Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Chapter Six: Distributive Justice and the Control of Global Warmin

V Bhaskar


Global warming raises unique questions about our responsibilities to future generations.' - Thus begins John Broome's recent book,

Counting the Cost of Global Warning, which argues that we have a responsibility to take action today so that the welfare of future generations is not adversely affected. The argument that Broome makes is a moral one, not one based on self-interest - after all, you and I will be dead and gone by the time that the greenhouse effect makes the oceans rise and alters our planet's climate in unforeseeable ways. The unborn are powerless, and if we choose to be guided purely by self-interest, we could bequeath them a wasteland without even being around to hear their reproach. Broome argues that our actions must be guided by ethical criteria, and that we must take remedial action to curb the emissions of greenhouse gases. The precise extent to which we do so depends upon how we morally evaluate alternative distributions of 'well-being' across generations. Broome does not provide any answers to the question of how much we must curtail greenhouse gas emissions. His purpose instead is to set out a framework, an ethical framework, within which such questions may be addressed.

Broome's dissection of inter-generational distributional questions is insightful and valuable. It establishes clearly that global warming is a moral issue. No discussion of it can be divorced from ethics, and from the ethical responsibility of today's denizens of the globe towards their descendants. However, one omission is striking - there is no mention of the heterogeneity of today's generation, of the different standards of living between the typical American and the typical Sudanese, or indeed between the Japanese and Bangladeshis of tomorrow. Nor is there any discussion of the different degrees to which these societies have exploited the global commons to date. What implication do these large differences have for the division of responsibilities? Do a Sudanese and an American have equal responsibility to future generations? What does 'equal responsibility' mean in the context of global warming?

Of all issues which impinge upon the welfare of future generations, global warming is perhaps the one which explicitly requires a treatment of intra-generational distributional issues. The global environment is a global public good, and greenhouse gases, which contribute to global warming, are an archetypal public bad. It does not matter whether a unit of carbon dioxide is emitted in Bangladesh or the United States - it contributes equally to global warming in either case. Consequently we must define, in global terms, the responsibility of current generations the world over to future generations. Inevitably in doing so, we confront the division of the burden of this responsibility within the current generation.

This paper begins where Broome's book ends. Given that we have a moral responsibility towards future generations, we have to consider what is the moral basis for distributing the burden of our responsibility within the current generation. How, in particular, this burden should be distributed between nation states (although in principle one can go further, and allow for heterogeneity within nations).

The primary concern of this paper, based on the ethical distribution of the burden of emission control, has been questioned. It is argued that questions of ethics are irrelevant, and the distribution of international burdens across countries will inevitably be determined by power-politics. I do not believe that this is entirely correct. The outcome of international negotiations may be determined by bargaining power; however, bargaining power is not determined entirely by material factors - it is also influenced by the perceived morality of one's position. The importance of ethical considerations may be greater than usual in the case of global warming, since we will feel a need to undertake any emission control only if we are concerned about future generations.

The remainder of the paper deals with a number of interrelated issues. The empirical evidence on the current distribution of global emissions of greenhouse gases is detailed. Next we ask whether there is a conflict between an efficient programme of emission reductions and an equitable one, and argue that traceable emission permits can ensure efficiency, allowing us to base the allocation of emission entitlements on grounds of equity. The next section sets out a simple model which explains how earlier generations may use the global environment in order to affect the intra-generational distribution of welfare in the future. The rest of the paper focuses on the moral issues. First we consider welfare-based theories and theories based on rights, and suggest that in the case of global warming, both theories give similar answers. A discussion follows of some of the specific criteria which have been suggested for allocating emission entitlements. Finally we address the question of historical responsibility: since the stock of greenhouse gases is due to past emissions, which are overwhelmingly by the North, should we correct for this differential exploitation of the global environment in the past, or should we let bygones be bygones?


