In 1990, the first intergovernmental climate change report concluded that carbon emissions in the atmosphere were causing the planet to warm to harmful levels  . Since the atmosphere is a common resource – or a "common concern good" in international customary legal terms – and has been proven to have limits, questions were raised over how to protect it  ,  . Climate change has been defined as a scientific issue, but it is not science's prerogative to tell us how to share the atmosphere. In the absence of 'clear and precise rules of international positive law'  , the United Nations Framework Convention on Climate Change (UNFCCC) commits to stop global warming 'on the basis of equity'  . There are of course many ways of interpreting what the UNFCCC means by equity. The Kyoto Protocol – the foundational climate text of the UNFCCC – is seen by many as an attempt to reach a more equal situation between countries in terms of emissions levels. However, in reaching for equality, it inscribes the principle of common but differentiated responsibility. It legally binds industrialised nations, who have done most to incur climate change through their historical emissions [note 1] , to 'act first' to reduce their emissions [note 2] . As seen in the latest round of climate negotiations in Durban, this foundational principle of common but differentiated responsibility agreed upon in 1997 is now being fatally challenged by developed countries who see that emerging economies are emitting just as much as they are  . They may ask, why, if we are attempting to create a more equal situation between countries, do we insist on differentiated responsibility to reduce emissions? Since this seems to be the main area blocking progress on achieving an international climate agreement, should we not forego this principle and agree to equal responsibility among countries  ? In this final year of the Kyoto Protocol – which expires at the end of 2012 – it is important to revisit the principle of differentiated responsibility. Through a re-reading of egalitarian distributive justice, I will be exploring the difficulties that the issue of climate change poses to an egalitarian distributive model  ,  ,  ,  , using Marx's theory of uneven development to highlight the importance of transitional retributive justice in achieving egalitarian justice.
Difficulties with distributive Justice
To understand the Kyoto Protocol's insistence on 'common but differentiated ' responsibility, it is important to first understand the principles of distributive justice. Distributive justice pertains to the distribution of burdens and benefits according to what is fair. What is considered a fair distribution is open to interpretation. Fair distribution as justified by a utilitarian or libertarian approach is based on the importance of liberty and opportunity, insisting that 'redistribution' is unfair and interferes with the natural order  ,  . In the UNFCCC arrangement, parties such as the USA champion this approach. A typical utilitarian argument is that those who can make best economic use of the distribution should be allowed a higher distributive portion because they will bring higher overall value. In contrast, developing nations emphasise the importance of an equal opportunity to grow, insisting that distribution must be based on historical emissions and a per capita allocation if it is to be just  ,  . This distinction of 'per capita' rather than 'per country' is important to the principle of egalitarian distributive justice, since each individual should have the same right to develop, and given the finite nature of the atmosphere, if some emit more, it means that others must emit less if we are to avoid catastrophic climate change [note 3] . Looking back historically in the way they propose is particularly important in terms of climate change because the emissions that were emitted in the past are still in the atmosphere now. Climate change is a result of cumulative emissions, therefore the past must be taken into account. If developing nations believe in an equal share to the atmosphere, why then is there an insistence on 'differentiated' responsibility to reduce emissions? Surely an egalitarian distributive approach to tackling climate change should see an equal division of burdens and benefits, not one that is differentiated. The reasons for differentiation, which I will elaborate on below, are two-fold: the first pertains to geography, and the spatial unevenness with which the effects of climate change are distributed. The second pertains to a historical and geographical view of cumulative emissions and uneven development.
A review of how distributive justice works in the environmental justice field begins to show us that distributive justice poses significant difficulties with respect to climate change. Environmental justice demands that environmental benefits and burdens are shared out equally  . A classic example is that of waste, where activists oppose the systematic construction of heavily polluting incinerators in poor neighbourhoods, asking for an equal distribution in the burden of waste management. The issue of climate change, however, cannot be dealt with according to the same egalitarian distribution because it lacks the same materiality. Although we have argued for an equal distribution of climate benefits in equal shares (the atmosphere) it is impossible to divide climate burdens in the same way (negative effects of climate change) and this 'spatial problem' has consequences for the theories of justice we have been presently exploring  . Climate science is highly complex and its effects are impossible to fully predict, but the fourth IPCC report clearly shows a geographical unevenness in the way it affects different parts of the world, with many countries in the Global South experiencing its worst effects  . Climate change takes an already unjust global distribution and exacerbates its undeserved inequality to the detriment of the least advantaged. Even in a situation where each country was equally responsible for climate change, egalitarian distributive justice would not in this instance help us achieve the most just outcome. How then do we go about finding a just solution? Can we be content with theories of distributive justice?
