Arguments
Comprehensive study: carbon taxes won't hamper the economy
Posted on 16 July 2018 by dana1981
Eleven teams participated in a recent Stanford Energy Modeling Forum (EMF) project, examining the economic and environmental impacts of a carbon tax. The studies included “revenue recycling,” in which the funds generated from a carbon tax are returned to taxpayers either through regular household rebate checks (similar to the Citizens’ Climate Lobby [CCL] and Climate Leadership Council [CLC] proposals) or by offsetting income taxes (similar to the approach in British Columbia).
Among the eleven modeling teams the key findings were consistent. First, a carbon tax is effective at reducing carbon pollution, although the structure of the tax (the price and the rate at which it rises) are important. Second, this type of revenue-neutral carbon tax would have a very modest impact on the economy in terms of gross domestic product (GDP). In all likelihood it would slightly slow economic growth, but by an amount that would be more than offset by the benefits of cutting pollution and slowing global warming.
Meanwhile, House Republicans are again on the verge of introducing a Resolution denouncing a carbon tax as “detrimental to American families and businesses, and is not in the best interest of the United States.”
The strong economic case for a carbon tax
The modeling teams examined four carbon tax scenarios, with starting prices of $25 or $50 per ton of carbon dioxide, rising at 1% or 5% per year. These are somewhat modest policy scenarios; CCL proposes a starting tax of $15 per ton rising at $10 per year, and the CLC proposes $40 per ton rising around 4% per year. The most aggressive policy considered by the Stanford EMF teams ($50 per ton rising 5% per year) falls in between these two proposals.
The carbon price each year 2020–2050 in proposals by Citizens’ Climate Lobby (blue), the Climate Leadership Council (red), and the four approaches modeled by the Stanford EMF teams (green). Illustration: Dana Nuccitelli
The modeling studies consistently found that for all four carbon tax policies considered, whether the revenue is returned via rebate checks of by offsetting income taxes, the direct economic impact is minimal:
in every policy scenario, in every model, the U.S. economy continues to grow at or near its long-term average baseline rate, deviating from reference growth by no more than about 0.1% points. We find robust evidence that even the most ambitious carbon tax is consistent with long-term positive economic growth, near baseline rates, not even counting the growth benefits of a less-disrupted climate or lower ambient air pollution
The last sentence is critical. The analyses consistently found that coal power plants would be the biggest losers if a carbon tax were implemented, and the costs associated with health impacts from other pollutants released by burning coal (e.g. soot and mercury) are substantial. Phasing out coal power plants results in significant health and economic benefits to society.
So does slowing global warming, of course. A working paper recently published by the Federal Reserve Bank of Richmond concluded that US economic growth would slow by an extra 0.2–0.5% per year if we stay on our current climate path (3–3.5°C global warming) than if we meet the 2°C Paris target. This compares favorably to a less than 0.1% per year slowing of the US economic growth rate under the carbon tax scenarios.
In short, climate change will slow American economic growth. If we don’t curb global warming, the economic impact will be larger. If we implement a carbon tax to help meet the Paris climate targets, the economic impact will be negligible, and will be offset by the benefits of phasing out dirty coal power plants.
Carbon taxes are effective at cutting pollution
The Stanford EMF studies also consistently concluded that a carbon tax is an effective way to curb carbon pollution, especially in the power sector:
carbon price scenarios lead to significant reductions in CO2 emissions, with the vast majority of the reductions occurring in the electricity sector and disproportionately through reductions in coal … Expected economic costs (not accounting for any of the benefits of GHG and conventional pollutant mitigation), in terms of either GDP or welfare, are modest
The analyses also found that the rate of increase of the carbon tax was more important than the starting price.
Jevon's revenge,
Tim Garrett, the author of the blog you cite, has had a lot of criticsm of his work. Since he also claims that humans cannot affect the climate, in contradiction to what anyone who reads the newspaper can currently see, perhaps his data is off. I note that your graph at 43 is truncated at the year 2000. Where I live it is 2018. A lot of progress has been made in producing more GDP per unit energy since 2000. Scaddenp's graph appears closer to what I would expect to see.
Since he is not mentioned at Desmog Blog he must be a minor character in the Climate Debate.
Jevons Revenge,
Perhaps the extra CO2 is coming from thawing permafrost, methane clathrate decompositon or released by forest fires. There is no reason to believe that nature will continue to absorb all we emit at the same rate.
This SkS article indicates that we may be near, or already past, a point where the climate tips against us and natural releases of CO2 control the temprature no matter what we do.
@michael sweet Thank you for sharing Cullenward et al's criticism! The last time I tried accessing it, all but the abstract was hidden behind a paywall. I look forward to reviewing these criticisms.
Jevons, the energy production has increased and it mostly comes from FF, then CO2 of course continues to rise. The ratio of power to wealth however is about how efficiently we generate wealth from that power. Those graphs show we are getting much more efficient at generating wealth from power. Also, the longer time interval gives a better perpective.
Looking at Garret's graph, I havent tried to reproduce from his sources, but I note that he is plotting power/wealth on a log axis which covers a much larger range than the actual data. In short, a well-known graphing trick.
Frankly, I think my plot is better.
scaddenp This, he claims to be a linear plot of the same data. Is the same trick present here?
LINK
[DB] Shortened URL and activated link. Reduced image width.
