Friday, April 22, 2011

Wow, just Wow.

The above is Table 3.1 of the IMF World Economic Outlook, with added orange boxes.  I've tended to be an inelasticity hawk, but these numbers are really eye-popping.  -0.007 for emerging market oil price inelasticity!   That implies a 10% increase in oil prices produces a negligible 0.07% decrease in consumption.   That's the short term number, but even the 20 year horizon is only -0.035.  Meanwhile, short term global income elasticity is 2/3.  So that 4% economic growth requires  2 1/3% annual increases in oil production to keep prices stable.

Can inelasticity really be this bad?  On a global basis, I've assumed -0.05 (eg here), but they are saying -0.02.  More later, I'm sure...

Update (3pm EST):

It's worth making a couple more points here.
  • Note that these IMF estimates are based on the 1990-2009 interval.  This was a period of generally low oil prices, except for a single price spike from 2005-2008 or so, and very widespread access to credit in developed countries.  
  • Access to credit is now more constrained, forcing some people to economize, rather than being in a position to borrow in order to bid prices of perceived necessities higher.  This will make demand more elastic.
  • Adaptation as a result of the 2005-2008 price spike was very limited (eg the impact on the US fuel economy averages was quite subdued).  Experience from the 1970s was that the really big changes came after the second shock (in 1979/1980), rather than the first (in 1973/74).  It may work the same way now.
  • In particular, if oil prices go high enough, often enough, we would expect that to trigger large scale switching of transportation options to plugin-hybrids, compressed natural gas, etc.  It will also promote more use of smaller cars, scooters, electric bicycles, etc.
For all these reasons, I expect elasticities in the next twenty years to be higher than in the last twenty.

Still, for the short term...

Update 2 (4pm EST):

And here's the elasticities when they extend their period of estimation back to 1965 (ie including the 70s oil shocks):

Still pretty small.  Note also that a lot of the adaptation in the late 70s and 80s was stopping using oil for power generation.  There's not much of that option left, which is one of the reasons things have gotten more inelastic since.


Geoff said...

In poor countries oil is used for buses, trucks, commercial vehicles, etc. Private personal transport use is small. Oil use is non-discretionary and very inelastic. It is however substituteable in the long term,in countries like Turkey and Argentina commercial use of natural gas for commercial transport is very high. With shale gas becoming very cheap and abundant, this will follow in many other countries, including rich ones.

James said...

Shale gas is cheap and abundant for now ... but I wonder how much of the slack in transportation energy would it have to take up for that to change.

I would love to see a Hubbert Linearization for shale gas. If anyone has seen one please feel free to post a link.

Or perhaps the history of shale gas development so far is too short for meaningful numbers to be produced from an HL?

Kenneth D. Worth said...

That's TEOTWAKI talk! Loss of Lybia + loss of 800k Saudi + other OPEC declines (~200k) due to "oversupply" (Kuwait's production numbers are down as well) + global growth = $250/bbl??

It was the Kuwaiti minister who, when asked how high prices could go, replied, "Very, very high..." I guess he wasn't kidding.

I was thinking that we would pass the old peak this time around, maybe get to $180) but it looks like we are having a confluence of major events (Lybia, big Saudi drop) that could make this second crisis much worse than it otherwise would have been.

In some ways, though, it will be for the good. The world needs $250/bbl to get the message across and the sooner the better. And a big price shock would be a consciousness changer which is frankly, what we need at this point. Otherwise, it will continue as BAU.

Eventually, we'll be able to get Lybia back on line and that will help boost production down the road when we'll really need it to moderate the declines elsewhere.

Stuart Staniford said...

James - HL would definitely be useless at this early a stage (presumably) of development of shale gas. It's not very reliable at all before the peak (and not necessarily 100% reliable after either, since it can't tell you about the possibility of whole new plays opening up).

Stuart Staniford said...