Distributive justice theories, whether normative or descriptive, help us analyse what is just, and give us models by which distribution can be improved  . However, our particular empirical case study determines that that they do not help us attain that theory of justice. Above, I have argued that egalitarian distribution of the atmosphere is the most just division, in view of the fact that the atmosphere, 'of all planetary natural resources it is the one that comes closest to being a pure public good'  . Since historically it has not been divided equally, we concede that corrective work must be done, which necessarily means departing from egalitarian distribution. 'Identifying a source of inequalities is not enough to warrant an egalitarian distribution: one must show that those who are worse off are not responsible for their predicament'  . Theories of corrective or retributive justice lend us the legal language of cause and responsibility, which I argue is key in achieving a just climate agreement. I will be using retributive theories to ask how fair these distributional differences are. What the root causes of these distributional differences? In Simms's words: 'there are key reasons why some people get more than others'  .
Marx discusses this phenomenon, which he calls the theory of uneven development, in volume one of Capital : 'Capital grows in one place to a huge mass in a single hand because it has in another place been lost by many'  . Since uneven development is thus actively produced, could we use the theory of uneven development to help us restore this balance through a theory of corrective justice? Though this does not mean we are abandoning egalitarian distributive justice, we are now reframing it as an end-state for climate negotiations, rather than as a way of achieving egalitarian distribution. Of course we know that every country does not, and has not in the past, had a fair share in the atmosphere. Energy specialists Arnulf Grübler and Nebojsa Nakicenovic have calculated cumulative emission for the entire world since 1800, showing a developed to developing ratio of CO2 emissions of 4:1  . Even World Bank figures, which look at emissions a-historically, show that today's per capita emissions are radically lower for developing nations than all other industrialised nations [note 4] . The climatic change that many countries in the global south are enduring is therefore not their responsibility. This acknowledgement helps us to explain the FCCC's departure from egalitarian distributive justice and differential treatment of Annex I countries and begins to show in what way, according to theories of cause and responsibility, this is in fact just. Though it cannot be helped that countries in the Global South are more affected by climate change for geographical reasons, an awareness of the causes of climate change may be able to prove that causal, if not moral, responsibility lies with industrialised nations, which, according to egalitarian distributive justice, would mean a carbon debt has been created. Grounded in principles of justice, many voices are now calling for this debt to be repaid, invoking the principles of retributive justice.
Climate debt: retributive justice
Leading the 'climate debt' argument, Andrew Simms, a front-line environmental campaigner argues that as with anything, if you use more than your fair share of anything, you run up a debt  ,  . If this debt is linked to the future of the planet's wellbeing, this debt must be taken very seriously. Andrew Simms refers to this as 'when sorry is not enough':
(There is) a growing awareness that there are global commons which provide 'public goods.' There are things to which we all have an innately equal claim, yet which have been very unequally used. If a global commons like the atmosphere, to which we all have an equal claim is being overused and corrupted by one group of people, they accrue an ecological debt to the wider community who also depend on the commons. 
The principles of distributive justice show us clearly that with regards to the equal sharing of the atmosphere, an imbalance has occurred. The particularity of climate change, as discussed above, is such that distributive justice cannot address this injustice. Standard retributive justice demands that compensation is paid to the injured parties. In this instance, however, this is neither an obvious choice, nor an easy one to administer. Indeed, a lively, if sometimes ferocious, debate surrounds our philosophical deduction, which seems to prove the fairness of 'common but differentiated responsibility' and goes to the heart of questions of justice: why should this debt this be accepted?
Aware that discussions of debt and retribution are sensitive subjects, it is first necessary to ask in what way retribution is just? The 'polluter pays' principle, which lies at the heart of environmental retributive justice, offers a 'responsibility based account of the proper distribution of costs: those who are responsible…should pay in proportion (to the harm caused)'  . Scientists have measured that in 1990, 75% of all CO2 emissions still in the air had been emitted by industrialised countries. Though the developed countries that are challenging the Kyoto Protocol can and do deny moral responsibility for emissions before 1990 they cannot, on scientific grounds, deny causal responsibility  ,  , [note 5] . This is a difficult position to contend with, and indeed few political scientists tackle this question directly for lack of a response, since in common and criminal law a crime is only a crime, actus reus , if it is done knowingly, mens rea . In the absence of international government, it is difficult, indeed impossible to pursue historical debt in a legal framework – though this did not stop the President of Tuvalu, Koloa Talake, from filing a law-suit in 2002 against the single biggest emitter and the biggest per capita emitter in the world (USA and Australia) [note 6] . On what basis then should developed countries accept a debt? Here, we come back to Marx's theory of uneven development for a normative response. 'They must pay because their riches pay back to a past in which the world was open'  . Marx's theory of uneven development helps us to argue the validity of a debt on the basis that developed countries have accumulated wealth to such an extent that 'only they can', both financially, and politically speaking, pay back the debt that will stop global warming.
