• Why are the ExternE numbers different from some other scientific sources?

    Some numbers may be based on avoidance costs, which is a very different methodology. Most other studies use a top-down approach analysis, which does not fully reflect the marginal cost approach referring to an additional plant or produced unit of electricity at a specific site. Other studies are also based on a kind of impact pathway approach, but may have used other models, assumptions or inputs. Thus a detailed comparison would be required.


  • Can I compare numbers of the most recent studies with those from older reports?

    One has to be prudent in comparing the numbers, because in some areas scientific understanding is changing fast and significantly, and this scientific development is reflected into the ExternE numbers. As a rule of thumb, the most important changes relate to the quantification of public health impacts from pollutants and the quantification of global warming impacts.

    Although for some case studies numbers have been upgraded, this is not a general rule, and there is no systematic upgrade of the numbers of older reports. A first short-cut attempt can be made by using new numbers per ton of emissions for the specific sector and country, and emission factors.


  • Why are ranges of numbers often displayed?

    There are two main reasons: to reflect variability in nature, technology, etc, and to reflect uncertainty in our knowledge. However, the ranges do not reflect changes of the results over time due to growing scientific understanding.


  • Can the numbers be used for policy preparation, even if you only give subtotals and uncertainties are large?

    Yes - although not as figures that have the same quality of accuracy as data gathered from the official national accounts. However, external cost data gained from ExternE are useful for a quantitative comparison of magnitude with each other as well as with other data expressed in monetary values, e. g. gross or net domestic products (which is done in "Green Accounting"). Such a comparison represents a trade-off ratio between marketed and non-marketed goods that is based on a consistent (and widely accepted in economic theory) approach of individual preferences and willingness-to-pay.

    Numbers have indeed already been used in several policy areas, such as economic evaluations of the draft directive on non-hazardous waste incineration, the Large Combustion Plant Directive, the EU strategy to combat acidification, the National Air Quality Strategy, the Emission Ceilings Directive, proposals under the UN-ECE multi-pollutant, multi-effect protocol and many more policies, green accounting research projects, and air quality objectives.


  • The uncertainty of ExternE estimates is quite high. Can decisions based on these estimates nevertheless be validly supported?

    It is a principle of the ExternE methodology to quantify the uncertainties and be explicit about their derivation, by contrast to many other studies of this kind. The expressed uncertainties of ExternE estimates are large. However, many policy decisions are remarkably robust against such uncertainties (Rabl, Spadaro & van der Zwaan 2005), i.e., whether input values at either upper or lower tail of the uncertainty distribution for component parameters are used, the ideal policy decision based on minimizing social costs would be the same or not significantly different. Even where this is not the case, there may be decision alternatives that can be ruled out with reasonable confidence. Without the information provided by ExternE the risk of a wrong choice or large social cost penalty would be much higher. Even where the uncertainties are too large for firm conclusions, the results of ExternE may be helpful for the decision process. An explicit and well-documented analysis provides a basis for informed discussion by laying assumptions and preferences on the table (e.g. about discount rates or value of prevented fatalities). Future research and scientific progress will help to reduce the uncertainties of externalities estimates.

    Rabl, Ari, Joseph Spadaro, and Bob van der Zwaan. 2005. “Uncertainty of air pollution cost estimates: to what extent does it matter?,” Environmental Science & Technology 39(2):399-408.


  • Not all damages have been monetised. There might be substantial impacts the costs of which are not included in the external cost estimates. How is this taken account of in ExternE.

    ExternE recognizes that there are gaps in what the method has quantified. Among those damages currently not included in ExternE estimates are nuclear proliferation, nuclear security, security of energy supply, visual intrusion and risk aversion. Quantification of damage costs requires both a link between exposure to an environmental or health burden and its impacts and monetary values for these impacts, which is not available for all burdens and impacts. For example, for many pollutants, there are no concentration-Answer functions, only thresholds (no-observed effect or onset of effect). Irreversibility of damages is another challenge. ExternE routinely considers new scientific information to see what gaps might be appropriately filled, with defensible quantitative estimates and reasonable levels of uncertainty. It is beginning to try to quantify additional important impacts, but these estimates are not yet sufficiently reliable to use.

    Potentially important gaps should certainly be reported together with the results. The problem is how to judge which impacts are potentially important, e.g. might have significant damage costs, and how to represent them. At the start of the analysis, ExternE used a screening process, analysing the ubiquity, irreversibility and persistency of a potential impact, and this screening process should continue. As with any assessment method, there may be other important impacts that have not yet been recognized as such (and ideal decision-making would take this eventuality into account).


  • Are there impacts, that can not be quantified and monetized in principle?

    In principle, any damage is monetisable in an externalities framework, e.g., through contingent valuation (as it is always possible to ask for the willingness to pay to avoid a certain described impact). Problems with acceptance may arise, if an issue is very controversial, as e.g. the risk of nuclear power. An important example of a criterion for which monetisation does not currently seem sufficient is the occurrence of a health impact that affects certain known persons (e.g. that live near an emission source) with high probability or certainty. In this case, a threshold should be implemented, that should not be exceeded. Thus ExternE recommends to use thresholds and combine them with the internalisation of external costs, that occur to allow for the inclusion of impacts that occur even when the threshold is not exceeded.

    As stated above, ExternE acknowledges that some damage types or decision criteria can not be expressed in monetary terms at present. With continuing progress in the art and science of monetary valuation and analysis of preferences the range of monetisable damages can be expanded. In fact, the work of ExternE has established a general framework for that, and impacts that would have been considered beyond the scope of monetisation in the past are now routinely evaluated in terms of social costs.


