Main moments
What is long term ethics?
What are the problems of society importantly to solve? Most agree that eradicating poverty, fighting injustice, overcoming war and stopping climate change are high on the list. Some would argue that fighting industrial animal husbandry is also at the top echelon.
The longtermism ethic, or `longtermism' in English, asserts that to positively influence the distant future are among our most important tasks. In other words, we should devote much of our resources to ensuring that the lives of those who will live thousands, if not millions, of years in time are allowed to live good lives, even at the expense of the welfare of those who live today.
To some, this certainly sounds like a search -- why should we put the distant future at the forefront, and why should we think that those who have not yet been born are as important as those who actually live in the here and now? On the other hand, it is not reserved for the few to worry about the future. We have the power to wipe out humanity forever through a full-scale nuclear war and subsequent nuclear winter. We are emitting greenhouse gases that will affect life on Earth for a long time and the institutions we build today could have long-lasting consequences for human and animal welfare. Many consider it our most important task to insure ourselves against catastrophic events and to safeguard the livelihoods of future generations.
The long-term ethic has its foothold at the University of Oxford and in particular the research community at the Global Priorities Institute and, until 2024, the Future of Humanity Institute. Nick Bostrom, Nick Beckstead, William MacAskill and Hilary Greaves, much based on the pioneering work of Derek Parfit in the book Reasons and Persons (1984), has argued in several books and articles that positively influencing the distant future should be our most important prioritization.
Not all long-term ethicists are proponents of this “strong” variant of the long-term ethic. IN What We Owe the Future (2022) taking MacAskill advocated that positively changing the distant future is only one of our top priorities. In this text, we will account for and defend the weaker thesis.
The essence of long-term ethics
The remove The future is different from it close the future. The near future is the one most of us relate to. It extends from tomorrow to perhaps a few generations in time. In contrast, the distant future spans millions of years in time.
According to long-term ethics, the distant future is of great moral importance because it can accommodate very many people who may have good or bad lives. Since it is a moral task to prevent suffering and promote welfare, and since our choices today affect how good the lives of the future are, it is of great moral importance to positively influence the distant future.
The long-term ethicist denies two allegations. The first is that we cannot influence the distant future and that we can therefore ignore it. Philosopher J.C.C. Smart (1973) believe we can safely ignore the long term consequences of our actions. Namely, they are like rings in the water, which get weaker and weaker and eventually disappear. Long-term ethicists, on the other hand, believe that we have the power to influence lives thousands and perhaps millions of years into the future. In this sense, the effects of our actions can be amplified over time, rather than disappearing.
The second claim that is denied is that the problems of the present must take strict moral priority over the problems of the future. Many believe it is implausible that those living now should devote their lives to increasing the welfare of the many billions that will exist in the future. But for long-sightedness ethicists, placing less emphasis on the welfare of someone just because they are not yet born is as serious a moral shortcoming as placing less emphasis on the welfare of someone just because they live far away.
There are several arguments for long-term ethics. One argument that has a long history is what we might characterize as an argument for the civilization contract. The conservative political philosopher Edmund Burke expressed this view thus in Reflections on the Revolution in France from 1790:
[Society] is a partnership in all science; a partnership in all the arts; a partnership in every virtue and in all perfection. Since the goals of such a partnership cannot be achieved only in one generation, it becomes a partnership not only between those who live, but also between those who live, those who are dead and those who are to be born (Burke 1790, §165).
We are part of a long, unbroken line stretching back thousands of generations. Everything we see around us, from edifices to technology, builds on the cumulative work of all the generations that have come before us. Projects such as creating a better world for everyone and understanding the universe, are intergenerational. According to the Civilization Contract argument, it would be a kind of moral violation of the intergenerational social contract if we had been the last link in this chain or not done our part to facilitate future generations to continue the civilizational edifice.
Oxford philosopher Toby Ord puts it this way in his book The Precipice as of 2020:
When I think of the unbroken chain of generations that have brought us forward to our time, and all that they have built for us, I am humbled. I am overwhelmed with gratitude and shocked at the size of the legacy, and the impossibility of giving back a single small part of the service. Because a hundred billion of the people to whom I owe everything are gone forever, and because what they created is so much greater than my life, than my entire generation... If we drop the baton and succumb to an existential catastrophe, we will fail our ancestors in many ways. We will fail to achieve the dreams they hoped for; we will fail the trust they placed in us as heirs; and we will fail in any duty we had to carry on the work they did for us (Words 2020, 56—57).
