TL;DR A policy can help determine which future people exist. In a genuine nonidentity case, a person living under the worse policy might never have existed under the alternative. That complicates a claim that the choice made that person worse off. It leaves room for moral reasons based on the conditions institutions create, the claims of people who will live under them, and comparisons between possible futures.
A council has a choice about extending a city. One plan builds quickly on a low plain and leaves future residents with a recurring problem of transport interruptions during floods. A second plan moves the neighborhood to a higher corridor and changes when homes and schools open. Both plans support lives worth living in this hypothetical. The second leaves people with more reliable access to ordinary services. The decision initially looks like a familiar comparison of better and worse consequences for the people who will live there.
Take it forty years forward. A resident named Jay lives in the low neighborhood built under the first plan. Jay has meaningful relationships and projects and also spends considerable time organizing work around disrupted transport. The planning office could reasonably have provided a more reliable arrangement. Imagine, however, that under the other plan Jay's parents would have met at a different time, their decision about when to have a child would have changed, and Jay as this particular person would never have existed. A different person might have enjoyed the more reliable city.
The example stipulates facts for philosophical analysis. Actual councils usually cannot determine precisely which people would exist under a different policy. Neither plan assigns anyone a lower worth, and a life that contains burdens can be valuable. The comparison concerns avoidable conditions and the identity of the people who encounter them.
Gregory Kavka examined paradoxes about choices that affect which future individuals come into existence. Derek Parfit developed the nonidentity problem in Reasons and Persons, including questions about policies whose effects reach people whose existence depends on those policies. The puzzle below uses an original municipal example to separate a comparative harm claim from several other ways of reasoning about the decision. Kavka, The Paradox of Future Individuals and Parfit, The Non-Identity Problem
Begin with a resident already alive
A child named Esme currently lives near the proposed construction. Under the low-plain plan the tram to school will be interrupted several times a year. In the other plan it remains dependable. Hold constant the facts relevant to the comparison and suppose Esme exists in both situations, relies on the tram, and is harmed by delays. The new scheme can make Esme worse off than Esme would have been under the other choice.
This is an ordinary same-person comparison. It identifies one individual and compares possible conditions for that person. It may need evidence about the delay and its consequences. The counterfactual existence problem is absent. If an alternative were authorized and would avoid the burden without comparable costs, the existence of the same resident across the comparison gives a familiar reason to favor it.
A choice can contain many same-person comparisons despite also affecting the identity of some future inhabitants. Current residents can face construction risks. People who exist before the plans diverge could later relocate to different parts of the city. Even farther ahead, some people might exist in both cases and experience the consequences differently. The nonidentity question concerns the people whose existence genuinely varies across the alternatives. It should not dilute concrete claims the council can already identify.
The control varies a second assumption. Suppose either street plan leads to Jay's existence, and the alternative would give Jay more reliable access. The sentence about leaving Jay worse off becomes straightforward in the same way as the sentence about Esme. Whether a particular resident exists in both scenarios changes the structure of a harm claim even if the relevant condition sounds identical. The council can distinguish this issue from a dispute about how serious the disruption is.
Trace the first changes that reach an identity
In the low-plain timeline Jay's parents meet while joining the original construction project. Their work and life follow the history described by the case. In the higher-corridor timeline the construction happens several years later, their work follows another path, and the conditions leading to Jay's conception never occur. A different set of residents comes to live in the higher neighborhood. Merely moving Jay's record to a different building would erase the premise of the example.
The dependence does not require the council to intend to choose future people. A decision can change schedules, places, and relationships as side effects. A change in when a child is conceived can change which particular individual comes into existence in the type of case Parfit considers. The simplified timeline fixes that relation so one can ask about it without needing a prediction about an actual family.
Jay can value existing and have a legitimate objection to a burdensome living condition. Thankfulness for a life does not entail approval of every decision in its causal history. Still, the available alternative at the time of choosing did not contain Jay with a better tram. Under a narrow comparative account in which harming a person means making them worse off than they would otherwise be, the most immediate comparison is unavailable.
The other obvious comparison is with never existing. The example specifies that Jay's life is worth living. It also leaves open complex questions about whether and how nonexistence can be compared with someone's well-being. None of that establishes that the low-plain policy was justified. The council needs a reason whose application does not depend on inventing an alternate version of Jay to live in a different history.
There is some discipline in what the two timelines leave out. Both plans are feasible, have the same relevant costs for the exercise, and produce the same number of future residents. The higher plan avoids the recurring disruption without a compensating downside under the setup. Population size and large competing benefits would bring additional dilemmas. Holding those factors fixed helps locate the concern rather than settling every choice about future people in one argument.