  • Is the risk of a nuclear accident evaluated in an appropriate way?

    Evaluating external costs for the risk of a nuclear accident is a topic of long debate (ExternE 1995, 1998; Krewitt 2002). Currently, ExternE evaluates costs of a nuclear accident by multiplying the assumed probability of an accident by its assumed costs, derived from modeling studies. This calculation does not account for risk aversion, which is pronounced in the case of nuclear accidents, radioactivity, and large disasters in general. There is no agreed procedure on how to deal with risk aversion within a quantitative framework. It would be possible to develop an assessment scheme, e.g. by using participatory approaches to explore the preferences of the population with respect to different risk types, but ExternE has not been funded to do so. In any case, nuclear energy is a technology where opinions tend to be very strong. It is conceptually important to include an estimate of the costs of a nuclear accident in calculating the external costs of nuclear power. However, such an average value, or perhaps any monetary valuation, is unlikely to be persuasive. In this case, the value of the ExternE methodology may be more a means for steering the discourse into specifics and demonstrating which assumptions have to be used to support different opinions.


  • What about nuclear proliferation and security in the event of terrorist attacks?

    Nuclear proliferation has been a topic of long debate in ExternE (e.g., ExternE 1995:521-541, ExternE 1998 v. 7:483-491), but ExternE has not attempted to quantify the potential damages. EU member states have presumably analyzed the risk of terrorist attacks, likely including some quantitative analysis. However, information on these is not publicly available, so it is not possible to quantify the remaining risks. Thus both issues are not included in ExternE’s external cost estimates and have to be treated as gaps. ExternE has done no work on nuclear technologies since 1998.


  • Are long-term effects of nuclear energy treated adequately, in particular, nuclear waste storage and other land contamination?

    In estimation of the impacts of future normal operation of storage facilities for nuclear waste is included in the analysis. The risk of leakages has been examined only for low and medium-level radioactive waste. It is expected, and assumed in the analysis, that future storage facilities for high level waste will be built and operated according to strict standards and any remaining risks will be limited to the local zone, which would be chosen in a low population density. Therefore, according to case studies, any leaks would affect only a small number of individuals and the associated damages would be small.


  • The choice of nuclear power plant is crucial in determining the external costs of nuclear power generation. Why not include Chernobyl-type plants in nuclear power results, rather than modern plants?

    ExternE is generally oriented toward the external costs of future plants, and Chernobyl-type plants would not be built today. Here the crucial issue is how the results are used and what they are said to represent. Results generated for a pressurized water reactor (PWR) type plant are not applicable to a Chernobyl-type plant.


  • Progress in the development of renewable energies is fast. Do available estimates of external costs of renewables reflect the latest state of the technology?

    The ExternE-Pol project has published results for wind, biomass and PV based on the most recent LCA (life cycle assessment) data, both for current and for advanced technologies (ExternE 2004). Of course, the prediction of damage costs for future technologies involves assumptions that can be debated. However, by explicitly stating the damage cost per pollutant, ExternE allows the user to recalculate the damage costs of future technologies for any other assumptions about the rate of emission of pollutants. In fact, this formulation of damage costs that allows users to adapt past ExternE work to any number of new scenarios and policy questions. Furthermore, the ongoing EC funded project NEEDS will generate estimates of LCA data for future energy technologies.


  • The current estimates of external costs do not take into account the potential of technical progress, that might lead to improvement and thus the benefit might outweigh the costs on the long run.

    ExternE aims primarily at providing a methodology that can be used to answer specific policy questions (though as we note below, it is also being used to generate more generic results). Ideally, the scope and content of the analysis has to be designed in relationship to the question, after which the methodology has to be applied. For the above example, one would first build future scenarios about energy demand and the development of energy supply technologies and then carry out cost-benefit analyses, using, as discussed just above, existing ExternE estimates of damage costs per unit of pollutant. The EC funded project NEEDS is currently compiling information about the possible development of certain energy supply technologies and will produce corresponding external costs.


  • How should the potential benefits of energy efficiency be included?

    If electricity prices go up because of policies meant to internalize damage costs, electrical energy efficiency will be encouraged because the higher prices will render it more cost-effective, other things being equal. The avoided damages count as benefits of energy efficiency.


  • The variability of wind power, in combination with regulations that require the grid to accept wind-generated energy when available, leads to substantial inefficiencies after which little of the benefits of wind remains.

    Again, this demonstrates that according to the questions asked, scenarios should be defined and then the ExternE methodology should be applied, instead of simply taking some results. In this case, two scenarios would be generated: an electricity generating systems with and one without wind energy. The difference in external costs for these two scenarios can then be allocated to the wind energy use.


  • Should mortality valuation be based on VOLY (Value of Life Year) or on the value of a prevented fatality (also called VSL = “Value of Statistical Life”)?

    Being based on accidental deaths, VSL is the correct measure for accidents and has been used for that by ExternE. But whereas most accidents involve a large loss of life expectancy (LE) per death, the loss per air pollution death tends to be small (the population average being on the order of months). Furthermore, the true number of deaths due to air pollution is not known whereas the loss of LE can be calculated; thus only VOLY can be used for the total mortality due to air pollution (Rabl 2003). This has not yet been universally recognized and some analysts, especially in the USA, continue to use VSL for air pollution.

    Rabl Ari 2003. “Interpretation of Air Pollution Mortality: Number of Deaths or Years of Life Lost?” Journal of the Air & Waste Management Association 53(1), 41-50.