Module 4 - Risk, Value, and Future Generations

Economists frequently make use of a social discount rate when they need to put a present-day value on costs and benefits that will happen later. ^[4] This is important, since, when we are deciding whether to invest capital now into a project that will reap benefits in the future, we need to remember that there is an opportunity cost to this investment. Since we can expect a return on investments, some unit of money M in the future is worth less than the very same amount today. Could the same apply to the welfare of future people? That is, could it be the case that some unit of welfare W of a present-day person is worth more than the very same unit of welfare of a future person? If so, this would have important implications for how much we should invest into securing and improving the far future. A very high discount rate might mean that we should invest little into securing the value of the far future, while a low or zero discount rate would suggest that we should invest much more. More germane to our purposes here, an appropriate discount rate might give us the result that, while we ought to invest in making the long-term future go well, we needn’t sacrifice as much as strong longtermists like Bostrom would recommend. 

In a rare show of near unanimity, however, almost all moral and political philosophers who have studied the question find the idea of a discount rate for pure time preference to be morally unacceptable.^[5] The first thing to note is that capital is very different from welfare. It makes sense for economists to apply a discount rate to future economic benefits for the simple reason that, due to the rate of return on capital investment, 100 dollars in ten years is worth less than 100 dollars today. The same is not true of welfare – what would it even mean to say that an hour of intense happiness in 10 years is worth less than the same amount of happiness today?^[6] As some philosophers have noted, it seems like a kind of prejudice to say that a person matters less because of her position in time. Furthermore, applying a discount rate for pure time preference has deeply counterintuitive implications of its own. Cowen and Parfit (1992) point out that if we apply a discount rate of 1%, then saving one life today is worth the same as saving a million lives 1400 years from now. If we apply a 10% rate, it only takes 145 years for that one life to outweigh a million saved in the future. Working backwards, it is impossible to believe that someone living in the 1870s would have been justified in saving the life of a single one of his contemporaries instead of saving 1 million people today. Or imagine, as Cowen and Parfit invite us to, that because Cleopatra had an extra helping of dessert one evening you must die of cancer on your 21^st birthday. How, they rhetorically ask, could this be justified? If it cannot be, then it is difficult to see how a discount rate on the wellbeing of future people is justified either. 

It seems unlikely that a discount rate will do what we want it to do. If we set the rate too high, it becomes wildly implausible, as we have seen in the last paragraph. If we set it too low, then it doesn’t resolve any of the counterintuitive implications of longtermism that we looked at above. The value of the future may force us, as Bostrom, Greaves, MacAskill, and other longtermists believe, to sacrifice a huge amount in the present to secure and improve the long-term future.

In the last subsection we looked at the possibility of applying a social discount rate to the interests of future people. In this section we will ask a more fundamental question – do we have any reason to create happy people? Longtermists would have us think that future people are just like us, differing only in their position in time^[7].  If that is true, then it seems like it would be sheer prejudice on our part to benefit present-day people instead of future people if we can benefit future people more than present-day ones at the same cost. Some philosophers argue however that this simple equation of present and future people is misleading, and obfuscates a morally significant difference. 

Let’s look at the word ‘benefit’ I used a couple of sentences ago. What does it mean to benefit someone? There are serious philosophical difficulties in giving an adequate account of what it means to benefit or harm someone, and several distinct explanations of what it means to do so^[8]. The most accepted account of benefitting/harming is a comparative one – we benefit someone when we make them better off than they were before our action, and we harm them when we make them worse off than they were before^[9]. This applies very naturally to presently existing people. If I break John’s nose it is clear that I have harmed him – he is worse off now than he was before I punched him. It is less clear how this comparative account applies to future people, however. How do I benefit a person by causing him to exist? I cannot compare the state she is in after I cause her to exist with the state she was in before I did so -there was no state she was in before I caused her to exist. This has led most philosophers to argue that we do not benefit or harm people by causing them to exist, or at least not in the sense that is normally considered to be morally important^[10]. 

These considerations have led some philosophers to argue that we have no moral reason to cause happy people to come into existence. Our moral reasons essentially involve the core moral concepts of harm and benefit, and if we cannot benefit people by causing them to exist, then we have no reason to do so^[11]. Narveson (1973) puts it like this – we are interested in making people happy, not in making happy people. Others such as Frick (2017, 2020) and Bramble (forthcoming) have accepted this line of reasoning. They agree that we do not benefit people by causing them to exist with good lives, nor do we make the world better by doing so. If they are correct, then it seems like our reasons to secure and improve the long-term future are far weaker than longtermists believe. 

Those who deny that we make the world better by creating happy people need to explain what is bad about human extinction. Longtermists and utilitarians have a good explanation of this – happy human lives are valuable, and, if we go extinct, the future will contain far less value than it would have had we survived. This explanation isn’t available to those who don’t accept that we make the world better by creating happy lives. One strategy, adopted by Frick, is to move the locus of value from happy human lives to humanity as a whole. Frick argues that humanity itself has intrinsic value, and, like other things that have intrinsic value such as languages, cultures, species etc., we have reason to preserve its existence. Note however that this does not entail that we have any reason to maximize the human population, as longtermists think. It is sufficient, on Frick’s view, that we ensure that humanity survives indefinitely, so long as it does so in a flourishing way. Frick may well be correct that humanity has intrinsic value, and that we ought to preserve it. It is difficult to believe, however, that humanity has intrinsic value, while human lives don’t. Furthermore, there are serious difficulties with denying that the expected happiness in a future life gives us a (defeasible) reason to create that life. For example, it makes it very difficult to explain why we are ever justified in procreating, at least if that justification must be grounded in the interests of the created person.^[12]

^[1] See Scheffler (2010, chapter1)

^[2] See Bostrom (2008) for an evocative hint of the kinds of lives that may be lived in the future. 