Human economic activity has, in the last 100 years, contributed to an increase in the concentration of 'greenhouse gases' - of which carbon dioxide (CO2), methane, nitrous oxide and chlorofluorocarbons (CFCs) are the most important. The increased atmospheric concentration of these gases gives rise to the 'greenhouse effect', whereby a larger proportion of solar energy is trapped by the atmosphere instead of being reflected back into space. This has the effect of raising global temperatures, which could have major effects upon global climate. The consequent melting of the polar ice-caps would raise sea levels, and could adversely affect low-lying areas such as Bangladesh, the Netherlands and island states. The precise size of the greenhouse effect and its impact is a matter for scientific debate, and far from settled. The economic costs of global warming are also a matter of controversy - Nordhaus (1991) for example argues that the costs are small whereas Cline (1992) is less sanguine. Nevertheless, the issue is too serious to be dismissed given the magnitude of projected increases in CO2 concentration. Carbon dioxide emissions increase rapidly with industrialization and the burning of fossil fuels. As developing countries industrialize, one can expect their levels of CO2 emissions per capita to gradually 'catch-up' with current levels in the developed world. If the situation continues unchanged, the Intergovernmental Panel on Climate Change (IPCC) estimates that CO2 concentrations will double (as compared to pre-industrial concentrations) by the year 2020, and will be more than quadrupled by the end of the 21st century. These large changes are likely to have non-linear impacts upon the global environment and the economy, which we can predict only approximately. Carbon dioxide exists in the atmosphere for a very long period of time. Consequently, changes in CO2 concentration will be long-lasting, and can only be reversed gradually. For these reasons it appears that precautionary measures to control the emissions of greenhouse gas emissions are warranted, and indeed the issue is on the international agenda today, although coordinated action to tackle the problem may yet be a long way off.

Of the greenhouse gases, carbon dioxide and methane are the most significant - the role of nitrous oxide in global warming has been downgraded in recent years, and agreement has been reached on the phasing out of CFCs. Carbon dioxide is resident in the atmosphere for an estimated period of between 50-200 years, while methane has a shorter time span of about ten years. Consequently, the long-term effects of emissions are substantially greater in the case of carbon dioxide. The primary source of CO2 emissions is industry, due mainly to the combustion of fossil fuels. Second in importance are emissions from biota sources, due to deforestation and other changes in land use pattern, which are mainly concentrated in developing countries. Estimates of emissions from biota sources are extremely divergent, and for individual developing countries total emissions can vary by a factor of two to four depending upon the estimate used. Table 6.1 gives an estimate of the distribution of industrial CO2 emissions world-wide, as well as an assessment of the contribution of various regions to current concentrations.

Table 6.1 Percentage shares in CO2 emissions, population, GNP, 1988
1988 Cumulative Population GNP
emissions emissions
North America 25.2 33.2
Western Europe 15.0 26.1
Eastern Europe + USSR 25.6 19.6
Japan + Oceania 6.1 4.8
Developed countries 71.9 83.7 23 84
LDCs 28.1 15.0 77 16

  1. Industrial CO2 emissions.
  2. contribution to current concentration of CO2 in the atmosphere, based on cumulative emissions since 1800, adjusted for decay.

Sources: Grubler and Nakicenovic (1992), Young (1991)

There is substantial uncertainty about the extent of anthropogenic methane emissions. Methane emissions in developed countries are relatively small, and are due to the use of fossil fuels and livestock/animal waste. In LDCs, methane emissions are due principally to livestock waste and rice cultivation. The extent of these emissions is extremely difficult to calculate, and there are a range of estimates in the literature. Table 6.2 presents the 1988 emissions of CO2 and methane in terms of gigatons/tons of carbon equivalent. It shows that the distribution of carbon emissions per head of population is highly uneven. The basic asymmetry is that the 'North' has a level of per capita emissions over four times that of the 'South'. There are further variations within the North, with North America having a level of per capita emissions over twice as large as Western Europe and Japan. The USSR and Eastern Europe have higher per capita emissions than Western Europe/Japan. Within the South, Brazil has a relatively high level of per capita emissions, and the exact figure could be higher due to uncertainties about the detrimental effects of deforestation.

Table 6.2 Regional CO2 and methane emissions 1988

Per capita
per unit GDP
North America 1.64 6.06 0.34
Western Europe 1.00 2.62 0.21
Japan + Oceania 0.40 2.82 0.22
Eastern Europe + USSR 1.70 4.25 0.75
Developed countries 4.73 3.97 0.35
LDCs 3.39 0.87 0.45


  1. Total emissions in gigatons of carbon equivalent
  2. Tons of carbon equivalent per head of population
  3. Tons of carbon equivalent per $1000 PPP equivalent

Source: Grubler and Nakicenovic (1992)

It is also instructive to consider emissions per unit of GDP, not so much because this is an appropriate way for allocating emission rights, but because this measures, very crudely, the efficiency with which output is produced relative to the global cost in terms of emissions. GDP is measured in dollars, and the conversion can be made either at the market/official exchange rate, or in purchasing power parity units. We have chosen the latter, which is a better measure of real income. Table 6.2 shows that the South has a higher rate of emissions per unit of GDP than the North, but this difference is not as large as one might imagine, being about one third higher. Eastern Europe and the USSR have very high emission/GDP ratios, more than double that of the OECD countries. There is also substantial divergence between OECD countries, with North America having an emission/GDP ratio of 1.5-2 times that of Japan/Western Europe.


