I argue that on normative grounds a positive (non-zero) social discount rate is defensible. Positive discount rates are justifiable in terms of epistemic access. The appropriate discount depends on the actions or policies being evaluated as well as the epistemic reasons.
This analysis depends two primary claims: one about our epistemic status and the other about decision-making under a limited epistemic status. The first is that we are not able to produce reliable credences in social outcomes beyond a 50 or 100 year timeline. The latter is that, if we have limited epistemic status about some outcome of our action, we are less responsible with respect to that outcome.
With these two claims, I argue that it is both normatively and rationally defensible to discount intergenerationally. This is due to the possibility of future radical societal reconstruction, including the possibilities (a) that future generations do not exist; (b) are not Earth-bound; or (c) have radically different needs (e.g. transhumanism).
I also consider objections from Parfit and Broome. Parfit's objection is that it is immoral to discount the utility of other individuals simply in virtue of the fact that they are temporally distant. Broome's objection is that discounting introduces objectionable time-relativity into judgments of goodness. I respond that intergenerational discounting is not a moral judgment of the value of future generations, but instead a decision-making heuristic. Thus, it is not problematic to have temporal relativity, since our epistemic position towards the past (certainty) is different from the past's position towards the present (risk or uncertainty).
