|Evaluation of the book "Smart
Solutions to Climate Change"
page for "Smart Solutions . . "
of economic models
Copenhagen Consensus on
"Smart Solutions to Climate Change. Comparing costs and benefits", edited by Bjørn Lomborg 2010.
The book title will be abbreviated SSCC in the following.
The book contains 8 main chapters and 13 shorter perspective papers, written by a total of 29 contributors. In the last part of the book, the presented opportunities are ranked by a panel of five prominent economists. Introduction and Conclusion is written by Bjørn Lomborg.
A survey of the chapters may be found here.
Purpose and agenda
The book presents the results of the "Copenhagen Consensus on Climate", a conference which was organized by Bjørn Lomborg in 2009. Lomborg-errors has a page on the conference here.
Officially, the purpose of the conference and the book is to gather information about how we can best cope with the challenge of climate change. The idea is to consider the range of policy responses that we have and to identify what are the most promising solutions - which are most realistic, economic and effective? "It would be morally unconscionable to spend enormous sums of money making a minor difference to long-term global warming and human well-being if we could achieve a lot more impact on the climate . . . with a smaller investment on smarter solutions." (SSCC p. 1). The idea was to use benefit-cost-analysis to find the most effective ways to respond to the problems.
However, one may ask if the real agenda behind the book is the same as the official purpose. In consideration of the agenda that Bjørn Lomborg has been following up to now, there may be a lurking suspicion that the whole thing was set up to lead to a pre-determined conclusion. Before this conference, all Lomborg´s activities have served to downplay those solutions which will interfere with the high profits in the fossil fuel industry, and to point out alternative solutions that will not affect these profits. Therefore, it is natural to study the book with this reasoning in the back of your mind. One may ask: Are there any indications that the set up is used to divert attention away from solutions that will go against the interests of the fossil fuel industry? This question will be the guiding principle in the following treatment.
What parts of the book are likely to be read?
It is no simple task to read the whole book, and few people will have the time, energy and motivation to do that. It seems certain that most people will just read the introduction by Bjørn Lomborg (4½ pages), the chapter on the economists´ ranking (12½ pages) and Bjørn Lomborgs conclusion (1½ page). Those who do just that, will get a clear message: "drastic carbon cuts would be the poorest way to respond to global warming". A global moderate tax on CO2 will be a very poor solution. So the conclusion agrees with the suspected agenda: that solution which goes against the interests of the fossil fuel industry is put in a bad light. On the other hand, that solution - climate engineering - which allows us to go on for some time with business as usual and make no sacrifices at all, is presented as extremely advantageous.
The role of the five economists
In the main part of the book, there are many caveats and assumptions, and many points of criticism in various perspective papers. In general, the five economists have not been careful to consider such caveats and criticisms, and have more or less taken the presented conclusions at face value. It seems for most of them that they have been very willing to accept just that evidence which points in the direction that Lomborg would want the most. Only one of them, Thomas C. Schelling, makes some critical comments which all five should maybe have made, but even he accepts the conclusion that cutting carbon emissions is a poor solution. Although three of the five are Nobel prize winners, they are also persons with subjective judgments, and it seems that Lomborg has gathered a team of persons who could be expected in advance to favour a liberalistic approach which interferes as little as possible with fossil fuel industry profits. One of them, Vernon L. Smith, even says that the need to reduce emissions of greenhouse gases is a distinctly speculative proposition (SSCC p. 388).
There are a number of other Nobel prize winning economists who would have chosen very differently and who would hardly have been willing to support Lomborg´s way of thinking. Thus, four Nobel Laureates in economics have endorsed the Stern Review which reaches very different conclusions (see here). So part of Lomborg´s idea may have been to select the right economists who would guarantee the wanted outcome.
Therefore, you should ask: did the panel of five prominent economists have to draw these conclusions from the evidence presented to them? Would a broader and more neutrally selected panel have drawn the same conclusions? Are there major objections to the presented evidence, and should these objections mean that other conclusions might be just as justified, or even more justified?
One of the first issues to consider is how the evidence is affected by the choice of economic models.
The economic models
To evaluate the impact of climate change on world economy, economists use computer models that integrate climate forecasts and economy forecasts. They are called Integrated Assessment Models. Nearly all analyses presented in the book are based on running such models. It is therefore crucial for all the results whether these models are satisfactory.
