Effects of Carbon Tax Redistribution Schemes on Energy Welfare of Households in Germany

Abstract

Increasing carbon taxes have the potential to disproportionately affect lower income households, who lack the financial and decision-making capacity to undertake investments to shift the underlying energy system. Carbon tax redistribution schemes aim to equalise the negative impacts experienced especially in lower income households and tenants. Through the development of a dedicated energy system optimisation model, our assessment evaluated the distributional impacts of carbon tax redistribution schemes on the evolution of energy use, emissions and energy welfare in different household types in Germany.

1 Introduction

The German ‘Energiewende’, or energy transition, aims to transform the energy system from fossil and nuclear-based energy sources to include more renewable energy sources and energy efficiency (Bundesregierung 2010). Efforts from the household sector will contribute towards achieving the targets in 2050 of an overall reduction in GHGs of -80 to -95% (compared to 1990 levels), 60% share of renewable energy in the final energy consumption, 80% share of renewables in gross electricity consumption, 50% less energy and 25% less electricity compared to 2008 levels, and 80% less primary energy demand in buildings compared to 2008 (BMWi 2018). This move towards a more renewable and energy-efficient energy system is well-founded and even required in order to avert the most detrimental effects of climate change, which are already felt around the world. Underpinned by the Paris Agreement, action on climate includes the transition to a low carbon economy, which entails alternative, clean technologies resulting in the reduction of emissions. The next 10 years will be critical for climate action, recognised by the fact that on the EU level climate milestone targets have been increased to a reduction by 55% by 2030. Given this urgency to act, Germany—within the European energy policy framework—has defined targets to achieve carbon neutrality by the mid-century with milestones in 2030 of 30% renewables in the total energy consumption and a reduction of energy consumption of 30% (BMU 2019; BMWi 2020). The national target to reduce emissions was increased from 55 to 65% in 2022 and sector targets will be redefined in 2024 to align with this more ambitious national objective (BMWK 2022). The strategies relevant to the household sector developed from these aspirations require the renovation of the building envelope, replacement of inefficient, fossil fuel-driven heaters with efficient, renewable-based heating systems, increasing energy efficiency of appliances, the installation of decentralised energy generation technologies and discouraging fossil fuel consumption through the introduction of a carbon tax (BMU 2019). However, achieving this roll-out poses a challenge since it depends on the co-investment by the private sector where a trend of decreasing willingness-to-pay is evident despite the overall consensus that environmental goals should be reached (Andor et al. 2017). It is estimated that the energy transition in Germany will require an annual investment of between 15–40 billion € annually for the next 30 years—although it should be noted that the full costs of the energy transition are unknown and estimates need to take into account the damage costs of not taking action as well as incorporate the savings (Agora Energiewende 2018). This highlights concerns about the economic impact of the energy transition in the debate and gives rise to the need to define what is meant by growth and include assurances for well-being (social and economic) within the limits of the environment.

At the heart of the energy transition, consumers are key to unlocking the potential to achieve energy and climate change targets in Germany. These renewable energy and energy efficiency targets of the energy transition will remain unachievable without the active participation of customers (households), as is ingrained in the European energy policy direction. Nearly, a third of Germany’s final energy use is directly attributable to households with two-thirds of this consumption met with fossil fuels. As such, households have a significant part to play in transforming the energy system. To encourage households to shift from fossil fuels and to invest in renewables and energy efficiency (including building renovations), financial disincentives, such as a carbon tax are added to fossil fuel consumption. However, the majority of German households lack the financial or decision-making power to make the necessary investments in energy-efficient and renewable technologies. Policies and national-level energy planning are typically based on energy system modelling assessments, which assume averaged households. However, basing assessments on a homogenous household sector may result in the overestimation of the contributions from this sector towards achieving energy and emission targets, and are not capable of addressing the energy welfare concerns of households.

Energy poverty is on the rise across Europe—and is further impacted by geopolitical insecurity such as in the case of the Ukraine crisis resulting in supply constraints and energy price increases and the economic impact of the COVID-19 pandemic, which poses particular challenges for the resilience in lower income households and could impact the energy transition pathway taken to reduce emissions or move away from fossil fuels. Assessments that do not include consideration for the heterogeneity of the household sector within energy planning will likely result in estimates falling short of the energy and climate change targets, a lack of active participation and a reduction in the energy welfare in the vulnerable population. Understanding the significance of this is crucial to ensuring all households have the opportunity to participate in the energy transition and are not disproportionately disadvantaged. The key challenges, therefore, are around the need to understand the differentiated needs and capabilities of households as well as the drivers influencing energy-related investment and consumption patterns and to incorporate these into a process, which will also allow a long-term assessment of the influences on the energy system with a view to achieving the energy transition targets. This characterisation of the household sector is essential to be able to assess the distributional impacts of energy-related policies on the energy welfare of specific types of households.

This chapter begins with outlining the significance of household energy vulnerability within the context of the energy transition in Sect. 2. This is followed by a characterisation of the household sector differentiated by socio-economic parameters and those related to the built environment in Sect. 3. Section 4 will summarise the discourse on carbon taxes and different redistribution approaches. Section 5 describes the methodology designed to assess investment and energy consumption patterns, emissions and the energy welfare of households. Section 6 discusses some results derived from the energy system optimisation model where scenarios explore alternative redistribution approaches, while Sect. 7 concludes with a short discussion.

2 Household Energy Vulnerability

Energy poverty is increasingly prevalent in the energy transition discourse. It is no longer a question of if the energy transition will benefit lower income households, but how to enable this (European Commission 2016; Sunderland et al. 2020; Ugarte et al. 2016). The lack of a common understanding across Europe on what energy poverty is results in fragmented approaches or discounting its significance entirely (Pye et al. 2015). This is a problem because there are indications of an increasing trend and current strategies risk leaving lower income households behind (Dobbins et al. 2019). Household energy vulnerability is commonly termed ‘energy poverty’ and understood as ‘a situation where households are not able to adequately meet their energy needs at affordable cost, and is caused by a combination of overlap** factors including low income, high energy prices, poorly insulated buildings and inefficient technologies and sometimes limited access to clean and affordable energy sources’ (Dobbins et al. 2019). The common policy approach to address household energy vulnerability is to provide financial support through the social welfare system. However, this does not allow for the possibility to address the cross-sectoral nature of the issue with policies directed towards alleviating the energy deficiencies related to the structural causes of household energy vulnerability.

