How Harmful Are the Average American’s Greenhouse Gas Emissions?

It has sometimes been claimed (usually without evidence) that the harm caused by an individual’s participation in a greenhouse-gas-intensive economy is negligible. Using data from several contemporary sources, this paper attempts to estimate the harm done by an average American. This estimate is crude, and further refinements are surely needed. But the upshot is that the average American is responsible through her greenhouse gas emissions the suffering and/or deaths of one or two future people. How Harmful Are the Average American’s Greenhouse Gas Emissions? In discussions of global climate change it is sometimes assumed that the consequences of a single individual’s greenhouse gas emissions are negligibly small. 2 I am not, however, aware of any serious attempt to justify that assumption. 3 The goal of this paper is to critique it. More specifically, I aim to estimate the degree of harm to human beings done by an average American (resident of the U.S.) through her participation in a greenhouse-gas-intensive economy. Any such estimate will, of course, make some debatable assumptions. Attempts to calculate the same quantity in different ways would almost certainly produce different results. Nevertheless, I hope to make some progress toward a defensible figure—and, if I fail, at least to contribute to the identification and elucidation of the difficulties of such an attempt. My method is to refine the meaning of the phrase “the harm done by the average American’s greenhouse gas emissions” through a series of stipulative definitions until a rough quantitative estimate of that harm becomes possible. The Average American To estimate the harm done by the average American’s greenhouse gas emissions, we need to know what the phrase “the average American’s greenhouse gas emissions” means. There is, of course, no such thing as the average American. So the phrase “the average American’s greenhouse gas emissions” must mean something like “the total American greenhouse gas emissions divided by the U.S. population.” Such an average is, of course, the mean, as opposed to the median or the mode. The mean might be a misleading statistic if, for example, the greenhouse gas emissions of some small number of Americans were huge and those of the rest were relatively modest. In that case, the median or mode might be more revealing. But the individual emission data needed to calculate the median or mode are simply not available, for it is difficult to assign particular emissions to particular people. The greenhouse gas emissions for a given person presumably include not only those that occur directly as a result of her personal consumption of energy—the energy she expends, for example, in driving, or in heating and lighting her home—but also those emissions associated with the work she does for a living and, perhaps, those “embodied” in the goods and services she purchases. But what proportion of these emissions ought to count as hers? If she regularly travels by bus, for example, what proportion of the bus’ emissions should count as hers? Or if she tends to buy food from around the world instead of seeking out locally grown food, what proportion of the emissions produced by the growing, transportation and storage of that food should be counted toward her total? There have been a variety of efforts to measure individuals’ carbon footprints, but they all depend on debatable—and, to some extent, arbitrary—answers to such questions. The difficulties, both empirical and conceptual, of ascribing specific emissions to specific individuals are impressive. We do, however, have fairly good quantitative estimates of total greenhouse gas emissions for the nation as a whole. We can, therefore, calculate the mean simply by dividing such an estimate by the U.S. population. This may not be not exactly what we want, but it is as good a statistic for the emissions of the average American as we can readily get. It is, to be sure, less a measure of personal emissions than of the emissions in which an individual is complicit by participation in a fossil-fueled economy, but that is still a statistic of some interest. This mean is in one respect likely to be an underestimate. For the U.S. imports more than it exports, buying many of its manufactured goods from such heavy greenhouse gas emitters as China. The greenhouse gases released in the manufacture and transportation of these foreign goods will not count toward our emissions, if we define our emissions only as those of our nation. But this omission tends to yield an underestimate rather than an overestimate of our emissions. A conservative estimate is fine for my purposes, so I will not attempt to correct for it. To obtain the greenhouse gas emissions of the average American, then, we will simply take the total greenhouse gas emissions for the American nation and divide by the population. But to be clear on what we are doing, we first need to define “greenhouse gas emissions.” Greenhouse Gas Emissions Greenhouse gas emissions are emissions by humans of gases that contribute significantly to global climate change. The most important of these are carbon dioxide, methane, nitrous oxide, and various halocarbons. But not all anthropogenic emissions of these gases should count. We ought, in particular, to make an exception for the carbon dioxide we produce in breathing. The contribution of human respiration to total anthropogenic emissions—though perhaps larger than one might expect—is still relatively small: something on the order of three or four percent. 