Global Climate Change—Who Ought to Pay the Bill?

: Global climate change is a signiﬁcant challenge for current and, particularly, future generations. In the public debate about the fair allocation of associated costs commonly the moral claim that the developed countries should burden the costs is expressed. To support this claim, often four moral arguments, based on the theory of justice, are raised: (i) the polluter pays, (ii) the historical responsibility, (iii) the beneﬁciary pays, and (iv) the ability to pay. The aim of the paper is to assess whether these principles impose a duty on the developed countries and whether a fair allocation of costs would be achieved.


Introduction
Almost all scientists agree that the global climate is changing, primarily due to human activities (see [1,2]). The anthropogenic drivers of climate change, i.a., increase in per capita consumption and, to a smaller degree, population growth, exceed adaptation efforts in the form of technological advances associated with carbon efficiency (see [3,4]). Although the full scope of the consequences of climate change is uncertain, severe effects on agriculture and food security (see [5,6]), biodiversity (see [7]), and poverty reduction (see [8,9]) are predicted and/or already observable. Furthermore, on a macro-scale it is expected that climate change significantly reduces global economic output and widens global income inequality (see [10]), although, on a micro-scale, firm profitability may in certain cases increase along with temperature, e.g. in the energy sector (see [11]). Most vulnerable to climate change are poor countries, primarily due to their geographical locations and to some degree due to limited adaption options [12].
The scientific debate about climate change is interdisciplinary and deals with a variety of topics. Natural sciences deliver specific knowledge about the sources, impacts and consequences of climate change-they are mostly concerned with environmental effects. Based on these insights, economists can analyse economic impacts, often on a macro-scale, and identify the most efficient solutions to climate change-based on efficiency, it can be assessed who should mitigate CO 2 emissions. Furthermore, national and international legislation provide a framework for taking action against climate change. The recent decision of the German Federal Constitutional Court [13] has been considered an important step towards this direction. Yet, the decision is based on a forward-looking perspective whereas global climate change also has a historical context. Thus, neither a legal perspective nor the efficiency criterion can be applied in order to evaluate who should burden the costs of climate change. This moral issue calls for the involvement of principles of justice and hence has to be analysed from a moral philosophical perspective. Put differently, the two essential questions to be answered are (i) how should the costs of climate change be allocated, and (ii) does a moral duty to pay exist and, if so, for whom and for what reason(s)?
In the public debate about the allocation of costs caused by climate change, commonly, four arguments are provided in order to make the point that the developed countries have a moral duty to pay: (i) they have a historical responsibility, (ii) which goes hand in hand

Global Climate Change-A Very Short Summary
This chapter provides a brief overview of global climate change. The aim is threefold. Firstly, the historical context with respect to past emissions and relative contributions to global emissions will be illustrated in order to lay the foundation for the analysis of historical responsibility. Secondly, the consequences of climate change will be described and thirdly, relevant types of costs will be discussed.
Greenhouse gas (GHG) emissions as well as atmospheric aerosol particles are driving global climate change and global warming. Especially, carbon dioxide (CO 2 ) and black carbon are the main contributors to global warming [14]. For the sake of simplicity, in the following, the development of global, regional, and/or country-specific emissions over time will be depicted for the case of fossil carbon dioxide (CO 2 ), which constitutes a major part of GHG emissions (73%). Thus, in what follows, the focus will be put on CO 2 emissions, although other GHGs such as methane (CH 4 ), nitrous oxide (N 2 O), and fluorinated gases and atmospheric aerosol loadings should not be neglected. With respect to CO 2 emissions, the main sources are fossil-fuel combustion (coal (44%), oil (35%), natural gas (22%)) primarily used for electricity and heat generation (36%), manufacturing (17%) and road transport (16%) [15,16]. These CO 2 emissions were also responsible for the increasing trend in GHG emissions in 2018 and 2019 [15,16]. As illustrated in Figure 1, global CO 2 emissions keep rising and reached a new peak (around 36 gigatonnes) in 2019. Furthermore, the CO 2 emissions' growth rate has accelerated again since 2016, mainly caused by an increase in global coal consumption [15,16].