Let me clear up the misunderstanding. Cullenward's bogus criticism only betrays that he didn't pay attention while reading Garrett's paper(s). The "Garrett Relation" that "Power/Wealth= Constant" means the following: "The global Current Power Consumption rate is directly proportional to the sum total of all GWP spending over all time (==Wealth)". Cullenward assumed he was saying that GWP is proportional to power. But that's false, it's the SUM TOTAL of ALL GWP over ALL TIME, that is proportional to current power. I am writing a paper with Tim Garrett as co-author which shows that this relation is even better obeyed than Garrett's original paper showed, as there are sublties to how inflation is corrected, how different currencies are calibrated, about the "shadow economy", and a bias in GDP reporting from big countries like China which skew official figures as well. Don't confuse the fundamental "Garrett Relation" with the much less significant CO2 vs GDP graph. That graph Garrett himself down-plays, but it's interesting as a visual proof at how badly we're doing in de-carbonizing. Back to the Garrett Relation; Jevons' Revenge (my term, I hope it catches on, to distinguish from the original "Jevons' Paradox" by Jevons himself, which is more restricted and so not relevant here) says that all improvements in energy efficiency result in MORE total energy being consumed. Because human civilization is driven (genetically?) by the goal to achieve the most rapid, efficient growth possible, and improving energy efficiency aids the expansion of civilization. The point of Jevons' Revenge is that any savings resulting from energy efficiency WILL BE SPENT, and it doesn't matter where, because the Garrett Relation shows that ALL spending sum totalled over all time, remains proportional to current energy consumption rates. It is explicitly in the mathematics of the CThERM model which includes the Garrett Relation and which is very well verified in historical data right up to the present. Now, the carbonization of energy is a parameter in the model which you can tune any way you feel is realistic, but Garrett has made a few projections under the assumption of (A) no decarbonization (that has been the history of the 21st century so far) and (B) decarbonize exponentially with a halving time of 50 years; meaning the CO2 emissions per joule of energy expended, is cut in half every 50 years. Both result, with a wide range of assumptions of the resiliencey of civilizataion to the decays caused by climate change, to increasing CO2 at different rates. If we REALLY got serious, we could do better than t(1/2)=50yrs, but so far, we're doing nothing remotely like this. FF's are still growing at an absolute rate that is about equal to that of renewables, so that the % of global energy from FF's has remained about 83% the entire 21st century so far.
Equally vital to understand, is that this "Garrett Relation" is ONLY applicable to GLOBAL power and global spending. California is meaningless. So is the U.S., So is Europe. Only the total globe is relevant. Trade in materials, energy, money across borders render looking at individual countries or continents meaningless. They are not "closed systems" and don't obey simple thermodynamics. But then, ths same is true of CO2, which is well-mixed and climate too is a completely GLOBAL phenomenon. We NEED to consider both CO2 and economics only in a global context. You'll fool yourself if you try to make generalizations based on regional data. There's a perfect correlation between the fraction amount of our manufacturing which we've outsourced to China and the apparent improvements in our efficiency of GDP per unit of power. Very convenient - China gets the bad rap and the CO2 guilt, and we get the rosy-looking energy graphs as their goods come back to the U.S. See.... LINK
Next, realize that achieving energy efficiency is not new. We've been continually and strongly and consistently improving the efficiency with which energy produces wealth. It is not in conflict with the Garrett Relation (GR). As long as global GDP is growing faster than energy efficiency, there need be no conflict between constantly improving energy efficiency and Power/Wealth=Constant. What's really interesting is that if you look at recessions, when d(GDP/dt is negative, then to be consistent with the GR then energy efficiency trends must reverse and get worse. What's interesting is that the "fake data" from China during their economic contractions shows exactly this. Overstated GDP (for political reasons) makes their local Power/Wealth look better than it is. Even taking the official World Bank figures at face value shows that energy efficiency improvements halt during each of the recessions of the past 30 years over which they give data. Again, those figures are in the pdf of the powerpoints presentation I have on this subject LINK
[DB] Shortened URL and activated hyperlink.
Here's my last comment for now. Even people who sort of accept that improving efficiency may help us grow, don't quite appreciate exactly how close the relationship is. Realize that it is the actual putting into practice the new energy efficiency that CAUSES the spurt in growth and hence the spurt in total energy consumption. We're like the not-too-bright donkey who's owner on his back is dangling a carrot on a stick and string a foot in front of his nose. Every gallop forward only carries the carrot farther in front to make a mockery of his efforts. Every instance of growth encumbers ongoing new power to support that negative entropy growth in the ordered system called Civilization, AND, it also makes us bigger, badder, better at accessing NEW energy reserves and we take full advantage of that. CO2 emissions grew 2% last year, and are expected to grow another 2% this year, and again in '19. Hence accelerates our energy consumption in total. That's what history shows, right up to the present. And for decades now, the fraction of that energy consumed which must be carbon emitting, has not improved. A constant fraction of an exponentially rising total primary energy consumption, is an exponentially rising CO2 emission rate, and an upward arcing atmospheric CO2 - which is exactly what we see right up through the present moment. We're not decarbonizing anywhere near fast enough to change this. And, if we tried, we'd have to work HARDER, and that means spending MORE energy to accomplish that mission, and that means MORE CO2 emissions to get to a day of smaller fraction of CO2 energy sometime in the future. We're between a rock and a very hard place.
I'm not arguing against trying to improve energy efficiency. But it's not our salvation. We've done it forever, and it's only gotten us to this very frightening place. We need improving energy efficiency AND Tibetan-monk level belt-tightening in our consumption at the same time. BOTH. Channeling every dollar not essential for survival, into decarbonizing.
The graph trick applies to the ratio of power to GDP. Use Log to make numbers small and then use axis range much large than range of value to claim something is constant.
scaddenp, it doesn't look like you even noted how flat the ratio is on the log plot. Over a time when GDP more than doubled and Power more than doubled, and we went through big oil price spikes, plunges, recessions, transformations from a US/Europe dominated world to an Asian one... the Power/Wealth ratio stays within a band only 16% wide and that is easily within the (unstated - what's wrong with economists anyway??) error confidence of the World Bank data. Yours is just not a valid complaint here. You want to see the graph with full spending, including the "shadow economy", and a conservative correction for the under-stated inflation? It's all within a band of only 13%, and the first and last year's data is on top of each other. See slide 238 from LINK Sorry, but I don't see how to insert a .jpg into these responses, only URL's.
[DB] Shortened link and activated hyperlink.
And note that my graph is linear, not log.
scaddenp Thank you for taking this seriously, and setting me straight in a calm, professional manner, despite my unabashed arrogance. I feel a freedom I have not felt since I first stumbled upon Garrett's work years ago.
indy222, you insert links and images via the toolbar in the "Insert" tab of the comments editor.
My problem is reconciling the graph I constructed from worldindata (which at its base is not too different in source from that used by Garrett) with Garrett's plot.
Also important from the climate perspective, is that while increasing wealth requires increasing energy use, the energy does not have to come from FF.
