Risking Future Generations

Abstract

Many of the policy choices we face that have implications for the lives of future generations involve creating a risk that they will live lives that are significantly compromised. I argue that we can fruitfully make use of the resources of Scanlon’s contractualist account of moral reasoning to make sense of the intuitive idea that, in many cases, the objection to adopting a policy that puts the interest of future generations at risk is that doing so wrongs those who will live in the further future.

Consider the following case, from Derek Parfit’s seminal discussion of obligations to future generations in Reasons and Persons¹:

The Risky Policy As a community, we must choose between two energy policies. Both would be completely safe for at least three centuries, but one would have certain risks in the further future. This policy involves the burial of nuclear waste in areas where, in the next few centuries, there is no risk of an earthquake. But since this waste will remain radio-active for thousands of years, there will be risks in the distant future. If we choose this risky policy, the standard of living will be somewhat higher over the next century. We do choose this policy. As a result, there is a catastrophe many centuries later. Because of geological changes to the Earth's surface, an earthquake releases radiation, which kills thousands of people. Though they are killed by this catastrophe, these people will have had lives that are worth living. We can assume that this radiation affects only people who are born after its release, and that it gives them an incurable disease that will kill them at about the age of 40. This disease has no effects before it kills (Parfit 1984,371-2).

Two features of this case make it particularly compelling. First, it calls attention to the fact that questions concerning our obligations to future generations often arise in the context of considering the moral acceptability of doing what creates a risk of compromising in significant ways the lives of those who will live in the further future (Parfit 1984,371–2). Second, as I see it, it focuses our attention on the intuitive idea that the objection to adopting the risky policy is that its adoption wrongs those who live in the further future. Though we are told that the risk does eventuate, resulting in thousands being born with an incurable disease that leaves them with only the possibility of truncated lives, knowing how things turned out is inessential for concluding that doing what creates the risk of this kind of catastrophe, simply to secure slightly better lives for ourselves and our immediate descendants, wrongs future generations put at risk.

This claim is of course controversial. Consider a more familiar case of unjustified risk imposition: you are outside, helping your young child her learn to ride her bike, when a driver comes driving down the street at high speed, zooming by before either of you has a chance to process what is happening. The driver’s reckless conduct imposed a serious risk of harm on you both, and that imposition, I believe, is grounds for taking yourself and your child to have been wronged. Nothing bad happened, but something could have.

Pre-theoretically, the objection to the driver’s conduct in this case looks to be analogous to the objection to adopting the risky policy. Say it is adopted, but it has no bad future effects. Those living in the further future are, intuitively, still wronged by its adoption. For though it had no bad effects, it could have. And just as we wrong our contemporaries by doing what puts their interests at risk in order to secure small benefits for ourselves, we have no pre-theoretical reason for thinking that we don’t wrong those who will live in the further future by doing the same to them.

The non-identity problem, however, appears to show this line of thought to be deeply confused. The reckless driver wrongs the bystanders because the risk she imposes on them could eventuate in their lives as actual injuries, and there are no good reasons to permit driving in a way that imposes a risk of that happening. But in the Risky Policy case, the particular individuals who exist in the further future, whose lives are truncated by the incurable disease, would not exist had that policy not been chosen. It makes no sense, therefore, to say that adopting the risky policy runs the risk of damaging their interests. But if their interests cannot be compromised by its adoption, it is hard to see in what sense doing so can be said to wrong them.

Perhaps it can’t; that just is the lesson of the non-identity problem. Philosophical reflection sometimes leads us to conclude that a powerful intuition is rooted in conceptual confusion, so ought to be set aside. Further, doing so might not look like much of a sacrifice. We would not, after all, be setting aside the idea that we have weighty obligations to protect the interests of future generations; we would just be adopting a particular understanding of their character. Rather than think of them interpersonal obligations, of the kind that bind presently existing individuals to one another, such as the obligation not to impose risks of harm on one another for trivial trivial benefits, we would understand them as serious moral obligations that are not owed to anyone, and so more impersonal in character.² Nothing of moral importance looks, therefore, to be lost when we abandon the idea that those who live in the further future are wronged by the risky policy’s adoption.

To so conclude, however, would be a mistake. The reason it would be a mistake can be illustrated by considering the intuitive objection to the claim that climate change mitigation efforts cannot be vigorously pursued, because doing so will be detrimental to the employment prospects of current generations. This way of thinking, the objection goes, is misguided because it too easily accepts that the interests of the living take priority to those who will live in the further future, simply because they are the interests of those currently alive.