^[3] See Parfit (1984) and Ord (2020).

^[4] See Mitz-Woo (2021).
^[5] For philosophers who are critical of such a discount rate see Broome (1994), Caney (2014), Cowen and Parfit (1992), Greaves and MacAskill (2021), Ord (2020), Parfit (1984), and Rawls (1999). A notable exception is Mogensen (2022) who argues for the plausibility of a discount rate based on partiality.
^[6] Note that we can legitimately discount for uncertainty. If we are sure that we will get one hour of happiness today, but the future happiness is uncertain, then it makes perfect sense to discount the future happiness relative to present happiness. This is not what we mean by pure time preference – pure time preference applies even when we are certain that the future benefits (or harms) will in fact ensue.
^[7] For example, MacAskill (2022) writes “Future people, after all, are people… they matter even if they live thousands of years hence” and Ord (2020) states “People matter equally, regardless of their temporal location”.
^[8] See Feinberg (1986), Shiffrin (2012), and Purves (2019).
^[9] For simplicity I ignore situations in which my seemingly harmful action actually benefits the person in question. Such cases require a more sophisticated, counterfactual account of harm. This account is still a comparative one, however, and so everything I say about benefitting/harming future people would apply to this account too.
^[10] See McMahan (2013) on ‘existential’ harms and benefits.
^[11] At least for their own sake. We may still have reasons to create new people that are grounded in the interests of already existing people.
^[12] See Benatar (2006).

Adams, C.J., Crary, A., and Gruen, L. (2023) The Good It Promises, the Harm It Does: Critical Essays on Effective Altruism. Oxford: Oxford University Press.

Benatar, D. (2006) Better never to have been: The harm of coming into existence. Oxford: Oxford University Press.

Beckstead, N. (2013) On the overwhelming importance of shaping the far future. PhD Thesis, Rutgers, New Jersey.

Bostrom, N. (2014) Superintelligence. Oxford: Oxford University Press.

Bostrom, N.(2013) Existential risk prevention as global priority. Global Policy 4(1): 15-31.

Bostrom, N. (2008) ‘Letter From Utopia’, Available at: https://nickbostrom.com/utopia 

Bostrom, N. (2003) Astronomical waste: The opportunity cost of delayed technological development. Utilitas 15(3): 308-314.

Feinberg, J. (1986) Wrongful life and the counterfactual element in harming. Social Philosophy and Policy 4(1): 145-178.

Greaves, Hilary, and MacAskill, William (2021) ‘The case for strong longtermism’, GPI working paper available at: https://www.globalprioritiesinstitute.org/wp-content/uploads/The-Case-for-Strong-Longtermism-GPI-Working-Paper-June-2021-2-2.pdf 

MacAskill, W. (2022) What we owe the future. New York: Basic Books.

McMahan, J. (2013) Causing people to exist and saving people’s lives. Journal of Ethics. Volume 17(1/2): 5-35.

Mitz-Woo, K. (2021) A Philosopher’s Guide to Discounting. In Budolfson et al (eds) Philosophy and Climate Change. Oxford: Oxford University Press, p.90-110.

Newberry, T. (2021) “How Many Lives Does the Future Hold?” Available at: https://www.globalprioritiesinstitute.org/wp-content/uploads/Toby-Newberry_How-many-lives-does-the-future-hold.pdf 

O’Brien, G. (forthcoming) The Case for Animal-Inclusive Longtermism. Journal of Moral Philosophy.

Ord, T. (2020) The Precipice. New York: Hachette Books.

Parfit, D. (1984) Reasons and Persons. Oxford: Oxford University Press.

Scheffler, S. (2010) Equality and Tradition: Selected Essays. New York: Oxford University Press.

Schubert, S., Caviola, L., and Faber, N.S. (2019) ‘The Psychology of Existential Risk: Moral Judgments about Human Extinction’, Scientific Reports 9:15100.

Tarsney, C. (2022) ‘The epistemic challenge to longtermism’, GPI working paper available at: https://www.globalprioritiesinstitute.org/wp-content/uploads/2020/Christian_Tarsney_epistemic_challenge.pdf 

Temkin, L.S. (2012) Rethinking the Good: Moral Ideals and the Nature of Practical Reasoning. New York: Oxford University Press.

The scale of the future and the amount of value it could contain makes it many times more important than the present and near-term future. Since that future is at risk, and we can take action to reduce that risk, we should do everything we can to do so, even if this means sacrificing the welfare of people and animals in the present. 

Though the scale of the future is huge, future people do not exist yet, and we cannot benefit people by causing them to exist. Even if we could benefit people by creating them, we would be morally justified in giving preferential treatment to those who are alive today, especially since we cannot know for certain how our actions now will affect the long-term future.

It is clear that we can benefit people by creating them. For example, I am very happy that my parents created me. Existential benefits are just as important as normal, comparative benefits, and creating happy people is both good for them and good for the world.

Our moral reasons relate to the core moral concepts of harming and benefitting. To benefit someone is to make them better off that they would have been. It is essentially a comparative notion. When creating a person there is no state they would be in had we not created them, and therefore there is no basis for comparison. This means we cannot benefit people by creating them, and so we have no moral reason to do so.

I agree that we do not benefit people by creating them, and we therefore have no moral reason to create them for their own sake. However, we still have reason to create people. This is because people are instrumentally valuable. They can further the projects that are important to existing people, they give people the happiness of being parents, and because humanity itself is valuable, and the only way to preserve humanity is to create new people.