I don't know which plot you're referring to. The CO2/GDP plot that JevonsRevenge put in, but which is not fundamental nor relevant for the Garrett Relation? The Garrett Relation plot from my grab of original data from the sources given in my pdf linked earlier, shows it to be quite similar to Garrett's for the data range he had then, and which is even flatter when biases are corrected for, as I describe in my pdf. You might check to see if the worldindata source is using MER (market exchange rates) or PPP (purchasing power parity) accounting to combine countries and get a global GDP. As I argue, Garrett correctly used MER accounting. Some sources may use PPP accounting, but that doesn't properly reflect the value of future civilization network building, key to understanding the energy implications contained in the Garrett Relation. You must use PRIMARY energy (different than some measures, e.g. electricity, other forms of processed energy) and you must use MER accounting. These are the measures which reflect the true cost and the true GDP relevant for energy consumption.
indy222, I will defer happily to you on any matter of economics, but the plot that I dont understand the power to weath relation, where my graph showed decreasing trends not a flat line. I am not reproducing the Garrett plot.
scaddenp, Understand what the Garrett Relation says.
the GR is not: Current GDP is proportional to Current power consumption rate
the GR is not: total summed GDP is proportional to total summed past power consumption
Both of the above are completely false....
The GR is: Today's global power consumption rate (about 18 TW) is directly proportional to the summed total inflation-adjusted Gross World Product over the entire history of civilization.
What my work shows, (so far, only published in my powerpoint I show to my students, linked previously, and includes a linear graph of the GR and components of it) is that this proportionality is even closer if you add back in the "shadow economy" that is missing from GDP figures, and if you attempt to correct for the bias in CPI inflation which, while not numerically well-agreed upon in studies, is at least well-argued and calculated to be higher than official figures (it saves governments COLA money to those massive unfunded liabilities). I've been conservative in using only the MIT Business School's "Billion Prices Project" correction. Garrett's original figures use officially reported GDP deflator values from the World Bank. Pretty safe, but may not do justice to the constancy of the (Current Power)/(total time-integrated GWP + shadow economy spending) ratio's constancy and how well it agrees with theoretical thermodynamic reasoning.Frankly, these corrections to the original ratio are both very small compared to the large changes in civilization wealth, power, FF use, oil prices... And the original GR as plotted by Garrett is still quite flat over time, within data errors.
Again, constantly improving energy efficiency is not inconsistent with the GR constancy. But if the GR is a true property of the human + physics nature of things, and can be relied upon to remain true in the future, it really points to a glaring inconsistency in the (UN politically directed) IPCC SRES scenarios, which basically invent, (whole-cloth!), economic and power trends cobbled simply to get a total specified climate forcing, without any appreciation of the connections between population, energy consumption rates, and GWP. And compound those flaws by then adding in massive atmospheric CO2 removal later in the century in order to produce the UN politically desired "carbon budgets" which clearly make no sense. Economic growth at the rates the IPCC SRES scenarios show are only consistent with MUCH higher atmospheric CO2 concentrations, unrealistic in that such high CO2 will cause such crippling climate change as to (in the words of James Hansen) make the world "ungovernable", and that is NOT an environment of economic growth, but instead rapid decay.
Michael Sweet, here's a link to the rather outrageous (and unfair to an author) treatment he's received on his work on Civilization as a thermodynamic system,
LINK
[DB] Shortened URL and activated hyperlink. You can do this too, using the insert/edit link tool, to avoid long links breaking page formatting.
Thanks for that clarification. However, I do not understand your point about the SRES scenarios. Climate models by nature evaluate "given forcing x, then is the climate you will get". The scenarios are to help evaluate the consequences of particular policy actions. If you do BAU, you get this - widely criticized as being too unlikely but you seem to believe the only one possible? - if we implement reductions on CO2, then we get this. With regard to the GR, nowhere do I see a suggestion that power consumption is reduced in the scenarios- only that we change from FF to other means of generation.
Of course, if want to argue that shouldnt attempt change because history suggests we cant, then that would be a perfectly self-fulfilling prophesy. In 1700, couldnt you make the same argument about the need for horses?
Indy222,
A blog post from Tim Garrott whining about how badly he was treated is hardly proof that he was poorly treated. His paper was published. The reviewers thought his work was crap and wrote opposing views. That is how science works.
Sounds to me like few experts thing Garrotts work is worth considering.
Indy222, If I have understood your statement of the Garrett relationship, then I am still not certain it is saying very much.
Take an exponential series. eg y1=a*EXP(b*t) (ie GDP)
Take y2 as a linear relation of y1 eg y2 = c*y2 + d (Power)
Divide the Integral of y1 wrt t by y2 and the result converges towards constant. Rate of convergence dependent on relative values of b and c.
And no, I havent looked at this rigorously from mathematical viewpoint but it looks suspicious to me.
In your example, when the large majority of the integral has already been accumulated, then the ratio can't change much. However, this is not the case with global GDP. It is HEAVILY weighted towards the present. Most of the sum total of all civilizations GDP since cave-man days, has happened since 1970. So the argument above doesn't carry much weight. It is indeed still remarkable that the ratio is so constant during a period when there has been such radical changes in society, in oil prices, in population, in technology, in everything.... and during a time when most of the total GDP every generated was in fact generated.
To Michael Sweet's disparaging comment. Clearly you have no interest in actually reading the criticism. Otherwise you would notice that the criticism showed the reviewer hadn't even bothered to understand that the Power/Wealth=constant) relationship defines Wealth as an integral over all time, NOT current GDP! That changes everything. Your comment likewise shows no interest in actually understanding this issue, but only band-wagon hopping on invalid criticisms. I too started out thinking Garrett must be a dismissable "apocalypto" when I hadn't actually read and digested his work, given that I heard about it from Uber Apocalypto (of no credibility) Guy McPherson. With "friends" like McPherson, who needs enemies? I think McPherson's hyping (w/o understanding even what a "heat engine" is) Garrett's work simply because it sounds apocalyptic enough to "support" McPherson's agenda, is a real problem for having Garrett's work digested for what is actually says and shows.
