A liability for the climate impact of foods for financing negative emissions

Abstract

Meeting the Paris Agreement's temperature objective requires the removal of significant amounts of carbon dioxide (CO₂) from the atmosphere, highlighting the urgent need for financing negative emission technologies. Given the difficulties in mitigating agricultural emissions, this paper proposes a “food climate liability fee.” The fee is based on the climate impact of food combined with an estimated cost-level of bioenergy with carbon capture and storage, and the generated income is earmarked for funding CO₂ removals. We explore the effects of implementing this fee using Sweden as an example, analyzing scenarios ranging from including all food-related emissions to a fee specifically on the harder-to-mitigate methane (CH₄) and nitrous oxide (N₂O) from high-impact animal-based foods. Targeting all emissions could generate sufficient financing for removals to balance emissions from Swedish food consumption. A more focused fee on high-impact foods could cover 85% of CH₄ and N₂O food-related emissions, of relevance if CO₂ emissions are addressed through other means.

Introduction

Meeting the Paris Agreement's temperature goal solely by reducing greenhouse gas (GHG) emissions is challenging, if not impossible. For more than a 50% chance of limiting global warming to 1.5 °C, the estimated remaining carbon dioxide (CO₂) budget is around 170 Gt for 2026–2100. Assuming 2025 emissions levels, the budget has exhausted already before 2030 (Friedlingstein et al., 2025). Consequently, most emission scenarios assessed by the Intergovernmental Panel on Climate Change (IPCC) extend this budget by introducing substantial amounts of negative emissions, that is, removing CO₂ directly from the atmosphere. The demand for CO₂ removal is driven by the immediate benefits of reducing atmospheric GHG concentrations, the necessity to offset residual emissions in order to achieve net-zero goals, and the requirement to address cumulative emissions that exceed the GHG budget of global temperature objectives, thereby enabling global temperatures to return to target levels (IPCC, 2023).

Removing CO₂ from the atmosphere comes with significant financial costs (see Section “Cost of achieving negative CO₂ emissions”). Given the large volumes of negative emissions needed, the question of who will finance these negative emissions is crucial (Bednar et al., 2024; Fridahl et al., 2023; Honegger et al., 2021; Lyngfelt et al., 2024). Generating income to finance government expenditure while reducing emissions is attractive, but incentivizing negative emissions, a public good not linked to marketable goods, is more complex (Fridahl et al., 2024; Lyngfelt et al., 2024). Without payment, there is no business case for large-scale CO₂ removal.

Substantial amounts of negative emissions will primarily occur in the latter part of this century, while a substantial sahre of the emissions they compensate for arise earlier. This leaves future generations with the cost of removing past emissions. Future governments will need to agree on how to share this burden, a challenging task given the high costs and lack of immediate returns (Lyngfelt et al., 2024).

The global food system significantly drives climate change, contributing between 21% and 37% of global anthropogenic GHG emissions (IPCC, 2023). To decrease food-related emissions in line with levels feasible with the Paris Agreement, decarbonizing the food supply chain has been pointed out as an absolute necessity, together with reducing CO₂ emissions due to land use changes in agriculture (e.g. Willett et al., 2019). However, apart from CO₂ emissions, methane (CH₄) and nitrous oxide (N₂O) from agricultural activities account for a large part of the GHG emissions from the food system. These emissions need to be tackled through production-side measures via improvements at the farm level, together with demand-side changes such as decreased consumption of meat and dairy products in favor of increased intake of plant-based foods (IPCC, 2023). To accelerate changes, various policy measures have been identified as necessary (UNEP, 2022).

However, a complete phase out of GHG emissions from agriculture is generally seen as unfeasible, which is why climate projections and national climate plans include future residual emissions from the sector. For example, Willett et al. (2019) project the remaining emissions of CH₄ and N₂O of about 5 GtCO_2eq in 2050. Within the examples of governments’ climate plans, Sweden's National Energy and Climate Plan states that “[s]ignificantly reducing greenhouse gas emissions from agriculture in a cost-effective and competitively neutral way is more difficult than in other sectors” (Government of Sweden, 2024: 241). Moreover, Austria's National Energy and Climate Plan highlights that “some sectors are hard-to-abate and cannot fully avoid their GHG emissions” and that “[t]his concerns in particular diffuse sources of GHG emissions in agriculture” (Government of Austria, 2024: 347). These residual agricultural emissions of CH₄ and N₂O, needed to sustain food consumption of the global population, would then need to be compensated for by additional CO₂ removals.