A detailed discussion of the models is referred to a special page on economic models on Lomborg-errors. On that page you will see that there is a long list of problems with the use of these models. For instance, the most widely used models, DICE and RICE, are criticized under the following headlines:
- Magnity of damages (they are probably underestimated)
- Catastrophes and tipping points (the risk of catastrophes is built in, but `tipping points´do not exist in the model)
- No dynamic technological change (there is no possibility for a price signal to stimulate development of renewable energy technology)
- No equity (see further below)
- Discount rates (discount rates are set much higher than the growth in the world economy, which profoundly affects the overall results).
Of special importance is that the FUND model shows a much more negative effect of climate mitigation on the growth of the world economy than nearly all other economic models. This is important, because teh conclusion in SSCC about the effects of carbon taxation are based on the FUND model.
Incongruent parameter settings
For the various analyses to be directly comparable, one should demand that crucial parameters were set at the same values. But such a coordination has not been performed.
An example is the parameter climate sensitivity (the temperature rise for a doubling of atmospheric CO2 concentration). This is set at 3° C in the DICE model (SSCC chapter 1) but at 2.5° C in the FUND model (SSCC chapter 2).
The most important parameter, however, is the discount rate (see also here and here on Lomborg-errors). There is no obvious `correct´ value, and when different experts use different values, this has huge consequences for what conclusions they reach. For instance, the discount rate determines whether adaptation or mitigation is the most profitable strategy (SSCC p. 239 and table 6.7). With a 0.1 % discount rate, mitigation is the best strategy; with a 3 % discount rate, adaptation is the best strategy. So those who prefer that we do not interfere with the consumption of fossil fuels and instead just adapt to climate changes, can have their preference verified by seemingly objective analyses just by choosing a high discount rate. Any expert who wants climate change to be a relatively minor problem will choose a high discount rate, whereas any expert who wants climate change to be a very severe problem will choose a low discount rate. Thereby, the apparently very objective economic analyses are in reality extremely subjective.
The modern version of the DICE and RICE models have an endogenously determined discount rate, which depends on the rate of growth of the world economy. This rate starts at about 5.5 % and gradually, over several centuries, drops to about 4.5 %. However, in chapter 7, the DICE model is run with a fixed discount rate of 5 %. In chapter 1, it is run with two start rates of 5.5 % and 2 %, respectively. The FUND model is run with a fixed discount rate of 5 %. In chapter 6, the WITCH model is run with discount rates of 0.1 % and 3 %. Chapter 7 uses discount rates of 1.4 %, 3 %, 4 % and 5 %. Chapter 8 uses discount rates of 3 % and 5 %. Chapter 5 is partially based on a report which used a discount rate of 10 %.
These differing discount rates mean that the different analyses cannot be compared directly. Strangely, the five respected economists seem to be completely unaware of these incongruences.
The economic models contain functions that for a given temperature rise define the size of damages due to increasing temperatures. In all models, such damages increase very gradually with temperature. This is shown in figure 6.2 (SSCC p. 225). Here, the damages predicted by several models are so small that they are not really important. For instance, at a 3° C global temperature rise, damages amount to only 0.5 - 1 percent of the global economic output, at a point of time when the global economy has grown by about a factor of three relative to present. Those damages that are included here are for instance changes in agricultural production and changes in the tourism industry. What is not included, is changes due to weather catastrophes, notwithstanding that the increasing risk of such catastrophes is actually the main reason for concern about climate change. Already now, estimated losses due to catastrophic weather are in the order of 0.5 % of current world GDP, and damages are increasing at a rate of 6 % a year in real terms (SSCC p. 224). But this main concern is simply not included in these models. Needless to say, the risk that we may pass some `tipping points´ beyond which run-away effects bring world climate completely out of balance is not built into these models - even though "the risk of extreme climate change is in fact the main reason why the mainstream of climate change scientists urge fast and strong action to rein in emissions" (SSCC p. 288).
An exception to what was said above is the DICE and RICE models. Here, the risk for catastrophes is built into the model. As a consequence, the estimated damage for a global temperature increase of 3° C is set at 2 % of world GDP, rather than 0.5 - 1 % of GDP as in the other models (figure 6.2 p. 225). Still, this is a very moderate damage, considering that we are talking of a temperature increase clearly above the threshold of 2° C when the risk for dangerous tipping points begins to play a role.
On page 1 in SSCC, Lomborg writes: "To make matters worse, the effects will be felt most in those parts of the world which are home to the poorest people who are least able to protect themselves and who bear the least responsibility for the build-up of greenhouse gases." Thus, he signals sympathy with the poor.
However, in all the analyses in the book, such sympathy is absent. There is no `equity weighting´. This means that costs and benefits are set proportional to the GDP per capita of the region considered. Thus, the loss of a human life or a hectare of land in the rich western world counts more than 10 times as much as the loss of lives or land in poor areas like Bangladesh or Africa south of Sahara. With this parameter setting, it is built into the models that changes in rich countries are much more important than changes in poor countries. And then, when the computer models are run, they show - lo an behold - that it is much more important to pay regard to the needs of the rich countries than the poor countries.
