Recent work in experimental moral psychology has uncovered a robust and unsettling pattern: subjects systematically assign diminished moral weight to persons separated from them by temporal distance, even when stipulated welfare consequences are held constant. Caviola and colleagues, building on earlier findings by Greene and Sinnott-Armstrong, demonstrate that this temporal discounting persists across cultures, education levels, and explicit philosophical commitments.

The phenomenon raises a question that economists long treated as settled and philosophers long treated as marginal. If a child suffering in 2200 matters less to our moral calculus than a child suffering today, is this a defensible feature of moral cognition or an evolved heuristic masquerading as principle?

The stakes are not merely theoretical. Integrated assessment models of climate policy, longtermist resource allocation, and pandemic preparedness budgets all encode discount rates—often implicitly—that translate directly into policy weights on the unborn. Whether those weights reflect rational temporal preference or cognitive bias dressed in equations determines what we owe to people who cannot yet vote, protest, or appear in our visual cortex.

The experimental record on temporal discounting in moral judgment is now substantial enough to warrant careful taxonomic treatment. Three distinct effects must be distinguished: pure rate-of-time preference, identity-based discounting of non-existent persons, and probability-adjusted discounting reflecting epistemic uncertainty. Studies by Wade-Benzoni and colleagues isolate the first by holding probability and identity constant, finding that subjects still allocate roughly half the moral weight to identical harms occurring a century hence.

Neuroimaging work complicates the picture. McClure's intertemporal choice paradigms, when adapted to third-party welfare decisions, show differential activation in the ventromedial prefrontal cortex and anterior insula depending on temporal framing. The dual-process signature suggests that future persons trigger weaker affective engagement, consistent with Greene's broader thesis that emotional salience drives much of what philosophers have construed as principled moral reasoning.

Crucially, the effect dissolves under reflection in certain conditions. When subjects are prompted to consider the symmetry of their position—that they themselves were once future persons relative to their ancestors—discounting rates drop significantly. This reflective equilibrium effect mirrors findings in the trolley literature: System 2 deliberation often corrects the parochial outputs of System 1.

Cross-cultural replications by Jaroslawska and colleagues find that the magnitude of discounting varies, but its directional bias does not. Even societies with strong ancestral and intergenerational frameworks—where lineage thinking might predict reverse discounting—exhibit reduced weight on the distant future. This invariance suggests the bias is not purely cultural but plausibly rooted in cognitive architecture optimized for ancestral fitness landscapes.

What experimental philosophy contributes here is not merely cataloging the bias but mapping its cognitive signature. Discounting tracks construal-level psychology: abstractness, hypothetical framing, and reduced perceptual vividness all predict steeper rates. Future persons are not so much rejected as never fully represented in the first place.

Takeaway

Temporal discounting of moral weight appears to be less a considered principle than a perceptual failure—future persons fade not because we judge them less, but because we model them less.

Once the empirical pattern is established, the philosophical question sharpens: does any defensible normative framework license treating future welfare as discounted welfare? Derek Parfit's analysis in Reasons and Persons remains the most rigorous attempt to canvass the options, and his verdict is largely negative. The arguments standardly offered—uncertainty, declining marginal utility under growth, special obligations to contemporaries—either reduce to other considerations or fail to justify the rates observed empirically.

Consider the uncertainty argument. It is true that we know less about future persons' preferences, identities, and circumstances. But uncertainty is symmetrical: future persons may be worse off or better off, and the expected value calculation does not warrant a pure rate-of-time preference. Epistemic discounting and temporal discounting are distinct, and conflating them, as much policy analysis does, is a categorical error.

The growth-based justification, central to Stern-Nordhaus debates, holds that future persons will be wealthier and thus marginal welfare gains matter less. This is a substantive empirical claim, not a temporal principle, and it cuts against discounting in domains—biodiversity, climate stability, existential risk—where future persons may face radically worse baseline conditions than ours.

Special obligations arguments fare little better under scrutiny. The intuition that we owe more to contemporaries than to strangers across time mirrors structurally the intuition that we owe more to compatriots than to foreigners. Both intuitions survive only insofar as we accept that moral weight properly tracks relational proximity—a view increasingly difficult to defend once one notices its convergence with parochial biases that ethics has historically worked to overcome.

What remains, then, is something closer to agent-relative permission: perhaps we are permitted, but not required, to weight nearer interests more heavily. Even this concession, however, cannot justify the magnitudes observed in policy discounting, which compound to render distant catastrophes negligible. The gap between defensible permission and observed practice is itself the philosophical scandal.

Takeaway

The standard justifications for temporal discounting collapse into either disguised empirical claims or repackaged parochial intuitions—neither of which carries the normative weight that policy practice silently assumes.

Climate policy provides the cleanest test case for these abstractions because the discount rate enters integrated assessment models as an explicit parameter. The difference between Nordhaus's 4.3% and Stern's 1.4% rates translates to order-of-magnitude differences in optimal carbon pricing—and downstream, to whether centuries-distant catastrophes register as comparable to modest near-term costs. The choice of rate is, in effect, a moral judgment dressed as a technical assumption.

Experimental findings on intergenerational discounting bear directly on how this debate should be conducted. If observed discounting rates reflect cognitive bias rather than reasoned preference, then aggregating revealed preferences into social discount rates—the standard welfarist move—imports the bias into policy. We get the climate policy our evolved heuristics endorse, not the one our reflective judgments would.

Tyler Cowen and Derek Parfit independently argued that zero pure rate of time preference is the only defensible position. The empirical literature now provides a complementary argument: insofar as positive rates are psychologically natural, they are also psychologically suspect. The naturalness of an intuition is, after Greene, evidence of its evolutionary rather than its rational pedigree.

This does not mean policy should ignore intuition entirely. Democratic legitimacy requires that policy track what citizens actually endorse, and citizens discount. But the experimental finding that discounting diminishes under reflection suggests a procedural remedy: deliberative institutions designed to engage System 2 reasoning—citizens' assemblies, longtermist impact statements, formal future-generations commissioners as adopted in Wales—may produce judgments closer to what reasoned moral cognition would yield.

The deeper implication for climate ethics is that the burden of proof inverts. Rather than asking why we should care about distant future persons, the empirically informed question becomes why our cognitive equipment so reliably fails to care, and what institutional scaffolding can compensate.

Takeaway

Climate discount rates are not neutral technical parameters but the operationalization of a moral psychology we have good reason to distrust—and good reason to design around.

The convergence of experimental moral psychology and normative ethics on the question of future generations produces a striking result: the bias is robust, its justifications are weak, and its policy consequences are enormous. This is not the typical experimental philosophy finding, where intuitions complicate theory. Here intuitions are the problem theory must address.

For researchers, the agenda is clear: refine the taxonomy of temporal discounting effects, identify the conditions under which reflection corrects them, and trace the neural substrates that distinguish defensible from indefensible discounting. For institutional designers, the task is to build deliberative architectures that approximate what unbiased moral cognition would conclude.

What we owe the future may ultimately depend less on what we philosophers argue than on whether we can construct cognitive and political environments in which future persons become perceptually real. The moral weight is there. The question is whether our institutions can hold it.