The objection written forty years later
Imagine Jay drafts an objection to the council. The draft says that the officials left residents with a foreseeable and avoidable limitation on access. It points to budgets and survey results that were available before the choice was made, and it asks why the planners accepted that limitation. The complaint is intelligible even if Jay also says, with full knowledge of the setup, that a different decision would have led to a different person living there.
An official responds that the city cannot say Jay would have been better off if it chose differently. That may be true under the narrow comparison. The reply answers a question about Jay's counterfactual welfare and leaves the original question about the institutional choice. A rule for justifying that choice might consider whether it was appropriate to establish a neighborhood with serious predictable weaknesses when a feasible alternative avoided them.
One account could protect the claim of whoever occupies a role in the future. Designers who will supply homes for later residents can owe whoever comes to live there reasonable conditions. This is a conditional obligation stated at the time of choosing, with the actual inhabitants later able to identify a breach once they exist. Its force comes from the conditions of building and administering a place for people. It does not depend entirely on evidence that the same complainant would have lived in the alternative.
There is a challenge for such an account. If the condition is too demanding, nearly every imperfect project leaves someone with a justified complaint no matter what options were feasible. If the standard is too weak, any life worth living may be treated as acceptable grounds to impose avoidable burdens. The view needs a defensible account of foreseeability, alternatives, duties, and the interests it protects. The thought experiment makes those requirements visible without selecting a universal threshold.
It can also distinguish suffering a burden from being wronged. A resident can suffer from a design despite an institution taking every reasonable measure in hard circumstances. A different resident may bear a foreseeable burden the institution had a duty and reasonable ability to avoid. Evaluating the conduct requires evidence about the earlier choice as well as the later experience. The population identity issue is one part of that judgment.
The planner who compares whole possible outcomes
Another member of the planning office says the higher plan produces a better future because the same number of people can enjoy the relevant benefits while avoiding the recurring disruption. This is an impersonal comparison in the sense that it evaluates the possible outcomes without requiring each benefited resident to have existed in the other one. The premise makes the contrast relatively strong. It gives both projects comparable costs, preserves other goods, and supplies a significant difference in a condition of life.
The account has its own questions. How should the evaluator compare different numbers of people? How should a modest benefit to many be weighed against a severe burden to a few? What if one plan raises well-being overall while imposing substantial inequality or violating a protected claim? A successful comparison in this example can be part of an ethical view without resolving all population ethics.
The argument becomes harder if the higher plan displaces a community that already exists. The council then has evidence of an ordinary claim from people harmed by their loss of a place. A forecaster should include that fact and revisit the apparently dominant alternative. Nonidentity in the more distant timeline cannot serve as a general permission to set aside comparison for existing people. The thought experiment deliberately separates the nearby burden to Esme so this connection remains visible.
A council can also use different levels of explanation for public reasoning. It can note that one option improves all relevant forecast conditions for inhabitants who would otherwise face chronic obstacles, with no identified balancing cost under the scenario. It can state whether the comparison relies on equal population sizes. If its decision depends on the total, the average, or fairness among future occupants, those theories deserve open discussion. Choosing an aggregation rule because it gives the desired answer is a poor substitute for that reasoning.
Neither planner needs to rank Jay and the possible other person as better or more deserving individuals. What counts in the argument is the avoidable problem in one world and the better conditions in the other. The identity difference explains why the comparative language about one future person's welfare is difficult. It leaves intact the value of either person's life if they exist.
A constraint a later public official can still uphold
A third account emphasizes rights, minimum conditions, or reasonable rejection. A city might commit to designing municipal access without an avoidable severe burden on whoever will live in a new neighborhood. This is different from a claim to make anybody exist. It concerns the terms on which an institution arranges the environment people will inherit. A prospective condition can apply without listing eventual names.
Suppose the worse plan leads to a life generally worth living while leaving Jay without a reliable route to a public service. If the council interprets the value of Jay's life as permission to ignore any constraint, it makes the constraints for future residents much weaker than those for established ones. The beneficiary of a sound rule can be whoever later bears the condition. The ability to identify them in advance is unnecessary for some classes of duty.
A person-affecting approach might try to preserve an account based on the interests of actual or possible people while rejecting the narrow worsening requirement. A resident might be harmed in a noncomparative sense by experiencing a serious avoidable setback. A person might also be wronged by conduct that sets back an important interest even where the alternative would have excluded their presence. Different theories of harm and wronging will treat these possibilities differently. The municipality case should neither collapse them into one view nor assume they fail because one counterfactual is unavailable.
Adding a minimum can also isolate a different dilemma. Imagine both plans meet every justified threshold and one provides a significant further improvement at the same cost. The specific rights-based objection to the worse plan may be weaker, yet the impersonal comparison still has force. If only the better plan achieves an acceptable minimum, the reasons may converge. The group should be able to say when it relies on a constraint and when it invokes a benefit beyond that threshold.