To Scaddenp; the point of Garrett's work vs the SRES scenarios, is that the SRES scenarios show fantasic economic growth and no consequences in CO2 harming the economy. No appreciation of what people do with savings, with efficiency gains. I call it "checkers thinking" vs "chess thinking" or "reflexivity" if you like George Soros' formulation of the concept. "Checkers thinking" assumes that if you improve the efficiency of some process, that the savings in energy is a pure gain and is not re-spent elsewhere. That's not what history shows. There's no appreciation in the SRES scenarios of how Draconian must be the constraints on global freedoms and desires in order to actually make happen the curves they simply cobble together as "representative". If we're to remain true to human civilization and human nature as it actually is, then realize that savings will be spent, or worse - leveraged with debt and THEN spent, so we're spending the present and future generations resources to fund what we spend it on. All spending makes "order" out of "disorder" which then must be continually supported with future energy to preserve against the 2nd law of Thermodynamics. Economic growth must be powered and that power must begin 87% as FF energy, and then try to decarbonize from there, but decarbonizing itself will take much energy to accomplish - FF energy. We're in a difficult place, to have to support 7.5 billion who have currently been supported by constantly increasing CO2 emissions, in some other non-CO2 generating way. It's not just me, now, that is recognizing the absurdity of these rosy scenarios. A new entry here is the work of Dunlop, Spratt, and Schellnhuber https://www.commondreams.org/news/2018/08/20/we-are-climbing-rapidly-out-humankinds-safe-zone-new-report-warns-dire-climate
To Scaddenp entry #68. Nowhere does Garrett or I adopt BAU as the inevitable. Again, the renewables fraction is a tunable parameter. So is the resiliance of civilization to the ravages of CO2-linked climate change. What is constrained, is that energy efficiencies permit and enable economic growth and that that growth WILL be engaged in. Neither as individuals nor as nations nor as a globe, do we take savings from our efficiencies and then burn those savings. And all spending results in encumbering future ongoing power consumption. Yes, I too once felt this must contradict the reality of improving energy efficiencies. But I was wrong. It does not. The details are in my .pdf already linked. The Garrett Relation's constancy is consistent with constantly improving energy efficiency as long as the rate of improved efficiency is less than the rate of economic growth. That indeed is what has happened. We put a portion of our economic growth into finding new efficiencies, but we prefer to enjoy a lot of our new-found efficiency-gained cash in other ways as well. And what about in recessions? The math says that improving energy efficiency will be scuttled in favor of just supporting against decay what we already have. And that too is what history shows. While individual country data doesn't really tell a true tale, so many people seem to want to consider the US, that I'll relate the following anyway. Since 1960 the efficiency with which a joule of energy can generate a dollar of GDP has improved a stunning 63%. Has this resulted in a reduction in our power consumption rate? Not even close! Our power consumption has gone UP by 300%. What do we do when we are given new cash from our efficiencies? WE SPEND IT! It's as simple as that, and hard to see how that will change w/o repressive government intervention - which would seem quite unlikely to be tolerated, frankly.
This thread has seen a lot of blather on the subject of Garrett (2011) 'Are there basic physical constraints on future anthropogenic emissions of carbon dioxide?' which has been seemingly shoddily treated by its publishers. Yet, if Garrett (2011) had merit, the author should simply have addressed the criticisms of Cullenward et al (2011) and Scher & Koomey (2011). I do not see evidence of such work. Indeed, the content of Garrett is very poor and should never have been published. In my view it is simple nonsense cloaked by scientific smoke & mirrors with no merit what ever.
While the major conclusion of Garrett (2011) is the GDP/Energy-Use relationship, its title sets out an that it is examining projections of CO2 emissions and proclaims "it looks unlikely that there will be any substantial near-term departure from recently observed acceleration in CO2 emission rates" a message that is evidently wrong. The GDP/dCO2 relationship is far from linear. The level has halved 1985-2017 relative to 1960-85. See here a graph (usually 2 clicks to 'download your attachment') of MLO CO2 increases ploted against World Bank global GDP. And even the GDP/Energy-Use relationship is far less linear than Garrett (2011) asserts, the relationship seeing an 8% drop 1985-2017 relatibe to 1965-85. An 8% drop is way outside the findings of Garrett (2011). So how can Garrett (2011) be anything other than nonsense?
I'm continually puzzled why people insist on mis-representing what Garrett's work shows, despite his patient rebuttal to Cullenward, Koomey, and despite my doing the same here. The graph that MA Roger links to above is GDP vs CO2. The Garrett Relation does NOT make a statement about GDP, but instead the total integrated global GDP over all history. And it does NOT make a statement about CO2, it instead says the current POWER (from any source, renewable or not) is proportional to time-integrated GDP. Stop constructing a phony straw man in order to justify dismissing the work, it only shows a lack of objectivity. It is the GR as stated above which is included in the forward projections curves, and those curves have a wide range of possible civilization resiliance to climate change crippling, and decarbonization rates. The point is that all of them show rising CO2 because we cannot accomplish anything, and for a given assumed growth rate of "Wealth"(=time integrated global GDP), more reslient civilization means HIGHER atmospheric CO2. Those curves are including decarbonization. Note that FF's have remained 87% of our global energy mix since 1973 right up through 2015. You have failed to justify why economists (not thermodynamics experts and not climate experts) Koomey and Cullenward's wrong-headed and dismissive commentary should not itself be dismissed. I've not found any substantive criticism of Garrett's work - only straw-man snipe'ing such as I am now seeing here.
As to Garrett's comment "It looks unlikely that there will be any substantial near-term departure from the recently observed aceleration in CO2 emission rates", there was a flattening in REPORTED emission rates in the '14/'15 area, but (1) China has been caught under-reporting their emissions (look it up!), (2) China goes through 5-year cycles of overbuilding (think "Ghost Cities") and then fallow periods, and you must average over those. (3) Look at the Keeling Curve, updated almost daily, at their website. It is as smooth an exponential as anyone could hope. (4) Even if CO2 emission rates take a meaningful turn downward and stay down as we more dramatically go towards renewables, that does not violate the work. The Garrett Relation is between ENERGY and TIME INTEGRATED GDP, it is not with FOSSIL FUEL ENERGY. So yes, a log/log plot shows how closely atmospheric CO2 has followed GDP, but Garrett himself puts no importance on this, it's really just a statement that since virtually all of our energy has been FF energy, then there's not much surprise at the close correlation. If we make FF's a strongly diminishing part of the energy mix, neither he nor I nor anyone else would be surpised to see that curve start to bend.
