Is Global Economic Growth Persistent?

This post is a commentary on University of Utah physicist Tim Garrett's recent paper How persistent is civilization growth? (pdf). Garrett models the global economic system as a physical system which is subject (like any other) to the laws of thermodynamics (see here). In my post Wealth And Energy Consumption Are Inseparable, I discussed a fundamental result in Garrett's work which is of the utmost importance as we consider the future path of global industrial civilization. I'll repeat that result here.

In a prior study (Garrett, 2011), I introduced a simple economic growth model designed to be consistent with general thermodynamic laws. Unlike traditional economic models, civilization is viewed only as a well-mixed global whole with no distinction made between individual nations, economic sectors, labor, or capital investments. At the model core is a hypothesis that the global economy's current rate of primary energy consumption is tied through a constant to a very general representation of its historically accumulated wealth. Observations support this hypothesis, and indicate that the constant's value is λ = 9.7 ± 0.3 milliwatts per 1990 US dollar. It is this link that allows for treatment of seemingly complex economic systems as simple physical systems.

This graph is from Tim's website and illustrates the relationship λ (lambda) between the world's total accumulated wealth (C, the integral) and our ever-accelerating energy consumption rate (a, measured in 10²¹ joules per year). λ = 9.7 ± 0.3 milliwatts per 1990 US dollar. That's how much energy is required to increase the world's economic wealth as measured in 1990 dollars. The growth rate 1.87% for energy consumption is an average for the period 1970-2006. The average growth rate for the total accumulated wealth was 1.82% over that period. Note that this an empirical result and thus stands outside any particular theory or framework, although it falls out of Garrett's hypothesis (thermodynamic model) that some constant like λ must exist.

In Garrett's persistence paper, the relationship between the energy consumption rate and the world's total accumulated global wealth is expressed in the equation and supporting data below.

Based on this constant relationship (over time) between energy consumption and accumulated wealth, Garrett posits that global economic growth is persistent, meaning we can expect it to continue in the future.

Here, the implications of λ being constant for long-range predictability are discussed in greater detail. The main implication is that global civilization has inertia. Eq. 1 [above] shows that the current rate of energy consumption a is intrinsically determined by the entirety of past economic productivity P, which, when adjusted for inflation, yields our current global wealth C. Because the past is unchangeable, civilization will carry its current wealth into the future, and also its associated rate of energy consumption a = λC. Unless there is very rapid decay from some severe external shock, near-term reductions in energy consumption and wealth are physically implausible. They would require civilization to somehow “forget” its past accumulation of wealth C.

In short, and modeled as a purely physical system, Garrett believes that global civilization has "inertia", meaning that we can predict its future behavior, at least in the foreseeable future, which he believes is something less than the global wealth doubling time of approximately 30 years at current growth rates (see his paper for details).

As a strong word of caution, persistence is never something to carry too far. As a guess, the rates provided in Table 2 [above] apply only for timescales significantly less than the wealth doubling time of 30 years. Essentially the future is unknowable, and unforeseen catastrophes or boons cannot be excluded. Equally, exponential growth cannot continue unabated because, sooner or later, civilization must face up to resource depletion or environmental degradation.

... as argued here, one can apply the principle of persistence based on recent trends, provided one is looking at quantities that are highly integrated over space and time. Just as one might make the purely statistical argument that recent trends in globally-averaged surface temperatures will continue into the near future, here it is suggested that near-term growth in economic wealth and energy consumption rates will also persist. The qualification is that the growth will not be super-exponential, as it has been in past decades, but more purely exponential. The forecasted growth rate is about 2.3 % per year.

Only a large internal or external shock would throw global economic growth off its current exponential growth path until resource depletion and environmental degradation gradually do the job . For example, an abrupt, steep, permanent decline in the crude oil supply would be such a shock. Alternatively, a huge, unexpected shift in the climate system (we pass a "tipping point") would also do the trick, as would a sudden collapse in marine ecosystems. So would a large meteor striking the Earth, a viral plague of Biblical proportions, and so on. Absent such a shock, growth will persist at the assumed growth rate (2.3%/year) because only such an extreme event would cause global industrial civilization to somehow "forget" its unalterable past accumulation of wealth in the next 10 to 20 years.

Garrett's view makes sense to me, and I agree with it. However, and crucially, I want to explain how my view of the problem differs from Garrett's. The two views are consistent and compatible, but, rather than model global civilization as a physical system as Garrett does, I prefer to "model" it as a human system. The end result is the same.

Consider this passage from the beginning of Garrett's paper. He is setting up the problem he will comment on later (conclusions above).

Meanwhile, economic scientists consider the evolution of civilization and its emissions to be driven by decisions made by individuals, organizations and governments. The judgement is that human perceptions and behavior control the rate at which civilization consumes fossil energy.