The institution can formulate the commitment in a way that remains usable across contact with new evidence. It names which service conditions need protection, whose decisions could undermine them, and when a review is required. It should preserve obligations that can be discharged after construction as well as questions appropriate at the planning stage. Repairs can help the people later living there even though the earlier counterfactual involved different people. A later institution can imagine a better future for actual residents from the present date.
The next vote occurs after the change in identity has already happened
Split the project into two genuine decisions. First the council chooses the site and sets the construction in motion. The first branch still affects who later exists. Thirty years later, before selecting a maintenance plan, a later board can choose between two ways of serving people who are already in the new district. Jay exists when the later options open. One maintenance plan gives Jay a much more reliable connection with no stated tradeoff in the example.
A subsequent official cannot automatically say that Jay's existence depended on every aspect of the arrangement. The existence dependence is relative to the alternatives under consideration at the earlier time. The maintenance choice presents new alternatives with Jay present in both. When evidence supports a worsening claim at the second vote, its application is compatible with identity dependence at the first. Separating the dates prevents a conceptual problem about one choice from excusing a different omission.
Now imagine that a planner defends the first vote by saying that selecting the higher corridor would have deprived Jay of a life. This proposes a further inference. The case gives no preexisting Jay whose opportunity to live was taken away by the choice. Nor does the existence and life's value of Jay under one plan by itself create an obligation to realize that plan instead of another. Such a rule would need to explain how to choose among mutually exclusive possibilities, each of which could contain valuable lives.
The same reasoning applies in the other direction. A description of the people who might come to occupy a different district gives no license to dismiss a known burden already placed on existing residents. A planner can distinguish positively valuing future lives from claiming that each possible person had to be brought about. That distinction helps keep the policy focused on the institution's actual duties toward people affected and the quality of the possible arrangements.
A later board can also learn from earlier uncertainty. Suppose the disruption was thought rare before opening, and years of observations show it interrupts ordinary life far more frequently. The evidence supports reassessment even if the original decision was reasonable given what was known. An ethical record can distinguish accountability for foreseeable choices from an obligation to respond once a real problem becomes clear. Applying one narrow counterfactual comparison to both would miss the difference between them.
One practical implication of the two-vote case is to preserve realistic future choices. The first city plan can leave room for an alternate connection, supply funding authority when a disruption crosses an observed threshold, and keep residents able to report unanticipated effects. Those measures do not settle questions about which people a plan brings into the world. They help institutions remain answerable to the actual people who inherit it.
The case where identities cannot be traced cleanly
In a real policy analysis, two feasible plans could plausibly affect who moves to an area, which families form, and which descendants exist. The official might have no reliable individual identity branch. They will have projections with different levels of uncertainty about access, cost, displacement, and use of public services. Those choices should remain available for evaluation. The absence of a credible map of every future birth is not a reason to invent one.
An evidence record can separate effects on identified existing people from projected conditions for future occupants. The future analysis may ask how robust an operational promise is across a range of population scenarios. If different futures expose different people to a severe burden, test options for reducing that burden without trying to decide which future registry should be treated as the one composed of real claimants.
There could be uncertainty about whether avoidable risks are present. A cheaper plan might be adequate if independent infrastructure arrives and inadequate if it never does. The council can say which assumptions make the plan acceptable, who can verify them, and whether an approval can be conditioned on durable provision. A high uncertainty can support a staged decision when the opportunity is real. It does not need to be mapped to a specific person's unobservable counterfactual life.
A planner should also consider whether an option can preserve choices for whoever arrives. Reconstruction budgets, space for future access improvements, and predictable management records can let later institutions repair errors. This has practical importance independent of any uncertainty about identity. People living with an earlier choice should not have to rely on precisely the infrastructure assumptions given to decision makers before they were born.
The paragraph that belongs in the council's explanation
The narrow comparative statement about Jay cannot do all the work. A stronger explanation describes the foreseeable condition, identifies the feasible alternative, and states why the selected rule is appropriate for the people any completed project will serve. It acknowledges specific costs to people already living and does not erase them with a forecast about the distant future.
Forty years later Jay receives a funded connection to a more reliable service and can take part in deciding how it runs. That remedy concerns the people now present. It is coherent with the claim that the past city decision affected who came to be there. One observation settles whose condition can be improved today. The other shows why the earlier decision called for reasons beyond a same-person comparison.
The nonidentity problem matters because language about a person being worse off is familiar and often appropriate. When the policy changes which person reaches the future, the language no longer settles the case. What remains explainable is the actual burden, the quality of the alternative, and the duty an institution has to the people who will live with its choice.