Democracy, Epistocracy, Lottocracy

oh, and don't forget plural and proportional voting

First, a welcome to Substack to my friend and former coblogger, Matt Zwolinski.  The founder of the original BHL blog now has his own substack, appropriately titled A Bleeding Heart Libertarian.  Check it out.

With that said, on to a topic I don’t write much about.  Elections.  Not discourse about elections, but the structure of elections.  I begin by admitting that I have long been attracted to both plural and proportional voting but have always retained doubts about them.

For those that don’t know, a plural voting scheme (at least as I am using the term here) is one wherein different people can have different quantitative input.  For a simple example, we might have a system where everyone (or every adult) has at least one vote, but that some have two votes and some have three votes.  It might be that passing a civics exam entitles you to the second vote and passing a more difficult civics exam entitles you to the third.  Perhaps political scientists, theorists, and philosophers automatically get three votes.  (Or more.)

For those that don’t know, a proportional voting scheme (at least as I am using the term here) is one wherein once a candidate gets a certain proportion of the votes, they are elected.  For example, instead of the US Senate and the US House of Representatives, we might have a US House of Proportionally Elected Legislators.  To gain entry into that assembly, for example, we might require that each candidate receive 250,000 votes.  (I base this on there being roughly 250 million adults and assuming the assembly would have 1000 members.  If we had a plural voting scheme, we would need to adjust this to accommodate the total number of votes.)  As a matter of process, this would likely involve something like what we now call Ranked Choice Voting since we would want everyone’s vote to matter—i.e., not to be disregarded just because the person they voted for already received the needed amount of votes (250,000 in my example). Given that, if the first person in their ranked choices already had enough votes, their next choice would be counted.  I should also be clear that as this intended to be a federal legislative body, my preferred way to have this system would be to take all ranked choice votes from everyone (or every adult) in the country without regard to their geographical location.  Members of the assembly  would then best be thought of as representing those who voted for them, regardless of their location.  The assembly itself would comprise representatives of all of the people, in proportion to how the voters voted.

Both plural and proportional voting (together) were proposed by Thomas Hare in 1873 (in his The Election of Representatives) and endorsed by one of my intellectual heroes, J.S. Mill.  Though proposed together, the two are separable.  I very much like proportional voting as I sketched it and suspect the biggest objections will be practical: people will not want to give up representation by state or by district.  I think that is a mistake.  I also think that with proportional voting in place, we could return to allowing the legislature to elect the executive directly.  This would have clear benefits: no need for national campaigns, more direct interaction between the legislature and the executive, presumably with the legislature itself being more willing to assert the control it should have than is now the case, and, I think, only really qualified individuals would be elected to the presidency (and, removed at the legislature’s will).  Of course, the practical/political objection here is potent: people think we should all vote for president.  Given how bad we are at that, I am not very sympathetic to the objection.  That, of course, is a large part of why I find plural voting attractive. (Though I don’t mean only that we are bad voters; I also mean we have a terrible system for getting candidates.)

Plural voting, at least as I laid it out here might be seen as a form of epistocracy.  While I don’t think I have ever called myself an epistocrat, I do think there is something clearly attractive about the idea.  Having the most socially, legally, politically, and economically wise people having a greater say in how things run when there are any collective action problems to solve (or just investigate) seems clearly worth considering. Still, there is also something worrisome about it—even the wisest amongst us are fallible and operate with their own biases.  If their views coincide with mine, then great.  If they don’t—well I suspect sometimes I’d think “they know better” and sometimes I’d think “idiots! I thought they were wise, but they are clearly wrong.”  This would not likely be any worse than the current election system in the US and is likely to be a good better, but it is still worrisome. ( I should note here that this post was largely spurred by this interesting piece.)

For those that dislike all I have said so far, I think there is one further system that might do as well if not better.  While I’m not prepared to fully endorse it, here it is: lottocracy.  Others have defended it (see, this by Alex Guerrero, for example).  For me, the basic idea is simple: the wisdom of the crowd, especially when the members of that crowd engage in civil discourse, tends to do better (in terms of gaining knowledge and attaining justice) than any individual or homogeneous group, so if we create a governing assembly (a new legislature, like that described above) that is composed of randomly selected adult individuals (again, say 1000) from across the country and have in place some basic rules regarding the mode of discourse in the assembly, they are likely to do better than what we have now and better than a straight epistocratic group.  I imagine the same political/practical objection that people have to proportional voting will be raised: we should all get to vote!  Why that should be (especially instead of “we should all get a fair shot at the lottery to be in the legislature!”) I am not sure.