Indy222 @76,
Let me add to your puzzlement. It may assist your gaining a resolution.
I am afraid I see no evident rebuttal of the criticisms levelled at Garrett (2011) in the two papers within Climate Change issue 104. You say you yourself provide such a rebuttal but you have have presented a lot in this thread and I have not forensically examined your commenting here, commenting which does need such examination as it is poorly set out. Yet given that, I would not recommend addressing Garrett (2011) in such a manner when it appears to have such basic errors elsewhere.
So addressing what appear to be very basic errors:
What relevant difference do you see between the World GDP data 1960-to-date presented by the World Bank and the GDP data used by Garrett (2011)? I see none. What technical issue have I managed to overlook?
I did point to why CO2 was an important variable in Garrett (2011). Look no further than the title! And you reinforce that importance by criticising "Koomey and Cullenward" for not being "climate experts". The CO2 relationship is certainly is no straw man.
As a sop to your criticism that Garrett (2011) solely considers energy use, I append to the graphic of CO2-rise v World GDP a graph of World primary energy v World GDP (two clicks to download, normally). The trace is evidently not straight but curving down towards the horizontal. The assertion in Garrett (2011) that the Garrett Relationship (Primary Energy = f[GDP]) is linear +/-3% is evidently wrong. Note that the final decade of data graphed slopes just 43% of the initial decade forty years before. That is rather a lot. (And note that within the first paragraph @26 your description of this Garrett Relationship and its employment is entirely garbled.)
Turning to your second paragraph, any difficulties in the assessment of annual CO2 emissions is not an issue here. You will note that I have been graphing MLO CO2, the ESRL data which is indistinguishable from the Keeling curve you recommend.
And while Garrett (2011) does not use CO2-rise in its Garrett Relationship, the whole relevance of that Relationship is set out in terms of rising atmospheric CO2 levels. The CO2/GDP ratio is is dipping down toward the hirizontal. And so too is your much-vaunted GR if it is analysed properly. I see no other conclusion.
Indy222, "However, this is not the case with global GDP. It is HEAVILY weighted towards the present."
Rate of convergence to a constant more about the size proportionality of y2 to y1.
If you have series that is exponential, and divide integral of series by a series that is positively proportional, then it will converge to a constant. As far as I can see the GR is nothing more than a statement of the mathematical structure. It adds no new information. It doesnt matter what errors there are in GDP or power, so long as exponentation and proportionality are maintained then ratio will be constant.
There's no appreciation in the SRES scenarios of how Draconian must be the constraints on global freedoms and desires in order to actually make happen the curves they simply cobble together as "representative".
Huh? Carbon tax or ETS is a draconian contraint on global freedoms and desires? If renewables are cheaper (and take away FF subsidies and they are in many parts of the world), then that will wreck economies to decarbonize? Does China care about "global freedoms" -what exactly do you mean?
In one breath, you are saying that climate change will damage economy, and next, they we will squander everything to maintain global economic growth and that is inevitable. I really dont get it. I dont think you get what SRES are for either. All the meeting minutes are online. I dont see any nefarious UN polical outcomes at play there. They are done to provide policy makers information on the climate that various forcing pathways will lead them to. Do you that there is a forcing pathway that is not within the range considered by SRES?
to MA Rodger: your 1st para: Garrett (and I now too, several times) have addressed why the Cullenword criticism was invalid. Garrett was not allowed to respond (which is pretty outlandish, and against policy, and with no explanation), after those criticisms were sandwiched on either side of his paper, but he responded in the (peer-reviewed) "No Way Out..." paper.
Again, GDP vs Power is only correlated, not a constant ratio. Nevertheless, there is tendency for that curve you plotted to have a shallower slope after the first '70's decade, but it was nearly entirely done early on, and not in recent decades. I address this by looking closer at the data. See http://www.cabrillo.edu/~rnolthenius/Apowers/A7-K43-Garrett.pdf slides 214 - 260. The slight down tilt in Power/Wealth disappears when you correct for: (1) under-estimated inflation (2) over-reported GDP (3) un-included "shadow economy", which makes up a continually smaller fraction of reported GDP and hence total spencing is not growing as fast as reported GDP.... all of which I document with good sources (not apocalypto bloggers, in case you're wondering). Look at the results for Power/Wealth in slide 241. The first point in the blue curve is identical to the most recent, and that is using only the most conservative and likely under-estimated correction to CPI inflation, from the MIT Business School. Instead considering the mid '90's Boskin Commission switchover to changing baskets, the estimate effect would be to remove the slight down tilt since ~'94 entirely. In generating the atmospheric CO2 curves, Garrett does NOT assume GDP is proportional to Power, Improving energy efficiency is not in conflict, not a noise issue, with the GR. As long as GDP is growing faster than energy efficiency, it's quite possible to have Power/Wealth remain constant. Now, you argue about "constant", and neglect that economists don't provide any error bars on their numbers. How "constant" must it be before it is usable as a principle from which to consider the future? You provide no answer, appearing to look for tiny deviations which neglect the closer-look, in order to throw out the entire thesis. That the ratio remains so closely constant, even with the data quality what it is, in a range of only 13% total top to bottom, and much less than 3% in determining the value of the ratio, while oil prices have gyrated wildly by an order of magnitude, while inflation has gone from 0 to 18% and back down, and up and back etc in the U.S., and we've gone from a declining CO2 per joule of energy to a halt in 2001 and rise and stabilize since then, and population rise by more than double, etc etc.... I find that impressive.
To scaddenp: "wealth" = time-integrated real GWP is growing at 2% / year in the 21st century, on average. That's the INTEGRAL growing at 2%/yr. Power consumption is not money. It's a totally different thing - it's the energy consumption rate to support that GDP. That Power/Wealth should be end up being so closely constant over most of the accumulated history of all GWP since pre-history, is making a worthwhile statement. Are you trying to say that we could choose ANY time integral and any non-integral which are both functions of time, and get a constant ratio to the extent seen in the GR? That makes no sense.