As a way to both reduce emissions from agriculture and to finance removals, this paper opts for a “food climate liability fee.” Such a fee would be proportional to food's climate impact and create revenues earmarked to finance the removal of a corresponding amount of CO₂ from the atmosphere. This fee would act as a disincentive to the consumption of food with high climate impact and ensure that the remaining emissions can be counterbalanced by removals. This approach aligns both with the producer liability principle, holding those who produce goods responsible for waste treatment, and the user pays principle, holding those who consume products responsible for their associated environmental impact.

This paper will present food climate liability fees using the Swedish market as an example. By estimating static effects on food prices and assessing the scale of negative emissions achievable under different conditions, we aim to illustrate the principle of a food climate liability fee and present a crude estimate of its food price effects. We also tentatively discuss uncertainties related to our proposal, as well as the feasibility of implementing the fee, a discussion which we hope will spur a dialogue on the need to finance removals in a low-emission future. The motivation for this approach includes the rapid depletion of the global CO₂ emissions budget, the necessity of sustained financing for negative emissions, the political risks of relying on funding negative emissions through national budgets, as well as the need to reduce emissions in agriculture, together with the difficulty of reducing CH₄ and N₂O emissions to zero.

Background

GHG emissions from the food system and policy measures to reduce these

Of the overall GHG emissions from the food system, agriculture and land use are predominant sources with between 70% and 80% of the emissions (e.g. Crippa et al., 2021; IPCC, 2019; Poore and Nemecek, 2018). The rest of the emissions correspond to those arising in pre-farm activities, for example, fertilizer manufacture, and post-farm activities, such as processing, refrigeration, packaging, and transportation. Of the agricultural emissions, the majority derive from biological processes such as CH₄ from feed digestion of ruminants, N₂O from fertilized soils, as well as CH₄ and N₂O from manure management (Crippa et al., 2021). Hence, in contrast to the energy and transport sectors in which emissions of CO₂ from the use of fossil fuels dominate, a large part of the GHG emissions from the food system is globally represented by CH₄ and N₂O.

Carbon taxes and emission trading systems (ETS) cover parts of the GHG emissions from the food system, such as CO₂ emissions from the use of fossil fuels in agriculture and the food industry, as well as N₂O emissions from mineral fertilizer production (World Bank, 2025). Suggestions to include agricultural emissions within the European Union (EU) ETSs have been brought forward (see e.g. Verschuuren et al., 2024) but the majority of GHG emissions from the food system currently remain untargeted.

Within initiatives at the country level, the government of Denmark is pioneering with a proposed fee on GHG emissions in agriculture, coupled with subsidies for investments in the agriculture sector's climate transition (Government of Denmark, 2024). As production fees and taxes for trade-exposed agricultural goods could create a comparative disadvantage, economic policy instruments on the demand side through taxes and subsidies have been suggested in several European countries (Lööv et al., 2013; SSNC, 2015; TAPPC, 2020a, 2020b). Such policy instruments have also been modeled in several studies, showing potential to decrease food-related GHG emissions as well as other environmental effects (Edjabou and Smed, 2013; Larsson et al., 2026; Moberg et al., 2021; Säll and Gren, 2015; Springmann et al., 2016; Wirsenius et al., 2011). Practical experience from climate-related demand-side taxes or fees on food is however still lacking.

Cost of achieving negative CO₂ emissions

Since completely phasing out agricultural GHG emissions is unfeasible, it is essential to design climate policies that can cover the cost of CO₂ removal to counterbalance the residual emissions. The cost for achieving these negative CO₂ emissions varies heavily depending on the choice of method (Fuss et al., 2018; Lyngfelt et al., 2024). Within the more prominent methods suggested, these include bioenergy with carbon capture and storage (BECCS), re-/afforestation, enhanced wetland and soil organic carbon content, biochar, enhanced weathering, direct air carbon capture and storage, and ocean liming (Shukla et al., 2022).

BECCS and re-/afforestation are key strategies in climate scenarios, with BECCS often dominating due to its high potential, secure carbon storage, and relatively low estimated costs. Biomass absorbs atmospheric CO₂ during its growth and by converting biomass into energy and capturing the resulting CO₂ for geological storage, BECCS achieves negative emissions. This technology is prominent in scenarios aiming for a 50% chance of limiting global warming to 1.5 °C by 2100, as highlighted in the IPCC's Sixth Assessment Report (Shukla et al., 2022).

BECCS technology is known, with experience in capturing, transporting, and storing CO₂. However, estimated costs for capture, transport, and storage of the CO₂ vary by site and range between USD 15 and 400/tCO₂. The lower cost reflects ethanol production near readily available storage sites, whereas the higher costs represent dedicated biomass-fired boilers further away from storage sites (Fuss, 2022). At a country level, Johnsson et al. (2020) estimated the costs of CO₂ capture at 28 plants in Sweden and found costs ranging between EUR 40 and 110/tCO₂, in which a major part of the biogenic CO₂ ranged between EUR 50 and 60/tCO₂.

