Candidate Pool Synthesis

So this topic is a bit askew to voting theory, but I just wanted to start/provide a space for discussions about the effects that voting systems can have on candidate pools and candidate nomination/acceptance, and vice versa. I personally think the space between ordinary social interaction and the formalization of social decision-making through voting is intriguing, for example, party formation and nomination processes. Here are some of the initial questions I have, maybe they can start a discussion:

What processes do you think are ideal for candidate selection, and what processes do you think are realistic? How many candidates do you think voters will typically be faced with on their ballots when the spoiler effect is eliminated and smaller parties and political entities have more incentives to invest in a campaign? How many candidates is too many? What are the pros and cons of large candidate pools, and what are the reasonable ways to manage them? Do large candidate pools have any unexpected affects on voting systems in practice? How much of a difference do you think a spoiler-less voting system will make against duopolization, and in what way?

I know that voting systems can theoretically manage any number of candidates. One issue that comes to mind for me is competition for media attention, voters being bombarded with too many political ads, and not having enough knowledge about any single candidate to make a well-informed decision. While plurality is garbage, it does have incentives in place to keep the candidate pool small—too small, but too big is also bad. Are there other voting systems that are reasonably good that also encourage the public to stick with candidate pools of sizes that are likely to have sufficient variety without being overwhelming?

Whatever thoughts about this that come to mind are welcome.

I think the best voting system is this:
Distributed Voting - Seats allocation

In this way we would have:

  • many candidates, since there are many districts
  • a lot of local representation, since in each district there are at least 3 winners and fractional seats are used
  • but at the same time it’s necessary for voters to know only the candidates of their district.

To all this we can add liquid democracy, which drastically reduces problems of corruption and demagogy.

I don’t think it’s off topic to talk about the effects voting methods have on nominating candidates. Personally, I think the worst thing about bad voting systems is this very effect. Plurality means you have a strong incentive to eliminate similar candidates because they split the vote. Duverger’s law says this results in two dominant (and relatively well balanced) parties. This results in the ugly partisan stuff we have playing out constantly in US politics.

Some methods are relatively immune to this. The simplest change to address this would probably be For and Against voting, which effectively cancels the effect out. The “for” votes get split, but so do the “against” votes. It’s not a particularly expressive voting method, but it addresses the worst problem. Approval does basically the same thing, if you consider “not approving” to be an “against” vote – both for’s and againt’s are split, cancelling out the effect.

In our domain name vote there were 18 voters and 11 candidates. While we had a Condorcet winner that agreed with the STAR winner, if you simply subtracted the last ballot, could see real problems STAR had with dealing with a lot of candidates, in that the first part of the vote can get split (i.e. determining who will be in the runoff). There would have been a situation where voters would have wished they could change their vote after the fact. Cardinal Baldwin and any Condorcet compatible method didn’t seem to have this problem.

Even with perfect methods (which we know can never really exist, but we can come close), there is the problem with large numbers of candidates that people can get overwhelmed. I don’t claim to know all the negative effects of this. Regardless I think the worst problems are solved by simply eliminating, or at least “cancelling out” the effects of vote splitting.

Do you think there may be some optimal degree of or functionalized vote-splitting (whatever a degree of vote-splitting is) in this regard? Considering the potential for candidate overwhelm I’m not sure if zero vote-splitting is actually ideal. I feel like there should be some rational basis for keeping the candidate pool from exploding. The most straightforward way to me to achieve that is if adding “too many” candidates of a similar flavor decreases the likelihood that any of them will win. It would be very nice if there was some kind of inverted U shape so that there is a small-ish number of candidates of similar type that optimizes the chances that one of them will win. Or at least as a fraction of the whole pool. Just some thoughts. I have no real ideas about how to achieve something like that.

Otherwise sort of ineffective arbitrary constraints on the number of “similar candidates” could be put in place. But that might just lead to more disingenuous campaigns.

I don’t think we should rely on vote splitting, coming from the methods, to limit the number of candidates, no. I think we should strive to eliminate the effect completely if we can.

There are other reasons we might want to reduce the number of candidates, such as the “psychological overwhelm” thing. But no I can’t get on board anything that intentionally causes any degree of vote splitting, toward that goal.

That would be like saying “maybe we should intentionally introduce some safety flaws in this car, so people don’t drive so fast”.

1 Like

Totally I understand your point. But we also have speed limits, right? Do you feel like this is a legitimate problem, and if so do you think we should have some analogous “similar candidate quantity” limits? I think it’s an interesting idea but I feel like that could get hairy, who decides which candidates are similar?

In terms of psychological overwhelm, the incentives might be arranged in a sort of prisoner’s dilemma. If parties are flooding the candidate pool, then other parties have incentives to do the same, since voters will be overwhelmed anyway.

Otherwise maybe if candidates from different parties were kept separate on the ballot and party location made easily navigable. Then every party would have its own incentive to keep its pool relatively small. I don’t know enough about how anything works lol.

Well I’m not a fan of parties at all.

But anything that encourages parties form so they can either eliminate candidates, or flood the candidate pool, is a negative, to me.

Currently, in most US elections, there is already some mechanism to reduce the number of candidates. For instance you have to get a certain number of signatures. But that is very different from parties forming expressly so they can game the election system.

1 Like

I actually don’t mind the petition system for getting on the ballot in practice in a lot of current elections. It’s not that hard if you’re serious enough about campaigning to stand a chance to win. It should apply equally to all candidates though; sometimes major parties are exempt in practice.

1 Like

I definitely agree, being on the ballot shouldn’t depend on partisan backing. @Marylander I do think petitioning is a reasonable way to constrain the candidate pool. Hopefully it will still be effective enough when there are good incentives for “third party” and clone-ish candidates to run for office.

I can imagine a huge explosion in the candidate pool soon after an alternative method is established. I wonder what it’s going to be like in Maine. Not that IRV is really very good, but voters’ and candidates’ expectations may be high.

1 Like