4 For obvious moral reasons, these emissions should not count. We may likewise ignore the fact that CO2 is not the only greenhouse gas emitted by the human body. Fortunately, our bodily methane emissions are comparatively miniscule. The power of a certain amount of a greenhouse gas to induce climate change is measured in mass units of CO2 equivalents. A kilogram of CO2 equivalent is the amount of additional atmospheric heat retention caused by the emission of a kilogram of CO2. Some greenhouse gases—methane, for example—are much more efficient at trapping heat in the atmosphere than is carbon dioxide. Thus it takes less than a twentieth of a kilogram of methane to produce one kilogram of CO2 equivalent. The total greenhouse gas emission, measured in CO2 equivalents, for the U.S. in 2008 (the most recent year for which data were available at the time of this writing) was, according to the EPA, about 6.96  10 9 metric tons. 5 This figure should be adjusted, however, for ongoing absorption of greenhouse gases by carbon sinks (e.g., photosynthetic plants). The EPA estimates that in 2008 carbon sinks in the U.S. removed about 9.4  10 8 metric tons CO2 equivalent from the atmosphere and thus sets the net U.S. carbon emissions at about 6.02  10 9 metric tons. 6 There were in 2008 approximately 3.04  10 8 Americans. Thus the net emission for the average American in 2008 was (6.02  10 9 )  (3.04  10 8 )  19.8 metric tons CO2 equivalent. The lifetime of the average American is about 80 years. Thus, at the 2008 rate, the lifetime emission of the average American is 80  19.8 = 1584 metric tons CO2 equivalent. The Harm Done by Greenhouse Gas Emissions How can we estimate the harms caused by our greenhouse gas emissions? We should note, first, that these emissions will not only harm many people; they will also benefit many. A warming climate can, for example, promote the growth of certain crops or make winters in cold climates mild. Shouldn’t such benefits be counted against the harms? They should if our aim is to calculate utilities. But my aim here is more limited: to estimate just the harms. My aim is limited in other ways as well. For one thing, among harms, I will consider only harms to human beings. Thus I will ignore not only potential benefits of greenhouse gas emissions for human beings but also both potential harms and benefits for nonhumans: animals and other organisms—and also such collective entities as species and ecosystems, if they can be harmed. Finally, among harms to humans, I will consider only the harms of manifest suffering (such as is caused by injury, disease, loss of livelihood, forced migration, etc.) and death. There are two reasons for these limitations—one practical, one moral. The practical reason is that without such limitations the problem is simply too hard. It is hard in part because harms to humans may be incommensurable with benefits to humans and with both harms and benefits to nonhumans. There is no commonly accepted measure by means of which we can aggregate, for example, harms to humans with harms to animals or species or ecosystems. 7 And even among human values, it is difficult to compare harms with benefits. There is no consensus, for example, on just how much benefit is worth the loss of a human life. By not trying to aggregate human harms with human benefits or with nonhuman values of any sort, we can avoid these difficulties. I am, in fact, not even going to try to aggregate the two sorts of harms to humans that I will consider: suffering and death. For there is no non-controversial way to rate them, too, on the same scale—to assign them both, say, definite monetary values or definite utilities. So I won’t attempt anything like this. If the harm of suffering and the harm of death are incommensurable, so be it. We may just tally them numerically—as in a casualty list from a battlefield: so many dead and wounded. By using only such raw statistics, I will avoid another issue that arises in attempts to estimate future harms: the problem of the discount rate. A discount rate is a function that revises utilities (both positive and negative) increasingly toward zero the further in the future they occur. I hold, as a matter of moral principle, that it is never legitimate to discount the utilities of human suffering or death. 