Within costs examples from the industry currently supplying these services, Norwegian Aker Carbon Capture offers carbon capture and storage (CCS) at a fixed price of EUR 70–150/tCO₂ including investment, operation, transportation, and storage. The lower capture costs only apply if waste heat is available for capturing CO₂, which is rarely the case, making a total cost range of EUR 100–150/tCO₂ more realistic. Other examples include the Swedish reverse auction contract with Stockholm Exergi, where Sweden provides state aid of EUR 159/tCO₂. However, the price paid by the government for delivery of BECCS does not reflect the full cost, and about two-thirds of the revenues come from other sources, such as the EU Innovation Fund and revenues from selling “contribution claims” to private actors through offtake agreements (Energimyndigheten, 2025). In Denmark, BECCS auctions rewarded contracts to three suppliers, ranging between EUR 130 and 348/tCO₂. Like the case of Sweden, the Danish auction allows for stacking private revenues on top of public support, which means that the specific cost for BECCS may be higher than the granted auction contracts (Energistyrelsen, 2024).

Novel combustion technology capable of capturing CO₂ without costly and energy-intensive gas separation processes could potentially reduce capture costs to below EUR 30/tCO₂ (Lyngfelt et al., 2022). The technology has now been successfully demonstrated at the 5 MW scale (Li et al., 2025). A capture cost of, for example, EUR 150/tCO₂ is higher than current incentives for CO₂ reduction, such as the average EU ETS price of EUR 65/tCO₂ in 2024 (Statista, 2025) although negative emissions are not currently included in the ETS.

To better comprehend the impact of a fee of, for example, EUR 150/tCO₂, it can be multiplied by the global carbon intensity, projected to be 0.17 × 10⁻³ tCO₂/EUR in 2027 (Lyngfelt et al., 2022). This yields 0.025, which means that mitigating total global emissions at that cost would amount to 2.5% of global GDP. However, the overall societal cost of a fee would be lower, as emitters in many cases could avoid the fee through emission reductions achieved at lower costs.

Policy options to finance the cost of negative CO₂ emissions

Even when assuming a relatively low cost of CO₂ removals, few existing climate policy instruments are capable of incentivizing these. The challenge lies in finding ways to compensate actors with the technical potential for BECCS to cover the private capital and operational costs associated with generating CO₂ removal, when CO₂ removal is a global common good.

Proposals for overcoming this challenge include, for example, obliging fossil fuel companies to pay for CCS in proportion to their extraction of fossil fuels (Allen et al., 2009). In 2021, this was developed and branded as Carbon Takeback Obligations (Jenkins et al., 2021), both to regulate incentives for fossil CCS and to extend the obligation to include a gradually increasing fraction of removals with BECCS. As such, this approach could incentivize overcompensation by requiring CO₂ takeback to exceed 100%. Lyngfelt et al. (2024) propose a similar scheme, but focusing solely on financing removals. Lyngfelt et al. (2024) argue that reductions of fossil CO₂ emissions, including the use of CCS, are already incentivized through existing and planned CO₂ taxes or cap-and-trade systems.

A challenge with takeback schemes is their simultaneous implementation with fossil carbon extraction, which does not address the need to go below net zero, especially in a future where a vast majority of current fossil fuel extraction should no longer occur. Overcompensation may not be sufficient to achieve the large removals needed later in the century. Several studies address the temporal dislocation of identifying duty bearers among current emitters and the need for future removals. Bednar et al. (2021) discuss saving carbon tax revenues in funds for financing future removal, concluding that protecting these funds from diversion would be challenging. Instead, Bednar et al. (2021) suggest “Carbon Removal Obligations,” treating these debts similarly to financial debts, issued by managing authorities to commercial banks at a base rate, with commercial banks issuing these debts to emitters and being held liable for insolvent debtors. Reinsurance of this liability increases the interest debtors must pay.

Rickels et al. (2021, 2022) propose establishing a Carbon Central Bank to procure and stock removal credits, acting as a clearinghouse between removals and the EU ETS market, and using removal credits to stabilize the EU ETS price signal. One rationale behind the proposal is that early incentivization of removal activities is necessary, while pre-mature introduction of removal credits on the EU ETS allowance market could undermine emission reduction incentives.