On the SRES scenarios, maybe our conflict is in just how realistic are. Garrett makes the observation that population, for example, is not an independent variable, it is a dependent variable - we grow to the maximum extent that we can. Improve energy efficiency, and we can grow FASTER, including population, and that's exactly what we do. It's the very creation of energy efficiencies which ENABLE growth, and then FORCE us to spend more ongoing energy to support that new growth, and enable FASTER exploitation of energy we can find because now we're a bigger industrial entity. Which we must, to support all past growth too against the forces of decay, into the future. It's what we do. It's how we're built. Now, how do you take a world which lives off FF's and convert it to renewables at the pace indicated, without generating far more CO2 than shown in the graphs? FF and CO2 IS our energy source right now, and we're stuck with it. All the efforts we've so far done have doubled power consumption rates and yet CO2 generating FF's comprise an unchanged 87% of primary energy supply since 1973. Could we do better? If we became a different species, perhaps, that CARED enough about the future to lower our FF consumption at the same time that we poured every dollar we could into decarbonizing. If we actually SACRIFICED our current comforts for the sake of future generations. But we don't want to do that, nor will we tolerate a government that tries to tell us that. Instead we install the kind of governments we see around us. But the GR shows that power consumption is closely proportional to Wealth, so all this spending is going to produce a lot of CO2 in the near term.
And too, none of this includes the inadequate climate physics in the "carbon budgets" cobbled by the IPCC at the insistence of the U.N. political representatives, whereby we grow more than we should now, in favor of kicking the can down the road so that later generations have to deal with figuring out how to massively pull CO2 out of the atmosphere and permanently sequester it. What kind of species does that? Listen to the work of Kevin Anderson, of Vaclav Smil, of many other scientists and energy experts and IPCC members complaining of the UN interference in the IPCC process and how the carbon budgets that emerged are just wrong. And the missing physics: The rising ECS with climate state. The permafrost melt from much more dramatic Arctic Ocean ice loss than early models predicted, and resulting CO2 and methane emissions, etc.
[DB] Sloganeering snipped.
The other important point, is that if you look at World Bank data on the efficiency of energy to generate GDP, you'll see Joules/$GWP steadily dropping. But not completely steadily (see my slide 253), it flattens during recessions. I don't have China data or articles going back to those past recessions, but just the near-recession of '15 in China shows the point - that GDP is significantly overstated, and so energy efficiency progress in fact reverses during recessions, as the GR says it must. In human terms - we abandon the luxury of further improving energy efficiency in favor of just trying to support what we already have with what energy we have. That paints a sobering prospect of what would happen in an engineered long term recession as Prof Kevin Anderson has called for. W/o a massive change in who we are as a responsible species, we won't get out of this as cavalierly as so many would like you to believe. Of course, it doesn't have to be an engineered (graceful?) long term recession - climate decay could force it on us unwillingly, as I think it probably will. Climate decay, and our exponentially rising debt finally becoming impossible to support.
Scaddenp at least gets the point that the GR has to do with TIME INTEGRATED GWP. Re-reading once again MA Rodgers, I see no appreciation of this, as he continually refers to power vs GDP(t) or some other GDP, not Integral(GDP) (t=0 to now) apologies for the very awkward quick-o attempt to put an integral in this !
[DB] Please curtail your all-caps usage (shouting is unhelpful).
" Are you trying to say that we could choose ANY time integral and any non-integral which are both functions of time, and get a constant ratio to the extent seen in the GR?"
Not quite. I am saying time integral of any exponential time series divided by time series that is positively proportional to that same series will converge to a constant.
y1 = a.Exp(b.t)
Int(y1) = a.Exp(b.t)/b
y2 = c.y1 + d
Int(y1)/y2 = a.Exp(b.t)/(b.c.y1 + b.d) = a.Exp(b.t)/(b.c.Exp(b.t) + bd)
as Exp(bt) get large, then approximates to a.Exp(b.t)/b.cExp(b.t) = a/bc
Actually you dont need to take integral either to get convergance but it will converge much quicker if you do.
Also, you seem to be insisting that decarbonizing will force a recession. I dont buy it.
scaddenp: Your criticism only works if the exponential has the same constants in the exponent throughout the time of interest. But that is not the case here. The exponential power has varied greatly not only during the times before fossil fuels, but even during just the past century, and even just the past decades. The growth rate of Wealth (hence Power) was moribund until the late 1900's, took another slight hiatus during WWII, and then had a huge sharp change beginning in 1950, then shallowed again as the 21st century got going and continuing till now.
Scaddenp #84: No, I'm not insisting that decarbonizing will force a recession. I'm saying that if the GR remains true, it means that a long term recession, no matter how initiated, will carry with it a trend for energy efficiency to stop improving, but instead to get worse. And I'm saying that when you look at the Federal Reserve's papers on the political biases in the China data (and also I've found in Mongolia data - only ones I've tracked down so far) you indeed see a dramatic over-statement in official GDP growth vs the actual GDP evident from more reliable proxies (again, it's in the slides on my pdf), and this supports that achieving energy efficiencies is either not done, or can't be done, during recessions. This makes trying to halt CO2 growth via an engineered long term recession, as Kevin Anderson has suggested as what's needed, must also carry with it a determination for us to force ourselves to spend what GDP we still achieve towards continued energy efficiences, against our usual tendencies. My hypothesis is that the Garrett Relation is telling us something fundamental about how human nature, and if it is deeply embedded in human nature to behave this way, then simply telling happy stories about energy efficiencies being our salvation - is just not realistic. We grow faster when we get more efficient, and that fights against what's needed - lowering CO2 emissions.
Indy222 , in 2010 Dr Garrett stated that he was "coming from a physics background, and being totally naive in economics ..." (And I myself, being an economics cynic, am likewise in danger of a Dunning-Kruger approach to the Dismal Science.)
In 2014 , Dr Garrett stated that "[his] core finding is that economic wealth or capital is not a static quantity that simply exists, but rather it requires continual energy consumption for its sustenance. Like a living organism, energy is required not just to grow civilization but also to maintain its current size."
Many decades ago, Dr Isaac Asimov & other commentators drew attention to the fact that the so-called advance of civilization has required both an increase in leisure time combined with the "power-multiplying" effect of harnessing draught animals & wind-power & water-mills — later, steam-power, internal combustion & electric motors.
Yet Dr Garrett's mathematical lucubrations seem merely to re-invent the wheel in this regard. So far, so good. But wrong-headedly, he seems to feel the CO2 tail is wagging the dog.
