What Is the Non-Identity Problem in Climate Ethics?

The non-identity problem is Derek Parfit's observation that the choices which shape climate policy also determine which people will ever exist, so a policy that leaves the future worse often harms no determinate person. This strains any account of climate obligation that rests on identifiable victims. Imagine a fourteen-year-old girl who decides to have a child now. Had she waited until she was twenty-eight, the child would have had a better life. The child who exists has a life worth living, and yet it is worse than the alternative. Who has been wronged? Not the child, Parfit argued, because a later pregnancy would have produced a different individual, drawn from a different combination of gametes at a different moment. The child cannot say, "You made me worse off than I could have been," since no version of this particular person existed to be better off. ## The Victim Who Cannot Be Named Most moral accounts of harm are comparative. An act wrongs someone when it leaves them worse off than they would otherwise have been. Carbon-intensive policy, read through this lens, appears to wrong nobody in particular. A regime that accepts a warmer, poorer world in 2150 is not a regime that makes the actual people of 2150 worse off than they would have been under a different regime, because the different regime would have produced different people entirely. Read the next headline about projected losses in 2100 and ask a plain question: who, specifically, is the victim? The question has no obvious answer, and that is the problem. ## Two Readings, One Puzzle The puzzle sharpens when the two standard frames are set side by side: | Question | Person-affecting reading | Impersonal reading | |---|---|---| | Who is wronged by delay? | A specific future person worse off than in the alternative | No one in particular; the outcome is simply worse | | Test of wrongness | Did this act make someone worse off than they would otherwise be? | Does this outcome contain less total or average welfare? | | Different-people outcomes | Usually no comparison available | Compared directly | | Verdict on policy delay | Often no wrong detectable | Wrong, because the aggregate is lower | The impersonal column rescues the verdict but at a price: it condemns acts without any person who has a grievance about them. The person-affecting column preserves the grievance but cannot register the wrong that most climate analysts take to be central. ## Three Exits and Their Costs Philosophers have tried to close the gap in different ways. Gregory Kavka's 1982 "paradox of future individuals" showed how readily intuitive principles collapse once future people are included. Parfit's own response, developed in *Reasons and Persons*, proposed a wider person-affecting principle, which ranks outcomes by the welfare of people who exist in them, and then tried to show that its verdicts converge with impersonal ones in most cases. Elizabeth Harman has argued that we can harm someone even in cases where, strictly, no alternative version of that person was possible, and that the apparent gap is narrower than Parfit assumed. Each exit buys coherence by spending something else: intuitive grip, or the clean separation between harm and outcome. > The harm climate change does is real, but it falls on people who exist only because of how we act. Our moral vocabulary keeps asking for a face, and the problem keeps declining to supply one. ## What Survives the Puzzle The non-identity problem does not show that we owe the future nothing. It shows that the grammar of victimhood is an incomplete grammar for climate obligation. Obligations can still be grounded in the wrongness of acts, in impersonal goods such as a livable biosphere, or in the rights of persons who will exist under any plausible policy. Stephen Gardiner's account of a "perfect moral storm" makes a related point: the difficulty lies partly in the structure of the problem, not only in our reluctance to act. Readers weighing these options should treat the puzzle as a diagnostic. Whichever framework one adopts, it must say what it thinks is wrong with a policy that harms no one in particular, and why that answer should persuade a reader who has been scrolling past the same numbers for years.

If you found this valuable, consider supporting our work.