A central concern in many proposed incentive structures is how to secure continued replenishment of funds to finance removals in a future with a limited number of emitters who could be held liable to compensate for emissions. Our proposal to explore liability for the climate impact of foods departs from the premise that a share of current CH₄ and N₂O and emissions from agriculture will remain in the future and need to be counterbalanced. It is therefore logical to explore options to raise funding for removals through taxing residual agricultural emissions, both to further incentivize reductions and to provide a secure and sustained source of funding for future removals in proportion to an expected significant source of residual emissions.

Materials and methods

To calculate relevant fees for food products on the Swedish market, we estimate the overall GHG emissions from the Swedish yearly consumption of food (see Section “GHG emissions from Swedish food consumption”) and then estimate the expected costs of the negative emission technologies to cover these emissions (see Section “Assumed costs for negative CO₂ emissions”). By dividing the expected costs over different sets of food products based on different rationales (see Section “Scenarios for dividing the costs of the negative emission technologies”), we then get estimations of the fees for the various food products.

GHG emissions from Swedish food consumption

We calculate the GHG emissions from Swedish yearly consumption of food by using estimates of GHG emissions per kg food from the SAFAD database (Röös et al., 2025) combined with consumption data from the Swedish Board of Agriculture (2023) and assuming a Swedish population of 10 million.

The SAFAD model builds on life cycle assessment (LCA) methodology and calculates the average GHG emissions of fossil and biogenic CO₂ and CH₄, as well as N₂O, directly associated with the production of 1 kg of foods and dishes available on the Swedish market. The data includes life cycle stages from cradle to consumer, that is, emissions that arise in pre-production, on farm (including those from land use and land use change), as well as in post-farm activities (transportation, packaging, and preparation) until the products are available for consumption at the household level, including food waste throughout the life cycle and including waste taking place at the consumer level.

To relate the impacts on climate change by the different GHGs, characterization factors of their global warming potential over 100 years (GWP₁₀₀) (IPCC, 2023) were used. Thus, emissions of biogenic CH₄ and N₂O in megatonnes (Mt) were multiplied by 27 and 273, respectively, to get MtCO_2eq, while fossil CH₄ was multiplied by 29.8. The climate impact data of all the foods are shown in Table S1 in the Supplemental materials.

We then multiply this data by the consumed amount of the food items, based on the statistics of direct food consumption by the Swedish Board of Agriculture (2023), using a 5-year average between 2018 and 2022 (presented in Table S1 in the Supplemental materials).

Assumed costs for negative CO₂ emissions

Based on the discussion in the Section “Policy options to finance the cost of negative CO₂ emissions,” we choose to exemplify the climate liability fee using BECCS as an example and assuming that BECCS removal of CO₂ is feasible at a cost of around EUR 150/tCO₂. We exemplify our case of a fee counterbalancing removals of residual emissions, using the concept of CO_2eq, and with a fee of EUR 150/tCO_2eq for food products. With more commercial experience, this cost is likely to be adjusted. At the time of writing, budding commercial experience indicates higher costs, while opportunities to learn indicate that these costs may fall over time.

Scenarios for dividing the costs of the negative emission technologies

There are a number of ways in which the expected costs could be divided. Here, we calculate fees for emissions based on four different scenarios. In these scenarios, a fee of EUR 150/tCO_2eq is applied to:

all emissions from all food products consumed on the Swedish market (“All GHGs, all foods”);

all emissions from food products with relatively high climate impact, that is, meat and dairy products, consumed on the Swedish market (“All GHGs, meat & dairy”);

emissions of CH₄ and N₂O from all food products consumed on the Swedish market (“CH₄ & N₂O, all foods”);

emissions of CH₄ and N₂O from meat and dairy products, consumed on the Swedish market (“CH₄ & N₂O, meat & dairy”).

In scenarios 3 and 4, the small amount of fossil CH₄ is not included, which we think is more logical to link to the fossil CO₂.

Scenario 1 could be argued to be “fair” in the sense of being “technology neutral,” that is, including all food products on the Swedish market in a non-discriminatory manner. However, the more food products that would be included in a fee system, the more complex the system would likely be, and more administration would be necessary, which would increase the costs of maintenance of such a system. Based on this, it could instead be argued that a fee can be imposed on the food products or categories causing the largest emissions. As mentioned, such products include many animal-based products, that is, meat and dairy products, which have been estimated to correspond to around 70% of the overall GHG emissions from the Swedish yearly consumption of food (Röös et al., 2024). As such, for scenarios 2 and 4, the fee would be divided over those food products, so these products would need to overcompensate for food products without a fee in order to raise enough funds to finance negative emissions that can offset the total climate impact of all food products.

