I got a very long comment from John Mashey caught in moderation, so I’ve decided to put it up as a guest post. John makes a number of important points, but doesn’t convince me that oil is essential to economic activity, for reasons I hope to spell out in a reply. In the meantime, readers are invited to chew on this. As always, but particularly for guest posts, civilised and courteous discussion please.
N FACE OF PEAK OIL&GAS, STANDARD ECONOMIC GROWTH RATE FORECASTS MAY BE
WRONG, and such are used in climate change economics, and if they’re wrong,
we have serious issues.
Maybe people can comfort me that neo-classical growth assumptions are right?
I worry along the following chain:
1) The Stern Report used GDP growth rates from the IPCC scenarios.
Basically, people are projecting *per-person* growth rates in range 1-3%
through 2030 … which is more-or-less predicting business-as-usual as it’s
been for decades, more-or-less.
2) The IPCC growth rates come from the usual sources (like World Bank, US
DoE, IEA, etc], which can be found in section 3.2.1.3, p180-181 of IPCC WG
III’s “Climate Change 2007 – Mitigation of Climate Change”:
Click to access ar4-wg3-chapter3.pdf
The 2000 IPCC SRES (p.301) has more-or-less similar assumptions, i.e.,
people will be noticeably richer in 2050, and richer yet in 2100.
The IPCC doesn’t make its own base economic forecasts, which I confirmed by
talking to Bert Metz, Co-Chair of iPCC WG III when he was here a few months
ago.
Stern of course had to use the 2000/2001 version, which among other things
(Fig RS.9) pegs oil to stick under US$20 J
Hence:
Std economics -> IPCC -> Stern, 1-3% indefinite growth, roughly
BUT WHAT IF THIS IS WRONG?
3) We will certainly hit Peak Oil within the next decade, and Peak Gas
within another two after that. Why does this matter? Neoclassical
economics doesn’t *seem* to think energy is very important, but for various
reasons[** below], that just doesn’t feel right to me.
I’ve been studying work by Charlie Hall at SUNY, and Robert Ayres+Benjamin
Warr at INSEAD, and Vaclav Smil.
4) These folks’ models make sense to me, and they are very, very scary, just
on energy alone. These folks think that the biggest single contributor to
economic growth is energy (or really, work = energy * efficiency). They
argue that energy is way more important to GDP growth than the 5% quoted
here and commonly elsewhere, and Ayres&Warr have done a lot to quantify how
much, i.e., getting rid of the “Solow residual”.
Most of our energy comes from fossil fuels [see Hall Balloon Chart below],
and fossil fuels are headed into Peak Oil and Gas. If we burn more coal to
make up for it, we almost guarantee melting Greenland and other bad things
within a few centuries, leaving later people to fend for themselves,
building dikes and sea-walls … with no petroleum. No amount of Internet
bandwidth or cheap Terabyte iPods will compensate for not having diesel fuel
when you need it, i.e., such goods are not substitutable.
See Kharecha & Hansen reference: we don’t run out of fossil fuels “fast
enough” to be safe.
See slide 46 of the Ayres PPT presentation below: with Peak Oil & Gas here
or coming soon, we will have a massive issue increasing efficiency and
building windmills and solar like crazy, just to keep the total GDP *flat*
in the US over the next 50 years, much less increase it at a few
percent/year/person. These are *different* predictions, of course, than
the standard.
The US DoE 2005 Hirsch Report (below) predicted a serious downturn if the US
didn’t go all out on efficiency 20 years before Peak Oil [we didn’t.].
I’ve visited Australia a dozen times, and you seem to have many similarities
with the US.. (and particularly with California, i.e., you and we both have
water issues, and here in CA, 20% of our electricity already drives water
pumps . and part of why CA has a lot of aggressive energy policies and sues
the Federal government so foten.)
Anyway, IF Hall & Ayres&Warr (and some others) have better approximations to
reality, all of us who use a lot of oil&gas had better be *investing* it in
more efficient buildings, more sustainable infrastructure, more efficient
vehicle fleets, lots of solar CSP, PV, windmills. We ought to be
considering any public infrastructure investments in the light of much more
expensive petroleum.
But, maybe there’s some flaw in what they’re saying? So, what does this
econ-savvy audience say? Suppose they’re right? What does that mean for
Oz? (and see the Rubin/Tal piece on implications of higher oil prices for
global trade.)