Holy crap: Scenario 2 in Chapter 3 includes a discussion of them doing an actual macroeconomic simulation of a post peak world in which oil production declines 2% a year! I've never seen any important government or international agency do that before:

Another alternative scenario considers the implications of a more pessimistic assumption for the declines in world oil output—3.8 percent rather than 1 percent annually—accompanied by a 4 percent annual increase in real extraction costs per barrel rather than 2 percent (Figure 3.11). This implies that, barring any increase due to the supply response to higher prices, oil production declines by 2 percent annually—a scenario that reflects the concerns of peak oil proponents, who argue that oil supplies have already peaked and will decline rapidly.30 In this scenario, the longer-term output and current account effects are roughly three to four times as large as in the benchmark scenario, meaning they increase roughly in proportion to the size of the shock. Declines in absorption in oil importers are now on the order of 1.25 to 3 percent annually over the period shown, while in oil exporters, domestic absorption increases by more than 6 percent annu- ally. Current account deterioration in oil importers is also much more serious, averaging 6 to 8 percentage points of GDP over the long term.
The most striking aspect of this scenario is, how- ever, that supply reductions of this magnitude would require an increase of more than 200 percent in the oil price on impact and an 800 percent increase over 20 years. Relative price changes of this magnitude would be unprecedented and would likely have nonlinear effects on activity that the model does not adequately capture. Furthermore, the increase in world savings implied by this scenario is so large that several regions could, after the first few years, experi- ence nominal interest rates that approach zero, which could make it difficult to carry out monetary policy.

Translation: your peak oil scenario breaks our simulation model...

Stuart Staniford said...

This Chapter 3 of the IMF WEO is a must read. Basically, it's a serious attempt by conventional macroeconomists to at least come to terms with the arguments put out by peak oil folks, and energy researchers like Ayres. They aren't persuaded, yet, but they are at least taking enough interest to put the kinds of assumptions that someone like me would have about how the world works, into their models, with a view to seeing what happens.

They aren't quite ready to simule the Export Land Model yet, though:

Finally, the simulations do not consider the possibility that some oil exporters might reserve an increasing share of their stagnating or decreasing oil output for domestic use, for example through fuel subsidies, in order to support energy-intensive industries (for example, petrochemicals) and also to forestall domestic unrest. If this were to happen, the amount of oil available to oil importers could shrink much faster than world oil output, with obvious negative consequences for growth in those regions.

Hypnos said...

At a lecture at Imperial College a high level economist from the oil business argued that while something like the ELM was a real possibility, he didn't believe forecasts of Chinese, Indian and MENA demand. He said that rising oil prices would force those countries to lift the subsidies on fuel, and consumption would drop drastically.

Iran, China and India have all moved in the direction of scrapping fuel supplies.

He cited the example of 1985 oil consumption forecasts made in 1979 I think by the IEA, which were off by several million barrels.

Hypnos said...

The wrong forecast I mentioned was not from the IEA but from the OECD: specifically it was in the 1977 OECD WEO. It forecasted demand at over 1000 mtoe for USA and rest of OECD and 400 mtoe for Japan. Actual numbers in 1985 were less than 800 for USA and Rest of OECD and 200 for Japan.

Harry J. Lerwill said...

My impression was that consumption responded to changes in income, not price, in that people will cut everything possible out of their budget in order to be able to drive to work. Lose that income (job) and consumption goes down.

Greg said...

They aren't quite ready to simule the Export Land Model yet, though

Stuart, I think they are simulating the basic ELM. They discuss the growing share of world oil consumption by oil exporters (App 2 p. 113), and (p. 103) say that the model incorporates a physical endowment of oil, which then grows with investment and technology improvement -- or shrinks, in scenario 2.

The graphs are hard to interpret. They are percentage differences in money values, referenced to the Report's baseline forecast (4% p.a. growth globally). That is, the x axis on each graph is the baseline's forecast for GDP (or oil exports). So the bottom figure in the "oil exporters" column in fig 3.11 (p. 107) shows how much greater oil exports will be in money terms, as a percentage of GDP, compared to the baseline forecast.

Now, we don't know (from ch. 3) the baseline's forecast for the price of oil. But if it were constant, year 20's 73% increase in oil exports and 780% increase in oil price would mean the physical volume of oil exported would be only a fifth of the baseline's. That matches ELM.