While there is consensus that energy should be affordable and is outlined as a key pillar of the energy policy architecture, there is no consensus on how the inability to afford energy should be addressed in Germany. Although a formal definition of energy poverty does not exist in the German context, there has been a noticeable increase in the number of households who struggle to afford adequate energy services such that approximately 11–21% of German households are estimated to live in energy poverty (Bleckmann et al. 2016; Heindl 2014; Pye et al. 2015). Energy affordability is a key component in defining energy poverty, which can have an even greater impact on low-income households, who typically live in less efficient buildings and are often tenants. Affordability remains a central goal within the framework of the goals of the energy transition to decarbonise the energy system, to increase energy efficiency and the contribution of renewable energies (Bundesregierung 2010). Germany is among the countries with the highest electricity and gas prices in Europe, with an estimated 17.4% of the population spending twice the median on energy in 2015. According to European estimates, 3–18% of the population in Germany are affected or at risk of energy poverty (EPOV 2021). Missing repayments led to more than 230,000 electricity and 24,000 gas outages in 2020 (BNetzA 2021), putting households in a cycle of debt and outages that further increases the difficulty of meeting basic energy needs (Bouzarovski et al. 2021). Due to the COVID-19 pandemic and energy price increases related to the war in Ukraine, evidence shows that these aspects are exacerbated and will negatively affect households further. This has the potential to further affect household energy wealth (Dobbins et al. 2019; Schultz 2022) and give rise to a unique opportunity to take action to resolve the multiple inequalities exposed by these crises (European Commission 2020). Nonetheless, while a definition is important to gain agreement and clarity, it is still possible to undertake an evaluation of the energy welfare of households relative to other households.

There is a need to be able to assess energy poverty by classifying the differentiated needs and capabilities of households. The exclusion of the consideration of this inequality is further compacted by the fact that current policies determining strategies for the household sector are based on modelling assessments, which assume a homogenous population and monitoring benchmarks for policies are gauged according to average households (BMWi 2018). This oversimplifies the assumptions for the household sector and leads to one technology (and therefore policies, measures and targets) identified as the most cost-effective solution to meet a particular demand. An average household does not adequately capture the observed technological diversity and the differences in investment decisions and consumption behaviour across different types of households and does not account for barriers to actual investment behaviour on the part of this sector. Therefore, there is a need to differentiate between the financial and decision-making ability of different households to be able to better determine how to meet the required investment demands leading to the achievement of sector-specific renewable energy and energy efficiency targets, especially when aiming to stimulate an increase in the numbers of prosumers, which is contingent on the mobilisation of capital from the private sector. To be able to determine how policies such as carbon taxes affect households and influence household energy vulnerability, it is insufficient to use methods applied to averaged households. A holistic methodology will account for the differentiated situation of households within the context of the energy transition.

3 Characterisation of the Household Sector in Germany

The characterisation of the household sector in Germany within the context of the energy transition is based on an evaluation of the drivers of energy-related investment and consumption patterns. Households are responsible for a significant share of the final energy consumption with 27.5% of the total final energy domestic consumption (excluding mobility) and 10.1% of the total greenhouse gas emissions (AGEB 2019; BMU 2018, 2019; BMWi 2021b). Households in Germany are expected to increase shares of renewable energy use in heating and self-generated electricity while also decreasing energy consumption (or increasing energy efficiency) in line with energy transition targets (BMWi 2019). The majority of household energy consumption goes towards end-uses for space heating and water heating met largely with gas and oil with the average household in Germany consuming 57 GJ annually, spending 1,644 € on consumption (direct household energy expenditure excluding mobility) and 564 € on investments (linked to indirect energy expenditure) in 2018 (BMWi 2021a; Destatis 2018). Expenditure is closely coupled to income and the analysis conducted will show that there is a mismatch between the expected financial capacity of the majority of households according to their income levels and that of the average household. Lower income households are disadvantaged on two fronts: lack of capital and lack of decision-making power. Income is a key factor for the ability to alter the household energy infrastructure and degree of reliance on fossil fuels. Key to ensuring the success of the energy transition’s objectives is the mobilisation of capital from the private sector. However, investment behaviour is not linear and not always rational. The investment and consumption behaviour of different actors is defined along socio-economic characteristics, preferences, financial capacities, techno-economic aspects and is guided by policy. The influence of these policies and measures on the investment and consumption behaviour of particular low carbon, energy efficient and/or renewable technologies can be further attributed to individual ideals and limited by purchasing power. This Section will describe the disaggregation of the household sector into distinct socio-economic profiles according to differentiated investment and consumption profiles.

3.1 Socio-Economic Disaggregation of Households

In order to be able to analyse the impact of policies and the opportunities households have as actors in the energy system to reduce fossil fuel consumption and emissions, and increase renewables, households need to be characterised by different socio-economic parameters. The disaggregation of the heterogeneous actors within the household sector were categorised considering the major drivers of energy demand based on socio-economic characteristics beginning with building type (single-family and multi-family homes), tenure status (owner/tenant), location (urban/rural) and then income group (disposable income, savings). The basic demographic development of Germany is the main driver of energy consumption as this determines the number and size of homes as well as the associated energy service demands (BMU 2019; Möller-Ühlken and Kuckshinrichs 2007). In 2018, the population increased to 80.3 million living in 40.7 million households with an average household size of 1.97 people per household (Destatis 2013; EUROSTAT 2020). In 2018, which is used as the analysis year for the case study for reasons of data availability, 58% of households in Germany live in rented apartments. Over 88% of households in the lowest income group are tenants, while the home ownership rate in the highest income group is over 75%. A total of 35% of households live in single-family houses, with the highest income group accounting for 30% of all single-family houses. Within this income group, however, more than 68% of all households live in single-family houses. The number of people per household also increases with income: 1.1 people per household live in the lowest income group, and 3.1 people per household in the highest income group (Destatis 2018b).

Income and expenditure are a central component of the socio-economic characterisation of households and determinant of their energy consumption. The disaggregation of the residential sector by income groups for the integration into modelling assessments is not often undertaken, but income has been recognised as a key driver of energy consumption as well as a limit for households to achieve a certain level of energy services in the home (Cayla et al. 2011). Disposable income determines the availability of capital which enables (or restricts) a household to invest in technologies and consume energy (Alberini et al. 2011; Cayla et al. 2011; Kaza 2010; Longhi 2015; Vassileva et al. 2012). Figure 1 shows the average shares of direct (operating costs of consumption) and indirect (investments) monthly energy-related expenditures and shares of income by income groups per household in Germany in 2018 (Destatis 2018b). The income groups are categorised in the national statistics database by household monthly income.

Fig. 1

Income and expenditure by income group per capita in Germany, 2018. Source own graph based on (Destatis 2018a) as given in (Dobbins 2022)

The indirect investments include expenditure on appliances, energy-related home maintenance and renovations. On average, each person spends a total of about 100€ per month on direct and indirect energy expenditures (193€ per household). However, the population per household distribution varies within income groups. The average person spends around 74€ representing 2% of their monthly income on energy consumption (direct energy expenditures and 146€ per household), but while a person from the lowest income group would spend 82€ or just over 11% (82€ per household), a person from the highest income group would spend around 67€ and or under 1% of their income (193€ per household). On average, people from all income groups spend less than ~ 1% of their income on indirect energy expenditures (energy-related investments). These expenditure patterns reveal that the spending on the upgrading of appliances and the home (indirect expenditures) enable lower energy bills. This reflects that higher income households have more disposable income to spend on energy-efficient technologies, thereby translating into savings on current energy expenditures. Investments on indirect energy (i.e. investments in household appliances or housing maintenance and renovations), increase with income (Destatis 2018a). Household sizes (people per household) also increase with income, which means that increased investments benefit a greater number of people. This is critical as the larger investment requirements do not scale proportionately with the number of people, so a single-person household requires one heating system, one building envelope renovation or one refrigerator much the same as a three-person household.

