Longtermist Political Philosophy: An Agenda for Future Research

Andreas T. Schmidt and Jacob Barrett examine institutional longtermism: when evaluating institutions, people should often give significant weight to their expected long-term effects. The chapter asks whether this position has a stronger foundation than longtermism about individual action, and whether familiar political values constrain or support it. Its arguments establish a research agenda rather than a settled ranking of political systems.^[1]

Institutions change what the argument demands

Institutions include laws, policies, norms and conventions, although the chapter concentrates on formal institutions. The definition gives the distant future significant weight without necessarily making it the overriding priority. It also leaves room for contexts where other considerations prevail. This is important because arguing that future people matter does not itself show that every decision should maximize expected future welfare.^[1]

The authors give several reasons to focus on institutions. They can act at scale, address many different problems, outlast particular decision-makers and alter the rules under which people act. Institutions can also coordinate action where isolated individuals have little influence. This may reduce some objections about individual demandingness or tiny chances of enormous benefits, without removing disagreement about future value or predictive uncertainty.^[1]

They distinguish concern over the next few thousand years from versions extending to millions or billions of years. The latter require more controversial assumptions about survival, population and our ability to affect enduring outcomes. A less blueprint-driven proposal is to make institutions better at learning and improving over time, rather than pretending to predict an ideal distant society in detail (§2).^[1]

Political values can pull in both directions

The chapter tests apparent conflicts with five values: justice, equality, freedom, legitimacy and democracy. A duty owed to a particular person can resist sacrificing that person for aggregate benefits. Yet duties to future people, or an undirected duty to promote just institutions, might themselves support long-term concern. The authors identify unresolved population-ethical questions: adding more people cannot simply be assumed to make a society more just (§3.2).^[1]

Freedom creates a similar tension. Restricting present choices in the name of distant benefits can threaten liberty. But concern for future people’s options and freedom from domination could motivate protecting them against irreversible constraints. Different theories of freedom, equality and obligation therefore change the argument; these values cannot be reduced to a single welfare score (§§3.3–3.4).^[1]

Legitimacy asks what justifies exercising political power over people. The chapter considers whether very high stakes could justify exceptional powers, but also warns that a perpetual emergency risks authoritarian capture and lasting value lock-in. It leaves open how a state could justify its decisions to people who cannot yet consent. Giving the future moral weight does not automatically authorize a present government to speak for it (§3.5).^[1]

Democracy and the future constituency

Democracy raises both a procedural and an empirical question. Existing voters may reject longtermist priorities, while future people have no vote. Whether they should be represented depends partly on who belongs to the democratic constituency: everyone affected by a decision, everyone subject to it, or some other account. The authors discuss objections to extending either principle across generations rather than treating future representation as already resolved.^[1]

They also distinguish democracy’s intrinsic value from its instrumental benefits. Elections can encourage short horizons, but democratic learning, feedback and resistance to capture may help protect the future. Uncertainty about long-run policy might favor experts if expertise is decisive, or diverse deliberation if nobody is reliably competent. The chapter presents these as conditional arguments, not a finding that a particular reform or regime maximizes long-term welfare (§3.6).^[1]

A research agenda for collective choices

The conclusion asks about international coordination, political duties and representation of future generations. Coordination could help address AI safety, climate or pandemics, while centralizing power could increase totalitarian risk. It also calls for engagement beyond contemporary analytic philosophy. These are questions for further investigation; the chapter’s large-population scenarios and cited institutional correlations are not independently established forecasts or causal evaluations.^[1]

The published chapter appears in Essays on Longtermism, with Oxford’s online publication dated August 18, 2025. This account uses an author-hosted copy of that published chapter; earlier manuscripts are separate bibliographic stages. For Social Alignment, it identifies a problem beyond an operator’s immediate preferences: whose interests institutions represent, which values constrain them, and how their authority remains open to revision. The Challenge of Value Alignment: from Fairer Algorithms to AI Safety examines a related problem of plural values and legitimate collective choice.^[1]