The assumed priority can, however, be defended by appealing to the idea that ordinary interpersonal obligations are normally understood to be authoritative, or to be a source of decisive reasons, with respect to how we regulate our conduct. In particular, interpersonal obligations are thought to give us reasons to regulate ourselves in certain ways that are not to to be tossed into the deliberative field along with other moral and non-moral reasons, where they fight it out to determine what there is all things considered reasons to do. Rather, the reasons interpersonal obligations give us are thought to take priority over other reasons we have. If we understand the obligations living individuals owe one another as interpersonal, and take obligations to future generations to be comparatively impersonal, the thought that the economic interests of living individuals take priority over the interests of future generations ceases to be so obviously mistaken.

An important reason, then, for characterizing obligations to future generations as interpersonal, then, has to do with our intuitively wanting to say that, first, many of our most important moral obligations are obligations we owe to one another; second, that the reasons such obligations give us take priority over other reasons we have; and third, that obligations to protect and promote the interests of those who will live in the further future are just as stringent as the obligations that currently living individuals owe to one another. Understanding obligations to future generations as interpersonal obligations that are not different in kind from those that bind living individuals to one another allows us to both make sense of all three intuitive claims and preserve the intuitive objection to adopting the risky policy.

There is, then, a case to be made for the appeal of an interpersonal understanding of our obligations to future generations. In what follows, I will draw on Scanlon’s contractualist moral theory, and certain points about the nature of risk, to argue for the vindication of such an understanding, and in particular, the cogency of the intuition that the moral objection to the risky policy is that its adoption wrongs those who live in the further future. An important reason for adopting the contractualist approach, I will suggest, is that it offers a particularly attractive account of the authority of interpersonal obligations that allows us understand living individuals as being in a kind of personal relationship with those who will live in the further future that is not different in kind from that which binds currently living individuals to one another.

Scanlon’s contractualism, as I interpret it, holds that one person wrongs another when “one’s mode of self-governance has ignored or flouted requirements flowing from another person’s standing as one to whom justification is owed” (Scanlon 1998, 271). Those requirements are fixed by principles no one can reasonably reject for the general regulation of conduct in different types of situations, of which there are indefinitely many.

Lying at the heart of this account is the claim that respect for the value of human life requires that one’s mode of self-regulation recognize the “capacity of other human beings, as rational creatures, to assess reasons and to govern their lives according to this assessment” (Scanlon 1998, 106). Recognition of another, in any set of circumstances, as having this capacity requires, minimally, that one’s conduct conform to what is permitted by the relevant principle, that no one can reasonably reject, for the regulation of conduct in such circumstances. A principle for the regulation of a certain type of conduct that no one (appropriately motivated) can reasonably reject is one that all individuals have reason, as assessed from each person’s own point of view, to license one another to be regulated by.

Whether a proposed principle is one no one can reasonably reject turns on comparing its implications with the implications of candidate alternative principles, as assessed from various relevant individual points of view.³ If a principle is one no one can reasonably reject, the strongest objection to it will not be as strong as those that can be pressed from other points of view against every plausible alternative. It is the principle whose implications are most acceptable to the person to whom it is least acceptable.

The account is avowedly non-consequentialist, locating the basis of a person’s claim to have been wronged in an intrinsic feature of the wrongdoer’s conduct: the wrongdoer’s failure to conduct herself in a way the wronged was, under the circumstances, entitled the wrongdoer conduct herself. Whether or not a person is harmed in being wronged has no relevance to whether she has been wronged, though the question of harm may be relevant to the issue what might be owed by way of redress.

We can understand, on contractualist terms, how a person can be wronged by the imposition of a risk of harm by appealing to the significance of what it is permissible to do as depending on what a principle, for the general regulation of conduct in the type of situation one finds oneself in, permits.⁴ Such a principle specifies what individuals who find themselves in that type of situation, whoever they are, may legitimately expect of one another’s conduct. Taking a particular course of action to be permitted by a principle no one can reasonably reject is to hold that there are good reasons that support the permissibility of any person similarly situated conducting herself in that way, and that the reasons supporting this standing permission are sufficient to defeat objections to it (Scanlon 1998, 197–201).

