[T]he existing estimates of [the] social cost [of climate change] are based on IPCC studies that so far have not included many . . . irreversible positive feedbacks. . . . So nobody has yet even asked what price should be attached to a century-long drought in the American West, or an enfeebled Asian monsoon, or a permanent El Niño in the Pacific, or a methane belch from the ocean depths, or a collapse of the West Antarctic ice sheet, or sea levels rising by half a yard in a decade. Though, on reflection, these are perhaps questions not best answered by accountants.1
1. Introduction: Catastrophes and Tipping Points
The basic science of climate change (CC)—our understanding of the way in which greenhouse gases (GHGs) warm the planet—is well established and understood. However, experts still have limited understanding of the nature and significance of various positive and negative feedback effects that could respectively accelerate or slow CC, and there is much disagreement among them about the accuracy and reliability of the models positing these effects, and/or the existence of data sets adequate to support their prediction. In particular, the possibility of powerful positive feedbacks is increasingly being taken by many experts to undermine the “gradualist paradigm” in thinking about CC, most notably as adopted by the Intergovernmental Panel on Climate Change (IPCC).2 According to this paradigm, the changes constitutive of CC are cumulative, smooth, and incremental, which makes them easier to predict and plan for. However, challenges to this paradigm have it that CC might happen through a combination of gradual changes and some very abrupt shifts, as we pass various “tipping points.” These events are points of no return, beyond which positive feedbacks causing runaway CC create a world not represented in any IPCC scenarios, and beyond the scope of much scientific imagination to date. The possibility of passing tipping points on the way to CC catastrophe (CCC) has led some normally cautious scientists to adopt the language of Armageddon in their attempts to get these possibilities into the public debate about CC; for example, NASA scientist James Hansen describes the current state of affairs as “a planetary emergency.”3
There are numerous significant tipping points, and new ones are emerging all the time.4 To provide focus, I shall limit discussion to a set of frightening positive feedbacks that scientists have awakened to only recently. Fifty-five million years ago the Earth experienced a period of extreme and accelerated warming called the Paleocene-Eocene thermal maximum (PETM), which triggered the largest mass extinction event since the “Great Dying” 251 million years ago (in which ninety-five percent of species went extinct). Many scientists believe that the PETM was largely caused by the release of trillions of tons of methane from beneath the oceans. This methane had been trapped in frozen, honeycombed sediments called “clathrates,” and was released as a result of global warming (probably caused by solar activity). At present, one to ten trillion tons of methane clathrates exist beneath the world's oceans, and could be destabilized by the rapidly warming oceans. The release of even a fraction of this methane could be catastrophic in at least two ways. First, it could cause massive underwater landslides that would create tsunamis comparable in size with, or larger than, the one that started in the Indian Ocean on December 26, 2004. Indeed, scientists now believe that the “Storegga slip” eight thousand years ago—which occurred off the coast of Norway and caused a tsunami of forty feet in Norway, twenty feet in Scotland, and sixty feet in the Shetlands—was caused by the explosive release of methane from a destabilized clathrate. Second, the release of methane on this scale could cause runaway CC from which it would be impossible to recover, given existing levels of anthropogenic warming;5 it is not science fiction to claim that in these conditions temperature increases could exceed 6°C (see section 4 for more detail). In this scenario the majority of life on Earth, perhaps including homo sapiens, could go extinct.6 Call this scenario “Methane Nightmare.”
I shall argue that a precautionary approach ought to be adopted by policy-makers addressing CC in virtue of the possibility that CC could cause us to pass key tipping points beyond which positive feedbacks are activated that could cause runaway and catastrophic CC. I shall present arguments for the Precautionary Principle (PP) as the approach to CCCs that should be adopted by policy-makers concerned (as they should be) with intergenerational justice, and I shall understand intergenerational justice in broadly Rawlsian egalitarian terms. What I mean by this is that it is appropriate to adopt Rawls's conceit of the original position when thinking about intergenerational justice. The impartiality modeled by this method for choosing principles of justice is required not only between members of the same generation, but also across members of different generations. Rawls incorporates this commitment into his hypothetical social contract by stipulating that intergenerational justice is to be governed by his “Just Savings” principle.7 Parties behind the veil of ignorance are to be thought of as contemporaries: they belong to the same generation (and know this), but they do not know which generation this is. He calls this the “present time of entry” interpretation of the original position.8 In this situation, parties would adopt a “maximin” rule to govern their choices between different sets of principles and, for Rawls, this holds with respect to intergenerational as well as intragenerational justice.9
Of course, there are few, if any, policy-makers whose decisions conform to, or whose reasoning reflects, Rawls's conception of intergenerational justice. But even if the non-ideal world of non-compliance with principles of justice endures forever, it still matters that we recognize the ways in which the world is unjust and could be made better. The defense of the PP offered herein is a (small) part of that endeavor as applied to the worst effects of CC on human social and political organization, with particular reference to its intergenerational aspect. In this paper, I shall distinguish between two Rawlsian arguments for taking precautionary action against the worst outcomes of CC. I shall show that although both arguments provide compelling grounds for taking such action, the argument making reference to the “strains of commitment” is more powerful than the argument making reference to the unjust distribution of advantage across generations that could be caused by failing to take precautions against CCCs.
2. The Precautionary Principle
One feature of CCCs that makes policy making with respect to them so difficult is that they are uncertain. In one basic sense, the meaning of “uncertainty” is clear: that which is uncertain is not certain. Given that nothing in the future is certain, and given that all policy is future-oriented, uncertainty (in this basic sense) is integral to the context of every policy decision. Uncertainty in this sense does not pose a special problem for policy-makers because it is consistent with a future event being uncertain that it is assigned a probability of happening (indeed, probability is a way of expressing uncertainty), and once a policy-maker has this information she can perform a standard risk assessment to guide her thinking about the event by multiplying its probability by its costs or impact.