Over history, countries' and/or regions' absolute emissions (see again Figure 1) as well as relative contributions to global CO 2 emissions changed to a considerable amount as illustrated in Figure 2. Not surprisingly, in the early 18th century, the UK was the main emitter of CO 2 . Throughout the 18th century also the United States and the later members of the EU-27 began to emit large quantities of CO 2. Consequently, in the middle of the 19th century these three emitters were responsible for almost all global CO 2 emissions. However, since the 1950s the relative contributions of the EU-27 and United States began to decrease and other countries, primarily China and India, rapidly and substantially increased their CO 2 emissions. Especially China has doubled its relative contribution over the last 18 years (from~20% to~40%) and has become the world's largest emitter of CO 2 due to its rapid industrialisation since the early 2000s. Over history, countries' and/or regions' absolute emissions (see again Figure 1) as well as relative contributions to global CO2 emissions changed to a considerable amount as illustrated in Figure 2. Not surprisingly, in the early 18th century, the UK was the main emitter of CO2. Throughout the 18th century also the United States and the later members of the EU-27 began to emit large quantities of CO2. Consequently, in the middle of the 19th century these three emitters were responsible for almost all global CO2 emissions. However, since the 1950s the relative contributions of the EU-27 and United States began to decrease and other countries, primarily China and India, rapidly and substantially increased their CO2 emissions. Especially China has doubled its relative contribution over the last 18 years (from ~20% to ~40%) and has become the world's largest emitter of CO2 due to its rapid industrialisation since the early 2000s.  However, when contrasting the absolute CO2 emissions per country with the per capita CO2 emissions (Figure 3), another situation emerges. In terms of per capita emissions, the United States emitted most over the last century. China is nowadays on the same level as the EU-27, whereas Russia is the second biggest per-capita-emitter. However, when contrasting the absolute CO 2 emissions per country with the per capita CO 2 emissions (Figure 3), another situation emerges. In terms of per capita emissions, the United States emitted most over the last century. China is nowadays on the same level as the EU-27, whereas Russia is the second biggest per-capita-emitter. However, when contrasting the absolute CO2 emissions per country with the per capita CO2 emissions (Figure 3), another situation emerges. In terms of per capita emissions, the United States emitted most over the last century. China is nowadays on the same level as the EU-27, whereas Russia is the second biggest per-capita-emitter. This overview has shown that global CO 2 emissions have been continuously growing since the industrial revolution, whereas country-specific emission trends have changed, especially in recent years. Having outlined the historic and present emissions causing anthropogenic global climate change, I will now turn to the consequences.
The latest reports of the UN Intergovernmental Panel on Climate Change [8,9,17] depict and predict severe effects of global climate change, and consequently, these reports have alarmed politicians as well as the general public about the need to take measures against climate change. Projected changes are manifold and comprise, i.a., increases in mean temperature, hot extremes, precipitation deficits and droughts, as well as heavy precipitation. Furthermore, climate risks are caused by an increase in extreme weather events, such as cyclones and floods [9,17]. Most of the effects are already observable. The years 2018 and 2019 were among the warmest since the recording of global surface temperature started (in the year 1880) [15,16]. Even if not the whole observed temperature rise can be traced back to human activities, it is estimated that the latter have caused approximately a 1 • C increase compared to pre-industrial levels [9]. As a result of global warming, the sea levels rise due to the melting of glaciers and arctic sea-ice. Furthermore, an increase in ocean temperature results in higher ocean acidity and a decrease in oxygen levels. Together, these consequences lead to severe effects on biodiversity (species loss) and ecosystems. Thus, irreversible damage is to be expected in the future [8,9].
The negative effects of global climate change on biodiversity and ecosystems involve a variety of climate-related risks for humanity. Generally, negative effects on ecosystems limit their functions and services to humans. Diminishing marine biodiversity and negative impacts on crop production are threats to food security. Moreover, water security is at risk, especially in dry subtropical regions where groundwater and renewable surface water are projected to decrease. Additionally, already existing health issues and diseases, such as malaria and dengue fever, will be worsened, e.g., by higher mortality rates and shifts in geographical range. This poses a special threat to low-income countries, which already suffer from these conditions and diseases. In sum, human well-being, health, livelihoods, food security, and water supply will be negatively affected by climate change. In general, these risks are global and faced by all societies. However, low-income countries face higher risks due to their geographical locations and lesser possibilities to adapt to climate change. Therefore, climate change jeopardises the goal of poverty reduction and may create new poverty traps [8,9].
The potential responses to climate change and associated costs can be categorised as mitigation and adaptation. The first one represents the opportunity costs that arise if certain actions causing GHG emissions are not undertaken. Put differently, mitigation costs represent the foregone benefits of actions not undertaken. If the act is nevertheless committed, there is a need to adapt to climate changes, which generates costs-the costs of adaptation. Examples for the latter are the construction costs of dams for flood-protection in coastal areas or additional expenses for combating climate-related diseases [18].