However, the most important argument for long-term ethics in modern philosophy is what we might call the scale argument (Greaves and MacAskill 2021; MacAskill 2022):
- The number of future moral subjects whose welfare can be affected by our actions is astronomically much larger than those subjects that exist now.
- If our actions affect two groups of moral subjects with the same moral value, and one group is vastly much larger, then the value of the outcome of the actions is mainly determined by their effect on the much larger group.
- The value of the welfare of moral subjects does not depend on when they exist (i.e., present and future moral subjects have the same moral value).
Conclusion: The value of the outcome of our actions is mainly determined by their effects on future moral subjects.
We consider premise 1 and 3 to be the controversial premises of the argument and those that therefore deserve more attention. Premise 2 should be reasonably intuitive. It defends a limited form of aggregation of welfare across individuals. If a group with vastly more moral subjects is affected by our actions than another group, the welfare of the first group will dominate the value of the actions. Thus, it does not require a purchaser to have unlimited aggregation, known from some consequential theories. The premise, for example, is consistent with the fact that the influence on a minority should have a disproportionate significance. The premise is also intended to open up that non-welfare-based events, such as loss of biodiversity that do not affect animal or human welfare, may have a major bearing on the value of the action's outcome, but not more to say than the welfare of a vast amount of moral subjects. Finally, it is important to emphasize that deontologists may subscribe to the scale argument, but still believe that one should not act to promote that outcome with the most value possible.
Can we help future generations?
Premise 1 of the scale argument states that the number of future moral subjects whose welfare can be affected by our actions in multiple choice situations is astronomically much larger than those subjects that exist now.
The first part of this premise should be uncontroversial. The future of humanity may be very long and contain a great many moral subjects. Toby Newberry (2021) from the Global Priorities Institute presents an “extremely conservative” estimate of 10! 28 possible future lives. That's as many as the number of grains of sand on Earth multiplied by one billion.
It is more controversial to claim that we can influence the distant future. If our actions cannot affect the welfare of future lives, neither will the value of our actions be determined by the distant future. This was J.C.C. Smart's point. We can ignore the distant and possibly unforeseeable consequences of our actions, because they get weaker and weaker the further in time they go, like rings in the water.
It is an empirical claim that the effects of our actions do not extend far into time. A moment's reflection shows it to be false. About 2-300,000 years ago, a lady in East Africa became pregnant with children. This 'Mitochondrial Eva' is the direct progenitor of every single human being that exists today. That this article is written and read is a direct consequence of this woman's reproductive actions. World history would have looked very different if Eve had refrained from reproducing.
The case of Mitochondrial Eve shows that our actions can have enormous impacts far into time. But that doesn't show that we can overstate the consequences of our actions. A better argument for us not being able to help future generations is therefore that even though consequences of our actions go far forward in time, so don't the knowledge our about them that. Since we are totally oblivious to the long-term consequences, we might as well ignore them when making decisions. This is the Epistemic criticism of the long term ethics (Tarsney 2023).
Epistemic criticism has something to do with it. Mitochondrial Eve had no idea all the consequences of her actions and we are in the same situation when it comes to many of our actions. Nevertheless, there are some broad categories of actions that we can know the consequences of for the welfare of future generations, with approximate certainty. These are those categories of actions that are irreversible or have particularly persistent effects.
One of these are actions that lead to there being no future generations, in other words, those that lead to the extinction of humanity. If a meteor hit Earth tomorrow and wiped out all life, we know for sure how it will affect the well-being of future generations; it will make them not exist It will mean that none of the countless good lives that could have been lived will be.
Perhaps the most important implication of long-term ethics is therefore that we should devote more resources to reducing the likelihood of human annihilation. Existential risk is, in this way, a kind of negative externality on future generations that current generations are underinvesting in reducing. If future generations could negotiate with us, one of their strongest wishes would have been that we make sure they exist.
Another category of actions with persistent impacts are institution-creating actions. John Adams, one of the founding fathers of the United States and their second president, wrote the following:
The institutions now being created in America will not wither for thousands of years. It is therefore of utmost importance that they start correctly. If they start wrong, they will never be able to get back on track, unless it happens by chance (Adams 1851, 296).
Just like annihilation, the choice of institutions could have long-lasting consequences. Perhaps the most important institutions from a long-term perspective are those that are self-correcting, like markets, democracy, the rule of law, and open debate, in other words, the liberal institutions. Without free exchange of opinions and criticism, and without institutions that allow us to adjust course when we are dissatisfied, suboptimal or outright bad values and notions can lock in and cause negative consequences for future generations for a long time.
