Been an essay by Kayode Adekoya —
Economist | AI Safety Advocate
[email protected]
Submitted to Essays on Longtermism’ Competition.Centre for Effective Altruism (CEA)
Summary
This essay engages deeply with Jacob Barrett’s chapter "Discounting for the Future" from Essays on Longtermism, examining how temporal discounting affects our moral and practical responsibilities toward future generations. It argues that while moral discounting - valuing future lives less simply because they occur later in time - is indefensible, practical discounting remains a rational tool for navigating uncertainty, risk, and opportunity cost.
Building on Barrett’s analysis, the essay develops a two-track framework that reconciles ethical impartiality with pragmatic realism. The first track asserts the equal moral worth of all persons across time; the second guides decision-making under empirical constraints. Applied to domains such as climate policy and AI safety, this model clarifies how longtermist principles can shape effective and just action.
Ultimately, the essay defends a longtermism that is both principled and practicable - one that rejects temporal bias while recognizing the limits of foresight. The future’s moral claim upon us is equal; our task is to honor it wisely.
Introduction
Discounting valuing future benefits or harms less than those occurring in the present is a foundational concept in both economics and policymaking. Yet when viewed through the moral lens of longtermism, discounting becomes far more than a technical tool; it becomes a question of intergenerational justice and ethical consistency. If, as longtermism suggests, future lives matter morally just as much as present ones, then any practice that systematically devalues their welfare calls for deep scrutiny.
In Essays on Longtermism, Jacob Barrett’s chapter "Discounting for the Future" examines this tension. He distinguishes between moral discounting whether the welfare of future people should matter less morally and practical discounting how policymakers might rationally use a positive discount rate to account for opportunity costs, uncertainty, and feasibility. Barrett’s analysis exposes the core dilemma: how can we act responsibly toward the future without becoming paralyzed by unrealistic demands in the present?
This essay argues that moral reasons for discounting the welfare of future persons are weak to nonexistent. Ethically, we should treat the well-being of future generations as comparable to our own. However, in the real world, uncertainty, opportunity costs, and political constraints justify a bounded, transparent form of practical discounting. I call this approach bounded practicality a decision framework that preserves moral impartiality across time while ensuring policy remains tractable and effective.
1. The Conceptual Divide: Moral vs. Practical Discounting
Barrett’s central contribution is to separate what is often conflated: moral and practical discounting. Moral discounting addresses whether the welfare of future persons is intrinsically less valuable than that of present persons. Practical discounting concerns the instrumental reasons such as uncertainty, risk, or opportunity cost that may lead decision-makers to apply a positive discount rate when evaluating policies over time.
This distinction matters because conflating the two leads to moral confusion. A policymaker might use a discount rate of 3% to reflect opportunity costs or uncertainty, but that does not mean she believes future lives are worth 3% less per year. Barrett’s framework therefore enables moral clarity: we can uphold moral equality across generations while still making rational, evidence-based trade-offs under uncertainty.
For longtermism, this distinction is pivotal. If moral discounting were justified, the case for prioritizing existential risk reduction or safeguarding civilizational flourishing would weaken dramatically. But if it is unjustified as I argue then the moral case for protecting the far future remains strong, even if our practical policies must operate under bounded rationality.
2. The Moral Case Against Temporal Discounting
Should we care less about a life lived 200 years from now than one lived today? Many find the idea intuitively appealing; proximity feels morally relevant. Yet most ethical theories and rational reflection reject this temporal bias.
2.1 Temporal Impartiality and Equal Moral Worth
Hilary Greaves, in her essays on the value of the future, argues for temporal impartiality: that the value of an individual’s welfare does not depend on when they exist. To discount purely because someone lives later rather than now is as arbitrary as discriminating based on location or ethnicity. Derek Parfit called this the “arbitrariness of time”: temporal distance alone cannot justify moral indifference.
Longtermism rests on this moral equality. The well-being of a person living in the 25th century is not inherently less valuable than yours or mine. If anything, given the sheer number of possible future people, the moral stakes of our present actions multiply enormously.
