Criticism of
"Copenhagen Consensus 2008"
    Home                                                                                        Copenhagen Consensus general page

The second "Copenhagen Consensus" conference was held during the last week of May 2008. Details can be read at the official Copenhagen Consensus web site.

This page on Lomborg-errors gives a few points of criticism of the conference.

In addition, a paper with a detailed criticism has been written by Christian Friis Bach from the Danish Church Aid.  The paper is called: The Copenhagen circle of solutions.
(The first few pages are a Danish summary. Go right on to the English text).

1) Uncertain estimates of costs/benefits due to climate change
The background paper on global warming, to be downloaded here, is authored by Gary Yohe, Richard Tol, Richard Richels and Geoffrey Blanford. 
   In this paper, the overall net effect of climate change on all human societies combined is judged to be positive up to about 2070 and negative thereafter. The positive effect that weighs most heavily in this estimate is the alleged reduction in winter deaths in rich societies (the value of a death is estimated from the GDP of the country where the death occurs, which means that deaths in rich countries weigh in much more heavily than those in poor countries). The sources given are the papers authored by Richard Tol, which have been dealt with in the criticism of the book "Cool it!". See this page on Lomborg-errors. Here you will read that it is doubtful if increased winter temperatures will lead to markedly fewer cold-related deaths. There will always be an overmortality during winter, no matter precisely how cold the winter is. If the decline in winter deaths will not be as marked as anticipated, the whole basis of the economic calculations crumbles away.
    As far as can be judged from the information given, the only important parameter on the minus side is the change in agricultural production. The global  agricultural production  is assumed to  decrease only after the average temperature rise has surpassed 3° C. The negative effects on forests and wetlands are clearly underestimated in the analysis. It is based on a 1990 estimate giving an average loss of 10 million $ per km².
   The calculated benefit/cost ratios are therefore extremely uncertain. With fewer positive effects of climate change, the negative effects will be dominating, and the benefit from fighting global warming will be larger.

2)  When will accelerated research and development be beneficial?
Lomborg repeatedly claims that we should rely on research and development to obtain improved low-carbon energy technologies, and apart from that, we should not interfere with economic growth by putting restrictions on the use of fossil fuels.
He anticipates that the new technologies will have been developed to a suitable level in 40 to 50 years from now, and only then should these technologies be applied. The immediate effect of such a strategy will be that for the next 40-50 years, nothing is done that can actually mitigate global warming.
   It is therefore interesting to see what the background paper by Yohe et al. says about such a strategy. In figures 3.1 and 3.2 of that paper, we see that a strategy involving only research and development is more efficient than a strategy involving only reduced fuel consumption up to about 2100. After that, research and development alone cannot efficiently impede CO2 emissions. That is, the strategy favoured by Lomborg has no effect for the next 40-50 years (up to about 2050) and is of little effect after 2100 - whereas the "mitigation" strategy (reducing CO2 emissions with carbon taxes or quotas) is useful before 2050 and especially after 2100.
   But of course, the best choice is to use both strategies at the same time.

3) Ranking of the research and development option
In the global warming theme, there is also a "perspectives paper" by Chris Green. In this paper, Green says that Yohe et al. have underestimated the difficulties of developing new technologies, and that much more money must be spent on research and development. He formulates an "incentive technology race" financed by a carbon tax, and calculates a beneift/cost ratio for this technology-based programme. The result is 16:1 (when using 4 % as the discount rate) or 28.5 :1 (when using 3 % as the discount rate).
   This result is so favourable that research in low-carbon energy technologies is ranked no. 14 in the Copenhagen Consensus ranked list of solutions. This is just below tuberculosis treatment (rank no. 13), which is the last programme on the list that can be affforded with the imaginary sum of $75 billion over 4 years.
   Higher ranks are given to treatments of malaria (benefit/cost 20:1), child diseases (benefit/cost 20:1), and heart diseases (benefit/cost 25:1). However, in these cases, a discount rate of 3 % was used. If the same discount rate is used for research in low-carbon technologies, this would have a benefit/cost ratio of 28.5:1 and thus be more favourable than treatment of the diseases referred to. But by using another discount rate for global warming issues (4 %) than for other issues (3 %), Lomborg and his team just manage to avoid spending some of their imaginary sum on something related to global warming.
   That is to cheat.