8 But since I will be dealing merely with numbers of the suffering or dead, and not with utilities, my estimate of the harm will not rely on this principle. The casualty numbers will be what they will be. Whether they should be discounted or not is a separate question, and I will leave it aside. This crude conception of harm makes the question of how much harm the average American is doing reasonably tractable. That is the practical reason for my definition. But practicality is not the only desideratum. We also want moral significance. I think we can have that as well. Elsewhere I have argued that the suffering and death that we inflict upon future people by our greenhouse gas emissions constitute unjust domination of those people, and that this is true regardless of whether we also provide compensating benefits. 9 That is my moral reason for not considering the benefits of our emissions: given that we are inflicting unjust harm, compensatory benefits cannot nullify the injustice. Thus to assess the gravity of the injustices in which we are implicated, we need not estimate complete utilities (which include both harms and benefits). The quantity of human suffering and death that these emissions produce is in itself morally significant. The IPCC has provided a rough estimate of some of the harm to present and future humans attributable to past, present and near-future anthropogenic greenhouse gas emissions: “climate change over the next century is likely to adversely affect hundreds of millions of people through increased coastal flooding, reductions in water supplies, increased malnutrition and increased health impacts.” 10 Some of the secondary effects of global climate change seem not to be accounted for in this statement. These may include the violence likely to be spawned as refugees migrate inland from coastal areas, and people struggle for scarce food and water. This IPCC prediction extends, moreover, only through the next century. Yet current and near-term greenhouse gas emissions will, according to the IPCC, “continue to contribute to warming and sea-level rise for more than a millennium, due to the time scales required for the removal of this gas from the atmosphere.” 11 And, the IPCC further notes that “the net costs of impacts of increased warming are projected to increase over time” 12 and that “there is high confidence that neither adaptation nor mitigation alone can avoid all climate change impacts.” 13 Since the population affected by current emissions spans at least a millennium, it is likely to be large. Assuming that there are on average three generations alive at any given time, if the average world population over the next millennium is 7.5 billion (a figure that we will likely reach well before mid-century), then there will be on average 7.5  3 = 2.5 billion people per generation. A millennium is about 40 generations. Thus a reasonable estimate is that at least 2.5 billion  40 = 100 billion people will live during times affected by our greenhouse gas emissions. This estimate is rough and uncertain, of course; but, given the Earth’s limited carrying capacity, the true figure cannot be more than a few times that, and, barring some enormous catastrophe, it is unlikely to be many times less. Of these 100 billion, a certain number will die and a larger number will suffer as a result of the various effects of global climate change. How many will that be? We have seen that the IPCC estimates that hundreds of millions of people will be adversely affected (which I take to mean “will suffer or be killed”) over just the next century. But the detrimental effects of climate change will extend for a millennium or more—ten times longer. These effects may diminish toward the end of this millennium. Nevertheless it is reasonable to suppose that the number who suffer or die as a result of our greenhouse gas emissions will be in the billions. The Non-Identity Objection It might be objected that this estimate overlooks a fundamental conceptual problem: it is impossible for our greenhouse emissions to harm people in the far distant future. For if (contrary to fact) we were somehow to quickly and deeply reduce our emissions, then many of those who in the actual future will suffer or die as a result of those emissions would never be born. Such a quick and deep reduction would require large-scale social reorganization that eventually would result in different reproductive pairings; hence the people who would inhabit such a reduced-emission future would not be those who will inhabit our actual future. But (so the objection contends) to harm people is to make them worse off than they would otherwise have been. Thus since the actual distant future people who will be harmed by our emissions would not have existed without those emissions, the emissions do not make them worse off than they would otherwise have been—and hence do not harm them. This objection is, of course, an instance of what Derek Parfit has called the “non-identity problem.” 