This is not the sense in which CCCs are uncertain. Rather, although CCCs are known to be possible (often because they have happened in the past), the processes that cause them are so poorly understood that it is not possible to assign a precise numerical probability to them, and thus a risk assessment cannot be made.10 Clive Spash's typology of uncertainty as “weak” and “strong” is helpful here.11 Weak uncertainty exists when outcomes and their probabilities (understood either as objective features of the world, or as subjective insofar as derived from persons' preferences) are known. The context for much policy making is weak uncertainty, and the dominant approach to policy making in this context is risk assessment. Strong uncertainty exists when outcomes are unknown and/or unpredictable, either because the outcomes are indeterminate, or because the state of knowledge with respect to these outcomes precludes the assignment of a probability to them.
The strong uncertainty of CCCs renders them an elephant in the room in CC policy debates. Of course, some GHG reduction targets—for example, the UK's commitment to cut CO2 by sixty percent by 2050—can be read as addressed to CCCs just in virtue of the fact that any reduction in GHGs can be seen as an attempt to mitigate CC, CCCs included. But such policies are certainly not justified by direct reference to mitigating CCCs; and a cut of sixty percent (let alone the paltry Kyoto targets) is unlikely to do the job. Furthermore, as already noted, the bibles of policy-makers addressing CC—the IPCC Reports—do not include CCCs in their scenario analyses, and work largely within the gradualist paradigm.
The unavailability of risk assessment as a method for policy making with respect to strongly uncertain CCCs has led an increasing number of thinkers (even those generally sympathetic to cost–benefit analysis, of which risk assessment is an example)12 to give serious consideration to the PP as a guide to policy making for such events.13
The PP is sometimes presented as equivalent to, or derived from, the principle “better safe than sorry.” So presented, it is hard to refute: Who would rather be sorry than safe? However, these presentations are inaccurate to the PP as it appears in policy documents.14 Rather, the PP removes constraints on reasons for action in policy making where scientific expertise is indispensable: It states that strong uncertainty about harm need, or must, not stand as a reason for inaction with respect to policy making that would adequately protect people from these harms. The principle has a weak and a strong formulation.
The weak PP: When evidence or information is insufficient to establish the nature and/or probability of harms caused by an activity, policy-makers are permitted to act in order adequately to protect people and other entities from these possible harms.
The strong PP: When evidence or information is insufficient to establish the nature and/or probability of harms caused by an activity, policy-makers are required to act in order adequately to protect people and other entities from these possible harms.
The weak PP (WPP) is a permissive principle: It states the reasons on which it is acceptable for policy-makers to act. The WPP permits policy-makers to take preventive action for reasons other than certainty that the action is necessary to prevent a harm. The strong PP (SPP) is a categorical principle. It states the reasons upon which policy-makers must act; that is, in order to protect people and other entities from possible harms, even when the nature and/or probability of these harms is not known. Given this uncertainty, reasons given in justification of action taken in the name of the SPP must appeal to considerations other than those of evidence- or model-based support for the belief that the action is necessary to prevent a harm. It is relatively rare to find examples of the SPP in policy literature; far more common are varieties of the WPP.15
The WPP is impotent as a constraint on the decisions of policy-makers because it is satisfied even if decision making on every policy issue in its scope is postponed until the relevant evidence and information is certain; that is, even if no policy-maker ever actually takes precautionary measures as set out in the WPP. When one does not do P despite having permission to do P, one acts consistently with that permission. Including the WPP in policy documents need have no impact at all on any commitments appearing elsewhere in them, or in any other documents.16
For this reason I shall focus instead on the SPP. This is a steely principle: It compels policy-makers to act in the face of strong uncertainty even though their action may in fact turn out to be an unnecessary precaution against a non-existent, or vanishingly unlikely, harm. Because such action is almost always expensive(in literal sterling terms), and costly (in terms of the intrusiveness, disruptiveness, and unpopularity of legislation to satisfy the principle, and its impact on other policy commitments), the SPP is generally dismissed as an unworkable and naïve proposal made by those in the grip of an ideology or suffering from distorted perceptions of risk.
One way of arguing for the SPP with respect to CCCs is with the claim that the SPP ought to be adopted by policy-makers in general when framing legislation for processes that create uncertain risks of harm. It is clear that this sweeping claim is a non-starter: As Sunstein puts it, “[t]he regulation that the principle requires always gives rise to risks of its own . . . hence the principle bans what it simultaneously mandates.”17 Applied at this level of generality, the SPP is literally incoherent: each course of action it requires of policy-makers it simultaneously prohibits.18
Although the SPP is indefensible as a general principle for policy-makers, it is nevertheless defensible with respect to CCCs, against the background of a commitment to deliver justice to future generations. In the next section I lay out a Rawlsian argument that makes this case, and I comment on its limitations. In section 4 I develop a second (and new) Rawlsian argument which is not subject to these limitations, and which delivers a more potent justification of the SPP with respect to CCCs.
3. The Playing Safe Argument
Famously, Rawls offers an interpretation of equality as maximin: to treat people as equals means (among other things) to ensure a distribution of (dis)advantage among them that makes the worst-off group as well off as possible. In addition, Rawls conceives of justice as intergenerational in scope, governing relations across generations as well as within them. These commitments can be made to work to justify the SPP with respect to CCCs as follows.
