ANDREAS MOGENSEN: (00:06) So the question I'd like to reflect on is: what sacrifices we should be willing to make for the sake of future people, in particular under what circumstances is it desirable to reduce consumption in the current generation for the sake of increasing consumption in the future? So the discount rate is the standard tool used by economists to answer this question representing the minimum riskless rate of return that must be earned by an investment project here assumed to be funded by reductions in current consumption in order for its implementation to be socially desirable. And Ramsey specifies a well-known formula for determining the discount rate for marginal investment projects funded by reductions in current consumption. The Ramsey formula states that r = δ + ηg… r here is the discount rate, g is the growth rate of consumption. The remaining parameters are so-called taste parameters. So we use η (eta) to denote the elasticity of the marginal utility of consumption which is generally understood as a measure of our aversion to intertemporal consumption inequality and then lastly we have δ (delta) which specifies the so-called rate of pure time preference. This is the proportional rate of decline in the utility discount factor which is the weight that we put on utility derived from consumption occurring at a given point in time. So that utility discount factor is typically normalized to 1 in the current time period, if it's value declines as a function of time and have a positive rate of pure time preference.

We care more about utility derived from consumption if it occurs sooner rather than later. And the choice of discount rate matters greatly for how we think about our obligations to posterity. So famously these sometimes hostile disagreement between William Nordhaus and Nicholas Stern concerning the extent of emissions abatement required in the near future is traceable principally to conflicting views about discounting and that disagreement is driven in large part by conflicting views about the values of the so-called taste parameters. So Nordhaus adopts a value for δ of (02:00) 1.5% whereas Stern rejects a positive rate of pure time preference as ethically indefensible paring δ down to a measure of the exogenous per year risk of human extinction. Now Nordhaus attempts to paint Stern's commitment to intergenerational impartiality as somehow controversial and idiosyncratic alleging that it stems from the British utilitarian tradition with all the controversies and baggage that accompanied that philosophical stance. However, Stern’s view is widely shared among moral philosophers including those who are hostile to these [inaudible 02:32]. So take for example John Rawls, arguably the most prominent critic of utilitarianism in 20^th century moral philosophy. Rawls tells us in no uncertain terms:

“There are no grounds for discounting future well-being on the basis of pure time preference.”

He tells us that:

“Justice as fairness.”

Agrees that, with Sidgwick, in rejecting time preference as a grounds for social choice. So the basic moral argument for rejecting a positive rate of pure time preference is stated quite clearly in a recent paper by Simon Caney who, as it happens, also rejects utilitarianism. So Caney says:

“A person's place in time is not in itself the right kind of feature of a person to affect his or her entitlements. For example, it does not make someone more or less deserving or meritorious. Similarly it does not in itself make anyone's needs more or less pressing. It is not the right kind of property to confer on people extra or reduced  moral status.”

So what should we think about this kind of argument? Well I'd like us to consider the following observation in general in order to be justified in caring more about some people than others. We need not regard the latter as having reduced moral status. So for example, I care more about my wife than any other person but I don't regard other people as having a lower moral status that she does. For example, I don't think there's a general requirement on moral agents to weight her welfare more than the welfare of other people. Therefore, the basic moral argument against a positive rate of pure time preference given by Caney appears to (04:00) be invalid. In order to be justified in caring more about the utility of people who are nearer to us in time and less about those who are further away, we need not regard location and time as conferring on people extra or reduced moral status. We can restate this point as follows: Caney's argument appears to presume that the only moral reasons that could exist for assigning a lower utility discount factor to S₁ than to S₂ must be agent-neutral for moral reasons such as that S₂ is more deserving than S₁ or is otherwise somehow imbued with greater moral status. However, some of the most important moral reasons that we have for caring more about one sentient being than another are agent-relative moral reasons, having to do with particular relationships in which we stand to some people and not to others. So my aim in this talk is to explore whether and what kind of pure intergenerational time preference might be justifiable by appeals to the existence of agent-relative reasons for valuing the utility of people located  nearer to us in time more highly. And this idea has actually been broached by a number of economists, Kenneth Arrow may have been the first, Wilfred Beckerman and Cameron Hepburn raised this in responding to the Stern Report and Stern himself notes it in his 2008 retrospective on the report. In fact in that article Stern described this as the…

“Only ethical argument for positive δ  that has some traction.”

But the idea has received surprisingly little attention from moral philosophers. The only sustained critical discussion of it of which I'm aware occupies approximately one page of text in Appendix F of Derek Parfit’s Reasons and  Persons.

So this talk is intended as explanatory. I come not to praise agent-relative pure time discounting nor to bury it. I'm just trying to see what traction we can get after thinking about pure intergenerational time preference in terms of agent-relative moral reasons. So I hope you’ll forgive or perhaps even somewhat enjoy the slightly rambling and inconclusive nature of my discussion. If anything I hope it will encourage you to think more about (06:00) these issues.