Energy efficiency is outlined as a fundamental step towards not only achieving the energy transition targets but also alleviating energy poverty in European legislation as this addresses some of the underlying causes of energy poverty (European Parliament 2018). Investments in energy efficiency increase with income, which underscores that the greatest energy efficiency potential often resides in the appliances used and buildings occupied by lower income households, who by default, often do not have the financial capital for the high-upfront costs of investments nor the decision-making power as tenants. The challenge will be to mobilise investment in the lower income and rental sector and this begins with the acknowledgement and inclusion of the variation in investment capabilities of households in energy system modelling leading to sector targets and policy measures.

The potential to afford the high upfront investment costs is examined by compiling the potential monthly financial savings accumulated per income group. Less than a quarter (22.2%) of all households save more than the average household with approximately 22.2% and 539€ per month in 2018, which could be considered theoretically available for investments in renewable energy and energy-efficient technologies or building upgrades. While the share of homeowners in the upper income groups has increased over time, the households which both have higher than average savings and are homeowners represents just 16.7% of all households, which underscores the limitations in the potential of these actors to be able to carry the burden of achieving the household energy transition targets.

3.2 Household Energy-Related Investment and Consumption Patterns

As income increases so does the size of the dwelling and the number of people per household. The socio-economic parameters and the living situation determine the energy consumption profile of a household and this correlates directly with the financing and decision-making ability of a household to make the necessary investments to reduce its dependence on CO₂-emitting energy sources and thus to a CO₂ price to be able to react.

Based on the population distribution into the actor groups, an energy balance was developed for 56 distinct profiles that took into account the differentiation of various actors or groups within the household sector where it was first necessary to characterise the drivers of energy consumption. While some of the drivers of household energy consumption are interlinked and cannot be distinguished which influence on energy consumption they have from one another, they can nonetheless be summarised into the following key categories: (i) demographics, (ii) income and expenditure, (iii) dwelling characteristics (including tenure, location, building type and heating structure), (iv) access and use of self-generation technologies, (v) appliance stock and use, and (vi) energy efficiency—current status and potential. Decisive to all of these demand categories is also the access to infrastructure for specific fuel types and the capability to react to price signals that may shift households to alternate between different options.

Level of urbanisation influences the energy demand and the types of technology investments made in residential buildings due in large part to the access to different energy sources, such as grid-based energy sources like district heating and gas, or some renewable energy carriers (Druckman and Jackson 2008; Kramer 2010; Satterthwaite 2009). The level of access, determined through the level of urbanisation, drives the dependency on specific fuel types to fulfil energy service demands, such as space heating and will, therefore, result in different consumption patterns (Arbabi and Mayfield 2016; Kleinhückelkotten et al. 2016; Kramer 2010). Home ownership is a significant determinant of energy consumption, which is typically characterised by greater living space and appliance ownership (Destatis 2014; Frondel and Kussel 2018; Li and Just 2018; Schlomann et al. 2004). The greatest potential for energy savings in the household sector lies in buildings, but one obstacle to increased uptake of decentralised energy supply systems and energy efficiency of the building envelope could be the ownership structure (Bird and Hernández 2012; Frondel et al. 2015; Kockat and Rohde 2012; Scott 1997). The building stock is the main determinant of the overall energy consumption in the household sector due to the significance of space heating consumption in overall household consumption. The energy consumption of the residential building stock is mainly characterised through the number of units per building (building type), the age and floor area. The building typology is commonly confined to two main building types: single-family homes (SFH) and multi-family homes (MFH) (Cischinsky and Diefenbach 2018; Diefenbach and Clausnitzer 2010; IWU 2012; Kockat and Rohde 2012; TABULA 2015).

Applying these energy drivers results in energy-related investment and consumption patterns by the socio-economic profiles. The energy consumption and CO₂ emission profiles of the different households form the basis for the assessment of the effects of CO₂ pricing and the possible relief through redistribution of the revenue. A significant share of fossil fuel consumption also comes from transport fuel consumption, which increases per capita as income increases, as shown in Fig. 2. The profiles for household-related transport are based on the German Mobility Panel (Ahanchian et al. 2020) and are calibrated according to (BMWi 2021b). Lower income households rely to a greater extent on fossil-based fuels, such as oil and gas, than higher income households and since lower income household sizes are smaller, the cost burden is condensed. This shows that the share of energy sources used for heating purposes in the energy consumption profile of low-income households is larger per person, while the share of energy used for transport is lower. This trend is reversed as income increases, which also leads to lower CO₂ emissions per person, however since the number of occupants per household increases as income increases, the total household consumption and CO₂ emissions increase with income. Compared to the average household, households with low incomes have higher per capita energy consumption overall and twice the energy consumption for heating. In contrast, the CO₂ emissions per household in households with higher incomes are well above the German average. These differentiated energy consumption patterns need to be considered when implementing carbon tax policies because of the impact on specific households and their ability to take action.

Fig. 2

Energy consumption for heating and transport per person and CO₂ emissions per household by income group, 2018. Source Own calculations based on (Dobbins 2022) according to (Destatis 2018b; BMWi 2021)

4 Carbon Taxes and Redistribution Policies

Since the climate crisis requires measures to reduce greenhouse gas emissions, CO₂ taxes and emissions trading systems have also established themselves as important measures to reduce emissions (Agora Verkehrswende und Agora Energiewende 2019; Vermittlungsausschuss 2019). Carbon pricing has the crucial steering function of signalling to consumers that using the environment as a carbon sink comes at a clear price. These can be applied to the supply sector to encourage electricity generation towards alternative-based energy sources, e.g. renewables. Similarly, the tax can be applied to the demand side where the consumer pays a tax per consumption of carbon-emitting fuels thereby leveraging a financial incentive for consumers to invest into more efficient technologies based on renewable fuels. While the theory is straightforward, the tax can have unintended distributional impacts (Bach et al. 2020). The carbon tax may disproportionately impact lower income households and tenants who lack the financial capacity or decision-making power to alter the structure designating the types of fuels and amount of energy necessary to meet household energy service demands. This relates particularly to the lower income and rental sector without the capital or decision-making abilities to make these required investments. This Section reviews the policies and discourse around the implementation of the carbon tax in Germany as well as approaches to redistribute the carbon revenue.