The SPP is justified with respect to CCCs because the worst consequences of not taking precautionary action are worse than the worst consequences of taking precautionary action, and choosing the former course of action is not consistent with treating present and future people as equals when we cannot assign a probability to each outcome, that is, when we are strongly uncertain of each outcome, as is the case with respect to CCCs. We can see this by adopting the Rawlsian conceit of the original position, and by imagining what parties choosing principles therein would say in justification of their choices to those whose interests they represent. What such persons could not do is to take a bet on the probability of the worst consequences of no precautions being lower than the probability of the worst consequences of precautions, given the nature of these consequences. Call this the “playing safe” argument; in order to make it in full I need to specify in more detail the nature of the consequences in question.19
The worst consequences of taking precautions against CCCs are that the costs of adapting to CC turn out to be lower than the costs of taking precautions. The consequences of CC may turn out to be nowhere near as dire as experts' worst fears, and it may be that tipping points and their CCCs exist only in nightmares and Hollywood movies, or are extremely distant in time, giving us more opportunity for cheaper precautionary action at a later date. It may be that the Earth is far more insensitive to our activities than we think. In this scenario, we spend a lot of money and time—with all the associated opportunity costs this creates—in taking precautions against consequences now that are in fact not necessary. Estimates of how much it would cost to implement adequate precautions vary widely; for example, Bjorn Lomborg puts the figure at $37.632 trillion, whereas the European Commission (working with a generous stabilization target of 550 ppm to be met by 2100) chooses $1–8 trillion,20 and the Stern Review estimates the cost at one percent of the world GDP in 2050, or $1 trillion.21 These figures are hotly disputed, but to make the case against precautions as powerful as possible I shall adopt the Lomborg estimate. Call this outcome “Unnecessary Expenditure.”
The worst outcome of not taking precautions is that the worst predictions of the IPCC turn out to be very conservative,22 that the gradualist paradigm is false, and that we are on the verge of various tipping points, with everything this entails. The worst outcome of a decision not to take precautionary measures with respect to CC is catastrophic ala Methane Nightmare.
With these outcomes identified, the playing safe argument from maximin for the SPP with respect to CC is prima facie powerful: Because Methane Nightmare is so much worse than Unnecessary Expenditure, we ought not to act as if the worst case scenario of taking precautions is more probable than this worst case scenario of not taking precautions, and thus we ought to take precautions against CCCs. Acting as if the worst consequences of taking precautions are more probable than the worst consequences of not taking precautions is ruled out because the harms that would be caused to present and future people by policy made on this assumption, if it turns out to be false, are much greater than the harms that would be caused by policy made on the first assumption, if this turns out to be false. And treating people as equals makes it impermissible to gamble with their interests in this way.
Stephen Gardiner's elaboration (following Rawls) of the conditions under which Rawlsian maximin reasoning delivers the requirement to take precautions is helpful.23 These are when (i) decision makers are in a state of strong uncertainty with respect to the probability of the events in question; (ii) decision makers are indifferent to gains above the minimum that a maximin strategy would guarantee; and (iii) the alternatives to a decision guided by maximin are unacceptable.24 We have seen that (i) is true of CCCs (I shall consider objections to this claim in section 5). Conditions (ii) and (iii) are questioned when it is claimed that it is rational or reasonable to gamble on the probability of Unnecessary Expenditure being higher than Methane Nightmare—and so not to take precautions—because if this gamble pays off the gains are so much higher than if the bet is not placed at all, or is placed on the probability of Methane Nightmare being higher than Unnecessary Expenditure.25 Call this the “skeptical stance.”
There are two comments on the skeptical stance. First, it does not take seriously the possibility that even if the bet at its heart were to turn out to be good, we may still have reason not to make it in virtue of how taking precautions against CCCs could have all sorts of other benefits that would make this choice attractive even if any CCC turns out to be highly improbable. This is the so-called “no regrets” strategy: the requirement to take precautionary measures against CCCs could stimulate technological innovation, efficiency measures, and new forms of political interaction that could greatly enhance the quality of our lives even if the precautions turn out to have been redundant with respect to CCCs. In other words, the opportunity costs created by not taking precautions might make it always less costly to take precautions, even taking into account the possibility that the probability of any CCC is tiny. Although the “no regrets” strategy is not part of Rawlsian maximin and so, strictly speaking, not part of the playing safe argument, it is certainly not inconsistent with it; indeed, it could be proposed as an interpretation of condition (iii) above: not taking precautions is an unacceptable alternative not only because of the possibility of CCCs, but also because of the possible gains associated with precautionary measures. Of course, whether taking precautionary action against CC is a no regrets strategy is highly controversial. Hence, if there is an alternative, or complementary, way to undermine the skeptical stance, then all to the better.
This brings me to my second comment, which relates to the type of justice to which the playing safe argument as it stands appeals. The argument is that CC policy-makers concerned to treat persons (always remembering that this includes future generations) as equals should be guided by the maximin rule to take precautions against CCCs, because Methane Nightmare is much worse than Unnecessary Expenditure. In what ways, exactly, is Methane Nightmare worse than Unnecessary Expenditure? Clearly, there are all kinds of losses—in biodiversity, to ecosystems, of habitats—that would occur under Methane Nightmare but not under Unnecessary Expenditure. Let me put these to one side in order to focus on a harm people in Methane Nightmare would suffer that people in Unnecessary Expenditure would not. In Unnecessary Expenditure we waste resources that could have been used to improve the position of the worst off in the current generation or, indeed, that could have been saved for the benefit of the worst off in future generations. These are considerations of distributive justice, and in line with the impartiality at the heart of the Rawlsian approach to justice, duties of distributive justice are intergenerational in scope.26 If we have a duty to ensure that the worst off—understood in intergenerational terms—are as well off as possible, and if Unnecessary Expenditure conflicts with this duty, then we have a reason to reject courses of action leading to it, and should avoid it, all else being equal.
