What I miss when I read about the morality of discounting is a disanalogy that explains why hyperbolic or exponential discount rates might be reasonable for individuals with limited lifespans and such and such opportunity costs, but not for intertemporal collective decision-making. Then we could understand why pure discount is tempting, and maybe even realize there’s something that temporal impartiality doesn’t capture. If there’s any literature about it, I’d like to know. Please, not the basic heuristics & bias stuff—I did my homework.
For instance, if human welfare was something that could grow like compound interests, it’d make sense to talk about pure exponential discount. If you could guarantee that all of the dead in the battle of Marathon would have, in expectancy, added good to the overall happiness (or whatever you use as a goal function) in the world and transmitted it to their descendants, then you could say that those deaths are a greater evil than the millions of casualties in WW2; you could think of that welfare as “investment” instead of “consumption”. But that’s implausible.
On the other hand, there’s a small grain of truth here: a tragedy happening in the past will reverberate longer in the world historical trajectory. That’s just causality + temporal asymmetry.
This makes me think about cluelessness… I do have a tendency to think good facts have a tendency to lead to better consequences, in general; you don’t have to be an opmitist about it: bad facts just tend to lead to worse consequences, too. The opposite thesis, that a good/bad fact is as likely to cause good as evil, seems quite implausible. So you might be able to think about goodness as investment a little bit; instead of pure discount, maybe we should have something like a proxy for “relative impact in world trajectories”?
What I miss when I read about the morality of discounting is a disanalogy that explains why hyperbolic or exponential discount rates might be reasonable for individuals with limited lifespans and such and such opportunity costs, but not for intertemporal collective decision-making. Then we could understand why pure discount is tempting, and maybe even realize there’s something that temporal impartiality doesn’t capture. If there’s any literature about it, I’d like to know. Please, not the basic heuristics & bias stuff—I did my homework. For instance, if human welfare was something that could grow like compound interests, it’d make sense to talk about pure exponential discount. If you could guarantee that all of the dead in the battle of Marathon would have, in expectancy, added good to the overall happiness (or whatever you use as a goal function) in the world and transmitted it to their descendants, then you could say that those deaths are a greater evil than the millions of casualties in WW2; you could think of that welfare as “investment” instead of “consumption”. But that’s implausible. On the other hand, there’s a small grain of truth here: a tragedy happening in the past will reverberate longer in the world historical trajectory. That’s just causality + temporal asymmetry. This makes me think about cluelessness… I do have a tendency to think good facts have a tendency to lead to better consequences, in general; you don’t have to be an opmitist about it: bad facts just tend to lead to worse consequences, too. The opposite thesis, that a good/bad fact is as likely to cause good as evil, seems quite implausible. So you might be able to think about goodness as investment a little bit; instead of pure discount, maybe we should have something like a proxy for “relative impact in world trajectories”?