Basing a fee targeting only the emissions of CH₄ and N₂O, while excluding CO₂, would be based on the rationale of CH₄ and N₂O being part of the hard-to-abate residual emissions from agriculture, while CO₂ will eventually be included in other measures for achieving climate targets and where new technology can be expected to curb emissions. However, as the current food market is globalized and a large share of the food products on the Swedish market are imported (Röös et al., 2024), it is not clear if, or when, emissions of CO₂ caused by food production in other parts of the world, would be included in such measures. Hence, it could still be relevant to include emissions of CO₂ in the negative emission technologies for the food sector. Imposing a fee on products based on their emission intensities of only CH₄ and N₂O could, on the other hand, be less complex as it would involve fewer emission sources.

Results

Incomes and potential emissions reductions from a food climate liability fee

Applying fees on all GHG emissions from all food categories (scenario 1) would generate EUR 2.8 billion annually based on current consumption patterns (Figure 1). This amount, based on the estimated cost for BECCS, would be sufficient to remove approximately 18.5 MtCO₂ from the atmosphere each year, corresponding to yearly emissions of Swedish food consumption. Limiting the fee to meat and dairy products (scenario 2) would yield an annual revenue of EUR 1.75 billion, corresponding to emission reductions of almost two-thirds of yearly emissions of food. As such, to offset all GHG emissions associated with Swedish food consumption, this would require a significant raise of fees for meat and dairy products by 50%, corresponding to a raise from EUR 150/tCO_2eq to EUR 237/tCO_2eq.

Figure 1.

Removals and incomes of the four scenarios.

Focusing on emissions of CH₄ and N₂O from all food products (scenario 3) would account for 8.9 MtCO_2eq per year (corresponding to about half of yearly emissions of food), generating EUR 1.34 billion annually. Further limiting the fee to including CH₄ and N₂O emissions by targeting only meat and dairy products (scenario 4) would reduce revenue to EUR 1.14 billion, that is, a reduction of 15% compared to covering all emissions of CH₄ and N₂O, corresponding to around 40% of the overall yearly GHG emissions from Swedish food consumption. To cover the difference between a fee targeting all CH₄ and N₂O emissions and only those from meat and dairy, the fees on meat and dairy would need to increase by 17%, from EUR 150/tCO_2eq to EUR 176/tCO_2eq.

Emissions of different food categories and corresponding fees

Looking at specific emissions per kg of food when including all GHG emissions, foods such as mutton, beef, instant coffee, cheese, crayfish, and canned fish were found to account for the highest emissions (Figure 2, see also the Supplemental materials), resulting in the highest corresponding fees per kg (Table 1). With regard to the cumulative emissions corresponding to yearly consumption of food, there was a significant contribution from meat and dairy categories, such as beef, hard cheese, sausages, pork, and milk, corresponding to 19.0%, 6.9%, 5.8%, 4.0%, and 4.1% respectively of yearly consumption (Figure 2). The dominance of meat and dairy categories was reflected and even emphasized when looking at both specific and cumulative emissions with CH₄ and N₂O emissions, excluding CO₂ emissions (Figure 3). Here, the contribution from meat and dairy products corresponded to 84% of the overall emissions. The same five food categories contributed to the largest shares of the overall emissions both with and without CO₂ emissions included, with beef accounting for 32% of the cumulative emissions, whereas hard cheese, sausages, milk, and pork contributed to 9.8%, 7.2%, 5.5%, and 4.4% of the emissions respectively. Notably, some categories in green in Figures 2 and 3 are compound foods, such as buns, pizza, and biscuits. It is likely that a fee targeting meat and dairy products would also capture a significant portion of these emissions, suggesting that the above share may be underestimated.

Figure 2.

Specific emissions for different food products versus their cumulative emissions, including CO₂ and fossil CH₄. The food categories are sorted based on cumulative emissions per capita. The colors indicate food product category: orange is meat products, yellow is dairy products, and green is other food products.

Figure 3.

Specific emissions of different food products versus their cumulative emissions per capita, excluding CO₂. The food categories are sorted based on cumulative emissions per capita. The colors indicate food product category: orange is meat products, yellow is dairy products, and green is other food products.

Table 1.

Fees per kg of selected food categories together with the corresponding cost per capita, with a fee corresponding to EUR 0.15/kgCO₂e.