[Neither a Deep Green nor Deep Brown, and not an economist]
SOME REFERENCES,
Robert Ayres and Benjamin Warr [INSEAD]
*Ayres&Warr, “Accounting for Growth: the Role of Physical Work”,
Click to access Ayres-paper1.pdf
*Ayres, PPT presentation [see p.46, especially].
http://www.bren.ucsb.edu/news/documents/robert_ayres.PPT
Ayres,”Lecture 5: Economic Growth (and Cheap Oil)
Click to access ASPO2005_Ayres.pdf
Charles A. S. Hall [SUNY]
*Charlie Hall’s Balloon Chart of EROI and energy sources
http://scitizen.com/screens/blogPage/viewBlog/sw_viewBlog.php?idTheme=14&idC
ontribution=1305
OR
http://www.theoildrum.com/node/3786
Renewables have a *long* way to go to replace fossils.
*Hall, et al “The Need to Reintegrate the Natural Sciences with Economics”,
2001
Click to access Need_to_reintegrate.pdf
http://www.esf.edu/EFB/hall/ home page at SUNY
Hirsch Report for US DoE, 2005:
http://en.wikipedia.org/wiki/Hirsch_report
Pushker Kharecha and James Hansen, Implications of “peak oil” for
atmospheric CO2 and climate
Jeff Rubin and Benjamin Tal, Soaring Oil Prices
Will Make The World Rounder
Vaclav Smil, “Energy at the Crossroads”, MIT Press 2003.
** Why does unimportance of energy not feel right?
1) I grew up on a farm. As farmers go from having no draught animals, to
having such, to having tractors, they get richer, because they command more
energy. The US went from 40% farmers in 1900 to 2% today, but mostly
because of cheap fossil fuels for machinery (and fertilizer), rural
electrification, with some help from plant breeding and scientific farming
improvements. In CA and other places, the main limit to farming is water,
which in our case, we pump around, using a lot of energy to grow tons of
food in deserts.
2) I used to work summer jobs for the US Bureau of Mines (i.e., coal).
3) When I was Chief Scientist at Silicon Graphics, I used to help sell
supercomputers to petroleum geologists, and an old friend of ours is Ron
Oxburgh, who used to be Chairman of Shell:
http://www.davidstrahan.com/blog/?p=40
4) As oil/gas prices go up, more marginal sources become financially
economical, but at some point, regardless of the price of a barrel of oil,
if it takes barrel of oil to get the next one out (i.e., EROI = 1), you’re
done, no matter how high the price is. [There may be some substitutability,
which is why they burn natural gas to extract oil from the Athabasca tar
sands in Alberta.]
John Quiggin 
Ernestine: (hoping part 1 of 2 gets through)
Part 2 of 2.
Now, back to your questions, as best as I can:
a) I have a giant stack of things to read, so I’ll add Arrow-Debreau-Mckenzie to that, thanks much. Prof. Arrow lives a few miles away, but I don’t know him. I know Ayres talks to him occasionally.
I think Figure 2 is a simple version of things like LLNL Flows, the “Exergy flows in the economy” from page 1 of the Ayres PPT, or “Global Exergy Flux, Reservoirs, and Destruction” from Stanford GCEP or
Fossil fuels “renewable but much slower than rate of extraction” – yes, that is half of the problem:
– if we had infinite oil, gas, and coal, we’d burn even more of it, i.e. demand is rising, especially given China and India.
– but, we’ll start to see supply peaking for the first two.
– and for coal and nontraditional oil, if we don’t stop burning unsequestered items there, there will be climate externalities we will not like, and that will be very expensive eventually, especially if this Peak effect squeezes the money available for investment. It’s worth looking at Hall on Oil, where he has a nice chart that shows overall money flows.
Anyway, can you point me at simple Arrow-Debreau scenarios/predictions of future GDPs? I’ll go look.
nuclear: I think the first column is just a gross characterization of fundamental types of energy. The GCEP version shows various specific types, which would fit Column 2 better.
I’ll go study, in the hopes of being able to ask better questions. However, note that carbon taxes and such are indeed considered “internalization of externalities” by environmental economists, but in the Hall / Ayres worldview, the additional concern is the extent to which wealth depends on energy. In this case, carbon taxes are almost more like an enforced-savings plan than an externalities issue.
Thanks again for your comments.
re: #44 Ernestine (I try again)
Thanks, this is the sort of discussion I’m hoping for, to better understand the economic models, what they mean, and what people think they mean.
Part 1 of 2, hopefully enough shorter to get through.