Is there a conflict between efficiency and equity in the distribution of emission entitlements? Prima facie, there indeed seems to be some conflict. Many of the developing countries have a low level of overall emissions as we saw in Table 6.2, per capita emissions are much lower in the developing world as compared to the developed world. However, the industries in these countries are often energy inefficient, use older technologies and are hence more polluting. This is illustrated by the higher level of emissions per unit of GDP in the developing world - a crude measure only since the composition of GDP is very different in different countries. Countries with a high ratio of emissions to GDP, and in particular with a high ratio of fossil fuel CO2 emissions relative to GDP, will tend to have lower costs of abatement. This is because their current technologies/practices are often energy inefficient relative to alternative technologies which are available. It is argued that this is the case in the former USSR/Eastern Europe and China and India, where the greater use of coal also contributes. Many developing countries could also reduce emissions by controlling deforestation; a measure which may be warranted independently, quite apart from its impact upon the greenhouse effect. Since greenhouse gases are a truly global pollutant, it matters not, from the point of view of global warming potential, where emission reductions are brought about. If it is cheaper to reduce emissions in developing countries rather than in developed countries, this would be an efficiency argument for focusing on reducing emissions in the former. However, would this not conflict with egalitarianism?

It is indeed the case that any system of non-traceable emission entitlement will bring about a conflict between equity and efficiency. There is no reason for these two principles to coincide, and indeed, for some very plausible principles of equity, such as equal per capita entitlements, the divergence can be very large. This criterion implies that developing countries could increase emissions, while developed countries make substantial reductions. However, an efficient system of emission reduction is one where emissions are reduced at lowest cost, in terms of output. If emission entitlements are distributed on a per capita basis, targets which are acceptable to developed countries will be well above what developing countries need, and hence the latter will be under no pressure to reduce the GHG emissions in the process of development.

Barrett (1991) estimates costs of reaching a given emission target for the European Community under alternative schemes for distributing this target among member states. He finds that uniform obligations are very cost inefficient, and in an illustrative calculation, finds that their total cost is almost 50 times greater than a cost effective policy. The cost difference is likely to be even larger in the case of the world as a whole, where differences in the marginal costs of reducing emissions are likely to be substantially larger than within the European Community.

A second problem with non-traceable emission entitlements is that distributional considerations will influence the global emission target. To illustrate this point, consider the case of equal percentage reductions. This will be extremely costly for low polluters and for the developing countries, and consequently they will press for higher global emission targets. Considerations of equity will therefore dictate a higher global target than that indicated by our earlier analysis. Take another case, of equal per capita emissions. This may be more equitable; however, countries with low levels of per capita income, but with a high level of emissions relative to income (such as many developing countries including China), will have no incentive to reduce emissions, even when these can be acheived at relatively low cost. Opportunities for cheap emission reductions will be foregone as a consequence.

This conflict between efficiency and equity can however be avoided by allowing emission entitlements to be traded. With traceable entitlements, a country which is obliged to reduce emissions must either do so itself, or persuade another country to do so, and buy its entitlement. A country such as the United States which has a high level of emissions, but which may also have a high marginal cost of reducing emissions, has the option of persuading China to control its emissions in exchange for monetary compensation. Indeed, if the market for traceable entitlements is competitive, it can be shown that the resulting distribution of emissions will be efficient - a target level of emissions will be achieved at minimum cost. Consequently the entitlements may be distributed on an equitable principle, since trade allows us to achieve efficiency independent of the initial distribution of entitlements. The system we refer to is of course one of traceable emission quotas. Each country is allocated a quota of permissible emissions on some distributional principle. Countries may however trade their quotas, so that a country which would like to emit more CO2 than its quota can buy quotas from a country which has a surplus of quota relative to emissions. If the market for quotas is competitive, the price of one unit of emission quota is the opportunity cost of one unit of emissions in terms of convertible currency, and this is the same for all countries. Consequently this system achieves an efficient allocation of emission reductions.

It must be noted here that a global target for emissions can also be achieved efficiently via a uniform global tax on emissions, which would be levied upon individual countries by an international authority. This is the carbon tax which has been discussed by a number of authors (Epstein and Gupta 1990; Hoel 1992; Whalley and Wigle 1991). This tax revenue can be redistributed to countries in ways consistent with distributional criteria. It can then be shown that the global tax achieves an efficient allocation of emission reductions provided that no country has a large share in tax revenues at the margin. A system of traceable entitlements is similar to a global carbon tax - indeed, in the absence of uncertainty and with an appropriate distribution of tax revenue the two systems are equivalent. We shall focus on quotas since they can be naturally related to our discussion of equitable entitlements.