Policy guides sources of primary energy, rates of human reproduction, individual wealth and lifestyles, and how efficiently energy is consumed to produce economic output. Global CO2 emission trajectories are determined by these choices.

Unfortunately, there is an exceptionally broad range of CO2 emission trajectories that is considered to be humanly plausible, and this greatly amplifies the uncertainty in the physics...

I could write a book explaining why Garrett has set up, in my view, a problem which doesn't exist, and if you consider DOTE a strange form of book, I already have. Let me lay it out for you—

Economists may believe that the evolution of civilization and its CO2 emissions are driven by "decisions" made by individuals, organizations and governments, but I don't see any "decisions" being made. All such "decisions" go one way and one way only—toward furthering economic growth.

Therefore, the "judgement" that human perceptions and behavior control the rate at which civilization consumes fossil energy is true in a trivial sense only in so far as further economic growth always requires more energy (a = λC), and fossil fuels are by far the largest source of that energy. Human "judgements" and "decisions" require Free Will, which does not really exist in this case.

Therefore, in the most important sense, not the trivial sense just described, policy does not guide sources of primary energy, policy does not guide rates of human reproduction, and policy does not guide individual wealth and lifestyles in the general case.

The idea that policy does guide these things, that human societies make choices, is one of the central delusions of our species. Striving for growth in populations and the economies which support them is a biological imperative, and will always occur if the means and opportunity exist to achieve it. For a similar view, see Craig Dilworth's Too Smart For Our Own Good.

And that, from the "human" point of view, is the situation as I see it for our ultimately self-destructive species. Note further something that I will discuss tomorrow—this "model" of how humans work is a hypothesis which can be falsified through observations of what humans actually do in the real world, just as Garrett's work establishes empirically that there is an inseparable physical link (λ) between the cumulative wealth human civilization has created over time and the rate of energy consumption required to maintain and grow it.

And with that, I bid you adieu.

Comments

[Ken Barrows]

The next question for me is: what is wealth? Is GDP a good proxy for "wealth?" For the link between wealth and energy consumption, it probably is.

[Ben]

Thanks a lot Dave. You made Garrett's work easy to understand.

[Oliver]

A televised debate between you and Garrett would be an interesting experience. :-)

[Greg Elsdon]

Dave, so far yours is the only reference to Dilworth's Too Smart For Our Own Good I've seen in a blog or any other venue not directly related to selling or reviewing the book.

Very good.

[Gregor Macdonald]

We can conclude therefore that all policy initiatives that aim to reduce energy consumption, including policy proposals typical to climate mitigation, will fail. Only policies which attempt to divert the "biological imperative" to consume energy from different sources can possibly have some hope of success. Finally, and perhaps more importantly, any mix of factors which severely slow industrial growth and its consumption of energy--such as aging populations, high debt levels, and a price revolution (higher) in energy costs--will accomplish far, far, more to reduce exhaust levels (waste) from our economies.
To wit: http://peakfish.com/2012/11/19/us-carbon-dioxide-emissions-from-energy-consumption-million-metric-tons-of-carbon-dioxide-1973-2012/

[TM]

I am confused. Is the claim being made that energy use and GDP are growing at similar rates? In that case, global energy intensity should stay constant. That is not the case. See http://www.economist.com/blogs/dailychart/2011/01/energy_use.

Over the last 40 years, global energy use has increased 115% (or 2% per year, roughly consistent with the data given here) and GNI 239% (3% per year, not consistent with the data above). (Check the Google Public Data Explorer). Of course, GDP or similar measures are inherently questionable. But what other metric is Garrett using? In any case, while the idea of a complete decoupling of economic growth from energy/resource use is probably illusory, the claim that there has to be a close correlation seems to me equally ill-founded. There clearly is a capacity for some dramatic reduction in energy intensity. Western Europe has much lower energy intensity than the US, and the difference is real. China is decreasing its (currently high) energy intensity.

http://www.slideshare.net/amenning/presentations

[TM]

Ok I overlooked that Garrett refers to "the world's total accumulated wealth" as the integral over GDP. If GDP grows at a steady 3% rate, which seems to be roughly the case, then the integral should look as follows:
P=P0 exp(pt), where p=0.03
C=Integral P= C0+P0/p exp(pt)
(Verify that C'==P).
Everything now depends on the integral constant C0. With C0=0, C would be growing at the same exponential rate as P. With a high C0, the growth rate would be smaller. I find the idea of "the world's total accumulated wealth" questionable if not meaningless, even more so than GDP itself. What about depreciation etc.? I suspect that Garrett is after a master theory of everything and I do not think there is good theoretical or empirical support for the idea of such a tight quantitative relation between wealth and energy use. There is a realistic approach between the illusory belief in limitless growth and a deterministic concept of "civilization" as a thermodynamic system.

http://steadystate.org/economic-theology/
http://steadystate.org/eight-fallacies-about-growth/