Admittedly, Thomas C. Schelling (SSCC p. 389) points exactly to this problem. But otherwise, there is paid very little regard to the needs of poor countries, and the sympathy shown by Lomborg in his introduction is therefore rather hollow. He could have demanded, for instance, that computer runs were made with equity weighting. He did not.
The question of carbon taxation
Probably the most important issue is whether there should be a tax on emissions of CO2, and if so, how large this tax should be. This requires a detailed discussion, and it has therefore been transferred to a special page on carbon taxation. The reader is requested to consult this.
The central chapter on this issue is chapter 2 written by Richard Tol. Tol´s conclusion is that it does not pay off to place any substantial tax on carbon dioxide emissions. This conclusion is subsequently accepted by all five prominent economists and adopted as a main conclusion in Lomborg´s final chapter. Tol and Lomborg have for a long time worked closely together, and it is quite possible that they have arranged by common agreement that this should be the conclusion.
However, Tol´s conclusion is biased and very weekly founded. It is based on Tol´s own FUND model. When running this model, Tol reaches an estimate of the social cost of carbon (SCC, i.e. the sum of future damages discounted to form a net present value per ton of carbon emitted), which is about $2 per ton of carbon. This is a very low value and actually an extreme outlier relative to all other estimates reported by Tol himself and by others. He gives no documentation for this estimate, and when asked directly about it, he answers that yes, it differs from other estimates, but that is just how the model output is. So obviously, the model, with the parameter values used by Tol here, produces an output which differs greatly from other studies. If we accept the model output at face value, then we must also accept that any tax in excess of $2 per ton of carbon will inflict a net loss and hamper economic growth. This is a strange conclusion, considering that several countries, including Denmark (Lomborg´s home country) and the Netherlands (Tol´s home country), have carbon taxes that are much higher than this, with no adverse effects on economic growth (very carbon intensive industries are exempted more or less from this tax).
Actually, the justification of this estimate does not exist. It does not fit in with other published values, and Tol has been unable (or unwilling) to provide an explanation when asked about it. In spite of this, when the five prominent economists rank all climate solutions, they accept Tol´s outlier estimates at face value without asking any critical questions, even though it diverts from even the low range of other estimates by an order of magnitude. This means that the overall conclusion of the whole Copenhagen Consensus on Climate is unfounded.
One may ask why the five top economists have not noticed this? How come that they failed to notice the discrepancy between the results of Tol and the results of others?
Another important theme is discussed in the chapters on technological innovation. In chapter 7 it is claimed that a `price signal´ from higher prices on fossil fuels will not be sufficient as an incentive to boost research and development in alternative energy sources. Instead, such research and development should be organized through some type of funding mediated by the government, and the funds should originate from a moderate carbon tax. Other experts claim, on the contrary (e.g. in perspective paper 7.1) that better funding of such research is not enough. To attain even a mild stabilization taget, a strong carbon price signal (that is a considerable tax) is necessary.
When the five prominent economists made their overall conclusion, they did not engage in any critical discussions of the above issues, and they failed to notice that several others argue that to attain our goals, "a strong carbon price signal is indispensable". One may ask: have they been led to ignore it or overlook it?
As a result, all carbon taxes dealt with by Richard Tol are designated as very poor solutions. This has led to official protests from two of the experts contributing to the book, Claudia Kemfert and Roger Pielke (see this web site). They write among many other things: "The two of us disagree about how best to price carbon . . . However, we strongly agree that putting a price on carbon is a necessary component in any portfolio of policies designed to decarbonize the global economy." This criticism was raised well in advance of the editing of SSCC, but has apparently not led to any corrections of the text.
Thus, it seems that Lomborg has succeeded to perform his trick whereby carbon taxation drops out of the portfolio of useful climate mitigation tools. Even though several experts (Bosetti, Popp, Pielke, Kemfert) state that considerable carbon taxation is ´indispensible´ or `necessary´ , and even though they refer to extensive literature confirming this view, Lomborg manages to have the five prominent economists overlook this completely. These economists fail to see the severe bias in Richard Tol´s chapter, and instead accept his conclusions at face value. In that way, carbon taxation is magically removed from the portfolio of useful options. We do not know if this trick has been performed to please the fossil fuel industry, but in any case, they must surely be very pleased.
