A case for medium-termism

There has been a lot of discussion of ‘longtermism’ recently—in TIME, The New York Times, and social media. Some people think longtermism is great. Others think it’s dangerous. I have concerns about how it is likely to be misused too. But mostly, I just think it’s philosophically mistaken.

The idea that we should care more about the long term sounds great, inspiring even—and I think it’s true, at least with a bit of moderation. But, on that note, I think something a lot less sexy sounding—medium-termism—is more defensible. More defensible, ironically enough, from a long-term perspective!

Longtermism gives moral priority to the long-term future. One major idea is that the number that people that might exist in the future could be vastly greater than the number of people living now. We might colonize other planets, or even create vast numbers of digital people in simulations who could live happy lives—happier lives, perhaps, than us.

So, the thought is, the future matters more from moral point-of-view than the present or near-future (i.e. the medium-term). Morally speaking, every person matters equally. But there are vastly more possible future people than people who live today or have lived in the past. So, the future should matter to us more than the present, the short-term, or even the medium term.

Never mind that digital people probably aren’t possible (consciousness appears to be fundamentally analog, not digital). But setting this aside, to see what I think is wrong with longtermism, imagine applying an analogous line of reasoning to your own life as an individual.

I’m 45. I might live to 110 years of age. I might have far more years in front of me than I have lived already. So, I should care far more about the long-term—for it’s possible that I’ll live to 110 and have so many more years to live.

But, wait a minute, how likely is that? It’s very likely that I’ll live to be 46, a bit less likely that I’ll live to be 50, a bit less likely still that I’ll live to be 60, a lot less likely that I’ll live to be 90. What about 110? Not impossible, but spectacularly improbable.

This is why we don’t plan our lives on the assumption that we’ll live be 110. We don’t prioritize the long term over everything else—because it’s so uncertain. When we are ten years old, we don’t plan to live to 110. We instead focus, if we’re prudent, on the medium-term—on taking steps to ensure that we are reasonably successful and secure at 35.

Then when we’re 35? Again, if we’re prudent, we plan once again for the medium-term: we try to save enough so that we can retire at 65 and support ourselves until our late 70’s (the likely length of our life). Etc.

It’s more prudent to live like this—to focus on the medium-term rather than the long-term—for two reasons.

First, to get to the long-term at all, you actually have to make it through the present and medium term. You won’t get to retire comfortably at 65 if you screw up your life at 35. So, when you’re 20, you should try to figure out how to get to 35 in good shape. Then, once you get to 35, you should try to get to 50, etc. So, you should plan for the medium term.

Second, the medium term is far more likely than the long-term. When we’re 10, we don’t know if we’re going to live until we’re 40. Chances are pretty good, but then again, we might not. So, taking into account likelihoods, one should try to enjoy the present and plan for the medium term. Then, if things work out (if we live to 40), we do it again: we plan to live to 65, and if things keep working out, we live a nice long life. So again, you should plan for the medium term.

The future of humanity hardly seems different. We, the human race, might not make it to 2050. Chances seem pretty decent that we will. But then again, it looks increasingly like nuclear war is a possibility and the world might be too hot by 2050 to make colonizing other planets or creating vast numbers of digital beings much more likely than zero.

The human race may or may not get to 2050, then, but the chances still seems pretty good. So, we should plan for it ensure that it happens. Then, if we get to 2050, we should try to get to 2100—etc.

But years 21,000 or 2,100,000AD? That humanity makes it to either them seems spectacularly unlikely from where we are now. So, we shouldn’t plan for it. If we’re prudent, we should try to make it through the present in one piece, plan for the medium term so that we make it there (i.e. 2050, 2100, 2150, etc.), and if and when we make it there, do the same again—so that we actually get to the long-term.

We shouldn’t prioritize the long-term, then. We shouldn’t be ‘longtermists.’ We should be medium-termists—for, if we get the present and medium term right, over and over again, then chances are as good that we’ll make it longer term…while not discounting the present and medium term!

That is, by focusing on the present and medium term, we have the best chance of living well in the present, medium, and long-term all together, treating them all as equally important.

This, I think, is how prudent people live. A prudent person doesn’t obsess over the distant and uncertain future. They recognize that the distant future is so uncertain that it can’t be planned for. So they try to live a good life as a whole—by living in the present, planning for medium term, and (with a little bit of luck) doing those things well so that it’s as likely that they live and flourish in the longer-term too.

So, if you’re a long-termist, sorry, you should be a medium-termist. Getting the medium-term right is the most prudent way to ensure that we have a long-term future to begin with. The medium-term is what we have in front of us. It is what stands between us and the long-term—a long-term that may or may not have vast numbers of people.