Ultimately, the choice is ours in how we power our civilization. We are not inevitably constrained to follow the current trajectory. The "Garrett Relationship" is largely a pointless exercise.
And we should also debate what is meant by "wealth". But that is a topic for another day.
indy222 @79,
May we park any rebuttal while the more basic criticisms of Garrett (2011) are addressed here?
The finer points of something being either a "constant ratio" or a "correlation" are not the issue. Anything presented as a linear correlation that ignores an evident acceleration is poorly presented and especially so if that acceleration throws the correlation beyond its stated confidence levels.
Your graph of Global Primary Energy on your Slide 241 shows no significant acceleration. Your two traces of World GDP both show a strong acceleration, GDP advancing over three-times more quickly at the end of the 35 years of data relative to the start. (That is a 3% annual acceleration.) Dividing GDP into Primary Energy does not disappear the acceleration. And achieving a better level of linearity for half the data through adjustment of inflation calculations is not the immediate step I would take to illiminate the unwanted acceleration.
Your final comments @79 throw a whole set of %s around, seeming to suggest that there is a lack of linearity in something but it is less than the noise within the input data so the lack of linearity can be dismissed as not relevant. I would insist on a more reasoned description before considering such an argument.
indy222 @82,
Concerning the use or otherwise of "time-integrated real GWP", note that GWP would be taken to mean Global Warming Potential in an AGW context (as you would expect here @SkS). I am unfamiliar with the acronym GWP as used but the expression from Garrett (2011) a = λC would put it (the quantity C) as some measure of Global Present Worth which becomes worthless were civilisation to disappear, and if Global Power Use (a) were to drop to zero, this disappearance would occur as the assertion is that λ is a constant. In Garrett (2011) the quantity C is actually defined in terms of P (Real Economic Annual Production) thus "C is civilization’s historical accumulation of real (inflation-adjusted) economic production of economic value P = dC//dt." Give this relationship, I am not sure why we see a Garrett graphic of Atmospheric CO2 levels against accumulative World GDP in-thread @55 above, the graphic I took for my guide when I plotted Annual dCO2 against Annual World GDP (as linked @77) which would be a more sensitive way of demonstrating the a = λC relationship. Perhaps I should have plotted da/dt = λP to test for the constancy of λ.
to MA Rodger: In my slide 241 showing the Garrett Relation with the refinments to the biases in the data.... you must pause and realize something Mr Rodger. Pause and realize.... One more time - I use the word "wealth" in the same way that Garrett does. Wealth = time integrated GDP. So when I label a curve "Wealth (GDP+shadow)" that means I'm plotting time integrated GDP, not GDP, and together with the spending that is not included in GDP, which Oztunali and many others refer to as the "shadow economy". Yes time-integrated GDP indeed goes up roughtly exponentially. The integral of an upwardly moving function is an even more sharply upwardly moving curve. GDP has, since the discovery of oil, been, on average, an increasing function. Recessions clearly slow that, but only for a year or so before it resumes typically with renewed vigor. And yet once again, neither Garrett nor I are arguing for a fundamental physics unchangable relation between atmospheric CO2 and global GDP. That graph is only possible when FF's are your ~sole source of energy and given the way CO2 source+sinks behave in the environment. I have already said in my ppt's that I expect that curve will shallow when (if!) renewables begin to make a dent in our FF use. How much renewables begin to make a dent is going to depend on a race - a race between when climate chaos begins to seriously degrade civilization in a rapid way (thus perhaps derailing our attention away from renewables conversion in favor of just hanging on to what we have), and when the "S" curve of adoption of renewables somehow kicks in and FF power plants already built become uneconomical to maintain.
As for "GWP" as Global Warming Potential, that's a communication point I hadn't considered. I'll tell you why I go back and forth with using it - I'm writing a chapter for a book on the economics of climate change right now, and being confronted with staying inside a word count (!) and "global GDP" is 2 words, and "GWP" is one word! Yeah. Sorry, I'll try to use "global GDP" here.
You can get lost in the weeds of arguing whether Power/Wealth is truly constant. The data that goes into it has biases, and there are no 1-sigma, 2-sigma or other statistical error estimates given by those nefarious economists to Power or GDP(!). Garrett spends a great deal of time making a case for a constancy based on seeing Civilization as a thermodynamics system. I think it has great insight. I've reframed it slightly differently by using "entropy" whereas Garrett frames more in terms of potential energy flows from higher to lower. Each way hopefully will trigger light bulbs in at least some people. Is the relation PROVEN, in some sort of uncontestable mathematical way, like the Pythagorean Theorem? No - civilization is ruled by human laws as much as by physics, and this is not a closed and perfectly defined and delimited and known logical system like mathematics (at least, not until we reduce biology and the resulting psychology to their quantum mechanical base layer!). Garrett has made a case, and checked to see whether the prediction Power/Wealth=constant holds in the data, and it does. I've looked for flaws in the reasoning and the data, as I started out as a skeptic myself. Instead of finding that the relationship is only a product of flawed data, I find that it is flawed data that makes the small deviations from flatness, and that if we remove those biases the relationship looks even stronger than Garrett thought. Very strong in the historical data. I've not yet seen a case to be made from good reasoning that the GR relationship will be broken. We must not be guilty of "magical thinking", to quote Garrett and to quote a good friend and NAS astrophysicist Sandra Faber. Again, constantly improving energy efficiency is not in either mathematical or logical conflict with Power/Wealth=constant. The complaints here in SkSci seem to be based on misunderstandings in what Garrett is saying. The rest of the reason why Garrett's work has not gotten a wider look may be because - his biggest proponent seems to be the Apostle of Apocalypse: Guy McPherson. As an aside, I've gone out of my way to emphasize criticism of McPherson's clear misunderstanding of Garrett's work and of climate (esp methane) in general and that his belief that all humanity will be extinct in 8 years is ludicrous.... and cruel to those naive enough to buy into his past life as a professor of ecology as sufficient justification to believe him. The last thing that I want, is for McPherson to praise my thoughts or work! I don't need friends like that. The future is grim enough without having the true situation dismissed baby/bathwater along with bogus NearTermHumanExtinction. My sympathies to Garrett.