Whereas carbon taxation was given bottom rank, climate engineering was given top rank. In chapter 1, Bickel and Lane claim that certain technologies, such as marine cloud whitening, can exclude so much sunlight that temperature rises due to greenhouse gases are practically annihilated. This will be quite cheap to do, and the benefit will be enormous. Calculated benefit-cost ratios range from about 2,000 to about 11,000. This compares with all other opportunities to mitigate climate change, which typically have benefit-cost ratios less than 5.
The five prominent economists choose to believe that this may indeed be possible, and therefore they give top rank to marine cloud whitening. Also injection of sulphate aerosols in the stratosphere gets a high rank. The economists choose to overlook and ignore the criticism or caveats that are presented especially in perspective paper 1.2. They do not share the attitude expressed by Galiana and Green on climate engineering proposals (SSCC p. 331) that "one hopes, given uncertainty about effects and effectiveness, they will never have to be used."
It should be noted that the proposal is to invest in research and innovation of climate engineering. For instance, concerning marine cloud whitening, it is uncertain if one can actually prooduce the seawater aerosol of the required size and concentration, and it is uncertain how one can disseminate these particles to ensure that sufficient numbers of them enter the clouds to be whitened (SSCC p. 17).
One should not underestimate the magnitude of the effort if one wants to change the planetary albedo as much as desirable (by 3 W/m²). 27.6 % of the world´s ocean area is covered by suitable marine stratiform clouds, and more than 40 percent of these clouds have to be constantly seeded (SSCC p. 40-41). So the nearly 1,900 vessels injecting droplets have to cover about 11 % of the world´s ocean area more or less on a permanent basis. The chapter has no estimate on to what extent this will interfere with global shipping.
Problems with benefit-cost ratios
It is a general idea of Lomborg´s project that all proposals are evaluated by benefit-cost-ratios. In ordering the proposals, the panel of prominent economists was guided predominantly (but not exclusively) by consideration of these ratios.
However, some of the contributors to SSCC are very critical of the calculation of such ratios. In chapter 4, the authors write (p. 143-144): "The uncertainty about these effects is profound, with no objective basis for the assignment of probabilities. Thus calculation of expected benefits is not a scientific possibility and any estimates of such benefits are highly speculative". Frank Jotzo writes in chapter 6.2 (p. 290): "The Copenhagen Consensus exercise places heavy emphasis on BCRs [benefit-cost-ratios]. These ratios come about as a result of highly contestable assumptions . . . Consequently, the estimated BCRs are highly unreliable as a guide for policy." In chapter 7.1 (p. 343), the authors write: "One should not forget that technological change is an uncertain phenomenon. In its most thriving form, ground-breaking innovation is so unpredictable that any attempt to model the uncertain porcesses that govern it is close to impossible."
Even if we choose to accept the calculation of benefit-cost-ratios, there are problems in how to understand them. This is due to the phenomenon of diminishing returns. If we do some effort to mitigate climate change, the benefits will at first be larger than the costs of implementation, and therefore we go on and extend our effort. The more we do, the smaller will be the marginal benefit-cost-ratio, and when this marginal ratio reaches the value of 1, we make a halt. Expressed in other terms, we move along a MAC curve (Marginal Abatement Cost), and we stop when we reach the breakeven point where gross costs and benefits are the same, and the net benefit is zero. When we optimize the extent of our effort in this way, the marginal benefit-cost ratio will by defininition be 1.
Some of the analyses presented in SSCC are based on such optimization, or depart from points of breakeven. For instance, dealing with the forestry issue, Sohngen departs from an optimal policy scenario provided by the DICE model, and then adjusts it to account for sequestration of carbon in forest growth. He finds that carbon prices fall substantially relative to the original scenario (by 56 %), but there is little benefit, because the goal (limiting temperature rise to 2° C) is the same, and thus climate damages are the same. Formally, Sohngen reaches a benefit-cost-ratio of 1.74, which is not impressive, but actually the merits of his project are much larger than indicated by this figure. As he states (p. 122): "A reduction in carbon prices by such a large amount would have enormous benefits for society by directly reducing compliance costs, and freeing resources for other productive investments." Thus, the forestry program seems very much worthy of implementation. But the benefit-cost ratio presented is very moderately positive, and so the panel of prominent economists do not rank this program very high (it attains rank 8 out fo 15). To be comparable to other benefit-cost-analyses, the author should have calculated the extra avoidance of climate damages that would have resulted from addition of the forestry program to the already defined basis scenario. He did not do that, and instead the analysis is based on an adjustment of an already performed optimization process. It seems that the economist panel has overlooked this point.