As concerns grow about the potentially regressive nature of carbon pricing for low-income households, there is an intense debate in Germany on how best to use the funds generated by carbon pricing to counteract this. The redistribution of revenues from carbon pricing is seen as a tool to achieve several goals. It should be ensured that the selected redistribution model mitigates the negative distribution effects, promotes investments in the energy infrastructure in the private sector and increases social acceptance of the CO2 pricing policy (Thomas et al. 2019; Vermittlungsausschuss 2019). In pursuit of balancing out social inequities experienced by some households as a result of the carbon tax, revenues derived from the carbon tax should fund the reduction of the EEG levy on electricity as a means to dampen these distributional impacts (BMF 2019; Harthan et al. 2020), or alternatively to redistribute to the population through a climate bonus (Bundestag 2022). General acceptance of the climate bonus is at risk if there is no steering mechanism in place to ensure that the funds go towards decreasing emissions (Barckhausen et al. 2022). Lower income households are prone to rebound when households invest in energy efficiency, or similarly, consumption may increase by increasing income because households are now able to afford previously unmet household service demands (suppressed demand) (Sorrell 2007). Therefore, it is critical to understand the impacts policies aimed at discouraging fossil fuel consumption and the progressivity of redistribution policies have on different households to better estimate the energy welfare of households in addition to the overall energy and emissions.

In Germany, in terms of CO₂ pricing, all combustibles and motor fuels that are not integrated into the European emissions trading system (EU ETS) (especially for use in heat generation for buildings and in transport) are regulated within the framework of the National Fuel Emissions Trading Act (BEHG). (Vermittlungsausschuss 2019). The entry-level price in 2021 was 25€/t CO₂. The price is set to rise to 55€/t CO₂ by 2025. Subsequently, the fixed CO₂ price system is to become a certificate trading system, in which the CO₂ price is formed freely on the market. A price corridor of at least 55€/t CO₂ and at most 65€/t CO₂ is planned for this in 2026. Based on a field report that will be presented to the federal government in 2024, a decision will be made on the further design of price corridors or fixed prices (Deutscher Bundestag 2019).

Germany is intensively discussing how the revenue generated can best be redistributed in the interests of revenue neutrality in the case of CO₂ pricing. Options include various measures such as funding for electricity price cuts, subsidies for low-income households, renters, heating systems and building renovations, and commuters, or even redistribution to the entire population or specific sections of the population. The government plans to increase the heating cost quota for social welfare beneficiaries in order to compensate for the addition of the carbon tax, and at the same time to evaluate the best way to implement the carbon tax for the rental sector such that tenants are encouraged to conserve energy while landlords are incentivised to invest in energy efficient or renewable heating systems as well as building renovations (Harthan et al. 2020). Finally, it was decided that the additional financial burden of CO₂ pricing should be cushioned, in particular by financing the commuter allowance and reducing the EEG surcharge on electricity (Vermittlungsausschuss 2019). With the coalition agreement of the new federal government (SPD/BÜNDNIS 90/DIE GRÜNEN/FDP 2021), it was also announced that a social compensation mechanism would be developed beyond the abolition of the EEG surcharge in the form of a climate bonus.

Several studies in Germany assess the potential impact of redistribution schemes on different household types and their fairness in terms of offsetting the cost burden of carbon pricing, with each study assessing a different focus, assumptions and redistribution options (Thomas et al. 2019). Here, the overview focusses only on the redistribution options assessed and the assessed impact on low-income households, as some studies also focus on alternative taxation options and other consumer groups. The most common types of redistribution are either general or targeted. An administratively simple option is to distribute the income equally among all people (Kalkuhl et al. 2021; Lange et al. 2019), the so-called per capita redistribution. Another possibility is to only give the income back to low-income households (Frondel 2020). A third option is to redistribute the income to households with particularly high energy costs or a significant burden from CO₂ pricing (Frondel 2020; Kalkuhl et al. 2021; Thomas et al. 2019; Thöne et al. 2019; Venjakob and Wägner 2021).

Each of these options has advantages and disadvantages, with similar trends seen across studies. The expected cost burden and relief by income group varies somewhat across studies, depending on available revenue, population and household distribution, and associated energy use and emission profiles. Sample household profiles are evaluated to show the impact of various parameters such as tenure, commuter status, building type, urbanisation, household composition or combinations thereof (Agora Verkehrswende und Agora Energiewende 2019; Bach et al. 2019; Bach et al. 2020; Frondel 2020; Kalkuhl et al. 2021; Lange et al. 2019; Thöne et al. 2019; Venjakob and Wägner 2021). In general, it is intuitively understandable that larger households with more people could benefit more from a per capita redistribution, since energy costs do not increase proportionally with the number of people per household, with particularly large households with children benefiting (Agora Verkehrswende und Agora Energiewende 2019; Frondel 2020). Higher income households would benefit less, since overall energy consumption increases with income—mainly due to the associated increase in energy consumption in transport—and the cost burden could exceed the redistribution amount (Frondel 2020). Households that do not rely on fossil fuels for heating energy would also benefit disproportionately as they would pay less CO₂ tax and get more back (Frondel 2020).

With a per capita redistribution, households with high energy costs could be more heavily burdened, such as households with low incomes and high consumption, but also, for example, commuters with high energy bills who depend on motorised private transport and often have no opportunity to use public transport to change (Venjakob and Wägner 2021). But low-income households that are unable to reduce their emissions could also be disproportionately affected (Frondel et al. 2018). According to these studies, the expected burden from the additional CO₂ pricing alone is higher for households with low incomes in absolute figures and in relation to income or consumer spending than in other income groups, especially those with high incomes. Accordingly, the notion of a regressive effect of CO₂ pricing is confirmed. However, the results of the studies vary with regard to the benefit of a per capita redistribution of the revenue from CO₂ pricing. Lower income households could benefit from an overall lower consumption of CO₂-emitting fuels (Agora Verkehrswende und Agora Energiewende 2019; Bach et al. 2020), unless they cannot invest to reduce their dependence on technology based on fossil fuels (Frondel 2020). In certain cases (e.g. long-distance commuters, one-person households, tenants, households with oil heating), households with low incomes may need additional, targeted financial support in addition to redistribution per person (Agora Verkehrswende und Agora Energiewende 2019; Bach et al. 2020; Frondel 2020; Kalkuhl et al. 2021; Thomas et al. 2019).

These approaches usually assess the effects of redistributing funds based on a per capita redistribution, measured using quintiles or deciles of the income distribution. However, such an approach distorts the effects of such relief measures, since the parameters relevant to the energy transition are not taken into account. The distribution of the population and households according to income is a key component for determining CO₂ emissions as a function of socio-economic parameters. When evaluating CO₂ pricing variants by income group in relation to the population, the lowest income group contains 2.6% of the population and 4.9% of households, while the highest income group represents 30.1% of the population and 22.2% of the households. As described in Chapter 3.1, the distribution of the population and households by income groups has implications for the investment and consumption requirements. The aggregation of the statistical income groups into income deciles in relation to the population means that the differences in financial possibilities (income) and energy consumption (based on factors such as the number of people per household or the type of building) are combined in one decile from different income groups, so that these are merged into profiles that do not adequately take into account the heterogeneity, particularly in the case of lower income households. Breaking down the population into income brackets instead of deciles provides a more detailed insight into household energy use and financial capacity, as the heterogeneous determinants of energy use, such as regional, technological, access and building factors, and financial capacity are directly accounted for, as well as the household size. Every household, regardless of the number of people living there, needs a heating system. With an increasing number of people in the household, the installation and consumption costs are also spread over more people.

