Skip to main content

Policymaking for posterity


Policymaking for posterity involves current decisions with distant consequences. Contrary to conventional prescriptions, we conclude that the greater wealth of future generations may strengthen the case for preserving environmental amenities; lower discount rates should be applied to the far future, and special effort should be made to avoid actions that impose costs on future generations. Posterity brings great uncertainties. Even massive losses, such as human extinction, however, do not merit infinite negative utility. Given learning, greater uncertainties about damages could increase or decrease the optimal level of current mitigation activities. Policies for posterity should anticipate effects on: alternative investments, both public and private; the actions of other nations; and the behaviors of future generations. Such effects may surprise. This analysis blends traditional public finance and behavioral economics with a number of hypothetical choice problems.

This is a preview of subscription content, access via your institution.


  1. If the income elasticity of demand for the amenity exceeds the absolute value of the elasticity of marginal utility with respect to consumption, this approach would be an underestimate of how to value an amenity to a future generation, and vice versa. This formulation is simplifying and assuming a generation lives but one period.

  2. With a 1% discount rate, delaying a loss by 300 years cuts its value to 7.7% of its original value.

  3. To be sure, merely assuming that there is a small probability that the world will end for some reason over which we have no control, for example a devastating war, can lead to effective declining weights as generations roll forward, but we are using this device to rescue the analysis.

  4. The original version of the problem has a switch that can shuttle the trolley to an alternate track, where one person will be killed. This variant was proposed by Thomson (1985); she also notes that more individuals are willing to throw the switch, which seems to represent much lesser involvement than pushing someone.

  5. Weitzman (1999, p. 29) recommends a discount rate of between 3% to 4% for 25 years, 2% for the next 50 years, 1% up to year 300, and then 0%. Gollier (2008) recommends a discount rate drifting down from 3.5% for the near term to 1% for a millennium hence. While these numbers may seem large, they are relevant in light of the EPA’s 10,000 year analysis of Yucca Mountain nuclear waste storage, which was recently exceeded by the newly announced 1 million year time horizon (see the introduction to this issue.)

  6. This begs the question of how to define a cataclysm. Few observers think global warming has much potential to wipe out human life forever, for example.

  7. Both sides also claim irreversibilities as a possible source of support. Gollier et al. (2000) observes that scientific progress tends to enhance the efficiency of actions taken in the future, which tends to promote the learn-then-act principle, which favors an initially smaller amount of emissions reductions. Gollier and Treich (2003) address when and when not to delay action from a quite different perspective, namely the risk aversion properties of the utility function as represented by its second, third, and fourth derivatives. We look at the attractiveness of the learn-then-act approach in our quantitative model below, though the potential motivation for waiting within it is quite different from those just mentioned.

  8. The structure of the model can matter as well, but we do not capitalize on this possibility.

  9. Normally it is assumed that the government itself should not be risk averse given its ability to diversify, even across generations, but given the potential costs associated with climate change (e.g., curbing emissions, damages, adaptation, and geoengineering), risk aversion would seem merited.

  10. If risk aversion is represented by a power function, its presence in effect raises the value of q, parallel to the way risk aversion on costs in effect raised the value of a.

  11. Stochastic shocks to the system, say a generation might bounce to poor and have to sell the furniture would complicate the problem, and would make it much harder to maintain the precedent.

  12. Their total altruism toward a generation incorporates both its felicities, e.g., benefits from consumption, and its altruistic concerns. The authors consider backward as well as forward altruism (Zeckhauser and Fels 1968).

  13. We reiterate the concern that having no diminution in weighting through time, as might come from an independent survival risk, implies that a small benefit going to all future generations has infinite value.

  14. As we observed in the introduction, any standard discounting approach using traditional discount rates would give far-future generations virtually no weight.

  15. We do not focus attention on any single generation, but think of the far future as a stream of future generations.

  16. The NIH 2008 budget was $29.5 billion. 2% of GNP would produce roughly ten times that amount. Presumably an increase in expenditures would take place predominantly in the for-profit sector, say achieved through some tax incentive, but still it would take a few years to ratchet up spending significantly without gross waste.


  • Arrow, K. (1999). Discounting, morality, and gaming. In P. R. Portney, & J. P. Weyant (Eds.), Discounting and intergenerational equity. Washington, DC: Resources for the Future.

    Google Scholar 

  • Arrow, K., Solow, R., Portney, P. R., Leamer, E. E., Radner, R., Schuman, H. (1993). Report of the NOAA Panel on Contingent Valuation.”

  • Cowen, T. (2008). Social discount rate. The New Palgrave Dictionary of Economics Online, 2nd ed.

  • Cropper, M., Aydede, S. K., & Portney, P. R. (1994). Preferences for life saving programs: How the public discounts time and age. Journal of Risk and Uncertainty, 8(3), 243–246.

    Article  Google Scholar 

  • Feldstein, M. (1964). The social time preference discount rate in cost benefit analysis. Economic Journal, 74(294), 360–379.