My criticism on the mathematical structure applies if the whole GDP can be a approximated by a single exponential function and if the relation between GDP and energy production is approximately linear. A quick fit of data that I have shows those conditions are met although linearity of GDP to Energy starting to change.
If you are willing to share the raw GDP and Power data on which are working (with whatever corrections you feel appropriate), I am happy to check the mathematical structure.
I will explain a bit why I think this point is important. Your argumentation has been based on GR holding, and regard this is an immutable "law of universe" (well a reflection of the law of thermodynamics). I am holding that if it is purely byproduct of underlying mathematical structure (and it formulation), then the only information expressed is the exponential nature of GDP and dependence on power. To use this information as a guide to the future, then need to consider whether GDP as an exponential is immutable and whether its relation to power is immutable.
Dr. Garrott copied this XKCD in his blog post complaining about how the journal treated him. Someone had suggested that it applied to his paper.
Perhaps if more peer reviewed papers and less blog posts were cited in this discussion we could reach a better understanding of the situation.
Since this is supposed to be a scientific board peer reviewed papers are strongly preferred over blog posts.
indy222 @89.
Yes, the message about the quantity being used as the denominator to calculate the Garrett Relation constant λ being cumulative global GDP has been recieved. I even discovered what your GWP acronym means - 'Gross World Product' (= global GDP) which was useful as the Wikkithing page on 'Gross World Product' gives a table of gGDP (1990 prices) from 2014 back to one million BC allowing a calculation of Cumulative gGDP. It has a slower accumlation than that shown by the graphic in Frame 241 of your presentation but the general scheme of things is apparent.
And while the cumulative trace sits higher up the graph because it is a "more sharply upwardly moving curve", the acceleration of an exponential function remains proportionately in tact. So Cumulatiive gGDP (C) will be accelerating at 3% per annum and as Global Energy Use (a) is showing no signs of acceleration but is instead rising in a linear fashion, the Garrett Relation constant λ cannot be constant for a = λC to hold. And that has far more significance that a "lost in the weeds" argument.
Of course, the Garrett hypothesis, in attempting to narrow projected Global Energy Use, it does not entirely rely on λ being constant (rather the slowness of that acceleration would suffice). But his analysis does need to set out the acceleration if it is to be considered useful.
And beyond that, while the Garrett hypothesis appears an interesting idea, I would question the theoretical basis of it on a number of grounds (which is of course off-topic for this thread).
The data Garrett used is from the usual official sources, and can be found in the "supplements" to his paper "The Long Run Evoluton of the Global Economy: Part 1 The Physical Basis"
I got GDP stats from UN data
with the link indicating it is using Market Exchange Rate (MER) accounting, which is appropriate. Perhaps your source is using PPP (purchasing power parity) conversion accounting? As you saw in my pptx, I argue that MER accounting is the proper accounting as it best includes an estimate of the future network building expected possible in that country. PPP accounting ends up putting a higher weight on poorer countries and since so many of them are growing faster, it ends up giving a spuriously larger global GDP growth. Garrett derived a conversion between global GDP using PPP vs MER accounting, in the link supplement, derived during the ~2 decades of time when they overlap. The much earlier global GDP estimates from distant history are from Maddison's classic work, but those are in PPP accounting and need a (very small) adjustment to MER accounting.
On the XKCD cartoon; I find it rather "rich" that an economist would toss that at Garrett, when the best that economists have come up with in characterizing a relationship is the Cobb-Douglas relation, which is so clearly a mere fitting function with arbitrary exponents and no thought of getting the physical units even right. It's a confession they don't know how to think about the subject. Also, it entirely leaves out the importance of energy - as if energy is merely something lying around free and its availability plays no role. Garrett's made an attempt to find a more defensible theoretical relationship between energy and economic growth. I find it interesting that the GR is so closely obeyed, and at this point a few percent here or there, compared to the massive changes in civilization along the way, appears not to justify a debunking even if perfect data were to still show them. Clearly in a system with at least some sort of human agency, I think it would not be expected that the GR would be obeyed to the last decimal even in a perfect dataset with no errors at all. When I ponder what does it MEAN that the GR is so closely followed? What I am thinking is this - that the human system and human genetic nature is finely evolved to strive valiantly for the maximum possible efficiency in GROWTH. So that even distant past spending counts every bit as much as recent past spending in the total "Wealth" which is so closely proportional to current Power. It's as if we didn't waste any efforts in spending in the past to bootstrap ourselves to today's massive civilization. That the guiding principle is GROWTH UBER ALLES, and that all energy we can lay hands on, as efficiently as possible, will be put to use to make that growth happen as quickly as possible. Clearly it is in each individual's advantage to grow to his maximum potential regardless of consequences to the environment, because environmental degradation, especially climate, is only evident from the sum total of vast others whom that individual does not control. Dumping my CO2 to the atmosphere has NO impact on climate. Only dumping 7 billion people's CO2 does. But dumping MY CO2 to the environment can have a big improvement in my family's financial well being. So of course - I do just that. These thoughts I believe are embodied in the evident close proportionality Power/Wealth.
scaddenp @91,
The data pointed-to @92 attached to Garrett (2014) 'The Long Run Evoluton of the Global Economy: Part 1 The Physical Basis' is non-controversial except the two-millenium record of World Energy Use (the quantity a) presented in the third of the Supporting Information links. This link does give a smooth exponential-type record for Cumulative gGDP which is the sort of thing you see in other presentations, like the graph below from HERE
Where things start looking very controversial is the data for World Energy Use in that third link. It stitches on a smooth exponential-type record for AD1 to AD1969 onto the as-expected 1970-to-date data. World Energy Use is not a smooth exponential for AD1 to AD1969. It remains essentially linear from the present day back to 1950. The graph below (gleaned from the internet, another referenced version Fig1.3 HERE) shows this linearity. Using CO2 emissions to calculate World Energy Use gives the same linearity (as well as reproducing the 1970-to-date record accurately).
I think indy222 is having problems addressing this evident mismatch between the exponential Cumulative gGDP data and the linear (back to 1950) World Energy Use data.
MA Rodgers. Interesting. I havent had time to look at it, and probably wont till next week, but my first issue is actually reproducing the GR from that data. A 5 minute play in Excel in lunchtime isnt the way to do it.