The playing safe argument as presented so far can be read as claiming that all else is not equal: the distributive injustice that would be created by Methane Nightmare is worse than that created by Unnecessary Expenditure, and we must choose between courses of action to which Unnecessary Expenditure and Methane Nightmare attach as worst outcomes, in which case we should aim to avoid the worst worse outcome—that is, Methane Nightmare—by taking precautions. Any disadvantage that accrues to the worst off as a result of this choice is justifiable to them in the name of equality: Given the badness of Methane Nightmare, choosing not to take precautions—given our uncertainty—would be to gamble with their interests by betting on the probability of Methane Nightmare being lower than that of Unnecessary Expenditure.
However, as we have seen, making the argument in terms of distributive justice leaves it vulnerable to the skeptical stance, as follows. This version of the argument turns on the requirement to ensure that the worst-off generation is as well off as possible in terms of tangible goods, which includes generations in the future who may suffer huge disadvantage if we do not take precautions against CCCs. It can then be objected that tangible goods can be priced, and that the cost of taking precautions now will make these goods unaffordable for future generations. People who take the skeptical stance characteristically specify why it is rational or reasonable to gamble on the probability of Unnecessary Expenditure being higher than Methane Nightmare (and so not to take precautions) in terms of monetary gains and losses: crudely, we will be richer, and make future generations richer, if we do not take precautions. Of course, most people who take this tack also insist that monetary gains and losses matter because we can do things with money, such as improving human health.27 However, as has been observed frequently, there are things we value (or ought to value) which cannot be compared in value to other valuable things, and so are not amenable to being priced. Consequently, such incommensurably valuable things do not appear in the calculations that inform rejection of precautionary approaches such as this.28 One of these things is intergenerational justice.29
The playing safe argument as made so far is limited to a conception of intergenerational justice in distributive terms. However, there is another way to make the playing safe argument that articulates the impermissible gamble at its heart in terms of a more fundamental vision of what intergenerational justice requires, and evades the skeptical stance. I make that argument in the next section.
4. The Unbearable Strains of Commitment
There is something bad about Methane Nightmare that is related to justice, but is not well captured in terms of distributive justice, and which is absent in Unnecessary Expenditure. Highlighting this feature of Methane Nightmare makes available to us an extra reason for taking precautions against CC that justifies this choice not just in virtue of an egalitarian commitment not to gamble with the interests of the worst off in accumulating as many tangible goods as possible, but also in virtue of a commitment to making it possible for future people to commit themselves to justice at all. This additional badness turns on the fact that Methane Nightmare would create unbearable strains of commitment for those living through and after it: it would make unreasonable any mutual expectations among future generations that they should propose and abide by fair principles of justice.30 If this would be true of future generations in Methane Nightmare, then we—the current generation—cannot adopt principles for action on CC that could have this outcome.
The strains of commitment are related to the idea of making an agreement in good faith; in Rawls's words, “not only with the full intention to honour it but also with a reasonable conviction that one will be able to do so.”31 Rawls claims that excessive strains of commitment cause people to withhold affirmation of principles of justice in two ways. First, people “become sullen and resentful . . . ready as the occasion arises to take violent action in protest against [their] condition”; and second, people “grow distant from political society . . . withdrawn and cynical [they] cannot affirm the principles of justice in . . . thought and conduct over a complete life.”32
Any attempt to implement principles of justice in Methane Nightmare—and many other CCCs—would probably have these effects on people. An indication of the temperature rises that might be experienced in Methane Nightmare is available by looking back at the PETM, to which it is thought that the collapse of methane clathrates greatly contributed.33 In this period, average temperatures increased by 5–10°C.34 One of the few models that address the effect of a global temperature rises of 5°C predicts severe desertification in already arid regions35—decimating most of the world's breadbaskets—combined with huge increases in precipitation at higher latitudes, making floods and storms more frequent.36 The net result of a 5°C increase would be worldwide famine, severe conflict over water in the world's desertified regions, and extreme territorial insecurity for those living in the dwindling inhabitable areas as an archipelago of refuges is put under increasing pressure by hungry and desperate people.37
At a 6°C increase, however, things likely become very much worse. Mark Lynas suggests that the best reference point for this amount of warming is not the PETM, but rather the end of the Permian era 251 million years ago, when the “Great Dying” extinction event, which wiped out ninety-five percent of the planet's species, occurred as a result of a temperature increase of 6°C.38 At the Permian-Triassic boundary, the oceans heated drastically and became devoid of oxygen, and so unable to support life, and also spawned superhurricanes which distributed heat to the poles (a further positive feedback).39 Destabilized methane clathrates beneath the oceans released huge quantities of CH4 which, if ignited, would have generated explosive, lethal blast waves traveling at two kilometers per second, and killing everything in their path.40 The rotting carcasses of animals, and decaying vegetation, in the oceans released large amounts of hydrogen sulphide, obliterating any remaining life there,41 killing most remaining land animals when released into the atmosphere,42 and destroying the ozone layer. For temperature rises beyond 6°C, predicting the effects is guesswork, but looking to conditions on Venus gives some indication. It is crucial to note that the temperature increases that caused the Great Dying probably took ten thousand years to effect, whereas—even on the conservative estimates of the IPCC—we could achieve such warming in one hundred years. As Lynas puts it, “[i]f we had wanted to destroy as much of life on Earth as possible, there would have been no better way of doing it than to dig up and burn as much fossil hydrocarbon as we possibly could.”43
The extreme scarcity of resources and almost unimaginable conditions in Methane Nightmare would make the joint pursuit of justice impossible.44 When there is not enough for each to have even the bare minimum for survival, to ask anyone to do anything other than pursue their self-preservation and perhaps that of their family—such as to act according to the demands of justice—is to ask them to accept death so that others can live. In normal circumstances, the sacrifices in self-interest required by justice are not unreasonable and, all else being equal, do not impose unbearable strains of commitment on those on whom they fall. But in circumstances as desperate as Methane Nightmare, where self-preservation dominates all other motivations, any justice-based request for self-sacrifice will at least endanger a person's chances of survival, if not actually require their death. What makes this request an unbearable strain of commitment is not just the content of the request in itself; rather, it is that this request be accepted as a requirement of justice, that is, as a measure against which those upon whom it falls can make no justified complaint.45 If future generations know that precautions could have been taken by their ancestors that would have averted Methane Nightmare, and would have made the request unnecessary, then sullenness, resentment, alienation from principles from which the request is derived, and violent protest against the request, are the least we can expect from them.46