15 But, contrary to the objection, it is quite possible for people to be harmed by conditions that are necessary for their existence. The climate change brought about by our emissions will, as the IPCC contends, “adversely affect” many people who will actually live. (Indeed it is adversely affecting many who are alive already.) It is true that if (contrary to fact) we were rapidly and drastically to cut emissions, many of the people who will be harmed would never be born. Still, our emissions will in fact cause many of them to suffer and many to die. Suffering and death are harms. Hence many for whose existence these emissions are necessary will be harmed by them. Moreover, the fact that our emissions are necessary for these people’s existence does not relieve us of responsibility for the harm. During the 1950s the drug thalidomide was widely prescribed, often for morning sickness, but also as a tranquilizer. Suppose that, as a result of the drug’s tranquilizing effects, a woman yielded to the sexual promptings of a man and became pregnant. Suppose further that since, notoriously, thalidomide also causes severe birth defects, the resulting child was deformed by the drug. Then the child was harmed by the drug, even though without it that child would never have existed. Moreover, the fact that the child would not have existed without the drug would not relieve its manufacturer (and, perhaps, other parties) of causal, moral and legal responsibility for harming the child. The case of our greenhouse gas emissions is analogous: though it is true that our emissions are necessary for the eventual existence of some of the people who will be harmed by them, that does not relieve us of causal or moral responsibility for harming those people. The Harm Done by the Average American’s Greenhouse Gas Emissions We are now ready to estimate the harm done by the average American’s greenhouse gas emissions. There is, as I have noted, no such thing as the average American. The phrase “the average American’s greenhouse gas emissions” is an abbreviation of the phrase “the total greenhouse gas emissions of Americans divided by the American population.” Let us likewise understand the phrase “the harm done by the average American’s greenhouse gas emissions” as shorthand for “the harm done by the total greenhouse gas emissions of Americans divided by the American population.” In 2008 U.S. emissions were about 19% of the world’s. 16 There were about 304 million Americans. Thus the average American in 2008 produced about 0.19 ÷ (3.04 × 10 8 ) = 6.25 × 10 -10 of global emissions. The total harm of global climate change results, however, not only from one year’s emissions, but from the total anthropogenic increase in atmospheric greenhouse gas concentrations that began with the industrial revolution. An individual can have a part only in those greenhouse gas emissions that occur during her lifetime—not those that occur before she was born. But she does, presumably, have a part in emissions that occur during her lifetime. 17 We assumed above that the lifetime of the average American is 80 years. For the sake of definiteness, let’s say that this life begins in 1960 and ends in 2040. From these dates we can calculate the increase in the atmospheric concentration of carbon dioxide during the average American’s lifetime. 18 The concentration when she was born in 1960 was 317 ppm. Today (July 2010) it is 390 ppm. 19 Projecting the current growth rate of about 2 ppm/yr into the near future we obtain an estimate of 450 ppm for 2040. 20 Thus total increase during her lifetime comes to 450-317 = 133 ppm. The pre-industrial concentration of CO2 was about 280 ppm. 21 So before this average American’s birth, the atmospheric CO2 concentration had increased 37 ppm. During her lifetime it will increase an additional 133 ppm. The total anthropogenic increase until the time of her death is therefore is 37 + 133 = 170 ppm, of which about 78% will have occurred during her lifetime. We noted earlier that the average American produces about 6.25 × 10 -10 of current (2008) annual global emissions. Let’s assume that her portion of the annual emissions stays roughly the same over her lifetime. 22 We have just seen that 78% of the total increase in atmospheric greenhouse gases to 2040 will have occurred during her lifetime. Thus her portion is about (6.25 × 10 -10 ) × 0.78 = 4.9 × 10 -10 —that is, about one twobillionth of the greenhouse gases contributing to anthropogenic climate change by 2040. To estimate the total harm done by the average American’s greenhouse gas emissions, we need, then, to take the total harm that will be done by global emissions to 2040 and divide that by two billion. We noted above that even if emissions are reduced to low levels fairly quickly—that is, even under the most optimistic of scenarios— billions of people are likely ultimately be harmed by them. If over the next millennium two billion people (perhaps two percent) are harmed (that is, suffer and/or die) as a result of current and near-term global emissions, then the average American causes through her greenhouse gas emissions the suffering and/or death of one future person. If the total is four billion, she is responsible for the suffering and/or deaths of two future people. These estimates, of course, are very crude. But they give us, I contend, some sense of the moral significance of our participation in a greenhouse-gas-intensive economy.