Opinions on whether a mitigation or an adaptation strategy is to be favoured diverge. Opponents of mitigation strategies argue that reducing GHG emissions is too expensive, while adaptation strategies would be more cost-effective and therefore preferable (see, e.g., [19]). However, it has been argued that mitigation involves only opportunity costs if it is assumed that economic activity without mitigation is efficient. Yet, considering climate change to be an unpriced externality, this assumption does not hold. Thus, internalising the costs of climate-harming activities would lead to a more efficient state compared to the status quo. Furthermore, in theory, the welfare of the current generation does not have to decrease if the costs are shifted to future generations through reduction of conventional investment. Put simply, smaller investments mean that current generations are compensated for emitting less GHGs by future generations [20,21] (pp. [37][38][39][40][41][42][43][44]. Nevertheless, usually the costs of compensation are considered to be a burden of current generations who are causing harm, meaning that future generations have to be compensated. Furthermore, it can be differentiated between ex-ante and ex-post compensation. The former is future-oriented, meaning that harm that is incurred in the future has to be compensated. For example, the development cooperation of industrialised countries that promote adaptation projects in developing countries is future-oriented. In contrast, ex-post compensation is backward-oriented and deals with reparation of already materialised harm [22].

Allocating the Costs of Global Climate Change-From Principle to Duty?
In the following, it will be discussed if a fair allocation of the costs of climate change can be based on one of the four principles: (i) polluter pays, (ii) historical responsibility, (iii) beneficiary pays, and (iv) ability to pay. To do so, firstly, it has to be defined what action is considered to be unjust with respect to climate change. On this basis, it can be defined who is responsible for a (potentially) unjust action-who the relevant moral agents are that have a duty to pay for the costs of climate change.
Generally, an environmental injustice arises if either environmental benefits are unevenly distributed or if an unfair burden in form of environmental harm is placed on one agent by another [23]. In the context of climate change, the first case builds upon an egalitarian concept of justice which conceptualises the atmosphere as a common good [24] (p. 28). The main moral principle underlying this is the principle of equal opportunity. This implies that every individual, independent of place and time, has an equal right to emit and benefit from it or to benefit from limited emissions. However, extensive use by some reduces the opportunities of others. An example for the second case is uncompensated costs of adaptation. In either case, the relevant action can be considered to be the emission of GHGs. Yet, not every emission of GHG is unjust and therefore, it has to be defined more precisely which emissions are unjust. GHG emissions can only be considered an unjust action if several criteria are met [20,25] (pp. 54-59): 1.
Harm is caused as a result of an action; 2.
The action aims at promoting own benefit; 3.
The harm is serious; 4.
The infliction of harm was not accidental, implying that the moral agent knows that committing the act leads to harming others; 5.
The emitter does not compensate the harmed; Sustainability 2021, 13, 13393 6 of 14 6. The harm is not reciprocal. Simply put, some emit more than others; 7.
The harm could have been easily avoided.
Now that the unjust action is defined, the question of the relevant moral agent remains. Certainly, individual actions cause emissions of GHGs, and to some extent, these emissions could be reduced without major sacrifices. For example, people could cycle more, use public transportation instead of cars, or reduce their meat consumption. Frequently, the argument of cost-efficiency is used against individual responsibility. The underlying idea is that individual changes in behaviour only have marginal effects and are therefore unsuitable to deal with the problem on a large scale. Thus, states (or state institutions) are often considered to be relevant agents, and hence, responsibility is ascribed to them. States do have the legal power to influence citizens' behaviour and could therefore address the problem of climate change. Nevertheless, the duty cannot be simply narrowed down to states because the actions of states are not the only cause for emissions. Corporations' actions cause environmental damage as well and international institutions, e.g., the World Bank, may influence the level of GHG emissions by promoting either sustainable projects or emission intensive projects [23]. Thus, it could be argued that all moral agents (individuals, corporates, states, and international institutions) bear responsibility. Yet, such generic ascriptions of responsibility are impracticable as duties cannot be allocated and, therefore, are likely to be dismissed [26] (p. 98f.). Therefore, in the following, it will be assessed whether the four principles are capable to ascribe specific responsibility and associated moral duty to a specific agent and thereby allocate the costs of climate change in a fair way.

