24 February 2011

Economists: Riddle Me This

Writing in yesterday's Financial Times, James Mackintosh says that,
The rough rule of thumb economists use is that a 10 per cent rise in the oil price equals half a percentage point off global growth.
I'm no economist, so I'm hoping that one will show up and explain to me how to reconcile this statement with oft-made claims that increasing the costs of energy (e.g., through a high carbon tax) will boost GDP, or less optimistically, reduce it imperceptibly.  I am particularly interested in pointers to the academic literature.  Thanks all!

And a note to the FT:  How about enabling video embed to your excellent videos like this one?


  1. Roger,

    Here's a thought:

    Increases in the price of oil (through normal supply/demand dynamics) enable this sort of thing:


    Carbon taxes can (theoretically) be used to offset other taxes that impede economic growth (e.g. payroll, corporate, income, etc)...

  2. Rules of thumb (including thumb sucking) are not necessarily good, especially for projections about the future. Remember that in 2002-2008 oil increased its price incredibly, but economic growth was very strong. On the other hand, cheaper oil in 1998-2002 did not stop the world's economy from decelerating, including recessions in several important countries like the US, Brazil, Russia and others.

  3. I haven't heard this "rule of thumb". There are two reasons why carbon pricing would have less effect. One is what Marlowe says above - carbon taxes are recycled within the economy whereas a rise in the oil price on imported oil is not. Second, there is a big difference between the effect in the short-run and the long-run. In the long-run new investments are made which adapt to the higher prices. For example, as the price of gasoline rises the new cars sold will be more economical. But it takes several years for the car fleet to adapt to the new higher price.

  4. -3-David Stern

    Thanks. I suppose your #1 depends upon what is done with relevant sovereign wealth funds of the relevant offshore energy company (state owned). Your #2 seems to imply a GDP dip until that adaptation occurs.

  5. Roger,
    A rise in the oil price has two sorts of effects:first it encourages people to use less oil in a variety of ways; but it also takes money out of the pocket of oil consumers and puts it in the pockets of oil producers. Now if oil producers spend proportionately less than oil consumers total world demand will fall. Those of us of a Keynesian bent expect that if nothing were done this fall in demand would reduce growth in the short to medium term. People with more faith in the efficacy of markets assume that there will be other reactions which will offset this effect. But even we Keynesians think that suitable macro-economic policy can offset the fall in demand.
    In contrast to this a revenue-neutral carbon tax only has the first sort of effect (provided that there is not too much difference in the saving propensities of the groups who gain and those who lose from the tax). I should add that this is very old fashioned analysis which may not go down at all well with North American academic economists, but is basically what most policy wonk type economist use in practice. A wonderfully clear expostion of this traditional view about the effects of oil price rises can be found in "Inflation, Exchange Rates and the World Economy" by Max Corden, see here

  6. This is looking very much like the broken window fallacy. The idea that raising the price of oil by whatever means doesn't make a difference because the money is recycled through the economy seems counter-intuitive to say the least. The price of everything depends on the price of energy. In the absence of general inflation, which will create winners and losers too, higher prices means less goods sold with all the consequences of that for employment and standard of living. But general inflation won't protect you from a real price increase. It just means the nominal increase is faster. That road leads to hyperinflation a la post WWI Germany or a bubble which will eventually burst. A burst bubble results in wealth destruction. You can make the case that it was, in fact, oil at $140+/bbl that burst the recent bubble.

  7. Incentives matter.

    If the price of oil goes up the only option is to reduce consumption or reduce spending on things.

    If taxes on oil increase the option of lobbying politicians for special treatment (i.e exemptions or kickbacks (aka tax reductions)) is available. In many cases, lobbying is cheaper than reducing consumption.

    Of course, kickbacks gained via lobbying completely undermine policy objectives.

  8. Freely agreed upon exchanges increase human welfare. Both parties are better off. Forced exchanges do not increase human welfare. Confusing the two by focusing on the fact that money moves in both examples is the classic Keynesian error -- confusing flows and stocks.