To make this point more concrete, let's imagine that the values and norms of the world as they were only 200 years ago were locked in for 10,000 years. We can imagine a counterfactual world where freethinkers, reformists, activists and philosophers were cracked down harder on by those in power. Then the following 10,000 years would contain more slavery, global persecution of homosexuals and other minorities and systematic oppression of women. Thus, avoiding locking in incorrect values, by maintaining and strengthening self-correcting institutions, is a priority if we think long-term.
However, we can also envisage more marginal institutional changes that will also be important. The think-tank Long Term, Where We Work, has advocated more institutional reforms that will be able to make the policy more compatible with our obligations to the generations that will come after us, as an agent of the future, to inform the Parliament of long-term threats and to strengthen the protection of posterity in the Constitution. These are just three of many possible reforms that will be able to increase the welfare of future generations.
We can therefore influence the welfare of future generations. But how much emphasis should we place on the welfare of people who will exist many thousands of years into the future? Should we even care about future lives?
Should we care about future lives?
Premise 3 asserts that present and future lives are equally worth, or more precisely, that the moral value of welfare is independent of time. In other words, we should pay no less attention to future welfare only Because it is in the future.
There may, of course, be other reasons for placing less emphasis on future welfare, such as uncertainty around whether it will actually take place. And there may be reasons to attribute less value to contributors to welfare, like capital, if it occurs in the future. This can be explained by the fact that capital has a opportunity cost. One million today is better than one million in ten years because capital has an expected return that is greater than zero. It can also be reasoned in that capital has disproportionately less effect on welfare the more we have of it already. More generally, we can say that many contributors to welfare have diminishing marginal benefit. Premise 3, however, denies pure time accounting, that is, to place less emphasis on future welfare simply because it is in the future. This is a view most philosophers, and a good deal of welfare economists, share (Lavik and Skjerve 2020). But it's also seemingly radical: if you can save a human being from death today or put the money in a savings account that will save two lives in 1000 years, should you rather do the latter?
Although the idea may have counterintuitive implications, it also has intuitive support. Imagine you are on a forest walk and smash a glass bottle on a trail (Parfit 1984, 356-357). Barely half an hour earlier, you passed some children you fear will cut themselves on the shards of glass if you don't clean up after you. In this case, it is clear that you should clean it up. But does this become any less clear if the children who are going to cut themselves are not just around the corner, but rather come running barefoot in a year? What about in ten or 100,000 years? It seems unlikely that the mere fact that the pain from the glass cut will take place in the future makes your reason for picking up the glass cut somewhat less strong. This is in line with the widespread view that we should be concerned about the consequences of climate change or nuclear waste, even though the lion's share of the negative effects of climate change and damage from nuclear waste accrue to people who have not yet been born.
One could object that this thought experiment only shows that harm is just as bad regardless of when it occurs and not that welfare has equal value regardless of time. In order for premise 3 to be valid, it must also be wrong to discount the benefits accruing to the future. But while it may seem tempting to disparage harm and welfare increases, it leads to absurd consequences. Say that, at no great cost to those of us who live here and now, we can create a giant nuclear power plant that can make energy virtually free for the entire world in the distant future, but which current generations are guaranteed not to benefit from. Unfortunately, the nuclear power plant also carries nuclear waste as a by-product, which will lead to the death of a future human being one year earlier due to radiation damage. This seems like an easy choice; we can drastically increase the welfare of all people on earth in the future, at no great cost to those of us who live today, as opposed to inflicting relatively little harm on one future individual. But if we only downgrade the value of future goods, and not harm, only the harm in the distant future has a say. The following we should not build the power plant. It shows that downgrading benefits and not downgrading harms leads to the wrong place.
That it is wrong to place less emphasis on future lives becomes even clearer if we look backwards: if we think it is right to downgrade future welfare by five per cent each year, for example, we are obliged to believe that preventing the death of one person five hundred years ago would have been as important as preventing some 39 billion deaths today. It's obviously absurd.
Nevertheless, one may object that life that exists today is more important than life that does not yet exist. As Jan Narveson (1973) put it: It is more important to make people happy than to create happy people. This view is often called a person-affecting vision. According to this view, nothing can be good or bad unless it makes something better or worse for someone (Parfit 1984). To make a living person happier will involve making that person better off. But creating a new happy person will not involve making a person better off. To say that someone is better posed presupposes that one compares someone's actual state to a non-existent state. And “the moment [one] begins to argue as from the abyss of the nonexistent, [one] expresses himself senselessly,” as is expressed in Henry Salts (1914) Famous formulation.
