[morgancmartin]

Well that would be a start but I think it's pretty obvious you would want to base subscribers' voting power on duration of subscription, community involvement/contribution, their standing in the community and so on. This way you would avoid other communities subbing solely to influence the outcome of an election.

I think it would be pretty neat to include metadata in the calculation of the voting power metric as well. Such as how each subscriber votes in relation to other subscribers and assign a higher score to those who vote inline with the community. The goal being to optimize for those subscribers who embody the spirit of the community. Of course that brings up issues of echo chambers but other kinds of metadata could be factored in to offset that too. I would have to think on it for a while to come up with something I felt was as meritocratic as possible.

[nucleardog]

The problem with a lot of that is "low effort" versus "high effort" content.

I've seen a lot of great subs and communities take a pretty dramatic shift when the influx of new users is too great. It's basically the "eternal september" problem -- when the new users start to overwhelm the old users, they're no longer able to be effectively inducted into the community and the community makes a hard shift, usually to lower effort content as the quicker it can be consumed, the quicker it can be upvoted or otherwise appreciated (and usually it's quicker and more accessible to create than truly valuable content).

Or more straightforward--any community which experiences significant growth will eventually devolve into memes.

The one mitigating factor can be moderation (e.g., Hacker News). But if the mods can be replaced as well, then there's nothing holding it back.