If social cooperation is possible in such circumstances, it is not according to principles of justice. In contrast, the badness of Unnecessary Expenditure is not related to any unbearable strains of commitment it would create.47 The putative reason for avoiding it is not that it destroys the possibility of justice, but that avoiding unnecessary expenditure allows for a more just distribution of goods by making more goods available for redistribution and/or saving now. That future generations are made less well off than they could have been because the current generation takes precautions against the CCCs that its activities could inflict on future generations is not obviously unjust (indeed, I have claimed that there are good reasons of justice for taking precisely this course of action). And that members of future generations know that they could have been made better off had the current generation not taken precautions does not create a strain of commitment for them that makes their joint pursuit of justice impossible. That future generations have fewer goods than they would have had had precautions not been taken is not in itself a reason for them to reject any requirements made of them by principles of justice in such reduced circumstances, especially given their knowledge that the decision of the current generation to save less in order to take precautions was made in conditions of strong uncertainty, and in the name of delivering justice to their descendants.
If not taking precautions against CCCs could impose conditions on future generations that would make the joint pursuit of justice by them impossible, then we—the current generation—cannot decide not to take such precautions. When thinking through our decisions from the point of view of justice we are required to adopt the standpoint of impartiality, which in part requires that we should not choose principles of social cooperation that could impose on future generations conditions which we ourselves could not agree to. To choose not to take precautions against CCCs such as Methane Nightmare violates this fundamental requirement of impartiality.48
5. Some Problems
The SPP is a highly controversial principle, and the science of tipping points and CCCs is young: bringing the two together delivers an approach open to many objections. I shall consider just two of these here. The first questions the fitness of the SPP as a principle for policy-makers in conditions of strong uncertainty, and the second addresses the scope of the SPP.
The first objection focuses on the SPP's requirement that precautions must provide adequate protection against uncertain harms.49 With respect to specifying what counts as adequate protection, there are two basic options:
- (a) that we devote all of our resources and efforts to taking precautions; and
- (b) that we devote less than all of our resources and efforts to taking precautions.
Prima facie, (a) is implausibly demanding because it would leave no resources for the pursuit of other important social goals. However, (b) is wildly vague, and gives no useful guidance to policy-makers. But being more precise about exactly how many, and what, resources are sufficient for adequate protection against a harm requires that we specify the probability of the harm: without this specification, the notion of what is adequate loses its anchor, and we must rely on guesswork. This fact about specifying what is adequate creates the following problem. Either we can specify the probability of harmful events to which the SPP applies, or we cannot. If we cannot, then the SPP provides no guidance with respect to making policy for those events. Given that we cannot specify the probability of CCCs, the SPP provides no guidance to policy-makers concerned with this aspect of CC. There is a way of making the same objection that comes at it from the other end, so to speak, as follows. We must know that any given CCC putatively governed by the SPP is not bound to happen—that is, that its probability is not one hundred percent—otherwise, we would not bother to think about taking precautions against it (they would be futile, given our certainty that it is inevitable) and would perhaps just throw a big party instead.50 In that case, we are not in a state of strong uncertainty with respect to the CCCs I have claimed fall within the scope of the SPP. In sum, the problem can be posed as a dilemma: Either our genuine strong uncertainty about CCCs makes the SPP inappropriate as a principle for policy-makers addressing them, or our application of the SPP to CCCs shows that we are not in a state of strong uncertainty about them.
In response, I think it is possible to specify two limiting cases to our strong uncertainty about CCCs that do not undermine the strength and nature of this uncertainty.51 These are that we know the probability of CCCs is not zero, and is not 100. That is, we know CCCs are not certain not to happen, and are not certain to happen, which means that we know that it is possible that precautions could make a difference, and that enacting them is not necessarily futile. Nevertheless, we remain strongly uncertain of the probability of these catastrophes within the range of anything more than zero, and anything less than 100. And given the catastrophic nature of the events, any probability in this range justifies precautions. This does not, however, mean that we are necessarily committed to devoting all our resources and efforts to precautions. If the SPP is, sensibly, qualified with “all else being equal,” then it is possible to claim that all else is not, in fact, equal because there are other important and urgent demands of justice that require resources to be achieved. In the face of strong uncertainty, it is these demands that set constraints on what counts as adequate for protection, and not any further, more precise, specification of the probability of various CCCs. The question of what other demands of justice there are that are sufficiently important and urgent to play this role is, however, properly a political question, to be debated by members of the demos, and to be decided using appropriate procedures for democratic decision making. Although experts can provide valuable information necessary to preserve the quality of such decision making, the decisions themselves are ultimately up to us.52
This relates to the second objection. There are various future events that are catastrophic to the same degree, uncertain in the same way, and would create similarly unbearable strains of commitment for future generations, as CCCs, but which are not related to CC and so would remain untouched by any precautionary policy adopted to mitigate CCCs; for example, comet collision. The Chicxulub crater in Mexico is the remnant of a ten-kilometer-wide asteroid that struck sixty-five million years ago, and is widely believed to have caused the extinction of the dinosaurs.53 Given that such a collision would no doubt do the same to us, should we be developing missile programs to defend ourselves against this possible future harm? If so, what about supervolcanic eruptions? Or invasions of malevolent aliens? Why should we devote our attention and limited resources to taking precautions against CCCs as opposed to other possible global catastrophes? Call this the “priorities” problem.54
The priorities problem can be addressed by reflection on the following point. First, the CCCs to which the SPP applies are anthropogenically caused, and this distinguishes them from comet collisions, supervolcanic explosions, and so on. The fact that we are doing things that could be (quickly) moving us closer to various tipping points gives us a reason for taking precautions against its catastrophic effects that we do not have with respect to other non-anthropogenically caused catastrophes: That we are, or could be, causing the problem prima facie renders us liable for finding the solution,55 and gives us a reason to prioritize finding a solution to that problem over others we are not, or could not be, causing (such as comet collision).