The per capita redistribution may be a simpler solution from an administrative point of view, but a redistribution per household, which seems similarly easy to implement, can better reflect the needs of the households. The question of how the revenues are to be distributed and whether this increases the social acceptance of carbon pricing depends on an analysis of the population and their living situation and how this correlates with the energy transition in terms of the financial burden of carbon pricing. Therefore, an overarching energy system optimisation model is a tool ideally placed to assess the long-term effects of policies on energy and emissions and should be developed to also incorporate the differentiated socio-economic disaggregation of the household sector.

5 Energy Welfare of Households within the Context of the Energy Transition

Typically, the household sector in Germany is represented in energy system optimisation modelling exercises as one homogeneously defined average household representing all households, disaggregated only by building type or location (BMWi 2018), which oversimplifies the situation and leads to one technology identified as the most cost-effective solution to meet a particular demand. The expected contribution from the household sector towards achieving the targets hinges on energy system analyses performed based on the profile of average households. Despite increased granularity of various attributes in the building sector (e.g. such as investor-specific barriers, ambience heat distribution and uptake of policies and measures), recent assessments have found that the building sector does not now nor will it meet the expected targets for 2030 (Repenning et al. 2020). These additions still do not allow an assessment of the energy welfare of households and so may still underestimate the impact on lower income households and overestimate the possible contributions from the household sector towards achieving the overall objectives of the energy transition. The TIMES (The Integrated MARKAL-EFOM System) model generator is a least-cost optimisation, bottom-up, technology-rich, linear-programming energy system model that can be applied to analyse the implications of a range of pathways for long-term energy investments and to identify least-cost measures to realise the climate and energy objectives of a particular region through the integration of relevant energy policies and technologies under a detailed technical and socio-economic framework (Loulou et al. 2016; Loulou et al. 2016). The TIMES modelling framework has a detailed representation of energy technologies and their linkages across sectors (or actors) and considers the interdependencies of the energy system. This enables the analysis of the competition and substitution effects between technologies and provides detailed results of the energy flows, capacity investments, emissions and costs. The TIMES framework is applied towards the development of a household sector model (the TIMES-Actors-Model (TAM) Households) with high actor resolution to enable the analysis of parameters around access and affordability, which are key to account for the differentiated needs and capabilities of energy-related investment decisions households make for building-related investments within an energy system model.

5.1 Disaggregation

Disaggregating a model to more specific user profiles is very data-intensive, especially in the case of this bottom-up energy system model, where each actor will need to be defined in terms of demands, technologies, buildings and the associated socio-economic projections. Disaggregation is also the cornerstone for integrating consumer investment and consumption behaviour, particularly with regard to develo** policies to improve the electricity consumption of households through energy efficiency measures (Gouveia et al. 2015; Jones et al. 2015; Sütterlin et al. 2011) or to account for other socio-economic factors, location, consumer or occupant-related behaviour (Druckman and Jackson 2008; Jaccard 2015; Leroy and Yannou 2018; Li and Just 2018; Reveiu et al. 2015; Tomaschek et al. 2012). The basis for modelling households as actors is the statistical investment and consumption behaviour by end-use for households in order to adequately capture and assess the socio-economic parameters (Destatis 2013).

As shown in Fig. 3, the final model disaggregation includes income group, tenure status and building type-specific profiles, energy service demands and technologies. The energy service demands are determined exogenously for each profile-defined building and are based on techno-economic assumptions for the development of technologies and the political and socio-economic framework as the key drivers for demand. This model is dynamic in that the population can shift into other income groups and buildings over time, thereby allowing a better representation of the shifts in energy demands precisely because the demands are directly related to the defined socio-economic profile.

Fig. 3

Reference Energy System for the household sector in TAM-Households.Footnote

Income groups are disaggregated by monthly income per household R1: < 900€, R2: 900–1500€, R3: 1500–2000€, R4: 2000–2600€, R5: 2600–3600€, R6: 3600–5000€, R6: > 5000€; Location by U = Urban, R = Rural; Tenure by O = Owner, T = Tenant, Building type by M = Multi-family home, S = Single-family home, Building age by E = Existing, N = New.

Source Own calculations as given in (Dobbins 2022; Dobbins and Fahl 2022b)

5.2 Budget Constraints

Income, expenditure patterns and available savings are key factors in affordability of household energy services (Alberini et al. 2011; Cayla et al. 2011; Kaza 2010; Longhi 2015; Vassileva et al. 2012). Available capital is essential to cover the costs of consumption as well as the investment costs of new or alternative technologies and measures. Modelling affordability is about: (i) understanding and incorporating the dynamics within income groups and within the profiles, (ii) reflecting the affordability of each profile according to the budget constraints, (iii) reflecting the present value of future cash flows through the application of appropriate discount rates and iv) incorporating the applicable co** mechanisms to meet needs with limited budget, such as extending the technical lifetime of technologies and/or buying second-hand appliances—which have lower upfront, but higher operating costs. The model restricts the financial ability of households to invest in the high upfront cost of appliances to better reflect the actual potential in overall capital investments by determining the overall available budget per profile based on statistical analysis of the disposable income, savings, GDP and typical investment patterns (Destatis 2018b; IMF 2019a, 2019b).

Based on (IMF 2019a), the GDP per capita in Germany is projected to increase by 81.4% between 2013 and 2060 from 36,948 €2015/cap to 67,0715 €2015/cap. With a total available capital (actual investment and consumption expenditure plus available savings) of 179 billion € in 2013, the distribution across income groups is projected to increase to 631 billion € in 2060 (Dobbins 2022). The majority of the wealth in the household sector resides in the upper two income groups. This available capital is further distributed per defined profile within each income group according to projections of the shares of households and population. These figures are used to define the budget restrictions for each actor group in the model described in the next section. The overall household energy budget is considered by including this into the assessment for households service needs. This additional disaggregation better reflects the holistic financial and decision-making power of specific actors in the household sector and is previously not reflected in modelling assessments for long-term energy planning in Germany. The investment limitations are represented with household budget constraints for each defined profile based on the available savings for each income group. This budget constraint represents the statistically available savings for each income group and is considered as the potentially available budget that households could invest in more efficient or renewable-based end-use technologies (heating, water heating, lighting and other appliances), retrofit the building and small-scale PV rooftop power generation (playing a role as prosumer).

