One concern with that system that came to my mind is the possibility that explicitly assigning one vote for a person’s own use and one for a person’s use on behalf of future generations could make the idea of voting in a more self-interested/short-termist way more salient, as well as making it seem more acceptable, or even like it’s the norm that they’re being encouraged to follow. It seems like it might therefore make people more likely to use “their own” vote in a more self-interested/short-termist way (which could just cancel out the increased longtermism of people’s “second votes”, rather than making things worse overall).
But that’s entirely speculative, and seems to me somewhat less likely than Demeny Voting causing people to (a) just vote twice for what they would’ve voted for anyway, (b) use “their own” vote like they would’ve anyway but their “second” vote in more longtermist way, or (c) use both votes in a more longtermist way. But it still seems like that’s a concern worth investigating. (It’s possible it’s been investigated already—I haven’t read any of the linked papers.)
One concern with that system that came to my mind is the possibility that explicitly assigning one vote for a person’s own use and one for a person’s use on behalf of future generations could make the idea of voting in a more self-interested/short-termist way more salient, as well as making it seem more acceptable, or even like it’s the norm that they’re being encouraged to follow. It seems like it might therefore make people more likely to use “their own” vote in a more self-interested/short-termist way (which could just cancel out the increased longtermism of people’s “second votes”, rather than making things worse overall).
But that’s entirely speculative, and seems to me somewhat less likely than Demeny Voting causing people to (a) just vote twice for what they would’ve voted for anyway, (b) use “their own” vote like they would’ve anyway but their “second” vote in more longtermist way, or (c) use both votes in a more longtermist way. But it still seems like that’s a concern worth investigating. (It’s possible it’s been investigated already—I haven’t read any of the linked papers.)