June 20, 2008

Will peak oil force the economy to decline disastrously?

Robert Hirsch is perhaps the most visible advocate of this idea.

He has published several studies. The last one suggests that oil consumption is related to GDP in a 1:1 ratio - in other words, if oil consumption drops by 10%, GDP will as well. Here is what he said recently: "So then if one calculates a range of 2 to 5 percent, some people think the number may be larger, 2 to 5 percent per year increase in oil shortage, one comes up with a rather disastrous indication world GDP will decline by 2 to 5 percent a year in tandem with increasing oil shortages."

Is this realistic?

No. We can see this from economic history: in the US, oil consumption fell by 19% from 1978 to 1983, and yet GDP grew slightly. Similarly, world oil consumption was flat 2004-2008, but GDP growth was quite strong, stronger than for the US (which itself grew 8% 2005-2008, with flat oil consumption). Oil consumption in the US fell much faster in 2008 and 2009 than GDP. Lately, in the 4th quarter of 2009, US oil consumption continued to fall by 1.1% over the previous quarter, while GDP grew by 5.8%.

Hirsch seems to have looked at the relationship between oil and GDP over the last 20 years, noticed that the ratio of oil increase to GDP increase has dropped from the previous 1:1 to roughly 1:2.5 (an analysis which he attributes to the DeutcheBank, but which can be derived straightforwardly from IEA statistics). In other words, in previous decades as the economy grew, oil consumption grew as quickly, while lately less oil has been needed. Hirsch drew the very strange inference that GDP has become more dependent on oil, rather than less.

An important and relevant researcher here is Robert Ayers . We see that he showed that GDP is related to applied energy (exergy), and only very loosely linked to energy, let alone to oil consumption. The research indicates that BTU's only explain 14% of GDP,and that the source of those BTU's can change (coal to oil to wind, for instance). Both energy efficiency and energy intensity can change. Further, oil is only one source of BTU's. Oddly enough, many energy commentators seem to misunderstand Ayre's research, and think that it supports the idea of a strong causal connection between oil consumption and GDP.

US (and world) GDP could grow much more quickly than it's energy consumption (even including electricity). The best example of this is California, which has kept per capita electricity consumption flat over the last 25 years, while growing it's GDP relatively quickly.

Ayres used "exergy services", which are not "very close to BTU parity". Exergy services are work performed. So, for instance, a Prius performs the same work as a similar vehicle with half the MPG, but uses half the BTU's. Strictly speaking, a Prius can perform the same work as a Hummer (transporting people), and use 20% of the BTU's. An EV also does the same work as a Hummer, and uses about 1/3 of the BTU's as the Prius, and 1/15 of the Hummer's...and so on.

Another source for this argument is here: http://www.postpeakliving.com/downloads/Sill-MacroeconomicsOfOilShocks.pdf from the Philadelphia Fed. It concludes that a 10% decline in oil availability would reduce GDP, on a temporary basis, by about a cumulative 2%. This means that GDP growth would be 2% lower than otherwise in very roughly the 2 years following the oil shock, then go back to it's historical growth rate. Interestingly, it finds no impact on inflation.

I would argue that the paper exaggerates the short-term effect, due to an over-emphasis on the 1980 oil shock (which was made much worse by a simultaneous change in Federal Reserve policy under Volcker), and underestimates the long-term effect of a permanent price/supply change due to the cost of imports, but the overall magnitude of the effect would be roughly the same.

----------------------

A reader asked - does this blog post add helpful information on this subject?

This strongly advocates for the idea that peak oil will "crash" the economy. However, it provides very little support. The discussion simply accepts Hirsch's arguments. It discusses both Hirsch's findings as well 4 other reports. Oddly, three of those other reports actually disagree with Hirsch's thesis of a strong causal connection between oil and GDP - like others, the post seems to misunderstand Ayres in the manner that is discussed above.

Here is an example of an unrealistically pessimistic perspective: "oil is peaking or will soon peak. I don't quibble over 2012 or 2020 — either date is a disaster for humanity because we can't get ready in time." source: http://thefraserdomain.typepad.com/energy/2008/02/yergin-climate.html .

On the contrary, a difference in timing of 8 years makes an enormous difference - an additional 8 years would dramatically reduce the cost and disruptions of the transition away from oil, allowing many car owners, for instance, to smoothly replace their old vehicles with ErEV's. This may explain the blog author's perspective - there seems to a confusion about short-term vs. longterm effects: he assumes a large short-term effect predicts an even larger long-term effect, whereas, as shown in the Philadelphia Fed paper, the reverse is the case: a large short-term effect will generally be followed by adaptation which will eliminate continuing impacts on the economy.

In one place he asks a correspondent to show proof that the US economy can survive a 20% reduction in oil supply. Here are two: the Fed paper above, which shows that a 10% reduction would only reduce GDP by 2%, and the 5 year reduction of 19% discussed above, where GDP managed to grow slightly.

One might ask about imports - hasn't the US out-sourced it's high-energy-consumption manufacturing to China?

Germany is a good counter-example to the argument that OECD countries are showing a higher ratio of GDP to energy because of out-sourcing, given that's it generally understood that Germany hasn't out-sourced it's manufacturing.