Because the permission is a general permission to act for anyone thusly situated, assessing whether it is permissible to act, here and now, in a way that has implications for a particular other (or others), requires considering the general question of what the reasons are that favor and oppose this type of conduct being generally permitted. The reasons that bear on this question concern the relevant considerations that could be appealed to from the different points of view of those who stand to be affected, where the only relevant considerations are those that bear on the principle’s implications for different persons being able to lead the kinds of lives they have reason to want to be able to live. Ruled out, then, are direct appeals to impersonal considerations, such as the aggregate value of a state of affairs.

Importantly, because what is licensed by a principle is the general permissibility, in a certain type of situation, of relating to another in a certain way, and there is no way of knowing how often it will obtain or who will find themselves in it, the points of view in question are not those of actual, particular individuals, but rather representative individual standpoints. A “standpoint,” as the term is being used here, is an abstraction, a way of referring to the reasons that persons in certain circumstances (assumed to have certain “normal” capacities for discernment, self-control, planning, and so on) typically have for caring about or wanting certain things (“generic reasons”), such as a degree of personal privacy, security against bodily injury, the freedom to nurture personal relationships with friends and intimates, and sole discretion concerning how one’s body is used.^5,6

Relying on standpoints is motivated by the thought that conclusions about the permissibility of conduct must be reasonably epistemically accessible to an individual thinking about what courses of action are open to her. Requiring that an individual take into account specific facts about the rational preferences, capacities, psychological dispositions, and so on of others in determining what it is permissible to do would place serious epistemic and cognitive demands on a person. This would constrain an individual’s ability to pursue her rational aims, making it unduly burdensome for her to live her life within the bounds of the permissible, and make it unreasonably difficult for both her and others assessing it to be confident of her conduct’s permissibility. These problems are avoided by relying, in thinking about what it is permissible to do, on relevant individual standpoints, characterized in terms of generic information about what individuals in the type of situation in question typically have reason to want.

Two implications of this generality built in to the assessment of principles governing how it is permissible for individuals to relate to one another are especially relevant for understanding how a person can be wronged by exposure to a risk. First, the reasonable rejectability of a principle is assessed ex ante. That some bad event could befall an individual if conduct is permitted under the terms of a proposed principle is not in itself a reason for objecting to it; what matters for assessing the force of the objection is the reason an individual has for not wanting this kind event to befall her, discounted by the likelihood of that happening. Second, because the individuals in question are representative standpoints, what needs to be assessed is not the likelihood of this event befalling this or that individual, but the likelihood of it befalling an individual described by a certain standpoint.

Assessing the reasonable rejectability of proposed principle for the regulation of a certain type of activity that involves imposing a certain risk of harm on others, then, is, very roughly, a matter of comparing the strength of the reasons that can be pressed from a certain standpoint for wanting to be able to engage in the kind of activity in question with the reasons that could be pressed from other standpoints for wanting to be protected from exposure to the risks imposed on others by the pursuit of the type of activity in question.

The reasons for caring about the permissibility of risky activity are, of course, reasons having to do with not wanting certain bad events to befall one. But just because the rationale for a principle prohibiting, say, engaging in a risky activity appeals to the reasons individuals have for not wanting to find themselves burdened in certain ways, it doesn’t follow that for a person to be wronged she must find herself burdened as a result of another’s impermissible conduct. Rather, whether one has been wronged turns on whether or not another’s conduct exposed one to an impermissibly high risk of harm.

On this view, then, there is no sharp line to be drawn between activity that imposes the risk of harm on others and activity that harms others: for purposes of understanding how it is that a person has been wronged by how another has conducted herself, conduct that results in another being harmed is just a case of a person having engaged in a type of risky activity and that risk having eventuated as a bad outcome for the wronged.

So far, I have only argued that we can make sense, in contractualist terms, of a person’s being wronged by the imposition of a risk by another. The challenge to understanding the objection to adopting the risky policy in interpersonal terms, however, does not have to do with whether or not a person can be wronged by the imposition of a certain risk by another. Rather, the challenge lies in making sense of who the “they” are who are wronged by the adopting of the risky policy. We do speak of “putting them at risk” when we explain the intuitive objection to choosing the risky policy. But perhaps that is just a misleading way of talking. As the particular identities of those who will exist in the further future is not independent of whether or not the risky policy is chosen, the philosophically sound thing to say appears to be that they were never put at risk by the choice of the risky policy.