There are many ways to specify person-influencing views. One variant is presentism, which states that the only individuals that matter are those that exist reaching. Presentism, however, has absurd consequences. Say that you are given the opportunity to choose whether the world, at a time in the future where there are no people who exist now, will consist of ten billion people who have wonderful lives, or ten billion people who suffer. Presentism says that both worlds are equally good, because future lives have no moral value. That obviously doesn't add up.
Another view is actualism. It says that the individuals who matter are actual individuals, i.e. those who will exist given our actions. But this view, too, has problems. Say you're given the choice between creating Amanda, who wants life that's so bad it's not worth living, or Berat, who wants an even worse life. Even if both possibilities are bad it is still obvious that we should create Amanda; after all, she wants a better life than Berat. But choosing to create Amanda will be wrong for the actualist. After the choice is made, the interests of Amanda, as an actual person, will count, while the interests of Berat, as just a possible person, will not count. The same applies if one chooses Berat. Whatever one chooses, one acts wrong—since one should not create life that we know is not worth living. Actualism thus gives rise to inconsistent implications.
A sight we can call Necessary lives (necessitarianism) states that individuals with moral value are those who exist independently of our choices in a given action situation. Those who only potentially exist, as a result of us choosing one way or another, we can disregard. But even this perspective carries unattractive implications. Say we can choose between A, B, and C, where Anna exists in all three outcomes, and Bob only exists in A and B. Anna has high welfare in C, very high welfare in B and bitingly slightly higher welfare than this again in A. Bob has barely positive welfare in A and very high welfare in B. Necessary Lives says that only the welfare of those who exist regardless of our choices has moral value. Since Bob does not exist in C, we can then discount his welfare completely, and we should only focus on how our actions affect Anna. So we should choose A, where Anna's welfare is highest and Bob has barely positive welfare. But this is crazy conclusion. We should rather choose B, where Anna is a bit worse than in A, while Bob is very much better.
The person-influencing view is therefore not plausible and we should look to what is called impersonal views, which allows that creating happy beings can be a good thing. The most popular variant is called total vision, which states that the value of an outcome is the total amount of welfare in that outcome. The Supreme Court supports premise 2. It should be said that this view also has problematic implications. The total view leads to what Derek Parfit (1984, 388) calls `the repugnant conclusion', or the repulsive conclusion: that it is better for a world with very many people living lives that are barely worth living, rather than a world with fewer people where everyone lives very good lives, as long as the total amount of welfare is greater in the scenario with many people.
In 2021, dozens of leading moral philosophers and welfare economists, including Norwegian economist Geir Asheim, signed a petition in the journal Utilities who advocated that the fact that a theory implicates the repugnant conclusion does not mean that it is wrong (Zuber et al. 2021). Although the conclusion is apparently repulsive, it is, according to its defenders, less repulsive than the alternatives. Swedish philosopher Gustaf Arrhenius (2000) has proven, in a handful of impossibility theorems, that there is not a theory of the value of different-sized populations that does not violate one or more intuitively obvious principles. And although philosophy is not a popularity contest, it is worth noting that impersonal theories, which support long-term ethics, are more popular than the alternatives among those who work on the question.
Conclusion
It may seem absurd to equate the welfare of people who will live thousands of years in the future with the welfare of those living now. But as the scale argument shows, the conclusion follows from reasonable premises. If one denies that positively influencing the distant future is one of our most important tasks, one must explain which premise is incorrect or why the conclusion does not follow from the premises.
That the implications of long-term ethics seem absurd is not reason enough to deny its moral imperatives. Previous expansions of our moral circle have all seemed absurd, like that people outside the tribe have the same moral value as people within, that women are just as valuable as men, that different ethnicity has no bearing on one's moral worth and that animal interests count on par with our own (Singer 2011). However, the long-term ethics need not be as controversial as previous expansions. Caring for the aftergenus has a long history, of which the Burke quote is an example. Human beings in all societies have lived in covenant with earlier and later lineages. The long-term ethics is a systematic theory that shows how far our obligations extend.
Sources
Arrhenius, Gustaf. 2000. 'Population Axiology'. PhD Thesis at Uppsala University, 222.
Huemer, Michael. 2008. 'In Defence of Repugnance'. The Mind 117 (468): 899—933.
Words, Toby. 2020. The Precipice: Existential Risk and the Future of Humanity.
Salt, Henry S. 1914. 'Logic of the Larder'. The Humanities of Diet.