	Fee, EUR/kg or EUR/L food product		Corresponding cost, EUR/capita and year
	Incl. CO2	Excl. CO2	Incl. CO2	Excl. CO2
Beef	4.7	3.8	53.7	43.3
Hard cheese	1.4	0.97	19.1	13.1
Sausage (Swedish)	1.2	0.74	16.0	9.6
Milk	0.17	0.11	11.2	7.3
Pork	0.86	0.45	11.3	6.0
Mutton	5.7	4.2	7.1	5.2
Other cheese	1.4	0.97	6.2	4.3
Buns a	0.45	0.22	6.5	3.2
Milk chocolate	0.36	0.19	5.7	3.0
Yoghurt	0.18	0.11	4.6	2.9
Wheat bread a	0.17	0.06	8.3	2.8
Chicken	0.35	0.12	7.3	2.6
Ice cream	0.43	0.28	3.9	2.5
Ham, pork	1.08	0.56	4.6	2.4
Rice a	0.51	0.37	3.1	2.2
Butter	1.10	0.74	2.9	2.0
Pizza, etc. a	0.56	0.22	4.9	1.9
Mayonnaise a	0.28	0.12	4.5	1.9
Cream	0.55	0.38	2.4	1.7
Biscuits a	0.47	0.26	3.0	1.6
Sum			186	119
Coffee beans a	1.37	0.18	10	1.3
Eggs a	0.19	0.10	2.2	1.1
Juice, orange a	0.22	0.04	3.7	0.65
Potatoes a	0.04	0.01	2.0	0.43
Salmon a	1.02	0.40	0.9	0.37
Pasta a	0.18	0.02	1.8	0.23
White sugar a	0.14	0.03	0.6	0.14
Tomatoes a	0.18	0.01	1.6	0.10
Cocoa powder a	0.19	0.03	0.4	0.08
Cod a	1.02	0.00	2.7	0.00
Total sum			277	134

Data are given both for emissions with and without CO₂.

a

Categories are non-meat and non-dairy, excluded from scenarios 2 and 4.

Discussion

Comparison of removals from a fee and current emissions in Sweden

The range of removals necessary to offset food-related consumption-based emissions in scenarios 1–4 amounts to between 7.6 and 18.5 MtCO₂ per year. In comparison, 36 Mt of fossil CO₂ emissions were generated from Swedish production in 2023. Additionally, 52 Mt of biogenic CO₂ were generated from biomass fuel combustion (Statistics Sweden, 2025), with about half arising from major point sources (Swedish Environmental Protection Agency, 2025). The 2020 Swedish government inquiry on negative emissions (SOU 2020:4, n.d.) estimated that the technical potential for BECCS in Sweden is well above 10 MtCO₂, and probably closer to 20 MtCO₂. The Swedish net-zero target for 2045 includes an 85% reduction in emissions compared to the 1990 levels. The remaining 15%, roughly 11 MtCO_2eq, is to be offset to cover emissions that are difficult to mitigate, which in the scenarios underpinning the government inquiry are mainly CH₄ and N₂O emissions from agriculture.

Effects and feasibility of a food climate liability fee

Environmental and social effects of a fee

In this paper, we exemplify how much income would be generated through a food climate liability fee and the potential corresponding removals achieved in each of the cases, assuming static consumer behavior. In reality, a climate liability fee would have dynamic effects on consumption that would result in reduced emissions from food consumption. As such, a food climate liability fee addresses both the need for continued emissions reductions, also in the agriculture sector, as well as the need to finance removal and balance residual hard-to-abate emissions from diffuse sources of CH₄ and N₂O in future low-emission agriculture systems. A further analysis of the dynamic effects of a fee in terms of consumption changes is, however, out of scope for this paper.

Within the body of research investigating the effects of climate taxation (see Ran et al., 2024 for a review), Moberg et al. (2021) estimated reductions of GHG emissions from a tax targeting all foods on the Swedish market to around a 10% decrease, whereas a tax targeting only animal-based products, or only beef, was found to yield emissions reductions of between 4% and 8% of overall GHG emissions from Swedish yearly consumption of food. Moberg et al. (2021) as well as Larsson et al. (2026) also discuss other tax options such as increasing value-added tax (VAT) levels on high-impact animal foods while subsidizing certain plant-based foods such as fruits and vegetables through decreased VAT levels, as both of these are associated with lower emissions per kg and are to be encouraged due to being more nutritious. For such scenarios, the net emissions reductions are estimated to be in the range of 4–6% (Larsson et al., 2026; Moberg et al., 2021), and would likely have other benefits such as positive health outcomes. Moreover, such policy packages could also generate higher acceptance, compared to implementing taxation solely on animal-based products or meat (Larsson et al., 2026).

Larsson et al. (2026) and Moberg et al. (2021) also discuss synergies between a climate tax on decreasing both GHG emissions as well as alleviating pressures on other environmental categories from food consumption, such as resource use and pressures on biodiversity. Potential goal conflicts stemming from a climate tax, and thus likely a food climate liability fee, could include those of decreasing pasture area and thus biodiversity in these areas, which could be further stimulated through policy instruments targeting maintenance of semi-natural pastures, for example, payments to farmers for managing such areas (Larsson et al., 2020).