This thought rests, however, on a confusion about the nature of risk. To illustrate what I have in mind, take a different example. Say prospective parents are told, before conception, that they must take a certain drug for at least six months before conceiving, or their child will run of a genetic mutation that is associated with a risk of severe birth defects. The parents choose not to accept this advice, conceive, and a healthy is child is born. It may be tempting to say that, in fact, their child was never at risk of being born with birth defects (say there is good reason to believe that the child carries none of the genetic mutations that might have manifested as birth defects). But that is a mistake. Risk, in the sense relevant here, is an epistemic concept, one that concerns the chance, or probability, that some bad event will occur, indexed to a particular epistemic position.⁷

To explain this point, take a common understanding of probability, the relative frequency view. It holds that the probability of a certain event occurring is fixed by defining some reference class that contains the object of risk assessment and determining what proportion of the people or events in that class have the relevant property that is of interest. Say I want to determine what the risk is of my going bald before the age of fifty. To do so, I need to define a reference class (e.g., men of a certain age, socio-economic status, and ethnicity) and then determine what proportion of men in the reference class go bald before age fifty. Casting the proportion of men in the reference class who exhibit the property of interest (going bald) in terms of frequency tells me what the probability is of my going bald before age fifty.

So understood, claims about risk are general, in that the risk to me is understood in terms of the risk to a person with certain properties, in certain circumstances. The reference class is that with respect to which that risk is fixed, and there are a myriad of possibilities concerning how the reference class can be characterized. How it is to be characterized is constrained by the purpose for which the assessment of the risk is sought. If I want to know the probability is of my going bald before the age of fifty, the reference class has to be characterized by considerations that both can be known, and whose bearing with respect to one another on the probability of the proposition in question is normatively defensible. It isn’t that a more esoteric reference class, such as one that cites all the relevant factors in the universe at a given time, is ill formed. It is just that knowing that the risk to me depends on all these considerations is not going to get me any closer to knowing what I am interested in knowing.⁸

In the case of the parents, then, what is relevant in determining whether or not their child was put at risk by their decision has to do with the evidence concerning the implications of their choice for the prospective child it was reasonable to have expected them to assume at the time of decision (in contractualist terms, the standpoint characterization). Given the information available to them at the time they were told their child is at risk, it is plausible to hold they put their child at risk. That claim isn’t falsified ex post information about the particular genetic identity of their child; they were just lucky that things turned out as they did.

There is no special problem, then, understanding how those who will live in the further future are put at risk by the choice of the risky policy, as claims about a person’s exposure to risk in ordinary cases does not appeal to particular facts about an individual’s particular identity to begin with. The risk, for instance, of a person qua person (as opposed to, say, left-handed musician) being killed in an automobile accident on a major highway is relative to the reference class of those driving on it over some span of time. Similarly, the risk imposed on persons qua persons who will live in the further future (which I assume to be the most morally relevant characteristic) by adopting the risky policy is fixed relative to the reference class of those who will live in the further future, along with what the risk is of the kind of earthquake occurring that releases the deadly radiation. It is tempting to think that, because a particular person would not now exist were it not for a certain choice made before she was conceived, it couldn’t be true that that choice imposed an additional risk of early death on her. But this thought misses the implicit generality of the description under which the person is being referred to when it is claimed that she was put at risk, one relevant for deliberating at about the permissibility of the choice of the risky policy at the time the decision of whether to adopt it has to be made.

What remains murky is whether sense can be made, in contractualist terms, of the thought that adopting the risky policy wrongs those who will live in the further future because it creates a serious risk that their lives will be significantly truncated. Skepticism that it can be is certainly warranted. On a straightforward understanding of contractualist reasoning, whether or not a principle that licenses the adoption of the risky policy is reasonably rejectable turns on a comparison of the burdens for individuals of generally permitting versus prohibiting the adoption of such policies.⁹ If permitting the choice of a policy will result in individuals coming into existence in the further future whose lives are not worth living, we can clearly say that such a principle is reasonably rejectable; we could then say that in choosing it, we wrong those who will live in the further future. But those who will come into existence if the risky policy is chosen are expected to have lives worth living; the risk is that their worthwhile lives will be truncated. So it is hard to see what objection they could have to the risky policy’s adoption. It is better for those adopting the policy (and their immediate descendants), and better for those who exist in because the policy was chosen, who would not have otherwise come into existence.