Greenhouse gas emissions are an example of market failure, and indeed of the failure of individual nation states, which requires remedial action at the international level. Individuals in the course of economic activity emit greenhouse gases which contribute to global warming. The costs of their economic activity are incurred by future generations worldwide. Although each individual may be concerned about the welfare of future generations, he/she cannot make sufficient impact as an individual to be able to take corrective action unilaterally. Indeed, since the greenhouse effect is a global problem even nation states are too small to internalize this negative externality. The global environment is perhaps best seen as a public good, which will be collectively bequeathed to future generations. Individual nations, by reducing their emissions, leave a safer environment for future generations worldwide. However, emission reductions in any one nation may be too small to have a significant global effect, and furthermore, can be offset by increased emissions by other nations. Collective international action in order to solve this problem is therefore imperative.

Consider the following stylized representation of the problem, which may clarify the interaction between inter-generational and intragenerational issues. Let there be N countries, with countries indexed by the superscript i. Current national income, y, is assumed to be a increasing, strictly concave function of the level of GHG emissions in country i, _ this reflects the fact that control of greenhouse gas emissions is costly, and the marginal cost of emission control is increasing, the more we reduce emissions. National income may either be consumed or saved; in the latter case t is transferred to the future generation. U represents the utility of the current generation in country i. This depends both upon the current generation's consumption, and upon the utility of its descendants, Vi. V' is increasing in the transfer received by the tomorrow's generation, t and decreasing in the total level of global emissions, G.


This stylized model incorporates two significant features. The current generation in each country is 'partially altruistic' in two senses. First, it cares only about its own descendants, and not the future generation the world over, so that its utility depends only on Vi. Second, its concern about its own descendants could be limited, so that vi could affect Ui only weakly.

Each country maximizes its utility function with respect to the level of emissions, gi, and the level of transfers, tit This gives rise to the first-order conditions:

dU/dgi = Uy(.) fy() + UV(.)VG() = 0 [5]
dU/dti = -Uy) fy() + Uv(.)Vy(.) = 0 [6]

[5] and [6] must be satisfied at an optimum for each country. From [4] we know that the partial derivatives of V depend upon G. the total level of global emissions. Hence each country's optimal choice depends upon the sum of emissions of every other country. A non-cooperative equilibrium is a pair (gi,ti), i =1,2,...N, such that [5l and [6] are simultaneously satisfied for each country. It is instructive to re-write the first order conditions as [71

VG(.)/VY(.) = - fg(.) [7]

This has a ready intuitive explanation. Each country chooses gi to equate the marginal rate of substitution of its descendants between income and emissions, to the marginal cost of reducing emissions, in terms of income.

It is easy to see that such an equilibrium is inefficient, because emissions are a public bad. Each country's emissions decrease the welfare of future generations everywhere, but the country only takes into account the effect on its own descendants while making its choice. However, our primary concern is with the ethical aspect of this problem. What are the implications of this doubly partial altruism for intra- and intergenerational distribution?

Consider first the implications of the fact that each country cares only for its own descendants. Its concern is expressed through a transfer in two forms - a transfer of the private good, ti, and a transfer of the public good, the global environment. The former can be directed exclusively to its own descendants, while the latter necessarily accrues to the future generation the world over. Since the country is concerned only about its own descendants, the transfer of the private good is more effective for this purpose. Consequently, partial altruism results in an excessive level of emissions. Each country will over-exploit the environment, thereby adversely affecting future generations in other countries. It compensates (or in fact, overcompensates) its own descendants for this adverse affect by transferring the
private good. In effect, each country uses the global environment to alter the distribution of welfare within the future generation - it takes welfare away from the other countries and gives to its own descendants.

If all countries are symmetrically placed, the redistributive effect cancels out. The inefficiency still remains since the level of emissions is too high, being used for this redistributive purpose. The redistributive effects do not cancel out if countries are asymmetrically placed. To take a simple example, consider two countries which are otherwise identical, except for the fact that the first (the affluent country) generates more income from each unit of emissions than the second (the less affluent country). In equilibrium, the first country will have a higher level of emissions than the second, and will also transfer more income to its descendants. In effect, the present generation in the first country can effectively redistribute welfare within the future generation, from country two to country one.

Although we may care for the future generation, we may do so insufficiently, so that V may carry very little weight in our utility function. What are the implications of this sort of limited altruism? The first is that we may transfer too little to the future generation, both in terms of the private transfer and by degrading the global environment. An ethically adequate altruism will imply that we would have to increase both transfers. In other words, we would be called upon to reduce emissions of greenhouse gases to a level which is consistent with the future generation having a level of welfare to which they are morally entitled, rather than that which we deign to give them.


Contents - Previous - Next