Let me first explain what I meant by “Are they wrong?” as that is an over-simplification. I normally think about any kinds of models in terms of their ability to describe the real world, and that isn’t really “right” or “wrong”, it’s “Is the fit good enough to be useful? is the fit so bad as to be dangerous?”
As far as I can tell, the GDP projections used by the IPCC, and therefore Stern are essentially neoclassical [please correct me if wrong], and the Ayres scenarios are *very* different.
Let me take the REXS model forecast part of US GDP 2000-2050, which is slide 46 of Ayres PPT 2006.
He offers 3 scenarios LOW, MED, HIGH (NOT predictions) in which US GDP grows to a peak, then flattens and falls, depending on the extent to which increases in technical efficiency ameliorate the downdraft from shrinking oil+gas.
Following shows US GDP ~2007 as about 22.5X compared to 1900, compared with real CAGRs or 1-3%. Even in the Ayres HIGH case, 2050 would be less than 1% CAGR. (This is all by eyeball, so don’t beat me up too much.)
CAGR 2007 2010 2020 2030 2040 2050
AYRES
LOW 22.5 22.5 20.0 18.0 12.0 10.0
MED 22.5 24.0 25.0 23.0 19.0 15.0
HIGH 22.5 26.0 30.0 34.0 34.0 32.0
1% 22.5 23.2 25.6 28.3 31.2 34.5
2% 22.5 23.9 29.1 35.5 43.3 52.7
3% 22.5 24.6 33.0 44.4 59.7 80.2
Hence, the scenarios used by IPCC & Stern are *very* different from Ayres & co, like between an airplane that flies, and one that doesn’t… which is why I’m trying to understand what these various models really mean.
RE #48 and #49. I didn’t actually specify a timescale for natural attrition. I was thinking sometime around 2100.
carbonsink, sure, I’m sure sometime in the next 200 years we will have resolved most of the environmental and resource-limit issues we have today. But that’s not much comfort when dealing with the next 50 years.
The earth’s population will reach 9 billion by then, and a considerable portion of them will be aiming for first-world standards of living. Obviously with current technology that’s pretty much impossible. So either technology is going to save us, or there is going to be a lot of very unhappy, and most likely starving, people. And either way, non-human species are generally going to have a rough time of it.
Putting my ‘doomer’ hat on … yes, and unfortunately the latter is far more likely IMO.
… and if you think I’m pessimistic, MontyA @ #28 is a real doomer.
In reply to #47 and 50 (the latter is included in the former), John Mashey,
I am using your section numbering.
1.
1.1. A model in economics which, IMO, is similar to the type of model you are referring to in paragraph 2 (‘goodness of fit and useful’) is the financial accounting model first developed by Luca Paggioli in the late 15th century. http://en.wikipedia.org/wiki/Luca_Pacioli . (At the time and for quite some time thereafter, the model’s prediction in terms of the change in wealth of the owners of ‘capital’ invested in an enterprise could be easily empirically tested by counting the ‘capital’ at the beginning of a period and at the end. ‘Capital’ consisted of physical objects that served as money (eg gold or silver). The model was particularly useful for the then existing forerunners of multinational firms, long distance traders, because the owners of ‘money’ could hold their managers ‘accountable’ – in a fiduciary sense.)
1.2. In paragraph 3 you provide a nice example of models of slightly more complex problems, using an illustration from the aviation industry. I can understand that you would like to know whether ‘Model A’ or ‘Model B’ provides a better fit and the usefulness of such models is self-explanatory for anybody who either flies in planes or is over-flown by planes or buys shares in companies classified as belonging to the aviation industry. However, you do not say how you would decide which model is the better one. I suspect that the computer models, to which you refer, are ‘calibrated’ at one stage. That is, the predictions of the model are compared with empirical observations under controlled conditions (say using a wind-tunnel in the case in question).
1.3. It seems to me you wish to draw a comparison between the importance of distinguishing between Model A and Model B in a specific aeronautical engineering problem with GDP growth (decline) scenarios as found in Stern versus Ayres & Co. But such a comparison would be invalid because the method of validation used in the aeronautical engineering problem could not be used for climate change models.
1.4. It seems to me you wish to use the term ‘neo-classical’ as discriminator for the Stern versus the Ayres scenarios. If this is the case, then I should say that I can’t see the usefulness of such an approach. The term ‘neo-classical’ is not a measurable quantity or some other well defined mathematical object but rather a label given to a sub-set of the literature in economics. I am not convinced that there is agreement within economics on the exact boundaries of this literature. I tried to indicate this in my first comment.