So, it—not the long-term—is what we should focus on and aim to get right. If we do, then with a little luck (and luck is a big part of long-term life, like it or not), then the long term will turn out as good as we can rationally plan for, given the limitations of our knowledge of it.

Comments

[Not a longtermist]

You define longtermism as the position that "the future matters more from moral point-of-view than the present or near-future (i.e. the medium-term)". Your argument against longtermism is that "It’s more prudent to live like this---to focus on the medium-term rather than the long-term".

These two claims are not incompatible. One is about morality and the other prudence. And for these two to be in tension, there needs to be a pretty controversial bridge principle to the effect that what matters morally is what is most prudent.

[Marcus Arvan]

@Not a longtermist: I don't think you're right about prudence. If you have a longtermist moral view, then you should want to advance it prudently rather than recklessly.

Here's another way of putting it. Suppose you think morality requires maximizing long-term good. In that case, to do *that* effectively, you to deliberate prudently about it--that is, in terms of what is the most likely to actually achieve that. If you behave recklessly in trying to advance the long-term good, you make it likely that you will fail.

This is why prudence is morally good. Virtually every moral philosophy (including Kantianism) takes prudence to a moral virtue. Indeed, it has been often called the keystone of all virtues for this reason. *Whatever* your moral aims are, prudence in seeking their realization is better than imprudence/reckless.

Which brings to your point: "These two claims are not incompatible." Indeed, this is my point: *longtermists* should be medium-termists. They shouldn't focus on the long-term because the best means of achieving the best results long-term is to focus on the medium-term.

I am suggesting, in other words, that longtermists should adopt a 'self-effacing' approach to their own moral values--that they should realize that medium-termism is the best practical approach to achieving long-term good and avoid catastrophe.

[Marcus Arvan]

On further thought, rather than saying that longtermism 'is mistaken', perhaps I should have simply said this: longtermists should be medium-termists!

[Redundant]

How new and original is this idea? This is the first time I heard of “longtermism” used as a philosophical principle/idea. Much of what I learned about “longtermism” in philosophy was called “inter-generational obligations” which already has a rich literature.

See. Obligations to Future Generations by R. I. Sikora & Brian Barry (1978) and Derek Parfit, “On Doing Best for Our Children”, in Ethics And Population (Michael D. Bayles ed., 1976).

I guess I’ll give him credit for putting a label on that position. But the arguments aren’t new. One of the problems with “longtermism” is that, even from a utilitarian perspective, it’s incoherent because it would require *each* generation to make (large) sacrifices for future generations resulting in a “jammed-up” tomorrow where the benefits of present sacrifices will never be realized and each generation will be worst off than they would have otherwise been.

Politically, it’s based on a distributive conception of justice and not a procedural one. Iris Marion Young provides an in-depth critique of such a model. See her work Justice and the Politics of Difference.

[Marcus Arvan]

Redundant: good worries about longtermism's coherence. I agree!

However, on Young and the distributive model, I recently had a paper accepted at a good journal arguing that Rawls's just society would eliminate all of Young's 'five faces of oppression' and that Rawlsian nonideal theory is a good approach to dismantling them, too--all in distributive terms. So, Young should embrace rather than reject the distributive model.

I should be posting the paper relatively soon, and hope you find it of interest!

[Redundant]

I will have to read it. But the general worry about the distributive model is that it can be reductionist and atomistic disregarding the relations *between* people. It also seems ineffective when we extend the distribution of material goods to non-material goods such as rights, opportunities, and self-respect. Non-material goods matter as well. I’m not entirely against the distributive model, but we should be aware of its limits and drawbacks, lest we end up repeating and reinforcing them. I'm looking forward to knowing what you’ll have to say about it in your upcoming paper.

[Marcus Arvan]

@Redundant: I don’t think any of those things are true. I think they have become widespread preconceptions about the distributive model, but the paper argues that they are not ultimately justified. It’s a very long paper (close to 80 pages) and it tries to address these things in detail, so I’m hopeful you’ll find it convincing. I should be posting it here and on PhilPapers sometime in the next several weeks, and I’d be curious to see what you think! :)

[Redundant]

I suppose it could be a pervasive stereotype about what distributive justice was, is, and does. I’m all for allowing such a model to be more useful in justice regarding pluralistic and diverse aspects of politics, people’s lives, and existence in general; pushing its boundaries and utmost potentiality. But when and where it is used unjustifiably narrowly, it warrants criticism. I do have high expectations of your paper if it is 80 pages though! ;)