The model takes into account the limitations in the available budget for each actor group through the implementation of profile-specific budget constraints. The budget constraints for each profile are calculated based on available statistics on income-specific typical investments in energy appliances, energy improvement investments and savings (Destatis 2013). The budget restriction is applied to each actor group through a user constraint on the investment and consumption costs (Ahanchian et al. 2020). This budget constraint is applied to all investments in owner-occupied households. Similarly, the budget constraint is included for tenants, but applies only to technologies which they have the decision-making power to replace and therefore excludes heating, water heating and PV technologies as well as building renovations. Instead, these investments include a higher discount rate to represent the apprehension of landlords to make costly investments in properties from which they may not derive a benefit, as outlined by the landlord/tenant dilemma (Bouzarovski et al. 2018; Griffiths and Causse 2010).

5.3 Incorporation of Policies and Measures

Policies and measures can be modelled as constraints according to particular targets (Senkpiel et al. 2020) and were modelled in TAM-Households in line with the policies and measures influencing energy use in the household sector, such as targeted greenhouse gas emissions. Methods to model energy-related policies and measures are largely adapted from TIMES-D (Fais 2015; Haasz 2017) and further developed within the Decentral project (Ahanchian et al. 2020). In TAM-Households, it was necessary to apply constraints (e.g. renovation rates, market shares for specific technologies or energy carriers) to achieve these targets for the whole sector or according to the profiles defined (e.g. homeowners, building type, location, income group). Measures, such as subsidies, grants (financial incentives) and taxes can be included through a price reduction on the fuels or technologies for specific actor groups (e.g. income group, homeowners). Specific policies and measures modelled include the decarbonisation targets and carbon taxes implemented in the scenarios. The decarbonisation target applies a zero emissions target in 2050 whereupon the model finds the least-cost pathway to achieving this target given other variables and constraints in the model, such as the budget constraints. Environmental taxes, such as carbon taxes, are added to carbon-emitting fuels and related to the consumption by each specific actor groups represented in TAM-Households.

6 Results

The majority of investments to achieve the goals of the energy transition in households will have to flow into increasing energy efficiency and into a higher contribution, directly or indirectly via secondary energy sources such as electricity, district heating or hydrogen, to renewable energies through building renovation, heating replacement and new means of transport. Low-income households are unable to meet these demands and risk being left behind in the energy transition. The majority of households lack financial capital or the decision-making power to make the necessary investments. Providing additional financial support to low-income households would enable them to pay their energy bills and provide a platform for infrastructure investment. The TAM-Households model was applied to analyse the impacts of financial support provided through the redistribution of carbon revenue. The method of including disaggregation and the budget constraints means that it is possible for a rich analysis of the impacts on energy and emissions for different household types. The results in this section explore the impacts of the carbon tax in general on different households according to their socio-economic parameters, followed by an analysis of the distributional impacts of various approaches to applying redistribution schemes.

6.1 Scenario Descriptions

Several scenarios are defined in the TAM-Households model and compared against a reference scenario. The reference scenario includes the TAM-Households methodology with the disaggregation and the budget constraints as well as all expected policies implemented underpinned by the same socio-economic development and price assumptions. In order to compare the effects arising through different redistribution approaches, four variations are contrasted with the reference scenario. While the current method to be deployed is to collect a carbon tax according to the consumption of fossil fuels and then, to redistribute this through a tax relief per kWh electricity consumed (Vermittlungsausschuss 2019), other options discussed include reallocating the funds through a ‘climate bonus’ at 100€ per capita annually (CB scenario) as explored in other studies (Kalkuhl et al. 2021). Another approach is explored in the CBLI scenario where the climate bonus is provided to the lower half of the population only but increased to 200€ per capita annually. However, as described previously the capital-intensive investments are bound to the household infrastructure, which must be implemented irrespective of the number of occupants. Therefore, two further approaches for redistribution are also explored. In the CBHH scenario, each household irrespective of occupants or income are provided the same subsidisation amount related to the equivalent of 100€ per capita and equals 193€ per household annually. One further scenario (CBLIHH) provides the per household subsidy but again only to the lower half of the population, which increases the allocation to 383€ per household annually. These scenarios are summarised in Table 1 and Fig. 4.

Table 1 Scenario description: Improving the energy welfare of households

Fig. 4

Overview of approaches to carbon redistribution schemes analysed

The effect that can be compared is the compensation of the carbon tax with a reduction in the electricity levy in the reference scenario (REF), or applying a carbon tax without the electricity levy reduction while redistributing the carbon revenue to the population as a means to compensate the impact and the means of defining the allocation (i.e. per capita or per household). The redistribution simulates a fixed amount of carbon revenue with alternative reallocation amounts so as to compare the impacts. It should be noted that as households invest in renewable energy and energy efficiency, the fossil fuel dependency will decrease and so will the carbon revenue.

6.2 Impact of Carbon Tax

The reference scenario details the impact of the carbon tax policy with a tax on the consumption of fossil-based fuels including funding a reduction on the renewable energy levy on electricity consumption. This levy relief should, in part, act as a compensation for any unfair effects on specific households. The carbon tax should incentivise households to invest in renewables and energy efficiency, while electricity levy relief incentivises a shift to renewable and electricity-based consumption (e.g. in heat pumps). These will impact households differently according to socio-economic parameters in 2025 and the annual financial expenditure and compensation per household is shown in Fig. 5. Lower income households consume more fossil fuels and less electricity and therefore pay more on carbon taxes than they receive in levy compensation. The carbon tax burden increases with income up to medium-income households, but decreases in the highest income groups. Higher income households are greater consumers of electricity and therefore benefit from a reduction in the electricity price since fossil fuels make up a smaller share of the total consumption resulting in lower carbon tax contributions. These trends indicate that the policy results in a disproportionate burden on lower income households.

Fig. 5

Comparison of annual expenditure on carbon tax and EGG levy relief per household by socio-economic parameters, 2025. Source Own graph as given in (Dobbins 2022)

Some studies point to different trends. Lower income households are assumed to consume greater shares of electricity which would mean households are compensated overall (Kalkuhl et al. 2021). Bach et al. (2020) agree lower income households would be disproportionately impacted by the carbon tax, but find that the electricity levy reduction compensates the financial loss. The difference lies in the input data for the assessments. Both studies assume constant electricity consumption per capita regardless of income level, while the energy balance in the present study developed a bottom-up calculation of all fuels according to socio-economic parameters thereby accounting for household size, appliance ownership, building type and access to technologies and resources. Lower income households typically have fewer occupants per household than higher income households, but some electricity-based appliances will consume the same amount of electricity without regard for the number of occupants. This consumption is distributed per capita, which reduces per-person consumption as the number of occupants in the household increase.

The disaggregation of the household sector also allows the analysis of the impacts for occupants by location (urban or rural), specific building types (SFH and MFH) and ownership (owners or tenants). A just allocation of the cost burden between landlords and tenants has been debated in parliament with various proposals discussed to ensure that landlords are incentivised to make investments. Given the share of households as tenants, this sector has a significant potential and role towards achieving emissions targets (Schultz 2021). While owners and tenants have similar levels of expenditure for the carbon tax, owners receive more compensation for the electricity levy reduction. Given the diversity of how the carbon tax impacts different household types, it is necessary to understand these discrepancies and as they relate to socio-economic parameters and the selected approach to redistributing carbon revenues.

