the disproportionate impact that rising energy prices have as income inequality grows.

  8. $4/gallon gasoline is not what it used to be

    Unfortunately, neither is the US economy. It doesn't have the strength to overcome many negative developments.

  9. @ #7

    Add to that the first couple quintiles have gotten larger in the last few years, that happens when 10 million plus people drop out of the labor force while population increases. Wages have stagnated or fallen especially at the low end while cost of living has not. Some of this is a direct result of the attempts of the FED to prop up the banks and world record housing deflation by papering over bad debt resulting in currency worth less especially when priced in energy. The relationship between money and energy is typically very misunderstood which is great for banks but for joe average it's a killer.

    Another graph which would illustrate what is going on is the decrease in energy leverage over the last decade along with a graph of the change in the individual sources of what makes up our aggregate energy usage. There is a price to be paid when you swap cheap high quality crude for oil sands for example.

  10. Looks like we commenters are having similar thoughts. GDP means nothing to me. I want to see a graph of spending on gasoline vs domestic earnings. There's a difference between how a variable affects the economy and how it affects households.

    In some sense, it doesn't matter whether I spend my money on gasoline or Twinkies or prostitutes - in any case the GDP remains the same. But it matters to ME that my property tax bill just went up, and my water and sewer bills just went up, and my electric rate just went up. And now ten more dollars per fill-up means that something has to give.

  11. You missing another element of the equation. Gasoline demand is way down. This is an indicator of miles driven see http://www.eia.gov/oog/info/twip/twip_gasoline.html scroll to the bottom. The average of the US fleet is 20mpg. 1 Million BBL/day = 2.2 million miles not driven/day. This represents economic activity not happening. The point is high gasoline prices have a much larger impact on the economy than just the incremental price.

  12. That is a curious graph. Since government spending translates directly into GDP, the more the government spends the smaller the fraction of GDP is spent on gas.

    If government spending doubled tomorrow, the percent of GDP spend on gas would decline significantly. How that decline has any real world meaning is yet to be explained.

    Since federal spending has increased from about 18% of GDP to 24% in the time frame of that graph.


  13. #9 - “There is a price to be paid when you swap cheap high quality crude for oil sands for example.”

    The price is based on the quality (grade) of the oil delivered regardless of the costs the producer incurs in getting it to a salable product. That Saudi Arabia incurs costs of under $4.00 per barrel while Alberta oil producers incur costs ten times that is immaterial. The buyer pays what the market demands at a given point in time for the grade of available.

    #12. Right on. The privat portion of GDP is smaller and should be considered.

  14. @ SC Mike

    Actually the quality ie leverage has quite a bit to do with the pricing mechanism and it does indeed ripple thru the economy at multiple levels. Like I said, the relationship between money and energy is poorly understood even though it should be quite obvious. Ask yourself this, what is money, what does it ultimately represent?

    The more energy that is used to extract the oil the less extra energy there is left as profit which in turn creates a receding horizons situation. This completely blows away the fallacy that increased prices will always allow for new sources of oil to become economically viable.

  15. I always hate the focus on cost of fuel but I understand it because you can't go anywhere without passing 10 signs telling you the latest gasoline price. It is very visible and I agree it probably better not to look at GDP but percentage of household income. The latest number I saw for gas expenditures for families was ~$4300 annually while the median household income was around $50K per year. So 8.6% of median income is a pretty high expense. Healthcare costs are rarely advertised and often hidden. A household's average cost is on the order of $9-10K in employee contributions to health insurance premiums and out of pocket expenditures while the employer kicks in about an equal amount as part of the employee's compensation package. The employer's contribution to healthcare puts downward pressure on wages. And remember, both employer and employee sends additional money to the healthcare industry via payroll taxes for Medicare.
    We seem to go nuts over the $4300 per year spent on gas while 4-5 times as much comes out a families' earnings to pay for healthcare. If you want to look at the real drag on the economy and employment, look at healthcare. One of the industries that seems to get LESS productive over time and get away with it with the help of governments.

  16. One more thing that is a bit misleading about the graph. The large drop near the end of the graph correlates roughly with a decline in work force participation. Fewer folks were commuting to work in that time frame.

    When you have a graph where a reduced work force make things look better, it is of questionable real world value.

  17. Sean 15

    Health care has the highest job growth in the economy. Hence expenditures have a multiplier effect on GDP. Fuel costs create few jobs and a further drag by decreasing economic activity. See my post #11.

  18. Roger, are there economic models that would tell us the job impacts of taking $18 billion from the US economy?

    Because crude oil prices are now up considerably, prices of distillate (diesel) and of products requiring oil feedstocks would also be higher, but let's leave them alone for now. It would be interesting to know how many jobs would be lost, according to econometric models, from a $1 per gallon increase in cost?