This conclusion is one, however, that we ought not to accept. First, recall that contractualist thinking sees guidance in one’s practical thought by principles justifiable to anyone on grounds no one can reasonably reject as what proper recognition of the value of human life requires.¹⁰ In particular, what is required is that one’s conduct be in fact justifiable to any other on grounds she cannot reasonably reject, whether or not any other either knows or cares that one’s conduct is so justifiable; that question turns on the balance of relevant reasons that individuals, characterized by the relevant standpoints, have for objecting to or welcoming conduct being regulated on the terms of the proposed principle.

Whether a particular person has been wronged, then, has nothing directly to do with how a person feels about how she has been treated. Nor does it have to do with what reasons she is inclined to press against or accept as justification for a principle licensing her being treated in a certain way. Whether a particular person has been wronged turns, rather, on considerations having to do with reasons that persons with certain capacities, who find themselves in her type of circumstance, typically have.¹¹

Intuitively, we want to say that it is particular persons who are wronged, not standpoints of interests. The point to which I am drawing attention is that, while this intuition is sound, it can be misleading. An individual has a claim to be wronged in virtue of an entitlement she holds against others that they not relate to her in certain ways. But she has that entitlement not in virtue of particular facts about her that mark her as a distinct individual from others. Rather, her entitlement, on the contractualist account, is one that she has in virtue of her circumstances being relevantly characterized by a more general description, one that could be instantiated by an indefinite number of others. That is, she in particular is wronged because, if you like, she ‘instantiates’ the the standpoint whose interests are decisive in fixing the conduct in virtue of which she has been wronged as impermissible.

Conclusions, then, about what, under the circumstances, it is permissible to do rest on the reasons that support the relevant principle for the regulation of the type of situation in question. Whether or not there exist individuals aptly characterized by the cluster of normatively relevant interests constitutive of the relevant standpoints is, from the point of view of person deliberating about what it is permissible to do, neither here nor there. What matters is that our current actions do have implications for those who, based on what we can reasonably now know, will live in the further future. Those people enter into practical thought as standpoints of interests to which any principle that licenses conduct that has implications for kind of lives available to those characterized by those standpoints must be justifiable.

In slightly different terms, one could say that to take seriously the importance of one’s conduct being justifiable to others on grounds no one can reasonably reject is to take seriously the importance of one’s conduct respecting the legitimate entitlements of others, regardless of whether or not one can always pick out a particular individual to whom the entitlement can be attributed. It is enough that there is reason to think that the cluster of interests that ground a restriction against conducting oneself in certain way could be predicated of particular individuals who will live in the further future.¹²That the particular identities of the future people is not independent of a choice we make now has no relevance for the question of whether or not we would be wronging those people in making that choice.

The reasoning I’ve employed to make sense of the idea that adopting the risky policy would wrong those who live in the further future overlaps in certain respects with a reasoning that employs what Parfit calls a wide person-affecting principle.¹³ This is so in two important respects. First, the skeptical challenge to using contractualist reasoning to make sense of the intuitive impermissibility of adopting the risky policy tacitly assumes that contractualist reasoning employs a narrow deontic person-affecting principle. This principle holds that an act cannot be wrong if this act would be worse for no one (Parfit 2017,2). My contention is that contractualist reasoning is not properly characterized in narrow person-affecting terms, so the challenge should be rejected.

Second, in rejecting the relevance of narrow person-affecting principles for reasoning about how it is permissible to proceed, contractualists are in agreement with defenders of employing a wide person-affecting principle in thinking about non-identity cases. A wide person-affecting principle,

makes claims about the intrinsic goodness of different outcomes…. When we apply the Wide Principle, we can consider each outcome on its own, and ask how good or bad this outcome would be for people. An outcome’s intrinsic goodness does not depend on its relation to other outcomes (Parfit 2017,23).

Narrow person-affecting principles, on the other hand, are not concerned with “intrinsic goodness or badness, [so] we cannot consider each outcome on its own. We cannot ask whether some outcome would be bet for people, or worse for people, since these relations are essentially comparative, holding only between different outcomes (Parfit 2017,23). Contractualism agrees with the wide person-affecting principle insofar as it takes the reasons people have for either welcoming or objecting to what a principle licenses to be based on non-comparative judgments. If there some likelihood what a principle licenses will result in those characterized by a certain standpoint bearing a certain burden, anyone appropriately characterized by that standpoint has a reason to want to be able to reasonably reject it. That is a non-comparative judgment. Whether the principle is one that can in fact be reasonably rejected turns on whether there is an alternative principle that will not burden anyone to a comparable extent. The later is a comparative judgment, but it is a judgment about the implications of alternative principles for people (those characterized by the relevant standpoints) rather than for particular individuals.