Polluter Pays Principle
Generally, the idea that the agent who caused a negative externality is responsible for her own action is considered to be fair. This is the basis of the so-called polluter pays principle. Caney [18] (p. 753) refers to this as 'micro-version' of the polluter pays principle because a single agent, the polluter, imposes a negative externality on another agent. The harm created by the negative externality causes an inequality between the two agents. As discussed above, further conditions have to be met in order to consider this action as unjust. Thus, it is assumed that the polluter imposes the costs on the other agent with the purpose of self-enrichment; that the harm is serious, uncompensated, not reciprocal, and could have been easily avoided; that the action was not accidental; and that the polluter had knowledge about the action's consequences. On this basis, the action is considered to be unjust. If at least one agent has been put in such a position of arbitrary inequality, the act of neutralising this inequality by imposing burdens on the polluter is a justified action [27]. This can either be achieved by the removal of the pollution or by the payment of a compensation which covers at least the costs of the externality [28]. However, to do so, the specific harm of a specific action has to be identified. This poses a challenge because, although a causality between the share of emission and the caused pollution exists, the resulting harm is very uncertain. Note that, in principle, this argument is only a concern for practical implication and not for moral considerations. Yet, identifying an action's specific harm is unreasonable in case of global climate change [18]. Therefore, Caney [18] (p. 753) extended the micro-version of the polluter pays principle, which considers only the action and associated pollution of a single agent, by the development of his so-called 'macro-version'. The latter takes into consideration that more than one agent may be involved in an action's performance which causes the pollution and, as a consequence harm, all polluters are then responsible for their individual share of the pollution. The challenge of identifying individual shares of global climate change remains. However, based on the macro-version it can be argued that the harm (global climate change) is a result of the combined actions of the polluters. Thus, even if individuals cannot be made responsible for their specific contributions to the harm, it can be argued that those who emitted more GHGs are more responsible than those causing less emissions. Therefore, the macro-version appears to be more plausible and more pragmatic with respect to global climate change as many agents emit(ted) GHGs and it seems difficult, not to say impossible, to quantify a specific harm or costs based on individual emissions.
However, the macro-version cannot overcome certain obstacles of the polluter pays principle, either. Firstly, in a globalised world characterised by international trade, the geographical location of production and consumption may diverge [29][30][31]. This implies that some developed countries outsource emissions [30,32], which poses difficulties in identifying the actual polluter. As outlined in Section 2, in recent years especially some of the rapidly developing countries have noticeably increased emission rates, e.g., China and India. Thus, the principle would make them pay for their emissions while neglecting the actual standard of living. Against this background, the polluter pays principle would not necessarily make the wealthiest countries pay and thereby, might hinder poverty reduction efforts. For this reason, Caney [25] argues to limit the principle to some extent in order to not jeopardise poverty reduction efforts and economic growth in developing countries. Shue [27] suggests that in these cases, a guaranteed minimum in ecological and material terms is legitimate. The underlying idea is that in a situation characterised by inequality while enough resources exist to guarantee everyone an adequate minimum, in this case, a certain standard of living, it would be unfair to not implement such a guaranteed minimum.
Secondly, based on the polluter pays principle, only the costs of anthropogenic climate change may be allocated whereas non-anthropogenic climate change, also referred to as natural climate change, is disregarded. While the influence of human behaviour on climate change is undeniable (see again Section 2), it is not the only source of climate change. However, natural climate change creates the same threats to human rights as anthropogenic climate change, which imposes a moral obligation to also prevent nonanthropogenic climate change. Yet, in case of natural climate change, a polluter does not exist and cannot be made responsible. Thus, an additional principle for non-anthropogenic climate change would be needed in order to identify duty-bearers [25].
In summary, the polluter pays principle in its pure form cannot be considered complete as (i) in its micro-version, it suffers from the inability to specify an action's specific harm; (ii) in its macro-version, the outsourcing of emissions hinders the identification of polluters; (iii) it ignores other aspects of justice such as global poverty reduction; and (iv) it disregards natural global climate change. While the first two points refer to practical problems, the other two points are moral concerns.