In relation to distributional effects from a climate liability fee, it is likely that it would be regressive, thus affecting low-income households disproportionately (as for a climate tax, see e.g. Larsson et al., 2026). However, Larsson et al. (2026) highlight that a policy package simultaneously targeting high-impacting foods, subsidizing low-impacting and nutritious food, while also redistributing the incomes from a tax, could be a possible pathway to avoid financial burdens for low-income groups.

With regard to health effects from a climate liability fee, previous research has found potential negative effects on health from climate taxation, where, for example, Larsson et al. (2026) found an increased risk of cancer from the reduction of meat consumption, as this was found to coincide with a reduction in fruit and vegetable consumption. Moberg et al. (2021) found that a climate tax could reduce overall energy intake, but without leading to insufficient recommended energy intake, whereas Röös et al. (2021) pointed out that a decrease in food consumption due to climate taxation could lead to deficient nutrient intake regarding, for example, vitamin D, folate, and iron.

By levying a fee at the point of consumption rather than production, a fee can influence dietary choices without causing negative impacts on the competitiveness of domestic producers (Larsson et al., 2026). Potential sectoral transitions could, in the long run, be necessary for producers within food categories being negatively affected, such as meat and dairy producers. To further safeguard the resilience of domestic food production, additional support could be needed through governmental subsidies for such sectoral transitions.

Practical considerations of a fee

Moberg et al. (2019) discuss administering a climate tax on foods, noting that current Swedish excise taxes, for example, those on tobacco and alcohol, are paid by producers or importers. Thus, supermarkets, restaurants, and other distributors of such products buy products that are already taxed. Moberg et al. (2019) suggest that a climate tax on food could also be administered as an excise tax, which would require taxes to be paid by major producers of foods such as mills, dairy companies, and slaughterhouses. This approach could be applied to collect a fee in proportion to the climate impact of food. Ultimately, however, the consumer would bear the cost of these fees.

Allocating funds generated by the fee toward removal activities

There are several ways to organize the suggested liability. One option could include investing the fees in a removals fund that could manage results-based payments based on certified removals. This would link the climate impact liability to finance removal procurement for a credit reserve with the primary purpose to stabilize prices in emissions trading systems (Rickels et al., 2022) or to create credit-demand in removals trading systems (Meyer-Ohlendorf, 2023), creating an independent foundation that use various investment vehicles to mobilize additional financial resources toward removals (European Commission, 2025), or enforcing retailers to pay deposit deeds corresponding to the emissions associated with the products sold (Lyngfelt et al., 2024).

If climate liability fees were introduced, the delay from the payment of the fee to the time the corresponding negative emission can be performed would need to be considered. If these fees were to be saved in public funds for decades, there are obvious risks for diversion. Organizing the liability around deposit deeds that create private property, protected by strong private property laws in many parts of the world—especially in high-income countries—could safeguard against diversion. Deposit deeds would be redeemed, with returns, when the owner of the deposit deed can present the corresponding certified negative emissions. Because of returns on investments managed by the trustee of the fund, the value of the deposit deeds will most likely increase with time, augmenting the incentive to make negative emissions and secure the deposit deeds from becoming worthless and not incentivize removals. The deposit deeds would also be tradable, which would be helpful as the companies that need to make the deposits are not likely to have a major interest in keeping them and, when possible, buying negative emissions. Thus, the deposit deeds can be bought by companies specialized in promoting future negative emissions. Further, the fact that the deposit deeds are the property of owners should be a significant protection against diversion. Such a deposit system, Atmospheric CO₂ Removal Deposits (ACORDs), has previously been proposed for fossil CO₂ emissions (Lyngfelt et al., 2024), but can also be used for GHGs like CH₄ and N₂O, translating the needed removals by CO₂ equivalents. The principle of ACORDs for food-related emissions is illustrated in Figure 4.

Figure 4.

Illustration of a deposit deed system for food-related GHG emissions. GHG: greenhouse gas.

Uncertainty in input data and modeling choices

Uncertainties in BECCS cost estimates

The costs revealed through the Swedish and Danish BECCS auctions (see Section “Cost of achieving negative CO₂ emissions”) highlight the significant uncertainties in BECCS cost estimates reported in previous research. While this paper proposes a food climate liability and illustrates fee levels using Swedish food climate impact data, it is important to note that the fee level is highly sensitive to CO₂ removal costs. These estimates vary significantly from the illustrative example of EUR 150/ton of removed CO₂, with reported costs in the range between almost a third of this assumption, while recent reverse auctions indicate that costs may be up to three times higher. Actual costs are likely to vary significantly depending on site-specific circumstances.