2.2 Uncertainty Is Not a Moral Reason
Some justify discounting by appealing to uncertainty: the future is unknowable, so we should weigh it less. But uncertainty is an epistemic challenge, not a moral principle. We do not morally discount the lives of those in distant countries simply because we know less about them; likewise, we should not devalue future people simply because we are unsure how our actions will affect them. As Barrett notes, uncertainty calls for probabilistic reasoning, not moral neglect.
2.3 Diminishing Returns and Opportunity Costs
Economists sometimes defend discounting on the basis that future generations will be richer, so the marginal value of additional welfare may be lower. But this is an empirical claim about expected wealth, not a moral argument. Diminishing returns can guide efficient resource allocation, but they cannot justify treating a future person’s happiness as intrinsically less valuable.
In short, none of the usual arguments for moral discounting hold. The default ethical stance temporal impartiality remains the only consistent position for a longtermist framework.
3. Why Practical Discounting Still Matters
If moral discounting fails, why retain any form of positive discounting in policy? Because moral ideals must operate within a non-ideal world. Policymakers face uncertainty, limited resources, and political constraints. Practical discounting acknowledges these realities without abandoning moral commitments.
3.1 Opportunity Costs and Investment Logic
A dollar invested today can grow and yield greater welfare in the future. Economically, this justifies applying a modest positive discount rate to compare policies. However, longtermist policy requires differentiating between investment efficiency and moral worth. A low practical rate say 1% rather than 5% may capture opportunity costs while keeping long-term benefits ethically salient.
3.2 Deep Uncertainty and Decision Fragility
The further we look into the future, the less certain we become about consequences. This epistemic fog does not erase responsibility but complicates prediction. Practical discounting, combined with sensitivity analysis and scenario modeling, helps avoid paralysis. As Nick Bostrom notes in "The Vulnerable World Hypothesis," the appropriate response to uncertainty is not inaction but resilience preparing for high-stakes scenarios even when precise probabilities are unknown.
3.3 Institutional and Political Feasibility
Democracies operate under short electoral cycles; institutions face immediate accountability pressures. Demanding absolute zero discounting may produce policies too idealistic to implement. A bounded, explicit discount rate offers a workable compromise: morally low enough to keep future welfare in view, yet pragmatic enough to be adopted and sustained.
Thus, practical discounting becomes a tool of translation bridging moral aspiration and policy execution.
4. The Two-Track Decision Architecture
To reconcile moral impartiality with practical constraints, I propose a two-track decision architecture for longtermist policymaking.
Track A: Moral Guideline (Ideal Reasoning)
Treat all persons, present and future, as morally comparable.
Evaluate interventions using a near-zero moral discount rate, especially for existential and catastrophic risks.
Prioritize actions with long-term leverage those that shape civilization’s trajectory, such as AI safety, biosecurity, or institutional reform.
Track B: Practical Evaluation (Policy Reasoning)
Apply a low but positive practical discount rate (e.g., 1-2%) for cost-benefit analyses under uncertainty and opportunity costs.
Perform zero-discount sensitivity tests: if setting the rate to zero changes the priority ranking of interventions, the project deserves special consideration.
Use robustness checks Monte Carlo simulations, worst-case analyses, and expert elicitation to test the stability of long-term value estimates.
Combine short- and long-term projects in diversified portfolios, ensuring resilience and political feasibility.
This two-track approach maintains moral clarity while acknowledging the constraints of policymaking. It ensures that existential risks remain top priority while routine development decisions remain economically coherent.
5. Policy Illustrations
5.1 Climate Change
Climate change is the textbook case for discounting debates. High discount rates dramatically reduce the apparent value of mitigating long-term damages, while low or zero rates justify aggressive mitigation. The two-track approach offers balance: catastrophic climate scenarios such as runaway warming or tipping points are treated under zero discounting, reflecting their existential stakes. More marginal projects, like short-term infrastructure investments, may use bounded practical rates to reflect opportunity costs. This ensures both urgency and realism in climate policy.
