5. Discussion, conclusions & recommendations

In this report we present for the Nordic region, country- and scenario-specific socio-economic damage costs of the air pollutants classified as short-lived climate forcers, including methane. For the first time, the damage costs consider the emissions’ effects on climate change, health, and the environment. Further, we compare these costs with climate damage costs of Nordic CO₂ emissions for the same scenario period. Finally, we calculate the socio-economic benefits of reducing SLCF emissions from what the UNECE Air Convention currently considers to be a likely emission pathway given current air pollution legislation (BSL), into an emission pathway that corresponds to a maximal introduction of technically feasible emission reduction measures (MTFR).

The results of the mid-value analysis show significant variation between the Nordic countries, where Iceland is an outlier. This is understandable since Iceland is an island remote from other populations and with relatively large SO₂ emissions. All in all, our results show that SLCF climate damage costs constitute a significant and increasing share of total SLCF damage costs for the Nordic countries (contingent on increasing global temperatures in the future). The damage costs due to health effects and effects on the environment still dominate the total damage costs, but climate damage costs provide significant additions to the total.

In addition, based on similar but not identical, CO₂ scenarios from the Nordic countries (WEM) and SLCF scenarios from CIAM & TFIAM (BSL), the calculations suggest that damage costs of SLCF emissions and CO₂ emissions will be of almost equal magnitude in the future. Comparing the BSL with the WEM scenario for the year 2035 (the last all-Nordic WEM-scenario year), SLCF damage costs constitute ~40% of the total CO₂ and SLCF damage costs by 2035.

Furthermore, the comparison of damage costs between the CIAM & TFIAM BSL and MTFR scenario shows that all Nordic countries would have substantial benefits from SLCF emission reductions corresponding to the difference between BSL and MTFR. Again, there are differences between the countries. The Icelandic emission reductions would imply socio-economic losses due to the net effect of Icelandic emissions on climate change. The Danish, Finnish, and Norwegian emission reduction benefits will to 13%–20% be due to climate change effects.

5.1. Discussion

As per usual, the results of our analysis are subject to scientific uncertainty and scenario-dependency. As we suggest from the uncertainty discussion, the most important uncertainty relates to the Nordic-specific climate impact of SLCF emissions. The literature is well aligned in its suggestion that the position of emissions with respect to both latitude and longitude will affect climate impact. More dedicated studies on these types of questions would be beneficial to establish the robustness of our results.

The scenario-dependency will affect the results and corresponding conclusions. Due to incompleteness in the data across the Nordics, we did not use national scenarios, but instead used the latest and policy-relevant scenarios from CIAM & TFIAM. For the applicable years, we have controlled the BSL scenario against nationally reported emission scenarios (Appendix I). To our understanding, at least on a national level, the BSL and national scenarios are sufficiently similar for our results to be policy relevant.

In addition to the scientific uncertainty and the ‘Knightian uncertainty’ about the future (Knight 1921), there are also uncertainties related to the economic parts of the results. First, as is standard when discussing damage costs of air pollution: the fact that we base our main analysis on the normative choice of VOLY instead of VSL for air pollution-induced health effects, implies that the relative importance of climate damage costs is higher than if we would have compared with the VSL metric. As a reminder, we chose VOLY since this is the metric preferred by inter alia the EU in air quality policy processes, and since it was not possible for us in this project to update VSL values based on the available information in the report used for SCM (US Environmental Protection Agency, 2023).

Second, as noted in Chapter 2.2.3, the mortality cost component of SCC and SCM is based on an estimate of the value of statistical life (VSL) from USA. This estimate is then extrapolated to countries in the rest of the world by normalizing GDP/capita. An improvement in future updates of the estimates of the damage of Nordic emissions would be to use new global VSL estimates that will soon be made available by the OECD. The authors of this report are engaged in the development of this OECD report (OECD, forthcoming).

Third, we are using economic damage costs of climate change derived from climate-economic models. These models rarely estimate ‘external costs’ of climate change in a manner similar to standard environmental economic theory. Instead, they often provide a numerical estimate corresponding to an intertemporal equilibrium point between emission abatement/mitigation costs and emission abatement/mitigation benefits. So if the emission scenario being analysed with a certain SCC/SCM value does not correspond to the temperature increase that was implied when originally deriving the SCC/SCM value, the results will be somewhat biased up or down.

Fourth, the climate change damage costs we use do not correspond to the governmental guideline values for damage costs of CO₂. For example, the Swedish Road Administration suggests using a CO₂ damage cost corresponding to ~650 €₂₀₂₀/tonneCO₂ (Swedish Road Administration, 2020). This value is more than twice as high as the SCC-values used in our analysis, even for 2050. In Norway, the Ministry of Finance, has a set of carbon price paths that are recommended for use in cost-benefit analysis. However, to our understanding, the scientific underpinning of the US EPA 2023 values is more solid than the values suggested by the Swedish Road Administration or the Norwegian Ministry of Finance. The Norwegian carbon values, for example, are based on costs of climate measures to reach climate targets and not assessment of damage costs. The carbon cost values we use may still, however, be incomplete in that they are not able to capture all the expected damages fully (as noted in Chapter 2, in US Environmental Protection Agency (2023) or for other reasons (P. Howard, 2014)).

If our results are used as policy advice, the following considerations need to be remembered.

The climate change effect of Nordic SLCF emissions is still in need of more verification with respect to magnitude of the impact,
The SLCF scenarios used are not perfectly aligned with current national emission projections for SLCF emissions,
The underlying values used to calculate climate, health, and environmental damage costs are similar but not fully aligned with the costs proposed by national agencies.

In addition, climate scientists recognize that the choice of climate metric for our type of analysis largely depends on the purpose of the analysis (Tanaka, Peters, and Fuglestvedt 2014; B. Aamaas, Peters, and Fuglestvedt, 2013). Which purpose that matches which metric or time horizon, has not been suggested to our knowledge.

Since GWP₁₀₀ is the metric most commonly used in climate policy instruments (such as in the UNFCCC Nationally Determined Contributions and the EU Emission Trading System), we have chosen this metric and time horizon. Most other metrics and time horizons that we know of as common in the literature, would have rendered a higher climate damage cost of SLCF emissions than we have presented in this report.

5.2. Conclusions

Based on our data, method, results, uncertainties, and discussion, the following conclusions can be suggested:

The SLCF emission damage costs currently utilised for air pollution decision support in the Nordic region are under-estimating the social damage costs of these pollutants,
Including SLCF climate damage costs would increase the socio-economic value of reducing emission. In our calculations that use conservative climate metrics, the increase would lie around 10–20%,
If the Nordic countries would be as effective at reducing CO₂ emissions as is projected nationally, Nordic SLCF emissions might become a problem as important, when assessed in monetary terms, as CO₂ emissions sometime around 2050,
More research, on especially the climate impact of Nordic SLCF emissions, would improve the robustness of the analysis of economic damage costs.

5.3. Recommendations

For the Nordic Council of Ministers group on air and climate we have the following recommendations:

Suggest to national governments that they complement their guideline values for socio-economic damage costs of SLCF emissions with estimates of their climate damage costs,
Improve the scientific underpinning of the climate impact of Nordic SLCF emissions,
Complement the global average social cost of carbon and social cost of methane estimates with values better adjusted for intra-Nordic circumstances and settings (inter alia by correcting so that VSL and VOLY values used for valuation of air pollution is the same as the ones used to value CO₂emissions),
Consider promoting an increase in national/Nordic emission abatement ambition levels since the socio-economic benefits likely are larger than previously known, especially for the SLCFs that are ‘warming’.