Importance of being feasible: international paretianism
This discussion has begun to show why I refer to retributive justice as 'transitional' but I wish to state my reasons more explicitly. The first is to highlight that this is not a punishment or a moral judgement, rejecting the use of punitive language such as 'reparations' [note 7] . Though we acknowledge and accept that communities in the global south affected adversely by climate change are rightfully angry about this debt, such morally loaded language must be avoided at the negotiating table if industrialised countries are prepared to discuss adaptation costs and mitigation targets on the grounds of a debt. Retributive justice that 'seeks to balance an injustice by rectifying the situation, or by regaining an equality that the injustice overturned' does not necessarily have to be punitive  . The second is that retributive justice is transitional in that once it has recreated a balance, achieving egalitarian distribution; it no longer needs to be applied. Grübler and Nakicenovic show that with an effective national carbon rationing plan, each citizen's carbon ration can be at an equal level by 2050  , according to the principles of egalitarian distributive justice. For an agreement to succeed, we must think in the long-term.
In my opening section I asked why it was so important that an international climate agreement is based on justice, asking whether justice and effectiveness were synonymous. On one hand, the Bolivian government submission to the FCCC insists that 'there is no viable solution to climate change that is effective without being equitable'  . On the other, taking negotiations in Durban as an example, it can be argued that an insistence on egalitarian distributive principles of justice are slowing down the process, with critics arguing that the most just agreement is one that actually exists  , and should thus be more feasible and paretian. I fully agree with this last point, and it is for these reasons – to mobilise wide cooperation – that I have argued for egalitarian and retributive justice  ,  ,  . As many critics of the Kyoto Protocol and the FCCC mention, developing country emissions are rising very fast, though they are still incomparable to historical emissions of industrialised countries (see Figure 1)  ,  .
In order to ensure that dangerous anthropogenic interference with the climate system is avoided, it will be important that fast industrialising developing nations such as China, India, Brazil and South Africa begin to accept emissions restraints  . In this analysis, developing nations are no longer viewed as climate victims but powerful brokers of a future agreement, a reality developed nations must swiftly come around to if a second commitment period is to be negotiated in Cancún and South Africa, the venues of the next two COPs  . Shukla explains that 'players reach a voluntary agreement only when it makes every player better off', which we have been referring to as a pareto improvement  . Thus, the only way that developing countries will sign up to emissions reductions is if industrialised nations do so first, and respect their targets. 'Justice in this context is vital for inviting wide participation from developing countries in the climate regime, a main criteria for the success of the regime'  . Negotiating on the basis of a just agreement is therefore the most efficient way of stopping global warming and saving the planet.
Through an in depth discussion of theories of distributive and retributive justice, I have attempted to show the importance of egalitarian justice to the climate negotiations in order to reach consensus on a legally binding agreement. A decade on, the geopolitical landscape and environmental situation has changed much. We must however ensure that the second commitment period under Kyoto retains the same strong language and principles of equity that were signed by 194 countries in 1997. When George Bush withdrew the USA from the Kyoto Protocol, his primary grounds were of its 'unfairness' over the 'disparity of treatment'  . The Senate's decision was clearly referring to the FCCC's analysis of equity on a 'per capita' basis, enforcing 'common but differentiated responsibility' that would employ corrective justice to achieve egalitarian distributive justice. This discussion has used theories of justice to explain this departure, and show its ultimate fairness, but not at the expense of making an international agreement unfeasible. On the contrary, my case, based on paretian principles, is that ultimately, an agreement that commits to paying back a valid climate debt is the only one that (fast industrialising) developing countries will sign up to, which should be the primary concern of industrialised nations, who, with the highest economic stakes, have the most to lose.
1 Historical, or 'cumulative' emissions are the total amount of emissions that a given country has emitted since the beginning of industrialisation, which is often dated as far back as 1850, though the date the UN uses is 1990. For a more complete discussion of historical emissions, see Gosseries 2007 and Shukla 1999  ,  .
2 It notes that 'the largest share of historical and current global emissions of greenhouse gases has originated in developed countries and that per capita emissions in developing countries are still relatively low' and so does not place reduction commitments on developing countries in the first commitment period in order that they have the opportunity to 'meet their social and development needs' (Ibid).
3 It is also important to highlight that per capita emissions are based on a division of a given country's emissions by the number of inhabitants in that country. This does not take into account that the emissions of most developing countries are released whilst producing consumer goods for developed countries. This is an additional injustice that, while much harder to calculate, requires attention while making calculations over emissions levels per capita.
5 1990 is often mentioned as a benchmark since it is the date the first IPCC report was published, and taken by the Kyoto Protocol to be the date beyond which could no longer claim ignorance of climate change.
7 See photo in Appendix 1 .
 Schlosberg, D 'Reconceiving environmental justice: global movements and political theories', Environmental Politics, 13: 3, 517-540, DOI: http://dx.doi.org/10.1080/0964401042000229025 , 2004.