Historical Responsibility Principle
Considering the historical GHG emissions described in Section 2, and in light of the idea that the polluter is a duty-bearer, it is sometimes argued that a historical responsibility exists based on the above-described egalitarian conception of the atmosphere as common good. Accordingly, the right to benefit holds regardless of the temporal or geographical scale. Disregarding the historical emissions would hurt this principle by favouring past economic growth and associated emissions. Thus, poorer countries, which have not experienced great economic growth yet, and future generations would be worse off. As anthropogenic GHG emissions are primarily associated with the beginning of the industrial revolution, costs should be allocated to countries that caused the highest emissions during history, mainly, the developed countries, based on historic accountability. Consequently, a Historical Emission Debt can be formally assessed as illustrated in Equation (1) [33]: This formula states that the Historical Emission Debt (HED i ) of a country i consists of the sum of the country's emissions from year j, which are still remaining in the atmosphere at the terminal period, y i (j), subtracted by the share of country i's population at year j, Pop i (j), and the world's population at year j, Pop w (j), times the world's emissions which are still remaining in the atmosphere at the terminal period, y w (j). The first period is s and the terminal period is e.
If a country caused high emissions in the past compared to the rest of the world, the HED is going to be positive. This implies that this specific country has to pay compensations to countries with a negative HED. Estimations of the HED have illustrated large differences between countries [34,35]. Many developed countries were found to have a positive HED (i.a., United States, Russia, Japan, Germany, Canada, United Kingdom, and Australia). The United States were found to have by far the highest HED, implying a continuous exceedance of the world average (per capita). Other countries such as India, Indonesia, Bangladesh, or Pakistan were found to have a negative HED [34]. Considering the HED's development over time, some countries experience an increase in HED (e.g., the United States, Russia, and Germany) while others have decreasing HED (e.g., India and Indonesia) [34]. Most notable is the development of China's HED, which was negative and historically low until the mid-2000s when the countries emissions (total and per capita) rapidly increased [34,35]. However, it has also been shown that the HED is sensitive to the considered period of time [34,35], as well as the examined range of emissions, e.g., also countries characterised by a high degree of deforestation and agriculture may have a high positive HED in addition to industrialised countries [34]. By formally assessing the HED, a new question arises: who is to charge if the pollution took place more than 150 years ago with the rise of the industrial revolution?
This question cannot be answered straightforwardly, even if the first intuition may be to consider the developed countries due to their historical emissions as polluters, or rather, duty-bearers. Nevertheless, historically, a variety of agents were involved in the process of emitting and associated climate change. The issue at hand is that many of the individuals and corporations which should be duty-bearers do not exist anymore. Due to the fact that the polluters do not live anymore, it would only be possible to transmit the responsibility for their actions to their descendants. From a moral perspective, this is problematic because the descendants did not undertake the unjust action, and from a pragmatic perspective, it appears to be challenging to define the polluters' descendants [28]. Thus, the only reasonable solution appears to be a shift in the responsibility to collective, transhistorical units, e.g., states. So, is it possible and just to define a state as duty-bearer?
As discussed above, the dimensionality of climate change is multi-layered. Climate change itself and its consequences happen on a global scale, and relevant public goods, e.g., the atmosphere, are global. Defining states as duty-bearers shifts the distribution of burdens to a national scale, whereas the states' citizens represent the individual scale. If the distribution of burdens remains on a national level, incentives to bear the costs are lacking on the individual level. Yet, considering that states are collective agents, they are capable of distributing the burdens on an individual level. Therefore, Miller [28] suggests a two-stage process: First, the costs of adaptation and mitigation are distributed among states. At the second stage, the states distribute these costs among their citizens based on nationally agreed upon rules. This approach overcomes the discrepancy between national and international responsibility while guaranteeing the nations' autonomy. Furthermore, the national policies for distribution within the states are legitimate based on the assumption of a foregone democratic process. Therefore, in the optimal case, social justice within the state is guaranteed, and incentives to reduce emissions are provided, e.g., through taxation of companies or private households. It can be argued that this two-stage process simply shifts the problem of missing incentives from an individual scale to an international scale because states may still act in their national self-interest. Thus, a sanction mechanism, or at least credible threat of a sanction, and a monitoring instrument would be needed. Miller [28] opposes that in a fair and voluntary agreement it is reasonable to assume compliance. Yet, even if states would voluntarily agree on burden sharing, the question remains if current states can be made responsible for unjust actions in the past.
