Discussion
The paper begins by comparing and contrasting Sweden’s emission reduction goals with the EU’s “Fit for 55” package. The authors highlight that nearly all of Sweden’s emission reductions in the Effort Sharing Regulation (ESR) sector are expected to be made in road transport. The authors’ calculations suggest that emission reductions using the biofuel blending mandate (“reduktionsplikten”) may be extremely costly. Moreover, Sweden runs the risk of requiring ESR quota units to meet its Land Use, Land-Use Change, and Forestry (LULUCF) obligations, which would be a relatively costly solution compared to much cheaper abatement measures, e.g. increasing carbon storage in Sweden.
According to economic theory, the most efficient way to reduce emissions in an economy is to ensure that the marginal cost of abatement in each sector is equal to the marginal damages from emissions. A regulator could achieve this ‘first-best’ outcome via a Pigouvian tax on emissions that is applied uniformly across all sectors. In reality, this is a difficult task due to market imperfections and other complications. In the case of Sweden, however, the disproportionate focus on abatement in road transport is also policy-induced since current policy interventions depart drastically from the equivalent of a uniform tax. Using a simple graphical approach, the authors illustrate the damage to economic welfare when the abatement of marginal costs is not equal across sectors.
The authors apply a Computable General Equilibrium (CGE) model of the Swedish economy in order to analyse the differences in GDP, abatement costs and economic welfare. CGE models have their drawbacks, as they have many moving parts and are, in some cases, sensitive to the chosen parameters. For example, any model would probably be sensitive to assumptions about future abatement costs across the various industries. However, I would argue that a CGE model is a suitable approach because it captures the many direct and indirect impacts of policies on various sectors and is the standard in this field. Their main conclusion is that more flexible alternative policies resembling something closer to a uniform carbon tax would boost GDP and economic welfare. Their results seem reasonable and are in line with their theoretical predictions.
I completely agree with the author’s argument that Sweden’s current approach to reducing emissions in the ESR sector is highly inefficient. The authors provide rigorous estimates of the cost of reducing emissions in the transport sector. In terms of abatement alternatives, they cite evidence for the relatively low cost of abatement in the LULUCF sector and projected ESR quota unit price. It should also be noted that they spend less time discussing alternative domestic abatement in the ESR sector, probably because these alternatives are less feasible. Based on the most recent statistics available (Statistics Sweden, 2023), emissions due to waste and the heating of buildings were only 2% and 3% of Sweden’s total territorial emissions in 2021, respectively, which leaves little room for further emission reductions. Emissions from Swedish agriculture in 2021 were 14% of the total and have remained roughly constant over time. The majority of emissions in Swedish agriculture not related to land use are caused by biological processes (Swedish Board of Agriculture, 2012), making abatement difficult to implement and monitor.
One potential criticism of the authors’ analysis is that in its focus on efficiency, it has ignored the importance of positive externalities that Sweden provides via ‘demonstration effects’. I would argue, however, that the previous government’s policy to increase the biofuel blending mandate in order to meet domestic emission targets in the road transport sector also suffered from a lack of demonstration effects, as it is highly dependent on imports. In the most recent statistics from the Swedish Energy Agency (2022), 92% of the raw materials for biodiesel in 2020 were imported. Similarly, 88% of the raw materials for ethanol were also imported. Such a high degree of import dependency makes an inefficient policy also very deficient in terms of demonstration effects.
One interesting result is the significantly negative effects on the agriculture and forestry sector in some of the scenarios. Even though these sectors are not directly affected by ESR policies, agriculture and forestry are both sensitive to the price of fuel. Although landowners in Sweden could stand to gain from increased demand for biofuel, the higher demand will probably be met by increased imports. In addition, the price of biofuel feedstocks is determined by international commodity markets. The CGE model used in the analysis does not explicitly model the land market, leaving a need for further research and a more in-depth analysis of the impact on the return to landowners.