There are other cases. Thus, in the chapter on methane, the authors note that there has never been an application of an economic model to analyse the costs and benefits of any single methane mitigation measure. Instead (p. 180), the authors estimate benefits of avoided damages by utilizing published values of SCC (social cost of carbon). However, such values infer that we optimize the efforts to reach the breakeven point. So the benefit-cost ratios are not directly comparable to those obtained in the ordinary way.
Unfortunately, the panel of economists is not aware of this problem.
The book contains a few examples of free benefits, also called `no-regret options´. This is the case for some of the methane mitigation opportunities, where substantial mitigation potentials can be realized at zero cost in several sectors (SSCC p. 187). For instance, heat production utilizing gas from landfills has a negative marginal cost, i.e. you earn money by implementing it. This is not just theory. For instance, it is carried out in Kristianstad in south Sweden (link here and here) to the extent that the city no longer uses fossil fuels for the heating of homes. The authors of the methane chapter write (p. 190): "Fist, we recommend tackling the "low-hanging fruit". Mitigation potentials at zero or even negative costs should be realized by removing institutional and social barriers."
How has the panel of prominent economists handled this proposal? They have ignored it. One would expect that free benefits are given top rank above all other proposals, but no, methane mithigation gets a fairly low rank. The economists write that "the best options to regulate CH4 in livestock and agriculture will face almost insurmountable obstacles in practice. " There is not a single word about free benefits - maybe some of the economists are free-market fundamentalists who agree with Lomborg in denying the existence of free benefits, although the city of Kristianstad is a living example that such benefits exist and are just waiting for us to utilize them. This may cast serious doubt on the way that the economists have evaluated the evidence presented to them.
In the introductory paragraph, I asked: Are there any indications that the set-up is used to divert attention away from solutions that will go against the interests of the fossil fuel industry?
The main interest of the fossil fuel industry would probably be to avoid any substantial carbon tax. In agreement with this, chapter 2 on carbon taxes is very biased and presents a result which is an extreme outlier. This unrepresentative result is then used as a basis for concluding that carbon taxation above a very low level will not pay off. This is based on running a model - the FUND model - which presents climate mitigation as having very negative consequences in comparison with nearly all other economic models. And the panel of prominent economists takes the output from this model at face value. One of the economists accepts the idea of having a moderate tax to provide funding for technological innovation, but even he fails to realize why profitable tax rates in chapter 2 are so much lower than those in chapter 7 on technological innovation. Apparently, the economist panel has simply not grasped how much chapter 2 differs from other literature on the subject. So if the intention was to deceive the economists on this point, this intention seems to have worked.
There is a discussion in various chapters about whether a high price on carbon is necessary to make sufficient technological innovation happen. The authors of chapter 7 argue that such a price signal has no influence, whereas several critics argue the opposite, with reference to a considerable body of evidence. It should be central for the whole field of climate mitigation if a strong price signal is necessary to reach the goals. But strangely, most of the five economists are completely silent on this issue which should be at the heart of any climate economist.
Another crucial issue is the choice of discount rates. The panel of economists must know that this choice has much influence on the output of the computer models. They could not avoid to see in chapter 6 that the choice is very important for the relative merits of mitigation and adaptation. In spite of this, they give no indications of acknowledgement that benefit-cost-ratios often depend strongly on discount rates.
Furthermore, it is strange that even if examples of free benefits are presented to them, none of the five economists suggests that these "low-hanging fruits" be picked.
And it is remarkable that they practically disregard the many reservations, caveats and criticisms in various perspective papers.
Referring to the panel of five prominent economists, I asked at the start of the page: did they have to draw these conclusions from the evidence presented to them? Would a broader and more neutrally selected panel have drawn the same conclusions? It seems that the answer to these questions should be no. And I asked if there are major objections to the presented evidence, and should these objections mean that other conclusions might be just as justified, or even more justified? It seems that the answer to these questions is yes.
Bjørn Lomborg writes on page 5: "I invite you to read the research and the Expert Panel´s findings, and form your own view on the best - and worst - ways we can respond to global warming." This sounds as if he does not want us to reach any particular conclusion, but are free to conclude what each of us feels most like. But actually, the book has a set-up which leads the readers to reach just the conclusions that Lomborg apparently wants us to reach. These conclusions are carefully presented just in those chapters at the end of the book that are so short that politicians and policy makers will actually manage to read them. So even if each of us is free to form our own opinion, those busy and influential people who only have time to read the short chapters that sum up the conclusions, those people will pick up the conclusions that Lomborg wants them to pick up.
Altogether, although the book contains a lot of useful information and interesting analyses, it has the same fault as anything else that Lomborg has published: it is severely biased.