I have been arguing that the skeptical challenge to using contractualist thinking to understand how adopting the risky policy would wrong future generations does not succeed; it incorrectly assumes that reasoning about a principle’s reasonable rejectability is conducted in narrow deontic person-affecting terms. But the challenge can be restated without that assumption, by seeing it as analogous to an early point made by Parfit, discussing a case in which a woman conceives a child with a life-shortening condition, though she could have easily delayed conception and conceived a child without it. Assuming that the short life is still well worth living, Parfit writes “I would not regret that my mother caused me to exist. And I would deny that her act was wrong because of what it did to me. If I was told that it was wrong, because it caused me to exist with a right that cannot be fulfilled, I would waive this right.” (Parfit 1984, 375).¹⁴ Waiving a right makes doing what would be otherwise impermissible permissible. So it is reasonable for the mother, in Parfit’s case, to take her choice to be permissible, foreseeing that her child will just waive any right it has that would otherwise be violated by her having chosen to conceive knowing her child will have a shortened life span.

That a person will not regret a choice having been made does not make that choice permissible. Nor is it normally thought to change the decisive reason a person has to avoid doing what wrongs another. But one might hold that the Risky Policy case is in this respect unusual. Even if we accept that the adoption of the risky policy wrongs those who will live in the further future, it will be true of those people that they can only exist as wronged. It is therefore plausible to hold that they will not complain of having been wronged, regret that the choice that wrongs them was made, or blame those who made it. This is significant if we understand, as contractualism does, the importance, or authority with respect to one’s practical deliberations, of conducting oneself on terms justifiable to others as grounded in the significance of doing so for the relationship one stands in to others being a valuable relationship, one characterized by mutual recognition of one another’s standing as persons. In non-identity cases, such as Risky Policy, it could be argued, the reason-giving force of conduct being impermissible lapses, because the conduct in question in no way poisons the relationship between the wrongdoer and those for whom the wrongdoing has implications.

To the extent this line of reasoning is forceful, it is because it seems plausible to hold that those who will live in the further future, provided they have lives worth living, and would not exist were in not for the adoption of the risky policy, will not regret its having been adopted. But that in itself is not sufficient to conclude that a relationship of mutual recognition (or respect) between those making the choice and those who will live in the further future does not require that the risky policy not be adopted. Why that is so requires a closer look at how contractualism characterizes the importance for the morally motivated person of avoiding conduct that wrongs others.

Contractualism portrays all human beings as standing in a certain kind of default relationship to one another, the ‘moral relationship’ (Scanlon 2008,139–41). Unlike a relationship of friendship, which is constituted by a certain history of interaction, attitudes towards one another, and each taking herself to owe certain obligations of friendship to the other, the moral relationship is both inescapable and one that a person just finds herself in; in these respects, it is like the relationship of being ‘sibling of.’ It is one that holds between oneself and another irrespective of one’s conduct and attitudes towards the other. All human beings stand in this default moral relationship to one another in virtue of a mutual vulnerability to one another’s attitudes that is characteristic of those who are both capable of, and care about, leading rationally self-governed lives. The same capacity that allows an individual to reflect on the reasons there are for wanting her life to go in certain ways, and actively govern it accordingly, enables a consciousness of others attitudes towards her, what Philippa Foot aptly calls the possibility of second order evil in human life: the consciousness of the disregard or indifference of others of the importance to one of living a rationally self-governed life, one that can be experienced as oppressive (Foot 2003, 168).

Like the relationship of being a ‘sibling of,’ the default moral relationship that holds between all individuals can be a distinctive kind of valuable relationship. For a sibling relationship to be a good relationship of its kind, both parties to it must understand themselves to have certain obligations towards one another that each takes to be important. Each must also exhibit a certain general pattern of other regarding concern and affection that is characteristic of a good sibling relationship. Similarly, the default moral relationship that holds between oneself and any other can be a distinctive kind of valuable relationship, characterized by mutual recognition of one another’s value as persons, if each complies with certain obligations owed to one another, as well as manifesting certain patterns of care and concern for others.

