Intergenerational Equity, Present Bias, and the Architecture of Long-Term Choice

Most of the highest-stakes decisions a society makes pay off, or impose their costs, over horizons longer than a single lifetime. Greenhouse-gas emissions today shape the climate of the twenty-second century; the depletion of fish stocks, aquifers, and biodiversity transfers scarcity to people not yet born; public debt and unfunded pension promises move resources between generations. Intergenerational equity is the principle that the interests of future people deserve moral consideration, and the practical question is how much (Solow, 1974). The most widely cited articulation is the Brundtland Commission's definition of sustainable development as development that "meets the needs of the present without compromising the ability of future generations to meet their own needs" (WCED, 1987).

Standard welfare economics turns this principle into a number - the social discount rate - and behavioural economics asks why, even where societies profess to value the future, their behaviour so reliably falls short. This paper joins the two: it locates the normative target in the discount-rate debate, explains the behavioural shortfall through present bias and the psychology of the distant self, and then asks how far opt-in versus opt-out policy design can move behaviour toward the target.

Formally, the weight placed on future welfare is summarised by the Ramsey rule, which decomposes the social discount rate into a rate of pure time preference and a term reflecting expected growth and the diminishing marginal utility of consumption (Ramsey, 1928). The two best-known applications to climate change reach opposite conclusions almost entirely because of the first term. The Stern Review sets the rate of pure time preference close to zero - roughly 0.1 percent - on the ethical ground that the mere date of a person's birth is no reason to value their welfare less, yielding an overall discount rate near 1.4 percent and a strong case for aggressive, immediate mitigation (Stern, 2007). Nordhaus, by contrast, argues that the discount rate should be inferred from observed market returns and savings behaviour, producing a rate of roughly 4 to 5 percent and a more gradual policy ramp (Nordhaus, 2007).

The gap between these positions is ethical rather than technical: it turns on whether discounting future generations for time alone is, in Ramsey's words, "ethically indefensible," or whether it is the appropriate reflection of how real societies trade present for future consumption. A third position has gained ground in between. Because the far-future discount rate is itself uncertain, and uncertain rates average out in favour of the lowest plausible value, the effective social discount rate should decline as the horizon lengthens - "gamma discounting" (Weitzman, 2001). An expert panel has since endorsed declining discount rates that give progressively more weight to distant generations, and the United Kingdom and France already apply such schedules to long-horizon public projects (Arrow et al., 2013).

Whatever rate a society endorses in principle, individuals do not discount the future at a constant exponential rate in practice. The behavioural literature finds present bias: a steep preference for immediate over slightly delayed rewards that flattens out for choices entirely in the future, well captured by hyperbolic and quasi-hyperbolic (beta-delta) discounting (Laibson, 1997; Frederick, Loewenstein & O'Donoghue, 2002). Present bias generates time inconsistency - plans to save, exercise, or cut emissions "next period" that are perpetually postponed - and so produces chronic under-saving and under-investment in long-term goods (O'Donoghue & Rabin, 1999).

A complementary psychological mechanism connects the private and the intergenerational case directly. People relate to their own distant future self much as they relate to another person, and those who feel less continuity with that future self save less (Parfit, 1984; Bartels & Urminsky, 2011). Interventions that make the future self vivid - for example, showing savers age-progressed images of themselves - measurably increase the willingness to save for retirement (Hershfield et al., 2011). If even the future self is treated as a stranger, the still more abstract welfare of future generations is discounted more steeply still, which is precisely why intergenerational problems are behaviourally hard rather than merely politically contested.

The distinctive feature of intergenerational equity is that one of the parties is absent: future generations cannot reward, punish, bargain with, or vote against the present. Experimental work captures this with the Intergenerational Goods Game, in which a sequence of "generations" each decides how much of a shared resource to extract and how much to pass on (Hauser, Rand, Peysakhovich & Nowak, 2014). When each individual decides alone, the resource is almost always destroyed, and the failure is driven by a minority of heavy extractors who override the restraint of the cooperative majority. When extraction is decided by binding democratic vote, however, the resource is reliably sustained - both because the cooperative majority can restrain defectors and because conditional cooperators are reassured that their sacrifice will not be wasted (Hauser et al., 2014). Consistent with the experiment, the authors find a robust cross-country association between democracy and energy sustainability.

The lesson is that sustaining the future is less a problem of average preferences than of institutions: because the beneficiaries are not present to enforce reciprocity, the present generation must build mechanisms - binding rules, enforceable votes, defaults - that commit it on their behalf.