The existence of this reason in part alleviates the priorities problem, but possibly varies in its potency across two cases marking either end of a spectrum. In the first case our status as the cause of possible CCCs is bad option luck for us, the current generation: In the full knowledge that increasing GHG emissions could move us closer to tipping points leading to CCCs, we are taking our chances and increasing our GHG emissions anyway. The correct analogy here is with the drunken driver who knowingly imbibes and then knocks down and kills an innocent child: The driver is fully morally responsible, and legally liable, for the death. Just as the driver either ought not to drink, or if it is too late and she is drunk, ought to take a cab, so we ought to enact precautions to protect future generations (the exact analogy for our circumstances vis-à-vis GHG emissions is probably the case of taking a cab). Note that if the problem of CC were not intergenerational in nature the drunk driver analogy would be incorrect. Instead, it would be open to us to claim that knowingly and intentionally increasing our GHG emissions creates risks of harm for no-one other than us and that—invoking something like Mill's harm principle56—we are thus under no obligation to abstain from, or mitigate the effects of, our risky activities. An accurate analogy would instead be of someone with no dependents or other special obligations to others who chooses to take heroin knowing the risks. But brute facts about the atmospheric life of CO2, and the speed at which the CC it causes occurs, mean that vulnerable future generations cannot be written out of the picture: They stand to us as the innocent child stands to the drunken driver.
The second case, lying at the other end of the spectrum, makes less headway with respect to the priorities problem. This is when being the cause of a (possible) harm carries no implications for blameworthiness and is nothing but a piece of bad brute luck for the agent. The analogy here is of a sober and careful driver who kills an innocent child through no fault of her own. That we think the agent's causal efficacy has some moral significance in cases such as these is shown by the fact that we would judge any such driver who experienced no regret at having killed a child—who simply shrugged it off as “one of those things”—as seriously morally deficient.57 However, explaining the significance of the agent's causal efficacy for moral responsibility, and (perhaps) legal liability, in such cases is far from straightforward.58 If the current generation stands to future generations as the blameless driver stands to the innocent child she kills, then it is less clear what difference it makes to the priorities problem that the CCCs under consideration are anthropogenically caused in contrast to other catastrophes which are not.
However, in my view, and in the view of most experts, we are in fact closer to the “bad option luck” than “bad brute luck” end of the spectrum—if not, in fact, at it—in which case the priorities problem is more tractable than at first blush. But note that even if we were closer to—if not at—the “bad brute luck” end of the spectrum, the priorities problem does not stymie the approach proposed in this paper, once we attend to what the problem actually is. The additional work that the priorities problem asks for is required to justify not extending the SPP to non-anthropogenically caused catastrophes. This lacuna in the account of the overall scope of the SPP does not affect the argument that it extends at least as far as CCCs. Thus, we do not need to wait for these further refinements before starting to think seriously about how to formulate and implement precautionary policies for CCCs.
In conclusion, I have distinguished between two Rawlsian arguments for the SPP with respect to CCCs. Although both are persuasive, ultimately the “unbearable strains” argument provides the most powerful categorical grounds for taking precautionary action against CCCs. Overall, I have argued that the nature of CCCs requires us to take drastic precautions against further CC that could lead us to pass the tipping points that cause them. This is the case notwithstanding the fact that we are in a state of strong uncertainty with respect to these events; indeed, our strong uncertainty with respect to them—given their nature—makes the case for action to prevent them even more persuasive, from the point of view of justice. Some people translate the strong uncertainty of CCCs into weak uncertainty in order to justify taking precautionary action using risk assessment.59 My argument is complementary to theirs. If divergent approaches to the uncertainty of CCCs nevertheless converge on the PP, then we have what Cass Sunstein calls an “incompletely theorized agreement” on a policy, that is, an agreement to which parties divided by often deep theoretical differences can nevertheless give their assent,60 which is all to the good from a political point of view. In the specific case of CC and its possible catastrophes, the fact that such an agreement has emerged provides a sliver of hope in the face of the increasingly dismal prospects for the planet uncovered by CC science as it progresses. At the level of principle, we are agreed, whatever justificatory reasons we advance. What is needed now—in the immediate present—is policy to promote our principles, and the political will to enact it.
I would like to thank Gillian Brock, Nigel Pleasants, Jurgen de Wispelaere, Jo Wolff, and the audience at the Priority in Practice Workshop, University College London, April 25, 2007, for very useful comments, suggestions, and discussion. Work on this paper was supported by a Leverhulme Trust Research Fellowship (Grant No. RFG/2006/0303), for which I am very grateful.
Fred Pearce, With Speed and Violence: Why Scientists Fear Tipping Points in Climate Change (Boston: Beacon Press, 2007).