Given the heterogeneous nature of the marginal removal cost of BECCS, any policy that operationalizes a liability for the climate impact of food consumption must include adaptive components. Unlike our illustrative case, which assumes a fixed cost of EUR 150/tCO_2eq, the policy should be responsive to actual costs as they are revealed through ongoing commercial experience and learning.

Uncertainties in calculations of the climate impact of Swedish food consumption

Assessing the climate impact of food products and diets involves both model and data uncertainties (Moberg et al., 2020; Röös et al., 2024). A large part of the emissions from the food system are variable due to factors such as climate conditions and soil characteristics. Additionally, calculations include various methodological choices which can substantially affect the results, such as which factor to use to characterize the impacts of different GHGs, and how to allocate emissions in production systems with several outputs. Moreover, as the Swedish diet includes foods from various origins, consideration should ideally be taken of specific production systems and conditions at the production sites, which may be inhibited by a lack of knowledge of origin as well as a lack of detailed inventory data.

To assess the climate impact of the Swedish diet, we use data on direct food consumption from the Swedish Board of Agriculture (2023), which represents the amount of food available for consumption. As such, this food could either be eaten or wasted, but we use it as a proxy for the amount of food that needs to be produced to sustain the Swedish diet.

Choice of emissions to be included in a food climate liability fee

In this paper, we demonstrate a food climate liability fee with four scenarios ranging from including all GHG emissions associated with yearly consumption of food, to emissions targeting only CH₄ and N₂O from the consumption of meat and dairy products. As discussed in the Section “Cost of achieving negative CO₂ emissions,” parts of the emissions arising from the food system are targeted by taxation on CO₂ emissions or included in the EU ETS. As such, for a potential future climate liability fee, it would be essential to discuss which emissions to include or exclude based on the emissions that are currently covered by policy instruments (see e.g. Gren et al., 2019 for a discussion on food taxation), together with the feasibility of a fee with regard to trade-offs between acceptance, administrative burden, and cost-efficiency (see e.g. Moberg et al., 2019).

Conclusions

This study proposed implementing food climate liability fees, that is, fees on food in Sweden, to finance negative emissions. These fees aim both to incentivize emission reduction via dietary changes—from high- to low-emitting food products—and to finance CO₂ removal to enable neutralize residual emissions from food consumption, particularly CH₄ and N₂O, which are difficult to fully eliminate. Therefore, the implementation of the fee would address two significant risks associated with the climate transformation of the agricultural sector. First, it mitigates the moral hazard of mitigation deterrence, in which emissions reductions are delayed or avoided based on modeled future removal potentials that may not materialize. Second, it reduces the risk of removal deterrence, ensuring that essential removal activities required to offset hard-to-abate agricultural emissions are undertaken by providing sufficient funding.

The fee was exemplified via four scenarios: scenarios 1 and 2 included fees for all GHGs, while scenarios 3 and 4 focused only on CH₄ and N₂O emissions. Scenarios 1 and 3 applied fees to all food products, whereas scenarios 2 and 4 targeted only meat and dairy products, due to their high climate impact per kg of food.

The analysis of GHG emissions from Swedish food consumption led to the following conclusions: yearly food-related emissions in Sweden are estimated at 1.85 tCO_2eq per capita or 18.5 MtCO_2eq, with 8.9 Mt from CH₄ and N₂O. Implementing a fee on all food products would be complex and burdensome. A simpler approach could focus on high-impact categories like meat and dairy.

A fee on meat and dairy, responsible for almost two-thirds of total CO₂, CH₄, and N₂O emissions (11.7 MtCO₂/year), would address all CH₄ and N₂O emissions (hard to reduce otherwise) as well as 30% of CO₂ emissions. This fee could generate EUR 1.14 billion annually, covering the cost of negative emissions of 7.6 Mt/year, equating to 85% of CH₄ and N₂O emissions.

The average yearly cost for a Swede would be EUR 114 per capita, potentially lower if consumption decreases due to the fee. The fee should target high-emission foods, raising funds for negative emissions and reducing demand for such food, which in turn reduces reliance on negative emission technologies for net-zero goals.

Supplemental Material

sj-docx-1-dcc-10.1177_29768659261449305 - Supplemental material for A liability for the climate impact of foods for financing negative emissions

Supplemental material, sj-docx-1-dcc-10.1177_29768659261449305 for A liability for the climate impact of foods for financing negative emissions by Emma Moberg, Anders Lyngfelt and Mathias Fridahl in Dialogues on Climate Change