6.3 Redistribution per Person

An administratively simple way to redistribute the carbon tax is to simply provide each person with an allocation. This first scenario considers this approach for an annual redistribution of the carbon revenue collected into a climate bonus given to all households at 100€ per person (CB) or 200€ only to the lower income half of the population (CBLI). The climate bonus is added in the model as additional available capital per household in the budget constraint, as shown in Table 2, which illustrates that with a per capita distribution the allocation per household increases with income as the number of occupants increases.

Table 2 Average annual additional budget per household by income group and redistribution scheme based on a per capita redistribution (€₂₀₁₅)

The final energy consumption does not change significantly across scenarios with the final energy consumption in the CB scenario resulting in 1,781 PJ (52 Mt CO₂-eq) and the CBLI scenario resulting in 1788 PJ (54 Mt CO₂-eq), compared to the Reference scenario with 1,736 PJ (54 Mt CO₂-eq) in 2030, as shown in Fig. 6. The REF scenario exhibits greater shares of electricity consumption compared to either Climate Bonus scenarios since the electricity levy relief is not provided and therefore disincentivises electricity consumption. The addition of the carbon tax without the EEG levy relief on electricity, however, results in higher demand to use the existing gas infrastructure together with a hydrogen or biofuel blend in favour of replacing existing technologies with alternatives in both Climate Bonus scenarios since households are unable to afford the high upfront investment costs for technologies. The shares of renewables shift only slightly from 20.6% in the REF scenario to 19.8% in both climate scenarios, while the share of fossil fuels reduces only slightly from 48.5% in the REF scenario to 47.6% and 47.2% in the CB and CBLI scenarios, respectively. Biomass continues to play a significant role in the fuel mix because the pricing is competitive in relation to the increasing carbon tax on fossil fuels.

Fig. 6

Final energy consumption in all households by energy carrier and scenario, 2030. Source Own graph as given in (Dobbins 2022)

In typical modelling assessments, it is not possible to assess the distribution of energy, emissions and costs on different household types and the added capital injection to households as provided through the redistribution scheme results in little differences to the overall emissions. Through the disaggregated assessment, it is possible to analyse the impacts according to the defined socio-economic parameters. Examining the energy consumption profiles of the lower four income households reveals significant differences in consumption across the scenarios, as shown in Fig. 7. The equal annual allocation of 100€ per capita in the CB scenario increases the average consumption to 34 GJ per household from 31 GJ per household in the Reference scenario. When the allocation is doubled and provided only to the lower income groups in the CBLI scenario, the consumption increases to 35 GJ per household. Fossil fuel shares reduce from 63% in the Reference scenario to 62% in the CB scenario and 59% in the CBLI scenario. While renewables make up 7.2% in the Reference scenario, these decrease to 5.6% in the CB scenario and increase to 7.9% in the CBLI scenario. This indicates a greater shift for the lower income households in the CBLI scenario towards renewables and away from fossil fuels compared to both other scenarios. In 2030, the end of the technological lifetime of the majority of space heaters is reached and requires replacing. Since insufficient budget has been accumulated in the lower four income groups to this date to afford infrastructural changes, the key bridging solution is to blend fuels, for example, with hydrogen or biofuels. In subsequent modelling periods, sufficient budget is accumulated for alternative investments and the use of hydrogen and biofuels disappears again. While the CBLI scenario shifts a greater extent of the demand to network supply, such as district heating and gas, and higher input fuels, such as biomass, the CB scenario reduces carbon-intensive fuels, including gas and oil, to a greater extent than in the other scenarios. None of the scenarios exhibit substantial investments into energy efficiency measures where the REF scenario incorporates an energy efficiency equivalent of an average of 0.07 GJ per household and each of the climate bonus scenarios an average of 0.1 GJ per household. However, an analysis of energy consumption alone does not render sufficient information about the energy welfare of households.

Fig. 7

Average energy consumption per household for the lower four income groups, 2030. Source Own graph as given in (Dobbins 2022)

The addition of the budget constraints related to the disaggregation in the methodology allows an analysis of the budget deficit experienced by households in meeting energy-related investment and consumption patterns. The budget deficit is translated into a quantification of the suppressed demand and provides an insight into the energy welfare of households. Suppressed demand is experienced extensively by lower income households in the REF and CB scenarios, and is reduced significantly in the CBLI scenario, as shown in Fig. 8. In 2030, 11.4 million people require an additional 84€ each in the REF scenario, with a redistribution of 100€ per capita in the CB scenario, suppressed demand reduces to 5.7 million people requiring an additional 52€ each. This shows that the additional budget supports the additional consumption of energy for previously unmet needs, but still remains insufficient to eliminate it. By increasing the redistribution to 200€ and targeting it to the lower income households only, the suppressed demand diminishes the number of households suppressing demand to 131,000 people requiring an additional 118€ each.

Fig. 8

Average suppressed demand per capita for the affected population by scenario, 2030. Source Own graph as given in (Dobbins 2022)

The trends in investment and consumption patterns highlight how additional capital influences suppressed demand in Fig. 9. The investment profiles in the REF and CB scenarios follow similar trends to 2040. While households in the REF scenario can only make investments once a sufficient budget has been accumulated, households in the CB scenario have additional budget but opt to increase consumption expenditure while making steady investment expenditures. In the CBLI scenario, both investment and consumption expenditure increase, which results in a greater degree of suppressed demand than in the other two scenarios.

Fig. 9

Investment and consumption trends in the average lowest four income groups by scenario, 2020–2050. Source Own graph based on (Dobbins 2022)

6.4 Redistribution per Household

Investments in building infrastructure will require households to be able to afford the high upfront costs and the key driver for these investment are the home and not the number of people living there. As such, these next scenarios compare an annual redistribution of the carbon revenue to each household rather than to each person. In general, with a redistribution of the budget per household instead of per capita, each household receives an additional annual budget of 193€ each (CBHH scenario). When these funds are redistributed to the lower income half of the population only, these households receive an additional annual budget of 383€ each (CBHHLI scenario). As highlighted in Fig. 8, the total budget deficit in the REF scenario totals 955 million Euros annually. A redistribution of 193€ per household surpasses this deficit, such that the suppressed demand is eliminated with an administratively more simple distribution across each household. With a redistribution across lower income households only, the model does not produce different results compared to the redistribution across all households, therefore, the analysis focusses on the redistribution across all households only. Compared to the REF scenario, the CBHH scenario results in overall shares of 47.1% fossil fuels and 29.3% renewables and 2.6% less emissions, as shown in Fig. 10.

Fig. 10

Total final energy consumption across all households by fuel type, 2030. Source Own graph as given in (Dobbins 2022)

A closer examination of the total energy consumption profiles by socio-economic parameters and fuel type is explored in Fig. 11. The overall shares of fossil fuels and renewables indicate that the share of fossil fuels decrease with income while the share of renewables increase with income, while owners (typically with higher incomes) consumer more renewables and less fossil fuels.