I borrow this term from the excellent discussion in Stephen Gardiner, “Saved by Disaster? Abrupt Climate Change, Political Inertia and the Possibility of an Intergenerational Arms Race,”Journal of Social Philosophy 40, no. 2 (2009): 140–62 (this issue).
James Hansen, “Tipping Point: Perspective of a Climatologist: Abstract.” Available at http://pubs.giss.nasa.gov/abstracts/inpress/Hansen_1.html (accessed July 5, 2007).
See Pearce, With Speed and Violence.
See Mark Lynas, Six Degrees: Our Future on a Hotter Planet (London: HarperCollins, 2007), 207–62.
See John Rawls, A Theory of Justice, rev. ed. (Oxford: Oxford University Press, 1999), 251–58.
Rawls argues that “maximin” is the decision rule that would be adopted by parties in the original position: it directs them to choose principles that govern a distribution of advantages wherein the worst-off group is made as well off as possible.
Cf. Kristian S. Ekeli, “Environmental Risks, Uncertainty and Intergenerational Ethics,”Environmental Values 13, no. 4 (2004): 426–27.
See Clive Spash, Greenhouse Economics: Value and Ethics (London: Routledge, 2002), 122.
See, for example, Richard Posner, Catastrophe: Risk and Response (Oxford: Oxford University Press, 2004), 148; and Cass Sunstein, “The Catastrophic Harm Precautionary Principle,”Issues in Legal Scholarship 7 (2007): Article 3, available at http://www.bepress.com/ils/iss10/art3 (accessed July 1, 2007).
For example, see Sven Ove Hansson, “The Limits of Precaution,”Foundations of Science 2, no. 2 (1997): 293–306; Per Sandin, “The Precautionary Principle and the Concept of Precaution,”Environmental Values 13 (2004): 461–75; European Environment Agency, The Precautionary Principle in the 20th Century (London: Earthscan, 2002); Derek Turner and Lauren Hartzell, “The Lack of Clarity in the Precautionary Principle,”Environmental Values 13 (2004): 449–60; and Radoslav S. Dimitrov, “Precaution in Global Environmental Politics,”International Journal of Global Environmental Issues 5, no. 1 (2005): 96–113.
Such as the Maastricht Treaty Article 130r, or the Rio Declaration on Environment and Development Principle 15. For a comprehensive list of international legal documents in which the Weak Precautionary Principle (WPP) and the Strong Precautionary Principle (SPP) appear, see European Environment Agency, The Precautionary Principle, 6.
For more on the WPP see Edward Soule, “Assessing the Precautionary Principle,”Public Affairs Quarterly 14, no. 4 (2000): 309–28.
This could be either a consequence of the nature of the principle (it is non-categorical), or a consequence of the unprincipled nature of policy-makers, or both. I take no stand here on the question of whether a principle must be categorical to count as such, or to qualify as adequate. As for policy-makers, it is almost certainly the case that they are often motivated by reasons not justified by good principles.
Cass Sunstein, Laws of Fear: Beyond the Precautionary Principle (Cambridge: Cambridge University Press, 2005), 14.
Note, though, that Sunstein does defend a Precautionary Principle for uncertain catastrophes, the paradigm of which he takes to be climate change catastrophes (CCCs). See Sunstein, “The Catastrophic Harm Precautionary Principle.”
Note that this argument does not turn on the controversial claim that maximin is the principle of rational choice in all circumstances of uncertainty. Rather, it is (a) limited to CCCs, and (b) it is an argument from egalitarian justice, which I do not take to be justified by reference to principles of rational choice.
Both figures given in George Monbiot, Heat: How to Stop the Planet Burning (London: Allen Lane, 2006), 51.
With a range of −1% to +3.5%. The Stern Review, Executive Summary (2006), 13–16, available at http://www.hmtreasury.gov.uk/media/8AC/F7/Executive_Summary.pdf (accessed February 21, 2007).
It is worth noting that the content of IPCC Reports is decided by consensus among the authors. Given their often deep divergences of scientific opinion, the net effect of this method is to produce conservative statements—the lowest common denominator of scientific opinion to which all authors can agree.
Note that Gardiner refers to this principle as the “Rawlsian Core Precautionary Principle.” See Stephen Gardiner, “A Core Precautionary Principle,”The Journal of Political Philosophy 14, no. 1 (2006): 33–60, at 47–48.
Ibid., 47. Gardiner also adds a criterion that specifies that the threats addressed by any Precautionary Principle must be “realistic” (p. 51), for example, not science fiction, purely imagined, paranoid, purely religiously inspired, and so on. Note that the science that informs Methane Nightmare is sound.
For discussion see Gardiner, “A Core Precautionary Principle,” 56–57.
Note, however, that Rawls himself does not conceive of his principle of intergenerational justice (the Just Savings Principle) as a principle of distributive justice. It is debatable whether he is wise in this approach. For an alternative interpretation of the Just Savings Principle, see Frederic Gaspart and Axel Gosseries, “Are Generational Savings Unjust?”Politics, Philosophy and Economics 6 (2007): 193–217.
See, for example, Bjørn Lomborg, “Climate Change Can Wait. World Health Can't,”The Observer, July 2, 2006.
See, for example, Frank Ackerman and Lisa Heinzerling, Priceless: On Knowing the Price of Everything and the Value of Nothing (New York: The New Press, 1994); and Elizabeth Anderson, Value in Ethics and Economics (Cambridge, MA: Harvard University Press, 1993).
Of course, economists and utilitarians—for whom all values are commensurable—will vigorously object at this point. Classic defenses of the existence of incommensurable values fit to meet these objection can be found in, for example, Isaiah Berlin, “The Pursuit of the Ideal,”The Crooked Timber of Humanity (London: John Murray, 1990); and Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986).