The statement you excerpted is about an explicit policy change by exporters to try to get more added (domestic political) value out of their oil, which would be over and above the normal action of ELM.

Stuart Staniford said...


I take the distinguishing feature of the ELM to be that the growth of oil exporters domestic consumption is utterly insulated from global prices. That's always struck me as an implausible assumption, and I think that's what the IMF is not simulating. Clearly strong growth in the oil exporters will have somewhat of an assumption of their consumption growth, even if their consumers were exposed to global prices.

Greg said...

But Stuart, table 3.2 (p. 113) shows the exporters' short term price elasticity to be [-0.028, 0.025] and the long-term to be [-0.368, 0.337] -- i.e., neither elasticity is significantly different from zero.

If they used that, the model is assuming that exporters' own consumption is unaffected by price.

(Table 3.2 is the only one that breaks out separate elasticities for oil exporters.)

Near the end of the main text, in the policy advice section, the chapter talks about the need for exporters to "strengthen the role of price signals" -- i.e., remove consumption subsidies.

Table 3.2 also shows a long-run income elasticity for exporters of +2.75 ([1.246, 4.552]), which is huge. So even if there were 'normal' price elasticity, it would be swamped by this. It still looks to me like they modelled ELM--just not "ELM-plus".

You're probably right that historical elasticities will no longer hold with a large change in price. That's always a problem with marginal methods--they're no use with saltations. The modellers do point this out, as you noted.

Som said...

My understanding is that long-run fuel price elasticities are of the order of -0.7 to -0.8. See

Justin said...

Scenario 3 on p106 is fascinating; that is,

"the third alternative scenario assumes that part of total factor productivity represents technologies that are possible and only remain usable when there is a ready supply of oil."


"The implication is that a negative oil supply shock resembles a negative technology shock."

This takes us into very different correlations between energy and growth in general and oil and growth in particular.

Further, I have one gripe with the statement on p93:

"High-income economies can sustain GDP growth with little if any increase in energy consumption."

Hmm. I suspect this was partly achieved through these countries outsourcing their primary and secondary sectors to developing countries as part of globalisation. Wonder if the statement would still hold if you account for the energy embodied in Chinese imports and so on.

Don said...

Harry said,
"My impression was that consumption responded to changes in income, not price, in that people will cut everything possible out of their budget in order to be able to drive to work."

Stuart et al, I've heard the argument made that people will and are defaulting on their mortgage payments in order to keep BAU. Not making those payments frees up a lot of cash for other goods, including gasoline. The combined effect of this process is a stronger economy than one would expect and much higher inelasticity for crude prices. Maybe we are seeing evidence of this now?

Perhaps as long as FASB allows banks to mark their mortgage "assets" to fantasy numbers, consumers ability to pay much higher gas prices will remain in force. I realize that your article is global in nature and that my comments are US centric but it would seem the the easiest way to cut crude demand would be to ratchet down demand from the world's largest consumer, the US, but I am not holding my breath for that to happen anytime soon.

As a final note, market "accidents" and "surprises", i.e. large fast price moves, almost always happen in the direction of the price trend...

Don said...

First KSA, now UAE...thank goodness demand is crashing...Oh look, there's the Easter Bunny!

"UAE producer cuts Murban crude supply by 5% in June"

Alexander Ac said...

Regarding the IMF and their ignorance of debt, here is

Diminishing marginal productivity of debt in the US economy

At some point the debt is NOT an option, and that point is right now...



jdl75 said...

The oil demand inelasticity is a myth, next price spike will again show that :
1) short term oil demand is very elastic, this is done through the process known as "recession" or "demand destruction"

2) long term oil demand can be made elastic, as the Europe vs the US per capita consumption show, this is done through high volume based tax on fuels.

2) is the only solution to avoid or mitigate 1), but for the US, as it is totally opposed to 2), going to 1) full speed is indeed coming soon

Som said...

Here's a good discussion of oil price elasticities:

The bottom line: the IMF's estimates are certainly wrong. They're too low, by far.