A strong argument against this idea is the creation and disappearance of states throughout history. How should, for example, emissions of the former German Democratic Repub-lic, Yugoslavia, or the Union of Soviet Socialist Republics be handled? Miller [28] argues that instead of states, nations should be made responsible for the acts of their predecessors. He argues that contrary to the situation of states, nations share culture, identity, institutions, and practices over time. Thereby, despite perpetual replacement of individuals (e.g., through death or emigration) intergenerational bonds exist. This implies that due to the membership to a certain nation, the present generations could be held responsible for past emissions. Neumayer [33] instead suggests that if a new country comes into existence, it should be responsible for past emissions on its territory. In case that data on past emissions are not available, the share of population or the Gross National Product can be used as a proxy.
In addition to the identification of intergenerational duty-bearers, the ignorance of previous generations represents a primary issue for the principle of historical responsibility and thus for the argument that the developed countries have a duty due to their historical emissions. Bearing in mind that human-induced climate change and the resulting consequences were not foreseeable for previous generations due to their lack of knowledge (see, e.g., [36]), the harm was not intended, and the ignorance is considered excusable [25]. Therefore, the question arises whether ignorance can represent a strong argument against a compensation of the suffering caused by climate change. Shue [27] argues that it has to be differentiated between punishment and responsibility. While it would not be fair to punish someone in this position, the agent can still be held responsible for the action's unforeseen consequences.
Jensen and Flanagan [22] remark that generally in the discussion about compensation, too much emphasis is put on historical emissions. In fact, global GHG emissions in 2019 were roughly 44% higher than in 2000 and 59% higher than in 1990 [16]. Thus, the current generations are mainly responsible for the existing harm. Therefore, relying only on historical responsibility would fail to allocate a large share of the costs of climate change.

Beneficiary Pays Principle
It has been illustrated that the polluter pays principle does not necessarily distribute the burden towards wealthy countries, whereas, based on the principle of historical responsibility, only a fraction of the burden would be distributed due to the higher relevance of recent and current emissions. Against this background, it has been discussed that the current generations should bear the duty as they have obtained a high living standard as a side effect of past emissions and, thus, are beneficiaries of past economic activities [27,33]. Due to the fact that they take part in a continuous economic system, they are also bound to the duty. Although in the discussion on the distribution of responsibility this argument was used to strengthen the polluter pays principle (see [27,33]), it rather imposes a duty to compensate on the agent who has benefitted from an action. Therefore, it is also referred to as the beneficiary pays principle. In contrast to the polluter pays principle, agents who have benefitted from the implementation of an action or policy through a third agent, who has generated a negative externality, have a duty to not act in the same way and to address the damage caused [18]. Ignoring this duty can be understood as transgenerational free-riding because the negative externality was created by past generations, whereas current generations are better off without undertaking an action themselves, apart from accepting benefits [37].
The polluter pays principle and the beneficiary pays principle are, in practice, mutually exclusive. Theoretically both could apply at the same time but only when two conditions are met. Firstly, all polluters benefit from their action and everyone who benefits is a polluter. In an intergenerational context this could be the case but not necessarily. Secondly, the benefit has to correspond with the damage in order for the damage to be coverable. This is an unreasonable assumption because the pollution and resulting benefits are unlikely to be perfectly correlated [23]. Therefore, a combination of the two principles has been suggested. As long as the polluter is still alive, the polluter pays principle applies whereas the beneficiary pays principle is considered when the polluter is no longer alive (see [27,33]).
However, this proposition is inconsistent because, in the second situation, benefitting results in duty-bearing whereas, in the first situation, it does not impose duty simply because the polluter is still alive [23].
Considering both principles as mutually exclusive, the question is whether the beneficiary pays principle alone suffices. As mentioned above, the fundamental assumption is that current generations are beneficiaries of past emissions. At first, it has to be defined what is understood by the term beneficiary. Parfit [38] (p. 487) defines beneficiary or rather a beneficial situation by 'the two-state requirement' which states that: "We benefit someone only if we cause him to be better off than he would otherwise at that time have been." Considering the intergenerational context of climate change, this would mean that the current generations benefit from past generations' emissions if they find themselves in a better state than they would have without these emissions. This is the fundamental assumption that has to hold in order to apply the beneficiary pays principle. Parfit's [38] socalled non-identity problem raises a strong objection against this assumption and illustrates the difficulty of weighting intergenerational interests. The non-identity problem assumes that choices undertaken in the present can affect the identity and the amount of people living in the future, although this does not necessarily have to be the case. On this basis, two overarching situations can be differentiated: same people choices and different people choices.