5.2 Artificial Intelligence Governance
In AI safety, expected value reasoning under zero discounting yields staggering moral importance: even a small probability of alignment failure implies enormous expected loss of future value. Yet research funding and policy operate under resource limits and uncertainty. A bounded practicality model prioritizes AI risk reduction as a moral imperative (Track A) while integrating it into feasible, incremental governance strategies (Track B) including regulation, interpretability research, and institutional resilience.
These examples show how the proposed framework operationalizes longtermism: neither utopian idealism nor short-term pragmatism, but a principled synthesis.
6. Addressing Objections
Objection 1: Zero discounting demands unreasonable sacrifices today.
Moral zero discounting need not entail reckless austerity. The two-track model distinguishes ideal moral valuation from practical policy application. Individuals and governments can still weigh opportunity costs and uncertainty while upholding future welfare as morally significant. The approach calls for intelligent prioritization, not absolute asceticism.
Objection 2: Positive discounting still privileges the present.
A low, bounded practical rate is not moral privileging but procedural humility. It acknowledges that our predictive and institutional capacities are limited. Moreover, by keeping the rate explicit and justified, policymakers can adjust it as knowledge improves. Transparency is key: the discount rate should be a parameter open to moral review, not a hidden bias.
Objection 3: We are too uncertain about the far future to act meaningfully.
As Thorstad’s essay on "Cluelessness" notes, deep uncertainty challenges but does not paralyze moral reasoning. Expected value frameworks, robust decision theory, and precautionary principles provide tools for rational action under uncertainty. Uncertainty should motivate risk reduction, not complacency.
Objection 4: Discounting ignores justice and distribution.
Longtermist discounting must integrate distributive justice. Projects that yield large aggregate benefits but exacerbate present inequality should be scrutinized. The bounded practicality framework pairs discounting with equity lenses, ensuring that protecting the future does not come at the cost of harming the vulnerable today.
7. The Broader Longtermist Perspective
Longtermism invites us to view humanity as a long-lived project our actions today can influence not just decades but millennia. Discounting, therefore, becomes a moral crossroads: either a convenient excuse for short-termism or a carefully calibrated tool for moral realism.
By rejecting moral discounting and adopting bounded practicality, we align with the longtermist ethos articulated by Hilary Greaves and William MacAskill: that the most important moral projects of our time may be those that safeguard the far future. Barrett’s analysis provides the structure; longtermism provides the motivation.
This synthesis echoes John Rawls’s "just savings principle"each generation owes the next the conditions for justice and prosperity. Extending this principle across deep time, we recognize that preventing existential catastrophe is not charity toward hypothetical people; it is a duty to sustain the moral community of all who can exist.
Conclusion
Jacob Barrett’s "Discounting for the Future" illuminates one of the deepest tensions in longtermist ethics: reconciling moral impartiality with practical governance. The correct response is not to abandon discounting altogether, nor to treat it as a moral license for myopia, but to reform it anchoring it in moral equality while calibrating it for real-world constraints.
Bounded Practicality the two track model proposed here embeds this balance. It preserves the moral insight that future lives matter just as much as ours, while enabling rational, transparent, and feasible policy choices. It aligns with the Effective Altruism principle of doing the most good we can, not the most good imaginable.
In longtermism, the moral horizon is vast, but our tools must remain usable. Between zero discounting as a moral ideal and low discounting as a practical rule lies a path that honors both the future and the present a path of reasoned stewardship for the long arc of humanity.
References
Barrett, Jacob.Essays on Longtermism: Present Action for the Distant Future edited by Hilary Greaves, Jacob Barrett, and David Thorstad. Oxford University Press, 2024.
Bostrom, Nick. “Existential Risk Prevention as Global Priority.” Global Policy 4, no. 1 (2013): 15–31.
Greaves, Hilary, and William MacAskill. “The Case for Strong Longtermism.” Global Priorities Institute Working Paper, University of Oxford, 2021.
Parfit, Derek. Reasons and Persons. Oxford University Press, 1984.
Rawls, John. A Theory of Justice. Harvard University Press, 1971.
Thorstad, David. “Cluelessness and Longtermism.” In Essays on Longtermism, 2024.