Now of course the neglected agent-relativity as a justification for pure time preference among philosophers would be understandable if Parfit had succeeded in showing this idea to be misguided and economists had simply failed to catch on. But I don't think that this is the case. So a particular variant of this idea that Parfit discusses is one that I'm going to call discounting for kinship. It's a view on which the application of a declining utility discount factor to the welfare of future generations represents the gradual weakening of partiality due to loosening bonds of kinship. So the basic thought is something like the following: the people who are born into the next generation are our children (or if we all expect to remain childless then maybe they are nephews and nieces). Now by the likes of common sense morality, we are each permitted and/or required to be strongly partial to the interest of our children and also presumably but perhaps to a lesser extent to our nephews and nieces. Plausibly, we are also permitted to be partial to our grandchildren and also to our grandnephews and grandnieces but to a lesser degree. For each succeeding generation we may think the degree of permissible and/or obligatory partiality declines as the degree of relatedness between present and future people declines. Therefore, we can permissibly weight the welfare of each succeeding generation less than that of the generation that preceded it. So that's the basic idea of discounting for kinship.

I should emphasize that this idea need not be understood as committing us to a view on which a merely biological relationship justifies partiality on family members. We should here understand the term kinship as referring to whatever relationship among those individuals whom we recognize as family members generates reasons of partiality. Different philosophers will have different reasons, different views about the nature of that relationship. Some might think that biology plays a very important part in it, others may think that it does not. So far as I can see this idea of discounting for kinship does not require us to take a stand in (08:00) this debate. One other thing that should be noted about discounting for kinship is that it is quite obviously not the mere passage of time that justifies the use of a declining utility discount factor. According to this view, what matters is not really distance in time but distance in respect of relatedness. However, in practice it will generally be true that the more distant someone of our descendants is from us in terms of time, the more distantly related are we in respect to kinship. And so as a simplifying assumption of our model it may seem warranted to adopt, to represent discounting for the kinship by adopting a positive rate of pure intergenerational time preference. So that's the basic view.

What are Parfit’s objections to discounted for kinship? Well in a sense he has none. Parfit does not object to discounting for kinship so much as using this idea to justify what he called “employing a standard Discount Rate.” I think it's easiest to grasp what Parfit has in mind when speaking about a standard discount rate if we just turn to the objections that he lodges against using discounting for kinship to justify what he takes to be the business as usual approach. So for example, Parfit insists that discounting for kinship cannot justify the use of a discount rate “to the infliction of grave harms.” His view is that when it comes to the infliction of harms of this kind, “special relations make no moral difference.” Therefore, even if we insist that diminishing bonds of kinship can justify some kind of diminishing concern for future persons, Parfit thinks that it cannot justify the application of a declining utility discount factor to all kinds of effects.

However, I see very little reason to believe this. So on its face special relations do make an important difference to the morality of harm. Consider for example famous Switch case discussed by Foot and Thompson. In this case it may be permissible to turn the trolley from the five and on to the one, but intuitively your reasons against doing so would be much stronger if the one on the side track were your own (10:00) child. We may think that… We might well think that it would be impermissible to turn the trolley in this case as Frances Kamm for one does. We tend to think that people who murder their parents, their siblings or their children do something especially horrific. So just to give one example of this intuition in Dante's Inferno, these people are consigned to the frozen lake Cocytus, which is nearest to Satan and described by the poet as “that final dismal hole which bears the weight of all the steeps of Hell.” So the view that special relationships make no difference at all to the morality of harm seems very hard to believe without some kind of supporting argument which Parfit does not supply.

The other objection that Parfit notes in arguing that discounting for kinship cannot justify a standard discount rate is that if we discount the kinship, then δ  must eventually drop to or asymptotically approach zero. So why does he think this? So we need to keep in mind that what is being discounted for time on this view is in effect the greater weight that attaches to the interest of certain people in virtue of their kinship to us. A kind of top up to the degree of concern that we owe to any sentient being living as such. Therefore, the practice of time discounting should not lead us to count the welfare of some distant descendant of ours for less than that of some complete stranger living now. If we end up doing that, we've clearly gone wrong somewhere. As Parfit says:

“We ought to give some weight to the effects of our acts on mere strangers. We ought not to give less weight to the effects of our actions on our descendants.”

The claim then is that if we discount the utility of our descendants using a constant positive rate of pure time preference, we end up violating this constraint. So this suggests a second factor that Parfit thinks of as constituting a standard discount rate, namely, the use of constant δ and that assumption is indeed standard in cost-benefit analysis. Here it is typically assumed that social preferences must be dynamically consistent. Roughly speaking this means that a plan must not be socially preferred at one point in time but socially dispreferred (12:00) at a later point in time assuming that no new information about the decision problem has come to light in the intervening period. But it's well-known that the declining rate of pure time preference violates that constraint.