The obligations in question are those specified by principles no one can reasonably reject. Conducting oneself in the way permitted by such principles is what is minimally required by recognition of, and respect for, the value of persons as rationally self-governing beings. A relationship of mutual recognition with another requires more, however, than each regulating her conduct in a way that conforms with the requirements of respect for the value of human life in general. It also requires each being disposed to regulate herself in a way that manifests certain attitudes towards the particular others with whom she has a relationship of that character; it requires that the parties to the relationship be disposed to conduct themselves in ways owed to others for the reasons that justify such conduct as owed to them, as well as being disposed to manifest attitudes towards one another such as being hopeful that things will go well for one another and being glad when they do (Scanlon 2008, 140).

With these points in mind, we can return to the question of how, assuming the risky policy is adopted, those who would not exist had it not been will be justified in morally evaluating that decision. As I have already granted, they cannot regret its having been made, because they do not regret their own existence.¹⁵ The question of whether or not they can regret the decision, however, is one that is indexed to their actual, particular, points of view, as living individuals looking back at a past decision in light of its implications for their lives. Whether or not its adoption was permissible, and the significance of doing so for the relationship between individuals, requires considering how those who adopted the risky policy regulated themselves with respect to the interests of future generations, under the description of those interests that it was reasonable to assume at the time the choice was made. That is, in morally evaluating the adoption of the risky policy, those who live in the further future must consider it, not in light of the implications it has had for them in particular, but in light of the implications it could have been expected to have for generically described ‘people who will live in the further future.’

I’ve already argued that there are good reasons supporting the conclusion that adopting a policy now that puts those who will live in the further future at risk of a truncated life, for the sake of small benefits for those who live in the next few generations, is impermissible, and these reasons are decisive grounds for not adopting the risky policy. From the point of view of those deliberating about whether or not to adopt the risky policy, these reasons are generic reasons, associated with certain standpoints, that bear on how respect for, and recognition of, value of others as persons requires that they regulate their conduct in light of the interests of others. Whether those who stand to be effected by its adoption are those who exist at the time of choice, those who will come into existence irrespective of that choice, or those who will only exist because a particular choice is made makes no difference to the significance of adopting the risky policy for the character of the default relationship between those who choose it and others. Choosing it displays a morally defective attitude towards the value of persons in general, one inconsistent with being in a relationship with others based on mutual recognition. It is inconsistent with the relationship between those who choose the risky policy and those who exist at that time it is chosen having that character, and it is inconsistent with the relationship that comes into existence between those who it and the particular others who come into existence in the further future because it was chosen being a valuable relationship of mutual recognition.

Those who live in the further future of course stand in a unique relationship to the adoption of the risky poicy; it being adopted had a significance for the character of their lives that it did not have for the lives of others. It is because it has this significance for them that, were we to now decide to adopt it, those who will live in the further future will be justified in blaming us for doing so, even if they will not be able to wholly regret that decision having been made. They can blame us because our failure to show appropriate concern for the welfare of others that was owed to them has had significant implications for the character of their lives (Scanlon 2008,147). Their blame will, in a certain respect, be less personal than interpersonal blame often is, because choosing the risky policy manifests an attitude towards the lives of those who will live in the further future, rather than an attitude towards the value of the lives of the particular individuals who exist in the further future. But this difference does not undermine their being warranted in blaming us for having chosen the risky policy.

I’ve argued that the contractualist framework provides us a powerful basis for making good sense of the thought that that choosing to pursue the risky policy may not be permissible because to choose it would be to unjustifiably risk, and thus wrong, those who will only come into existence in the further future if it is chosen. The argumentative strategy does bear some resemblance to what is sometimes called the ‘de dicto’ approach to dissolving the non-identity problem (such as that advanced by Caspar Hare, though unlike Hare’s, the argument employs no comparative claims about better and worse lives), and for that reason might be seen to fail what Boonin calls the ‘modesty requirement’: that the proposed cure for the non-identity problem should not be less plausible than the disease.¹⁶ But that charge misunderstands the argument. It does not introduce any new apparatus to escape the snares of the non-identity problem. Rather, it starts with the general contractualist analysis of how a person can be wronged by the impermissible imposition of risk and then draws attention to its applicability to the case of choosing between policies that have implications for those who will live in the further future. The innovation, if there is one, is not that of introducing a neat trick to enable the side stepping of the non-identity problem, but that of drawing attention to the importance of interrogating the assumptions about moral permissibility (and how claims about what is permissible relate to notions of guilt, regret, blame, and compensation) that are assumed both in the framing of the problem and by challenges to attempts to dissolve it.