The European Commission has recently suggested a new greenhouse gas reduction target for 2030 that is perceived as too low by many NGOs. Moreover, some claim that the Commission has hidden the fact that higher emission reductions than proposed would lead to GDP gains. Brigitte Knopf from the Potsdam-Institute for Climate Impact Research argues that GDP gains are a very specific finding with a non-standard model and that emission targets can only be set politically and not by science.
Do we have a Modelgate in Brussels? Did the European Commission hide secret results of computer simulations about the transformation of our energy system? Is climate change mitigation cheaper than told by those people in Brussels? The answer is No. Contrary to the accusations raised by Brook Riley of Friends of the Earth, the European Commission’s projections are surprisingly in line with other independent research results. This is shown by a comprehensive multi-model analysis we recently published.
I will comment on two main aspects mentioned in Riley’s post. The first aspect is his claim that the EU Commission’s own Impact Assessment shows a positive effect on the GDP with more ambitious climate targets but that the Commission has hidden this fact.
The Commission uses four different macro-economic models from three institutions to assess the impact of the different climate targets and additional targets for renewable energy and energy efficiency. Two of the models, GEM-E3 and PACE, show GDP losses in a scenario with 40 percent GHG reduction compared to the reference case. The losses are rather small with a range of -0.1 to -0.45 percent. By contrast, the model family E3ME/E3MG shows GDP gains in the range of +0 to +0.55 percent. All this is very transparently stated in Table 16/17/18 (p. 84), Table 27 (p. 101), and in the overview Table 40.
With E3ME, an additional analysis is done to assess the impact of energy efficiency policies and renewable targets, again showing gains. However, no message can be deduced that GDP gains would increase with higher targets or additional policies, as claimed by Riley, because here he compares two fundamentally different types of models. But it is scientifically not sound to compare directly the results for 40 percent GHG scenario (GHG40) from one model to a scenario outcome with higher reductions of 45 percent and additional energy efficiency policies and renewable targets (GHG45EE RES35) from another model.
Different models, different answers
The only statement that can be made based on the Commission’s assessment of the GDP effects of climate policy is that different models give different answers: two show GDP losses and one class of models shows GDP gains. But in contrast to Riley’s claim, it is rather the model family that shows the GDP gains, E3MG/E3ME, which is exceptional in the modeling world. In general, models show GDP losses due to climate change policies relative to a baseline or reference case.
It is important to note that setting targets always includes value judgments and ethical considerations.
This is different in E3ME/E3MG: A major driver of the mitigation strategy in this model class is the recycling of revenues raised from the auctioning of carbon permits (in the sectors covered by the EU Emission Trading Scheme) and of carbon taxation (in the non-ETS sectors). In these models, the revenues are used to lower indirect taxes such as labor taxation. But this means that in this model you would also see GDP gains if you tax the use of apples instead of carbon, as long as the revenues are recycled in the same way. This specific kind of revenue recycling is also the reason why in these models the employment effect of climate policy is positive.
In addition, the standard models assume a perfect reference case as the benchmark and evaluate mitigation policies against this benchmark. By contrast, E3MG is a simulation model without perfect foresight and optimization, where resources are not fully employed or optimally utilized in the reference scenario. The increase in investment induced by climate policy can therefore achieve net GDP gains. These gains are then attributed to climate policy.
However, it is important to note that these gains can in principle be achieved also by policies other than climate policy, such as taxing apples. So the gains cannot be simply attributed to climate change mitigation, as often done, but are rather an effect of assuming an imperfect reference case with idle capacities that can be put into use in the mitigation scenario. It is crucial to understand the differences of these econometric simulation models such as E3MG/E3ME compared to the great majority of the other models that show GDP losses. This is not splitting hairs, this is fundamental, because the latter models do not assume inefficiencies in the baseline and can therefore accurately compare a case with and without climate policy.
But there is still a positive message: also in the standard models, GDP losses due to climate policy are rather small for Europe. In an independent model comparison we came to the conclusion that upscaling greenhouse-gas emissions reduction from the current policies in the reference scenario to 40 percent GHG reduction by 2030 would cost less than an additional 0.7 percent of economic activity. However, costs may increase by 2050.
Figure 1. GDP reduction as a percentage of a scenario with 80% reduction by 2050 compared to a reference scenario with 40% reduction. Source: EMF28 model comparison, Knopf et al. 2013
Out of line with the science
The second aspect I want to comment on is the author’s claim that “a 40 percent reduction is out of line with the climate science.” What can climate science really say when it comes to setting targets and what does climate science actually say about the EU target? First of all, it is important to note that setting targets always includes value judgments and ethical considerations. The latest IPCC report has clearly pointed this out, stating that:
“Many areas of climate policy‐making involve value judgements and ethical considerations. These areas range from the question of how much mitigation is needed to prevent dangerous interference with the climate system to choices among specific policies for mitigation or adaptation. Social, economic and ethical analyses may be used to inform value judgements and may take into account values of various sorts, including human wellbeing, cultural values and non‐human values.” (IPCC WGIII SPM)
This implies that climate science cannot answer the question of what the “correct” target for the EU is, be it 35, 40 or 55-60 percent reduction, because it is a political decision. The only answers that scientists can give are if-then statements. If the EU wants to achieve 80 percent reduction, what is then the (cost-efficient) milestone for 2030?
It is certainly legitimate to call for more ambitious targets than proposed by the EU Commission. One should only be careful in how to use model results in this context.
The Commission comes, based on its own assessment, to the conclusion that this is 40 percent GHG reduction. With our independent analysis we come to the conclusion that for achieving 80 percent GHG reduction by 2050 a CO2 reduction by 2030 of 47 percent with a model range of 40–51 percent is required. This 47 percent CO2 reduction corresponds to a GHG reduction of roughly 43 percent. This indicates that a reduction of GHG emissions of 40 percent by 2030 — as suggested by the Commission — could, in principle, be in line with the long-term effort to reduce emissions by 80 percent by 2050. The model median, however, would suggest setting a more ambitious target.
This means that 40 percent GHG is not “out of line with the climate science”. But this also implies that if the EU wants to go for 95 percent reduction by 2050, then cost-efficient emission reductions have to be even higher by 2030. And on the other hand, this does not imply that the long-term target cannot be achieved anymore if we do less than 40 percent GHG reduction by 2030. It only means that it will become more costly. But there is no imperative by climate science what the “right” target is, science can only inform the debate.
In summary, the Impact Assessment of the Commission is balanced in the sense that it assesses different types and classes of models. It is sound work also in terms of scientific standards. It is transparent in reporting the outcomes, although not on every detail about model input and model fundamentals. And the main conclusions, especially those of the EU “Energy Roadmap 2050”, can be independently confirmed by a model comparison with 12 different models. So there is no reason to call this Impact Assessment and the Commission’s communication about it a “Modelgate”.
However, it is certainly legitimate to call for more ambitious targets than proposed by the EU Commission. One should only be careful in how to use model results in this context.
Brigitte Knopf is head of the research group ‘Energy Strategies Europe and Germany’ and deputy head of the research domain ‘Sustainable Solutions’ at the Potsdam Institute for Climate Impact Research (PIK) and is on Twitter as @BrigitteKnopf
This article is a response to the article by Brook Riley that we published on 29 April: “Welcome to Modelgate: Brussels’ justification for a small climate target is based on a big lie“. See also the comments below this article for more discussion on the topic.