    And the idea expressed in the comments that diminishing living standards today will somehow result in enhanced living standards somewhere down the road because government will enact other wise policies is an exercise in wishful thinking that ignores the many lessons of history. Must be part of that hope and change thing.

  9. I am no economist but I'm going to take a swing at this one. You'll forgive me if my views seem a little naïve or simplistic.

    All taxes added to the cost of oil are recycled into (renewable/alternative) energy infrastructure as opposed to corporate profits. This energy infrastructure reduces the need for foreign oil and subsequent drops in demand should result in proportional drops in price (provided there isn’t an increase in demand elsewhere in the system, i.e. China). Products shift from oil/gas to electric fuel sources driving innovation and sales.

    With greater investment in alternatives/renewables comes greater investment in R&D also. This should drive the prices down to where it is competitive enough to stand on its own without the tax. A continued (although reduced) “tax and provide” policy from oil to alternatives/renewables ensures greater energy independence moving forward.

    I like to think of it in terms of rent vs own. Consumption of foreign fossil fuels represents paying rent. Building renewable/alternative domestic infrastructure is like paying a mortgage. Rent is cheaper in the short term but when the mortgage is paid we are much better off in the long run.

  10. Roger, you asked for something authoritative. This is the best I have been able to find:


    It's about 10 years old, but it deals directly with the relationship between energy prices and the economy. Here is an excerpt:

    "Viewed from a long-term perspective, inflation, measured by the rate of change in the consumer price index (CPI), tracks movements in the world oil price. Not only do oil and other energy prices constitute a portion of the actual CPI, but downstream impacts on other commodity prices will have a lagged effect on the CPI inflation. (Figure 4)

    "Looking from the 1970s forward, there are observable, and dramatic changes in GDP growth as the world oil price has undergone dramatic change. The price shocks of 1973-74, the late 1970s/early 1980s, and early 1990's were all followed by recessions, which have then been followed by a rebound in economic growth. The pressure of energy prices on aggregate prices in the economy created adjustment problems for the economy as a whole. (Figure 5)"

    Please follow the link and then click on the embedded link for figure 5. You can see that there is a distinct inverse relationship between inflation and GDP.

    Obviously, inflation is affected by more than just energy prices, so inflation won't precisely mirror the cost of energy. But it's hard to argue that higher energy prices don't increase inflation and therefore reduce GDP.

  11. I imagine that an economist would discuss the income and substitution effects of a price increase in a particular fuel.

    If a carbon price were collected in a revenue-neutral manner, consumers would retain the choice to spend that money as they please. Whether an income effect (consume less of everything) or substitution effect (consumer alternative, cheaper, items) predominates depends upon the consumption alternatives available to the consumer, which can be seen in the demand elasticity for a particular fuel.

    The idea of a technology-neutral carbon price is that carbon emissions reductions (from a particular fuel) would be premised upon that fuel's carbon intensity.

    Oil is not very carbon intensive on a per energy unit basis. I think that I remember reading somewhere that a $100/ton carbon price would result in approximately a 20 cents/gallon price increase in gasoline. Coal, on the otherwise, would be expected to see a substantial price increase.

    Looking at demand elasticities:
    - short-term demand is fairly inelastic for oil, due to the lack of readily available alternatives. An income effect thereby dominates when the price of oil rises.
    - short-term demand is much more elastic for coal, due to the availability of alternatives for energy generation. The claim is that a substitution effect thereby dominates when the price of coal rises.

    An increase in the price of carbon would therefore be expected to have the following effects:
    - Not much effect on the price of oil. An income effect would be felt, but would be small.
    - An effect on the price of coal. An income effect would be felt, but would be smaller than a substitution effect.

    The final claim would be that less carbon-intensive generation industries which gain consumer demand due to the substitution effect are more labor intensive than the industries which close consumer demand due to income or substitution effects.

  12. (While writing this I was thinking that this was a "green jobs" question - I'd like to delete the last portion re: labor intensity.)

  13. Even a revenue-neutral carbon tax would exact substantial transition costs. You can't make people's homes uneconomic without creating a lot of pain. Same for their jobs and businesses.