Saving is the concrete act by which resources are moved forward in time. At the individual level it is retirement provision; at the collective level it is the accumulation of physical, human, and natural capital, and the abatement of long-lived pollution. Present bias implies that all of these are under-supplied relative to people's own stated long-run goals, and the empirical record on retirement preparedness bears this out (Benartzi & Thaler, 2007). Public pension systems formalise the transfer as an explicit intergenerational contract: pay-as-you-go schemes tax today's workers to support today's retirees on the promise that the next cohort will do the same, which makes their solvency directly an equity question between generations. Generational accounting was developed precisely to measure the fiscal burden each cohort passes to its successors (Auerbach, Gokhale & Kotlikoff, 1991). The behavioural and the equity framings thus converge: a society that under-saves shortchanges both its own future selves and the generations that inherit its capital stock and its liabilities.

If present bias and inertia keep people from acting on their long-term intentions, then the default - what happens when a person does nothing - becomes decisive. The cleanest demonstration is organ donation: countries with an opt-out ("presumed consent") default register vastly higher effective consent than otherwise similar countries with an opt-in default, with much of the difference attributable to the default itself rather than to underlying attitudes (Johnson & Goldstein, 2003).

In retirement saving, the same lever is automatic enrolment. When a large employer switched from requiring employees to opt in to its 401(k) plan to enrolling them automatically with the option to opt out, participation rose sharply (Madrian & Shea, 2001). The Save More Tomorrow programme refined the idea to address present bias head-on: employees commit in advance to allocating part of future pay raises to saving, so that contributions rise without any felt loss in take-home pay; participants' saving rates rose from around 3.5 percent to roughly 13.6 percent over about four years (Thaler & Benartzi, 2004). Evidence from Denmark shows why such automatic contributions are so effective: most people are "passive savers" who do not respond to price-based incentives such as tax subsidies, so directly raising the default contribution does far more to increase total saving than subsidies do (Chetty, Friedman, Leth-Petersen, Nielsen & Olsen, 2014; Madrian, 2014).

The most direct application to intergenerational sustainability is the green-energy default. When the renewable tariff is made the standard option that customers must actively opt out of, uptake of green electricity rises severalfold and, in large field studies, the great majority of households and firms remain on the green default with effects that persist for years (Pichert & Katsikopoulos, 2008; Ebeling & Lotz, 2015; Sunstein & Reisch, 2014). Here choice architecture acts on behalf of future generations almost by construction, lowering emissions through the inertia of the present generation rather than against it.

Defaults are powerful but not a panacea, and three qualifications matter for the intergenerational case. First, defaults anchor: under automatic enrolment most participants stayed at the low default contribution rate and the default fund, so a poorly chosen default can lock in inadequate saving even as it raises participation (Madrian & Shea, 2001). This is exactly why escalation designs such as Save More Tomorrow are needed, and why the choice of the default value is itself a consequential policy decision. Second, effects often shrink when interventions move from academic trials to population scale, so estimates from small studies should not be extrapolated uncritically to national policy (DellaVigna & Linos, 2022). Third, the distributional incidence varies: automatic contributions tend to help the inattentive and the liquidity-constrained, whereas subsidy-based approaches disproportionately reward those already saving (Chetty et al., 2014).

There is also a normative debate that intergenerational stakes sharpen rather than settle. Opt-out defaults are the leading example of "libertarian paternalism": they steer behaviour while preserving the freedom to choose otherwise, which their proponents argue is both effective and respectful of autonomy (Sunstein & Thaler, 2003). Critics counter that engineering outcomes through inertia can substitute the planner's judgment for the individual's and complicates the welfare analysis when the "true" preference is unclear (Bernheim & Taubinsky, 2018). For purely intergenerational problems this debate runs into a hard limit: future generations cannot be enrolled, defaulted, or consulted at all. Every default that affects them is necessarily chosen for them by the present generation, so choice architecture cannot sidestep the discount-rate ethics of Section 2 - it can only implement whatever weight the present generation has decided to give the future.

Behavioural economics does not replace the discount-rate debate; it situates it. The Stern-Nordhaus disagreement, and the emerging case for declining rates, set the normative target for how much the present should sacrifice for the future (Stern, 2007; Nordhaus, 2007; Arrow et al., 2013). Present bias and weak future self-continuity explain why behaviour falls short of any such target (Laibson, 1997; Hershfield et al., 2011). The cooperation experiments show that, because the future is unrepresented, closing the gap requires binding institutions rather than goodwill alone (Hauser et al., 2014). And opt-out defaults - in pensions, organ donation, and above all green energy - are a demonstrably effective way to enlist inertia in the service of the long term, provided their defaults are well chosen and their limits at scale are respected (Madrian & Shea, 2001; Thaler & Benartzi, 2004; Ebeling & Lotz, 2015). The throughline from the earlier papers in this series holds: the same imperfect, present-biased cognition that attenuates belief updating and complicates trust also governs how generously one generation provides for the next - and good institutional design, not the assumption of far-sighted rationality, is what makes intergenerational equity achievable.