4) Inconsistent use of discount rates
The experts were instructed in advance to use two discount rates, viz. 3% p.a. and 6 % p.a. That is, the benefit/cost ratios are calculated twice, once with the low rate, and once with the high rate. In practice, however, only the results obtained with the low rate (3 %) have been used in the conclusions. So, the high benefit/cost ratios for the many programmes in the fields of health, nutrition and diseases were calculated using a discount rate of 3 %.
   The only exception is in the field of global warming. Here, Yohe et al. use a discount rate of 5 %, gradually declining over 100 years to 4 %, whereas Green uses 4 %.
   Why did the climate specialists not use the prescribed discount rates? As to Yohe et al., the explanation given by the economist Richard Tol here is as follows: 

"On the discount rate: I do not know what the other papers used. We used a consistent discount rate — all calculations, and all reporting was done with the same discount rate. The models that we use would require extensive recalibration for a different discount rate.  .  . As we used dynamic optimization models fitted to observations, we had to stick to the discount rate we had. As the rest of the Copenhagen Consensus used simpler methods, they should have used our discount rate."

So Richard Tol thinks that all other specialists should have used the discount rate that gradually declines from 5 % to 4 %, because his group could not easily adapt to the prescribed 3 %, whereas it would have been relatively easy (?) for all others to adapt to his group´s discount rate. In any case, the result is that the disocunt rates are not comparable.

As to Green, he performed his calculations with a 4 % discount rate, but also included results for a 3 % discount rate.

Now, when all data for all items were summarized and compared, all other projects were represented by the benefit/cost ratios obtained wit a discount rate of 3 % (the results with a 6 % discount rate were not used in the final evaluation). Only the climate projects were represented with different discount rates. And these rates were higher than those used for other issues. Which is against the usual thinking that the longer the time perspective, the lower must the discount rate be. As the climate issue has the longest time perspective, it should have the lowest discount rate.
   This is especially remarkable in the case of Green´s project. Here there existed a version where a 3 % discount rate was used, but in spite of this, the version included in the final ranking was the one using a 4 % discount rate. And that matters quite a lot. As stated in the previous paragraph, for Green´s project, 4 % yields a benefit/cost ratio of only 16:1, whereas 3 % yields a benefit/cost ratio of 28.5:1. Which would have brought Green´s climate project near the top of the ranking list, above the efforts against tuberculosis, malaria, child diseases and heart diseases. The climate project would have got a priority near the absolute top. But the slight change in discount rate, from 3 % to 4 %, sent the climate project far down along the ranking list, out of the range of projects that are granted money.
   Incidentally, this is the same situation as in the Copenhagen Consensus 2004. There, the discount rate used for the climate issue was 5 %, whereas that used for HIV/AIDS was 3 %.
   So there is an obvious reason why the climate issue always is ranked last. It is systematically treated with a higher discount rate than the other issues.

It would of course be interesting to hear Lomborg´s comments to this criticism. And indeed, he has been forced to comment upon this in a debate in a local Danish newspaper in February 2009. See here.

The climate issue is also this time treated differently from the other issues. In the Copenhagen Consensus 2004 conference, a discount rate of 5 % was used in relation to climate, but only 3 % in relation to HIV/AIDS. In the 2008 conference, this trend was kept: 5 % (gradually declining to 4 %) was used for climate issues, but 3 % for all other issues. A discount rate of 4-5 % is definitely much too high for issues on very long time scales (nobody can sustain a return on investments in production of 4-5 % per year over periods of several hundred years). But the use of a low rate like in the Stern report was not accepted in the prescriptions .
   So, like in 2004, it was predestined in advance that climate would end up as the lowest rank. This should not wonder. In all probability, the whole idea of the Copenhagen Consensus conferences is to "prove" that nothing should be done to mitigate climate change, at least not now.
   There is no indication that Lomborg is under direct influence of the oil industry, but he always acts as if the purpose were to help the oil industry.