Sunday, November 17, 2019

A Puzzle About Future People

We ought to do something about climate change. One reason for thinking this is that if we continue the way we’re going, we will make future generations much worse off, and so we owe it to them to e.g. enact policies that will dramatically reduce greenhouse gas emissions. But such policies would have all sorts of other effects. Factories would be moved, rebuilt, or shut down. People would find jobs as regulators, inspectors, and lobbyists that they wouldn’t have otherwise. People would drive different cars, get different degrees, and buy different products. And all of these things would affect decisions about when, where, and with whom to procreate.
This generates a puzzle (the Nonidentity Problem). To see why, let’s zoom in a bit. Suppose that, if we do not implement new emissions policies, Baby X will be born to a couple in Sacramento a year from now. However, if we do adopt new policies, the couple will instead move to Washington, D.C. for new jobs with the EPA. They’ll agree to wait a few years to settle into these new jobs before having children, and they will have a different child, Baby Y, in 2023.
If we do not adopt the new emissions policies, who will be made worse off? Baby Y will not be made worse off, since Baby Y will never exist. The couple will raise Baby X as an only child in Sacramento, and Baby Y will never be around to suffer whatever consequences result from continued emissions at current levels. Nor will Baby X be made worse off by not adopting emissions restrictions, since, if we did adopt them, Baby X would never exist. On the (perhaps optimistic) assumption that Baby X’s life will not be so bad as to be worse than never existing at all, Baby X will be no worse off under our current policies than under the new ones.
The point generalizes. There are the people, the Xs, who will exist if we do not clean up our act, and the people, the Ys, who will exist if we do. Given the fragility of the circumstances under which a particular person is conceived, they are mostly not the same people. So, whom do we make worse off if we refuse to change our current policies? Not the Xs, since they will only exist if we maintain the status quo. And not the Ys, since they will never exist to suffer the consequences of our refusal to change.
We ought to do something about climate change. Maybe we even owe it to future generations to do something. But not because they will be worse off if we don’t. So, why?

Brandon Carey
Philosophy Department
Sacramento State


  1. Muzak and potatoes for everyone!

  2. Brandon,

    You say "There are the people, the Xs, who will exist if we do not clean up our act, and the people, the Ys, who will exist if we do." But it doesn't seem to me that this is an exhaustive specification of the people who will and will not exist. There are also people who will exist whether we clean up our act or not. And this group seems to me to be quite a bit more substantial than the ones you focus on. So, can't the answer to your question be that we should do it for the people who will be born either way, since they will be worse off if we don't?

    Perhaps another way out of the puzzle is to claim that it doesn't matter who specifically will be made worse off. Rather, what matters is whether, as a result of our action or inaction, future populations will in fact be worse off than previous populations. It is not so much that we owe future populations anything. It's just that, ceteris paribus, a future population that is suffering less is morally preferable to a future population that is suffering more.

    1. Hi Randy,

      Those are both things we can say, but I’m skeptical that either accurately accounts for our obligations. Let’s call the folks who will exist regardless of the choice we make in a situation like this ‘the Zs’. The Zs will be worse off under some decisions than others, so we can say that e.g. we are obligated to do something about climate change because failing to do so would make the Zs worse off. For metaphysical reasons, I think the Zs are a very small chunk of the total set of future people, but that’s not especially important to the puzzle because our obligations are not just determined by what happens to the Zs. If they were, we would have no obligation not to do something that benefits the Zs through the exploitation of a bunch of Ys that will only exist because of our decision. But we do have that obligation. The reason for this obligation can’t be that it makes some future person worse off than they would be otherwise, so what is the reason?

      I don’t think it can be (only) that future populations, regardless of their composition, ought not be made to suffer more than previous populations, in part because I find it plausible that we are obligated to do more than the bare minimum necessary to not increase suffering. But also, more fundamentally, I just don’t know how to fill out the ‘ceteris paribus’ condition of your last sentence. What are the morally relevant factors that we ought to hold fixed when comparing the suffering possible future populations? An answer to that question would be pretty informative about our obligations to future people.

  3. This is a fun puzzle, Brandon. It's the kind of thing that gives rise to suggestions like ones Randy makes, which leads to Parfit devising a new hypothetical to show their limitations. Rinse, repeat. Ahh, it's the stuff of a great philosophy seminar.

    I wonder about how much it really matters, though. I thought most people just say, "yeah this is really kind of hard, but we can just assume that there is a 'theory X' that solves the non-identity problem in a way that avoids the repugnant conclusion." We're still left with the practical problem of what we should do now about future generations (whoever populates them).

    In that case, I want to return to something I said in the comments section of my previous post. I mentioned William Nordhaus's DICE model.

    The main idea is to estimate the social cost of carbon. The costs of its production and use aren’t fully reflected in market prices and so we should impose a market-correcting Pigouvian tax on carbon. The Dynamic Integrated Climate and Economy (DICE) model calibrates the optimal carbon tax relative to a baseline of a significant delay in doing something about warming.

    The results he published compared many policy approaches. I’ll summarize three he presents on p. 89 of his 2008 book, A Question of Balance:

    50-year delay: incur $1.55 trillion in abatement costs to generate $3.69 trillion in reduced damages from emissions.

    Optimal carbon tax: incur $2.16 trillion in abatement costs to generate $5.23 trillion in reduced damages from emissions.

    Limit temp. to 1.5C (the UN ICPP proposal): incur $27.03 trillion in abatement costs to generate $12.60 in reduced damages from emissions.

    Making the world about $14 trillion poorer isn’t a good way to make more people’s lives better, including those coming along in the future.

    1. Thanks, Kyle. Maybe it doesn’t really matter, except that it’s fun, and fun really matters. Here’s my best defense of its relevance to the question of what we ought to do, though:

      If the reasoning in the puzzle is sound, it entails that our obligations are not determined solely by facts of the form ‘so-and-so will be worse off if we do A instead of B’. So, there are other facts that matter in determining what we ought to do. One approach to solving the puzzle is to say what those facts are. A successful solution of this kind will tell us at least one kind of fact that matters in answering the question of what we ought to do, and so the puzzle matters in virtue of prompting solutions that tell us about important kinds of facts.

      If the facts that mattered were only about total welfare, then we would be stuck with the repugnant conclusion, which suggests that global economic consequences are insufficient to determine our obligations to future people. So, it might be that we ought to make the world $14 trillion poorer, even if it is a bad way of making people’s lives better. It depends on the other relevant facts, whatever they are.

      (Also, this is Brandon. For some reason, Blogger is not recognizing my Google account.)

  4. My stock reply to the non-identity problem: it doesn't matter. The fact that there is no specific person that is made worse off is irrelevant; I've made the world a worse place in some way, and that's enough to condemn me. I am the only definite person necessary. If I knowingly do something that makes the world a worse place I'm a bad person, even if there is no definite person who is made worse off.