Fig. 11

Final energy consumption per household by fuel type and socio-economic parameter, 2030. Source Own graph as given in (Dobbins 2022)

The results are analysed further to compare the cost burden from the carbon tax paid on gas and oil fossil fuel consumption and the compensation received from the climate bonus and are presented in Fig. 12 as a percentage of net household income. In the CB and CBLI scenarios, the climate bonus is redistributed per person, while the CBHH scenario redistributed the carbon revenue per household and the reference scenario has no redistribution. The effect of CO₂ pricing remains regressive even with a per capita redistribution. The reference scenario indicates that the majority of households spend more on the carbon tax than received in compensation. In the CB scenarios, the carbon tax burden outweighs the compensation for lower income households while higher income households benefit. When the redistribution is targeted to compensate only lower income households, these households benefit substantially while higher income households receive less compensation than they pay. Redistributing the carbon revenue to all households benefits all households, with lower income households benefiting to a greater extent than higher income households. This better aligns the redistribution with the needs of the households, so that redistribution per household in the lowest income group achieves a net benefit of + 1.0% of the net household income (CBHH), compared to 0.2% with a per capita redistribution (CB). The average household benefits in the CB and CBHH scenarios and are negatively affected in the CBLI and REF scenarios.

Fig. 12

Carbon tax cost burden (positive values) and redistribution compensation (negative values) for the scenarios by income group, 2030

These variations across income groups change with parameters such as the heating structure (oil and gas heating), tenants without decision-making power or residents of multi-family homes. The result is whether the climate bonus has a progressive or regressive effect. Linking the redistribution of carbon pricing revenues to the number of households irrespective of the number of people living in a building provides a better opportunity of involving low-income households in making investments.

6.5 Discussion

A key challenge in the energy transition, which demands action from the private sector is to ensure that emissions are reduced. Carbon pricing is a common policy to disincentivise fossil fuel consumption and incentivise investments for renewables and energy efficiency. Lower income households and tenants do not have the financial or decision-making capacity to make the necessary investments to shift their underlying household infrastructure. Carbon revenue redistribution schemes aim to compensate households that may be disproportionately affected by carbon pricing policies. This study compared four approaches to redistributing carbon revenue to a reference scenario with no compensation scheme. This additional capital to the available budget² per household in each income group, which increases with income due to the larger household sizes. The lower income groups remain below the average additional budget in the CB scenario due to the amount of debt and inability to accumulate savings while the higher income households have greater incomes and savings at disposal.

The common methodology to assess the effects of CO₂ pricing and the redistribution of income based on the population, underestimates the social consequences. Higher income households receive net benefits with a per capita redistribution while lower income households pay more in carbon taxes than they receive in compensation. Targeting the redistribution to lower income households only provides the necessary support these households require to shift the underlying household energy infrastructure. This relates to the types of investments required to achieve the household sector energy transition targets. Regardless of the number of people living in a house, each home will need only one building-related investment, such as for renovation or a heating system. Therefore, progress towards achieving the goals and supporting low-income households can be better achieved through a redistribution program per household. This also increases the social acceptability of CO₂ pricing in contrast to a per capita redistribution, in which households with low incomes are disadvantaged on average.

This study showed that the focus is specifically on the effects of carbon pricing and the redistribution of the revenues generated to low-income households. For this purpose, the usual, purely arithmetic redistributions, which concentrate on a per capita redistribution, are being switched to a needs-based approach that would better reflect the financial and decision-making abilities of the households. This would ensure that the assessment of the impact of carbon pricing options on low-income households is more in line with their needs and requirements for participation in the energy transition and that they are not disproportionately affected without the possibility of mitigation measures to benefit.

7 Conclusions

This chapter described a methodology developed to better assess the heterogeneous needs and capabilities of households with a view of supporting lower income households within the energy transition process in Germany. The method presented here incorporates disaggregation and budget constraints to better represent the heterogeneity of the household sector in relation to their needs and capabilities around energy-related investment and consumption. The method was applied to evaluate the impact of policies on household energy and emissions as well as the energy welfare of households.

Carbon taxes have long been implemented as a means to reflect the environmental damage incurred through the combustion of fossil fuels, but these can disproportionately impact lower income households and tenants who lack the financial capacity or decision-making power to alter the structure designating the types of fuels and amount of energy necessary to meet household energy service demands. CO₂ pricing is an important measure to support the energy transition in order to be able to reduce CO₂ emissions efficiently and effectively. However, it can have negative affordability and social impacts on low-income households, particularly those who are already struggling to meet their basic energy needs. The redistribution of the revenue from CO₂ pricing should cushion the effects of the resulting increase in energy costs and provide financial support for investments in energy efficiency and renewable energies in households. Because the household sector is so diverse, it is important to consider the different needs and capabilities when assessing the effects of revenue redistribution.

The most common methods assess the impact of redistribution, particularly per capita redistribution, using income deciles based on population distribution. The crucial problem is that income deciles based on population distribution aggregate the heterogeneity of household income groups and thus do not correctly reflect and overestimate the financial possibilities of households in the lower income categories. A redistribution of funds per person benefits higher income households more, since higher income households also have more people per household. On the other hand, an alternative redistribution of income per household or per household living in buildings would benefit households with lower incomes more, which is of crucial importance since investments have to be made in the building. However, low-income households often struggle to meet their basic needs, and additional funds from the redistribution of carbon pricing revenues could accordingly lead to higher consumption rather than investment in the building. By linking redistribution to investment in the building, low-income households would be better able to absorb the long-term impact of energy and carbon price increases. Overall, this analysis of CO₂ pricing variants shows that a redistribution based on households and not population recognises the important role of the building as a target for investments in the energy transition. Social acceptance of carbon revenue redistribution schemes can better be guaranteed when investments are channelled into investments that will reduce carbon emissions.

While this study has provided a methodology that vastly improves the overall understanding of investment and consumption patterns in different household types cased on their socio-economic parameters, there are some caveats that should be noted. The geopolitical energy security challenges have shaped consumer energy prices that affect lower income households to a greater extent. The role and influence of prices in energy-economic modelling exercises should be considered. The prices in this study represent pre-pandemic and pre-war energy prices as long-term energy models reflect megatrends as opposed to (hopefully) short-term disruptions. However, under these price assumptions, it is already possible to establish the trade-off investment decisions households make. Higher fossil energy base prices would further justify investments into renewable energy and energy efficiency. To better reflect holistic household budgets, this household model could be coupled with a transport model since transportation is a significant contributor to greenhouse gas emissions and vary considerably.

Designing carbon revenue redistribution schemes must take into consideration not only the impact they have on the energy-related investment and consumption patterns of households but also on energy welfare. This will require methods that take consideration of the differentiated needs and capabilities of households to better ensure that households are able to undertake investments that will shift the underlying household energy infrastructure.