We can see it easily in energy use vs. GDP graphs. Till around 1970 it was linked very closely, after 1970 not so closely anymore and after 1990 not at all.

A chart of oil vs GDP would be even more dramatic, given Germany's fast declining oil consumption.
-------------------------------
I'm willing to believe that a dramatic decline in energy availability would over-stress our economy, but that would require decline rates much, much higher than we're seeing. For better or worse, we have plenty of coal for the next 30 years; and ASPO projections are for only a 11% decline in all-liquids by 2030.

That's not nearly dramatic enough to cause anything to collapse outside of the very poorest countries.

How do you forsee this transition actually playing out? The DOE has said we need a 20 year head start to mitigate serious consequences of PO. I see almost no evidence of widespread planning for PO and a 20% decline over the next 20 years?

First, that wasn't the DOE, it was a study by Hirsch, paid for by the DOE. The DOE doesn't endorse it's conclusions.

2nd, the study was badly flawed by it's assumption that liquid fuels had to be replaced. It didn't consider EVs at all.

3rd, a transition has indeed begun. It started roughly in 1994, when the Clinton administration started the PNGV program, which led fairly directly to the Prius. Now, a new generation of EVs has been developed, and are coming out in 2010. We could ramp up the Volt and the Leaf to very large numbers in 10 years.

4 comments:

Half Empty said...
This comment has been removed by the author.
Half Empty said...

Ayers concludes: "If economic growth is to continue without proportional increases in fossil fuel consumption, it is vitally important to exploit new ways of generating value added without doing more work. But it is also essential to develop ways of reducing fossil fuel exergy inputs per unit of physical work output (i.e. increasing conversion efficiency)."

He is saying there will be no growth without corresponding increases in fossil fuel consumption unless we find more ways to add value. That certainly implies no growth if fossil fuel production declines.

Ditto the second part of his conclusion: growth would presumably turn negative if the rate of increase of conversion efficiency for fossil energy is not faster than the rate of decline in the availability of fossil fuels. Note that Ayers implies that we are still waiting for some of the ways of reducing fossil fuel exergy inputs per unit of physical work to be developed.

I would keep a close eye on the UK economy, where the price of BTUs, rather than their immediate availability, is likely to be a key factor in the coming deep recession. The price effect, of course, is as much a part of Peak Oil as the long term decline in available energy.

Nick G said...

"He is saying there will be no growth without corresponding increases in fossil fuel consumption unless we find more ways to add value."

Adding value is one part: increasing energy conversion efficiency is the other. I.e., moving from a Hummer to a Prius. I agree, the way he phrases it suggests that both value addition and efficiency are necessary simultaneously, but I think in this context that's a misinterpretation: part of the conclusion of this paper is that the majority of the economy is still related to work and efficiency. Clearly he expects/recommends an increase in very low-energy GDP, but he doesn't see it in the data yet.

"That certainly implies no growth if fossil fuel production declines."

I think the correct interpretation, as I noted above, is the efficiency is the more important of the two, and can stand alone. If everyone were to switch from a 22MPG vehicle to a 44MPG vehicle, GDP would remain the same, but light vehicle fuel consumption would drop in half. That's not to say we couldn't have GDP decline slightly, if these vehicles were cheaper, but quality of life of would remain the same.

"Ditto the second part of his conclusion: growth would presumably turn negative if the rate of increase of conversion efficiency for fossil energy is not faster than the rate of decline in the availability of fossil fuels."

True, but that's a big "if". We saw in the 78-83 period that we could increase the GDP:oil ratio by 4% per year. We should keep in mind that the 78-83 period was also the period during which the Fed decided to wring inflation out of the economy, so we were doing two large jobs at once. It's very likely that we could have done better, otherwise. I would argue that we could do 25% in one year, under a wartime mobilization effort.

"Note that Ayers implies that we are still waiting for some of the ways of reducing fossil fuel exergy inputs per unit of physical work to be developed."

I'm not sure what you mean - could you elaborate? I suspect that you're over-interpreting his words. I think he just meant to say "improved efficiency is vital!". We have plenty of tools at hand (e.g., moving from Tahoe's to Prius's). Of course, we do need to continue to develop plugins, EV's etc, but the technology is here, it just(!) needs to be ramped up.

"I would keep a close eye on the UK economy, where the price of BTUs, rather than their immediate availability, is likely to be a key factor in the coming deep recession. The price effect, of course, is as much a part of Peak Oil as the long term decline in available energy."

I don't think we can separate price rationing from availability. On the other hand, I would argue that the effect on trade balances is much more important than the lack of BTU's. If the BTU's consumed dropped by 20% over 5 years, but no BTU's were imported, I think the effect on the economy would be rather small, no matter what prices did: higher prices would transfer income from one sector to another, some people would do much, much better than others, but the economy on the whole would be OK. On the other hand, if a lot of BTU's are imported at very high prices, that could harm the economy - how much would depend on the efficiency of petrodollar recycling. It might just end up that Abu Dabi owned Big Ben.

Nick G said...

I would add that for the purposes of Ayre's paper the distinction between fossil and non-fossil sources wasn't important. I think he would agree that it would be appropriate to add a 3rd action recommendation for sustainability and elimination of the dependency of GDP on FF's: replacing fossil-fuels with low-CO2 energy sources.