In the first case, current choices cannot affect the identity nor the amount of future people. This is referred to as same people choices. In the second case, the identity and the number of future people can be affected, this is called different people choices. In the latter case, it has to be distinguished if only the identity but not the amount of future people is affected (same number choices) or if both is affected (different number choices) [38] (p. 356). The actions associated with global climate change can be considered to have a strong impact and thus are different number choices [22]. Following Parfit's [38] argumentation, this implies that these actions cannot harm future generations because without these actions, even if they accompany environmental damage, the future people would not have come into existence. Considering the above given definition of a beneficial situation, it cannot be concluded that emissions of past generations have benefitted current generations because of the impossibility to define a benchmark if other identities were in existence. Thus, the beneficiary pays principle is inadequate for an intergenerational context and does not impose a duty on the developed countries to bear the burdens of past generations [18].

Ability to Pay Principle
The ability to pay principle does not consider any historical emissions. Thus, as a forward-looking principle, it leaves any potential debts or duties from past generations aside. As the name implies, this principle allocates responsibility based on greater ability to pay. Shue [27] (p. 537) defines ability to pay in the following way (second principle of equity): "Among a number of parties, all of whom are bound to contribute to some common endeavour, the parties who have the most resources normally should contribute the most to the endeavour." Thus, the more 'excess capacity' a country has, the more it should contribute to combating climate change [39] (p. 561). In order to define fair contributions-or to be precise: payments-in accordance with the ability to pay, a rate of contribution has to be chosen. This rate should consider the correlation between wealth and payment in an adequate way. A flat rate of contribution implies an equal share for every agent which seems inadequate considering that it does not take into account the ability to pay. Furthermore, it may situate agents beneath the earlier discussed guaranteed minimum if the rate is chosen too high. Therefore, a progressive rate is more suitable to achieve a fair distribution of costs [27]. This poses an issue already observed for the polluter pays principle, namely, that the extent of harm has to be assessed in order to define fair contributions [25].
Irrespective of the issue of defining a fair rate of contribution, the moral question at hand is: Is it fair if someone, in this case, the advantaged or rather wealthy, who did not actively cause any harm is assigned the responsibility for existing and/or future damages and harms? To answer this questions, two different scenarios have to be considered [20,25] (p. 43ff.): business as usual and efficiency without sacrifice. In the business as usual scenario, emitters still behave as before to the detriment of those suffering from the negative externality. Thus, agents who have not undertaken the action themselves but are exposed to the consequences have to pay, in the context of global climate change, most likely the poor due to their high vulnerability. In the efficiency without sacrifice scenario, GHG emissions are reduced, which implies harming the emitters who miss out on benefits. However, the externality represents an inefficiency, and thus, emitters could be compensated by agents who were beforehand suffering from the externality. From an efficiency perspective, this is favourable over business as usual because some are better off without anyone being worse off. Still, again, an agent who was not the emitter might have to pay for the compensation. Thus, the ability to pay principle cannot be invalidated by the argument that the advantaged did not cause the harm, as the ones suffering from the externality also did not cause it but would have to burden the costs otherwise [25]. Thus, a third scenario, efficiency with sacrifice, in which emitters reduce the GHG emissions without being compensated is most favourable. As in the efficiency without sacrifice scenario, the externality is corrected, but the distributional effects are different. Resources are shifted towards the future, and these investments will generate higher benefits than currently possible. Thus, total benefits are higher with the same input of resources suggesting that a sub-optimal emission regime leads to a pareto-improvement [20].
However, Kelleher [40] criticises this welfare-economic perspective with its focus on (pareto) efficiency and its definition of costs and benefits in terms of willingness to pay and willingness to accept. Considering the above-described non-identity problem, this conceptualisation becomes problematic. The benefits of the internalization would have to be measured in terms of the willingness to pay of future generations. Thus, the willingnessto-pay of future generations in the business as usual case serves as a benchmark. However, as discussed above, the relevant actions are the different number choices. Therefore, changes in current emission behaviour or policies would have an influence on the identity and the amount of future people. This implies that future people would not be willing to pay for a reduction in GHG emissions as this would mean that they will never come into existence. Only if the life of future people would not be worth living would a willingness to pay be reasonable [20,40]. Thus, climate change cannot be oversimplified to a paretoinefficiency, and therefore, efficiency without sacrifice is implausible in the context of climate change [40].

Discussion and Conclusions
Climate change is a significant challenge due to the associated threats and costs for current and, especially, future generations. As discussed, a major driver of climate change is human activities, to some extent past activities, but to a larger degree activities undertaken by current generations. It has been argued that climate change represents an injustice which needs to be addressed. Therefore, the capability of four principles, all based on the theory of justice, to ascribe specific responsibility and the associated moral duty to agents in order to allocate the costs of climate change in a fair way has been assessed.