  19. -18-Tom G

    Thanks, here is a quick back of the envelope, plz vary assumptions as you'd like, say a job is $50k (includes benefits/health/etc) then $18B is 360,000 jobs. Apply economy-wide multipler of your choice, say 1.5 and we are in the neighborhood of 500,000 jobs. in an economy of 135 workers that represents an increase in unemployment of 0.4%. YMMV ;-)

  20. When Americans whine about the price of gas it's always good to remind them what it costs in other countries. I just did the sum converting pounds/litre to dollars/US gallon and it's about $8 here. Similar elsewhere in Europe.

  21. 15 Sean

    I always hate the focus on cost of fuel but I understand it because you can't go anywhere without passing 10 signs telling you the latest gasoline price.

    There are psychological price points. My primary commuting vehicle is a motor scooter. My 'normal' fillup is 1.4 gallons. 3.57/gallon or less I can 'fill up' for $5. Whether the price is $2.50 or $3.50 I wouldn't notice. $3.50 or $4.00 I would notice. At $4.00 I have to pay with a $10 bill rather then a $5.00 bill.

    The average 'mid size' car has something like a 16 gallon tank. A $50/bill is no longer enough to pay for a 'fillup'. Now they need to pay with a $100 bill and get some change back.

  22. It's a shame that budget cuts will curtail EIA reporting (see: http://www.greenbiz.com/news/2011/05/04/budget-cuts-hit-eia-agencys-data-collectionefforts). EIA should also now be reporting the rise in exports of US petroleum products and fuel, including the reported projections of gasoline exports (see e.g. "U.S. Nears Milestone: Net Fuel Exporter" at: http://online.wsj.com/article/SB10001424052970203441704577068670488306242.html, and "Gasoline: The New Big US Export" at: http://money.cnn.com/2011/12/05/news/economy/gasoline_export/index.htm).

  23. Mike Henry,

    When health care costs add $5/hour to manufacturing and take another nearly $10/hour of purchasing out of the economy, its growth is at the expense of every other sector in the economy. Most people don't realize that the stimulus contained had 5-10x more money for health care (through Medicaid) than it did for actual infrastructure. It's no wonder healthcare has the highest growth rate, there's always more money being directed its way.

  24. Sean 24

    You still need to balance that against what GDP is generated through health care. An employer sees these costs as part of the total compensation to the employee. If the employer withdrew health benefits the employee would want to be compensated. Any expenditure should be looked at in terms of whether it creates wealth or not - GDP growth. Lets look at a comparison of Germany's health care costs and the USA's. They have have equal life expectancy. Germany spends 11% of GDP on health care and the USA 16%. Lets say the USA reduces health care cost by 6% to equal Germany's. What does the USA do to recoup a loss of 6% of GDP? Contrast this to oil prices which simply take from GDP growth.

  25. Roger, thanks for the back of the envelope, half a million lost jobs estimate for a #$1/gallon increase in gasoline prices. There is likely a pretty wide range in your estimate, but it does explain a link between rising energy prices and non-negligible increases in unemployment.

    FYI, a new NBER paper (Climate Policy and Labor Markets, Deschenes, 2010) examines the relationship between full time employment jobs (FTE, a subset of total employment) and concludes that "The preferred estimates in this paper suggest that in the short-run, an increase in electricity price of 4% would lead to a reduction in aggregate FTE employment of about 460,000 or 0.6%."

    FTE employment is about 77 million, vs. total US workforce of about 135 million. If the 4% were applied to the total US workforce, the 4% increase in electricity prices would suggest an increase in unemployment of about 810,000 workers.

    We can see why people are concerned about increased electricity prices, or increased gasoline prices.

  26. Life expectancy is a worse metric for examining health care spending than GDP is for pretty much any economic activity.

  27. Mike,

    I'm not of the school of though that spending is spending. A capital expense and an operating expense is very different. It's wise to borrow for one but not the other.
    The 6% difference you mention is pretty close to $900 billion. If you distributed it evenly per capita that's $3000 for each person or $6000 for each worker. (I suspect each of those people will have some good ideas how to spend the windfall.) $900 billion is also getting pretty close to the annual national debt. If the government was not taking the bulk of the free capital out of the economy, perhaps it would be easier to finance new businesses or if the government no longer needed to keep increasing the money supply, the price of food and fuel might not be rising so quickly. (There are many that feel the price of oil and gasoline is more a reflection of the devaluation of our currency than a reflection of energy market conditions.)

  28. 23. Brian said...
    Brain - 23
    EIA should also now be reporting the rise in exports of US petroleum products and fuel, including the reported projections of gasoline exports (see e.g. "U.S. Nears Milestone: Net Fuel Exporter

    The EIA does report that - it's right here http://www.eia.gov/dnav/pet/pet_move_wkly_dc_NUS-Z00_mbblpd_w.htm

    Also the article from the WSJ is a bit misleading. The US has for the last couple of months been a net exporter of 'refined' petroleum products. We are still a very big importer of crude oil.