However, the view that dynamic consistency is a desideratum on social preferences is, I think, not altogether compelling. It seems especially implausible when a change in social preference is explained by changes in the composition of society such that later generations have aging relative reasons to weight the interests of people differently from earlier generations and therefore they end up having conflicting social preferences. I think if we believe that persons are in some sense suitably unified over time then dynamic consistency may represent a sensible requirement on individual choice behaviour. Although it should be noted that Parfit’s own views about the unity of a person and what matters in survival rule this out, as shown in a recent paper by Arif Ahmed. In any case I think social collectives being successively constituted by wholly distinct people with different loyalties and different preferences do not have the degree of unity across time needed to make a requirement of dynamic consistency plausible. Assuming we find this persuasive, Parfit’s objection seems to fizzle. It may be true that the standard practice of discounting assumes a constant δ and it may be true that if you use a declining rate of pure time preference, you will be dynamically inconsistent. However, time consistency, I think, is not a very plausible constraint on reasonable social preferences.

So let me now add two of Parfit’s discussion and three observations of my own about the apparent limitations of discounting for kinship. These fall on, I think, relatively naturally from Parfit’s remarks about the need for defining discounts. So firstly, recall this observation that the application of a declining utility discount factor to the well-being of our descendants should not lead us to value the welfare of some one of our descendants less than that of some currently existing stranger. On its face I think this represents an important methodological observation that can help (14:00) to constrain how we should think of the utility discount factor as changing over time. It tells us that we should never end up in a position where we value the life of one of us existing now more than any of our distantly related descendants existing at time t unless we also value one of us existing now more than n currently existing strangers. Now because our intuitions about prioritizing among currently existing people might be clearer and more confident than our intuitions about prioritizing among people existing at very different points in time, this principle can guide us helpfully in selecting a reasonable time schedule for the utility discount factor.

Furthermore, it seems plausible that following this prescription is going to call into question what might otherwise have seemed to be relatively modest suggestions for pure intergenerational time discounting. For example, suppose that we are wondering how much weight we should put on the utility of our descendants living 500 years from now. We might wonder whether it would be appropriate to weight their welfare as if we had applied a constant rate of pure intergenerational time preference of 1.5% compounded annually. Now we know by now that we should not in fact be using a constant discount rate, nonetheless, we might wonder whether the application of the right declining pure time discount rate would lead us to value the utility of people living 500 years from now as if we had been discounting at 1.5% annually. Well given the methodological constraint that I just mentioned it turns out that we are barred from doing this unless we value an increase in the utility of one currently existing family member more than an equivalent increase in the utility of 1,710 currently existing strangers (assume that I've done the maths correctly). And assuming that I have done… Well opinions might differ but at least to me this would seem to require an excessive degree of clan loyalty and will therefore [inaudible 15:53].

So here's the second observation. Note that discounting for kinship entails that we (16:00) should apply a declining utility discount factor only with respect to the welfare of descendants. So discounting for kinship might justify us in caring more about those of our descendants who are nearer to us in time but it provides no justification for caring more about the welfare of unrelated strangers on the basis of their location in time. Within the context of assessing optimal climate policy, this may prove especially significant. It's widely recognized that harms from climate change will form principally in developing countries who have historically contributed least to the current environmental crisis. These countries arguably have the weakest obligations to pursue aggressive abatement policies as recognized by their exemptions from the group of Annex 1 countries bound by the Kyoto Protocol. So let's be our [inaudible 16:45] notes. From the perspective of developed countries or people in developed countries a key question for the design of optimal climate policies is whether it is better to benefit future inhabitants of developing countries by mitigating greenhouse gas emissions or to fund development projects that help poor inhabitants of developing countries in the here-and-now.

To drive home the weight of the choice that we face here, Lomborg in his 2001 book, calculates that the cost liened to Annex 1 countries who are complying with the Kyoto Protocol was enough to fund the extension of basic health education, water and sanitation to every person living in the developing world. Now from the perspective of many or perhaps most current citizens of developed countries, both current and future inhabitants of developing countries are equally unrelated to them. For such people discounting for kinship would provide no basis for the choice of a positive rate of pure intergenerational time preference when deciding between abatement and development projects insofar as their concern is directed towards the inhabitants of developing countries who will be hardest hit by climate change. So this suggests that a zero rate of pure intergenerational time preference is the only appropriate option for addressing one of the most important prioritization decisions faced by many inhabitants of developed countries when (18:00) deciding on their priorities in relation to global climate change.