Again, economists and utilitarians committed to denying the existence of incommensurable values will not be moved by such an argument, as they will reject the incommensurable value of intergenerational justice that lies at its heart. However, it is important to note that, in this case, all roads could lead to Rome. Economists and consequentialists can make alternative arguments for the SPP with respect to CCCs that do not depend on the incommensurable value of doing intergenerational justice. See, for example, Posner, Catastrophe, chap. 3; and Sunstein, “The Catastrophic Harm Precautionary Principle.”
John Rawls, Justice as Fairness: A Restatement (Cambridge, MA: Harvard University Press, 2001), 102.
See Gerald Dickens, “The Blast in the Past,”Nature 401 (1999): 752–55.
Scott L. Wing, Guy J. Harrington, Francesca A. Smith, Jonathan I. Bloch, Douglas M. Boyer, and Katherine H. Freeman, “Transient Floral Change and Rapid Global Warming at the Paleocene-Eocene Boundary,”Science 310 (2005): 333–36.
See Syukuro Manabe, Richard T. Wetherald, Christopher Milly, Thomas L. Delworth, and Ronald J. Stouffer, “Century-Scale Change in Water Availability: CO2 Quadrupling Experiment,”Climatic Change 64 (2004): 59–76.
Chris Huntingford, Rhiannon L. Jones, Christel Prudhomme, Richard C. Lamb, John H.C. Gash, and David A. Jones, “Regional Climate-Model Predictions of Extreme Rainfall for a Changing Climate,”Quarterly Journal of the Royal Meteorological Society 129 (2003): 1607–21.
See Lynas, Six Degrees, 223–30.
See David Kidder and Thomas Worsley, “Causes and Consequences of Extreme Permo-Triassic Warming to Globally Equitable Climate and Relation to the Permo-Triassic Extinction and Recovery,”Palaeogeography, Palaeoclimatology, Palaeoecology 203, nos. 3–4 (2004): 207–37.
See Gregory Ryskin, “Methane-Driven Oceanic Eruptions and Mass Extinctions,”Geology 31, no. 9 (2003): 741–44.
See Jesper K. Nielsen and Yanan Shen, “Evidence for Sulphidic Deep Water During the Late Permian in the East Greenland Basin,”Geology 32, no. 12 (2004): 1037–40.
See Lee R. Kump et al., “Massive Release of Hydrogen Sulfide to the Surface Ocean and Atmosphere During Intervals of Ocean Anoxia,”Geology 33, no. 5 (2005): 367–400.
Lynas, Six Degrees, 254; emphasis in original.
Note that this is not to say that there could be no justice between people in such circumstances, or between us and them; rather, it is to say that they would be in no position to achieve justice. See Brian Barry, “Circumstances of Justice and Future Generations,” in Obligations to Future Generations, ed. R. I. Sikora and Brian Barry (Philadelphia, PA: Temple University Press, 1978), 204–48.
For reflections bearing on this point, see John M. Taurek, “Should the Numbers Count?”Philosophy and Public Affairs 6, no. 4 (1977): 293–316.
For more on this question, see my “Getting Motivated in the Last Chance Saloon,”Critical Review of Social and Political Philosophy (in press).
I grant that this claim does not hold in all possible worlds. For example, when the costs of precautions are one hundred percent of GDP every year, then the harms of Unnecessary Expenditure could exceed the harms of Methane Nightmare. However, this is not the case, even according to the highest estimates of the costs of precautions. That the claim is true in the actual world is good enough for me.
For discussion of whether placing the strains of commitment at the heart of political justification in this way is consistent with a contractualist approach, see Jean Hampton, “Contracts and Choices: Does Rawls Have a Social Contract Theory?”The Journal of Philosophy 77, no. 6 (1980): 315–38.
Thanks are due to Jo Wolff for pressing this objection with me.
Mike Otsuka made the point to me in this way, although I think the addition of a big party is mine.
I am grateful to Jurgen de Wispelaere for suggesting this way forward to me.
Note that there may also be further categorical principles relevant to CC justice which, like the SPP, ought to frame the political debate but are not themselves justified by reference to the deliberations of the demos. One such set of principles relates to corrective justice, whereby (roughly) compensation is owed by an agent of harm to those she harms. See my “Corrective Justice and Climate Change,”Annual Review of Law and Ethics (in press).
See Posner, Catastrophe, 24–29.
Perhaps the SPP would be insulated to some degree against the priorities problem by incorporating into the account Gardiner's “realistic outcomes” criterion (see note 27). This would, for example, probably rule out alien invasions as events against which CCCs compete for resources necessary for precautionary action. However, it would not rule out everything. Comet collisions and supervolcanic explosions are very real possibilities.
Note that I do not limit liability with the requirement that an agent of harm is in fact the cause of that harm; I think that risk creation, where harms have not yet in fact been caused, can create liability in some cases. See my “Corrective Justice and Climate Change.”
This principle states that a person ought to be free from interference by the state and by others unless her actions affect the interests of others and cause them harm. See John Stuart Mill, On Liberty (Cambridge: Cambridge University Press, 1989), chap. 1.
For a classic discussion, see Bernard Williams, “Moral Luck,” in Moral Luck (Cambridge: Cambridge University Press, 1981), 20–40.
For discussions that bear on this hard question, see Judith Jarvis Thomson, “Remarks on Causation and Liability,”Philosophy and Public Affairs 13, no. 2 (1984): 101–33.
See, for example, Posner, Catastrophe, chap. 3; and Sunstein, “The Catastrophic Harm Precautionary Principle.”
See Sunstein, Laws of Fear, 2.