1.5. I should say up-front that I am not interested in discussions of specific climate models involving GDP for several reasons.
a) This topic has been discussed at length on this blog-site.
b) I have no interest in discussions of specific climate change models involving GDP. As far as I am concerned, these models have a role in international policy formation and the exact numerical values produced in the various scenarios are not crucial. I leave detailed discussions to experts in macro-economic model building.
c) In support of my position that the exact numerical values of the said models are not crucial, I’ll give you a ‘real world’ example, from the aviation industry. In 1982, the National Acoustics Laboratory (a government owned organization) published their incredibly thorough job in calibrating an aircraft noise model. This noise model entered the decision making process for expanding an airport in Sydney in the early 1990s, via a complex communications strategy which I call ‘PR-filter’, and involving ‘organisational restructuring’ (splitting one authority into two). It took me a while to discover it within the PR-filter or words. There was (and still is) a lot of resistance on part of residents to the aircraft noise generated by aircraft movements to and from this airport. At the environmental impact statement stage, there were lengthy discussions among various experts, some using secondary material, about technical matters (ie Model A or Model B). However, none of these technical matters turned out to be crucial. The model was mis-applied by those in charge of the Environmental Impact Statement, the Federal Airport Corporation, an organization which was set up in preparation for privatization. The model was calibrated for distances up to about 15 km from the edge of the runways. But this was less than half of the distance of residential areas over-flown by aircraft landing from the north of the airport. I discovered that no noise measurements were taken, no survey was carried out during the EIS stage for areas more than 15 km to the north of the airport and residents were ‘fobbed off’ being told ‘they are outside the aircraft noise affected area’. Furthermore, I discovered the noise model assumes landing takes place over a flat plane while the ‘real world’ topography in Sydney is such that the land rises to the north. This monumental planning error doesn’t go away if the reality is distorted by such vague notions as ‘psychological modifiers’ (eg there are published studies which try to convince people that their reaction to aircraft noise is so ‘bad’ because they are scared of a crash. What nonsense. While aircraft don’t tend to crash onto houses regularly and they don’t crash because people on board are scared, there are records of crashes into residential areas, hence assigning a zero probability to such an event would be irrational in some sense. But the risk of a crash is a separate matter from the aircraft noise pollution.) The consequences (noise pollution) would go away if the planning error would be corrected but this is now ‘too expensive’ for those who now benefit from the error; the owners of the airport.
2.
2.1 The Arrow-Debreu model belongs to a different category of models. I mentioned it in relation to the paper by Hall et al which specifically talks about the need to integrate natural science with economics. As I have mentioned, this is really the area I am interested. It is the methodology of theoretical research in the Arrow-Debreu model which is relevant. There is no prediction model of the type you are interested. I could say that the scenarios in Stern and the scenarios in Ayres, are possible in this general framework. The important point is that the existing empirical data on ‘economic activity’ does not contain all information because markets are incomplete. That is, past time series GDP do not contain costs data for pollution. There is nothing one can do about this; inventing data retrospectively is not a good idea.
2.2 ‘Nuclear’. I’ve checked with people who have appropriate scientific qualifications and, so far, all of them said they would not put ‘nuclear’ in column 1 of Figure 2 in Hall et al but they would put ‘uranium’ and related resources from which nuclear power can be generated in column 2. This also corresponds to the schematic representation in the Stanford GCEP publication you referenced in your reply.
3. “…carbon taxes are almost more like an enforced savings plan than an externalities issue.� As far as I am concerned, it is not a question of being an enforced savings plan or a way to deal with externalities but it is looking at the same thing from more than one angle. Obviously, if oil and gas would have been used up at a slower rate then there would be more left now (lack of saving in the past) and the CO2 problem would be correspondingly smaller (externality). The need for taking externalities into account is now overwhelming – not only physical externalities but also monetary ones (eg the financial system cannot be said to be stable).
I enjoyed the conversation. This is as much as I can contribute. Best of luck with your endeavors.
re: #56 Ernestine
Thanks so much. You have been very helpful and have given me much to chase.
My airplane analogy wasn’t intended to go off into the details of models, climate or otherwise, it was just an analogy to say that if two models (whether computer or otherwise) seem fairly different, in some cases it matter very much whether either of them match the real world.
I’ll see if Charlie Hall’s newer charts move nuclear and ask him about it next time.
Thanks again.
Closing comments due to spam infestation