So here, the third observation which follows on very naturally from the previous. Under discounting for kinship there is no unitary discount rate. What I mean is, precisely because reasons for discounting are presumed to be agent-relative there will be no shared rate of pure time preference such that each current existing individual ought to discount the welfare of future people at a certain rate per period no matter which future people they might be. If reasons for discounting or reasons for pure intergenerational time preference are agent-relative, then we cannot ask whether and to what extent it may be justifiable to discount the welfare of future people in general. Instead we need to specify for whom it may be justifiable to discount the welfare of which people relative to what other people in light of their location in time. But this is not how discounting is standardly conceived by economists concerned with long-term policy settings such as climate change mitigation. There it is standard to assume that there is some kind of utility discount factor that should be applied to future generations as a whole whose rate of change is under debate. This way of thinking we might suppose makes very little sense if we assume that reasons for pure time discounting are agent-relative and must therefore vary in character from person to person. However, this may be a little bit too quick so I think there are some contexts in which economists are especially concerned about the justifiability of pure time discounting where this question about what utility discount factor should be applied to future generations as a whole might be sensible because in these contexts the assumed perspective may not be that of some particular individual but nor is it the sort of purely impartial point of view of the universe. It's something in between. Something like the point of view of all of currently existing humanity. So let me explain what I mean. Consider in particular, again, assessments of optimal climate policy. So as Stern conceives of the problem at hand (20:00) here, it concerns:

“…social decisions by the world community now, bearing in mind consequences for future generations.”

The thought is that climate change is global in its origins and in its impacts and so an effective response requires international understanding and cooperation. Now although he famously chides Stern for adopting what he calls:

“…the lofty vantage point of the world social planner.”

Nordhaus in fact adopts a similar frame. For example, the discounted utilitarian social welfare function in his DICE model is assumed to represent the collective preferences of the world as a whole. In this debate, therefore, it may be plausible to suppose that the key question on the discussion is something like… What the world community now should do, as opposed to what you or I should do or what some particular country or bloc of countries should do. It is a question about what we ought to do together where “we” seems to range over everyone who's currently [inaudible 20:53]. Thus, insofar as there are reasons of partiality in play here, those reasons may belong not to some particular individual but to this much greater collection of agents. To help clarify this idea an analogy might be helpful. So suppose that Kasei is Hiroko’s son and Nikki is Nadezhda’s son and they are otherwise strangers to one another and we'll imagine that Kasei and Nikki are drowning and so is a third person who is a stranger to all of them. And we also see  that neither Hiroko nor Nadezhda can save any person on their own, but together they can save exactly two of the three people who are drowning. Then intuitively they together have the most reason to save Kasei and Nikki, but note that Nadezhda considered on her own, may have no reason to prefer that Kasei and Nikki are saved as opposed to Nikki and the stranger and similarly Hiroko considered on her own may have no reason to prefer that Kasei and Nikki are saved as opposed to Kasei and the stranger. In the same way…

(22:00) Sorry! Before here rather then, is that there are reasons of partiality that they together have which pick out the pair of Kasei and Nikki as uniquely important but which no individual considered in isolation has. So that's the key kind of thought and the sort of extension then of this… The thought is that in the same way the world community now in acting together in response to global climate change or other similar challenges may have reasons of partiality that belong to us collectively as opposed to being the reasons of partiality of some particular individual and those reasons of partiality may pick out the next generation as being uniquely important to us collectively and subsequent generations as being less important. So the thought is, if we together who have greater reason to care about the next generation than about generations more distant from us because those who are born into the next generation are more closely related to us whereas succeeding generations will be more weakly related to us. They call this way of understanding what discounting for kinship means in the context of global problems requiring internationally coordinated action, global collectivism. So many questions can and should be raised about global collectivism. I've only just given you a sketch of this idea, so you might wonder, for example, who are the members of this world community now? Should we think of this as being a collection of all the individuals who currently exist or should it be instead the collection of all states that currently exist? Global collectivism seems to presuppose the existence of collective reasons, reasons that are not the reasons of some particular individual, of a group of individuals.

How are we supposed to make sense of that? Should we think that the whole of currently existing humanity constitutes some kind of collective agent, have these reasons? Should we think that this way of talking involves so-called plural predication which outruns the expressive capacity of first-order predicate logic? And what exactly is the relationship between the reasons that we together have to be especially concerned about the welfare of the next generation and the reasons that individual people have to be especially (24:00) concerned about the welfare of their children? These are all good questions. Unfortunately, there is not enough time remaining for me to answer these questions as a part of this talk. Instead this the point at which my exploration of agent-relativity as a justification for pure intergenerational time preference will have to draw to a close. So I hope this brief ramble has been worth your while and that at least some of you will be encouraged to think more about the promises and perils of partiality as a justification for a positive rate of pure time preference. Thank you.

[Applause]