The analysis of the principles (polluter pays, historical responsibility, beneficiary pays, and ability to pay) has shown that in its purest, form neither of them is able to allocate the costs of climate change in a sufficient manner. For every principle, shortcomings have been illustrated which would hinder the fair allocation of costs: The polluter pays principle is unable to account for non-anthropogenic climate change and does not necessarily define the wealthiest as duty-bearer. Yet, this is considered to be fairer than forcing a payment on a poor agent. Furthermore, based on this principle, developing countries with currently high GHG emissions would also have to pay for costs arising as a result of climate change. This might undermine other efforts to address global problems, e.g., poverty reduction. As a result, a problematic valuation or weighting of present and future problems would be needed.

2.
Historic responsibility is incomplete because most of the relevant agents are not alive anymore. The definition of duty-bearers is disputable because no unambiguous definition of transhistorical units can be made. Thus, a clear allocation of costs is considered impossible. Additionally, the ignorance of past generations and the definition of the temporal boundary of ignorance would have to be considered.

3.
In an attempt to overcome the dilemma faced by the polluter pays principle, the beneficiary pays principle defines agents who have benefitted from a third agent's emissions as duty-bearers. Nevertheless, the amounts of benefits and costs are likely to differ, which would result in uncovered costs of climate change. Furthermore, based on the non-identity problem it was illustrated that only under strict assumptions it can be argued that beneficiaries actually exist. 4.
The ability to pay principle fails to specify the costs in an appropriate way as the approach relies on the willingness to pay of future generations. Again, the inability to apply this principle is based on the non-identity problem. Ultimately, the willingness to pay has to be expected to be zero in different choice situations.
As illustrated, none of the four principles in their pure form appears to be defensible with respect to a fair allocation of the cost of climate change either due to the incomplete identification of duty-bearers and/or the inadequate specification of duties. Thus, the moral claim that the developed countries should burden these costs may be better supported by other means. Yet, the above presented analysis was limited to a rather narrow dimension of justice with a focus on harm and burden-sharing and the rather strong assumption that one principle is able to provide a complete account of the distribution of costs associated with climate change. Departing from the idea that a single principle is able to do this, hybrid models may combine the advantages of the individual principles in order to sequentially define relevant duty-bearers and burden-sharing (see, e.g., [39]). Furthermore, the dimension of justice may be extended towards rights-based approaches and capabilities (see, e.g., [41,42]), as well as harm prevention (see, e.g., [43]). With this in mind, the adoption of a more forward-looking approach with a broader scope of justice may also consider the precautionary principle (see [44]), as the harm of climate change is considered to be serious and irreversible. On this basis, the recent German Federal Constitutional Court's decision argued in favour of a 'fundamental right to a future in accordance with human dignity and a fundamental right to an ecological minimum standard of living' [13] (p. 27/78).
Additionally, when shifting the focus on human rights and the future, the non-identity problem must not necessarily overturn the discussed principles. Firstly, it relies on the assumption that future people are not worse off than they would otherwise have been as long as their existence is worth living. However, if it is assumed that climate change will harm future people to such an extent that human rights are threatened, independent of their identities, a duty of justice would nevertheless arise. Otherwise, people would be unable to live a minimally decent life [28]. The bare existence cannot be considered to be a better state than non-existence. Secondly, as discussed above, collectives, e.g., states or nations, can also bear responsibility for actions. As climate change represents a global problem across borders, for pragmatic reasons, it seems feasible that collective entities rather than individuals should bear the major burden. Thus, if individuals are held collectively responsible for actions, the collective could receive the associated duties of justice.
Lastly, the boundaries of the justice dimension, to be precise, the scope of moral duties, could be extended to include duties of goodness in addition to duties of justice. In contrast to duties of justice, duties of goodness are not owed to specific people but to generally improve the world or specific affairs. Therefore, harmful emissions are a violation of the duties of goodness to promote opportunities of good lives [20].
In conclusion, the question of how a fair allocation of costs of climate change can be achieved is not trivial to answer, and there is still a need to combine various non-mutually exclusive conceptions of justice to achieve a fair and complete account of the distribution of costs associated with climate change.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in Figures 1-3 are openly available and published online at OurWorldInData.org, [45] (accessed on 9 December 2020).