    Article  Google Scholar 

  • Foot, P. (1967). The problem of abortion and the doctrine of double effect. Oxford Review, 5, 5–15. Reprinted in Bonnie Steinbock and Alastair Norcross, eds. Killing and letting die, 2nd ed. (New York: Fordham University Press, 1994), pp. 266–279.

  • Frederick, S., Loewenstein, G., & O’Donoghue, T. (2002). Time discounting and time preference: A critical review. Journal of Economic Literature, 40, 351–401.

    Article  Google Scholar 

  • Gollier, C. (2008). Discounting with fat-tailed economic growth. Journal of Risk and Uncertainty, 37 (2/3).

  • Gollier, C., & Treich, N. (2003). Decision-making under scientific uncertainty: The economics of the precautionary principle. Journal of Risk and Uncertainty, 27, 77–103.

    Article  Google Scholar 

  • Gollier, C., Jullien, B., & Treich, N. (2000). Scientific progress and irreversibility: An economic interpretation of the ‘precautionary principle’. Journal of Public Economics, 75, 229–253.

    Article  Google Scholar 

  • Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291.

    Article  Google Scholar 

  • Lichtenstein, S., Slovic, P., Fischhoff, B., Layman, M., & Combs, B. (1978). Judged frequency of lethal events. Journal of Experimental Psychology, 4(6), 551–578.

    Google Scholar 

  • Nordhaus, W. D. (2007). A review of the Stern Review on the economics of climate change. Journal of Economic Literature, 45(3), 686–702.

    Article  Google Scholar 

  • Phelps, E. S., & Pollak, R. (1968). On second-best national saving and game—Equilibrium growth. Review of Economic Studies, 35(2), 185–199.

    Article  Google Scholar 

  • Portney, P.R., and Weyant, J.P. (Eds.) (1999). Discounting and intergenerational equity. Washington, DC: Resources for the Future.

  • Schelling, T. (1957). Bargaining, communication, and limited war. Conflict Resolution, 1(1), 19–36.

    Article  Google Scholar 

  • Stern, N. (2007). The economics of climate change: The Stern Review. Cambridge, UK: Cambridge University Press.

    Google Scholar 

  • Thaler, R. H., & Benartzi, S. (2001). “Save more tomorrow: Using behavioral economics to increase employee saving.”

  • Thomson, J. J. (1985). The trolley problem. Yale Law Journal, 94(6), 1395–1415.

    Article  Google Scholar 

  • Weitzman, M. L. (1999). Just keep discounting, but…. In P. R. Portney, & J. P. Weyant (Eds.), Discounting and intergenerational equity. Washington, DC: Resources for the Future.

    Google Scholar 

  • Weitzman, M. L. (2007). A review of the Stern Review on the economics of climate change. Journal of Economic Literature, 45(3), 703–724.

    Article  Google Scholar 

  • Weitzman, M. L. (2008a). Some dynamic economic consequences of the sensitivity inference dilemma. Mimeo: Harvard University.

    Google Scholar 

  • Weitzman, M. L. (2008b). On modeling and interpreting the economics of catastrophic climate change. Mimeo: Harvard University.

    Google Scholar 

  • Wigley, T. M. L. (2006). A combined mitigation/geoengineering approach to climate stabilization. Science (314), 20 October 2006, 452–454.

  • Zeckhauser, R., & Fels, S. (1968). Discounting for proximity with perfect and total altruism. Discussion Paper Number 50, Harvard Institute of Economic Research.

Download references


We thank Guan Yang for his research assistance with the literature, and Nils Wernerfelt for both research assistance and the computation of the results of our uncertainty and learning model. Kip Viscusi provided helpful comments.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Richard Zeckhauser.



1.1 Topographical renderings of first-period reductions and the effect of second-period damages

In both Figs. 1 and 2 on the graphs on the right, as a increases, the value of R 1 first increases, then decreases. Below is another graph for q = 2, showing R 1 along the vertical axis as a function of values for H and a, given that b = 0, γ = 3, and q = 2. Figure 3, consistent with the right side of Fig. 2, shows that R 1 is maximized for some value of a around 4. R 1 decreases as a shrinks to 2 or rises to infinity. The graph is qualitatively similar for q = 1.

Fig. 1
figure 1

Period 1 reductions as a percent of total reductions, q = 1

Fig. 2
figure 2

Period 1 reductions as percent of total reductions, q = 2

Fig. 3
figure 3

Topographical rendering of Fig. 2

We set aside second-period damages thus far; that is set we b = 0. Obviously making b > 0 will increase R 1 for any value of H. The next question is how do positive second-period damages affect dR 1/dH. This depends upon the parameter values in question. For example, in Fig. 4 (γ = 4, a = 1.5, and q = 2), for higher values of H, the value of b matters more; for lower values of H, b matters very little. Thus, given a large amount of uncertainty, the marginal effect of second-period damages on dR 1/dH is greater, albeit decreasing.

Fig. 4
figure 4

First period reductions given second period damages

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Summers, L., Zeckhauser, R. Policymaking for posterity. J Risk Uncertain 37, 115–140 (2008).

Download citation

  • Published:

  • Issue Date:

  • DOI:


JEL Classification