Saturday, October 15, 2016

A Potential Voting Strategy

Suppose you are confronted in the US Presidential election with precisely three candidates (with no option for a write-in, just because). You find both Candidates A and B, the two favorites, such that you cannot cast your vote for them, and you are trying to figure out whether you should vote for Candidate C or just refrain from voting altogether. You’ve been told both that a vote for C is a vote for A, and a vote for C is a vote for B; you’ve also been told that a lack of vote is a vote for A, and a lack of vote is a vote for B. Suppose you want to find out what the best strategy is: to vote for C, who will almost assuredly not win, or not vote at all, to register your displeasure at the system, to retain your moral integrity, because you’re apathetic, or any combination of these.

Conveniently, in our stipulated story, there are only 99 voters in your state (whew, now that’s low turnout!). Also convenient is the fact that you know that Candidate A has received 50 votes (50.5%), Candidate B has received 48 votes (48.5%), and Candidate C has received 1 vote (1.01%)[1]. Consider the following four scenarios:

1.     You vote for Candidate A.

This has the effect of increasing A to 51%, decreasing B to 48%, and decreasing C to 1%. Your actual impact on the election is minimal, given that A wins whether you vote for A or refrain from voting at all.[2]

2.     You vote for Candidate B.

This has the effect of decreasing A to 50%, increasing B to 49%, and decreasing C to 1%. Your actual impact on the election is minimal, given that A wins whether you vote for B or refrain from voting at all.

3.     You refrain from voting at all.

This has the effect of maintaining A, B, and C. Your actual impact on the election is not known, given how we’re using “actual impact” and the fact we haven’t investigated C yet.

4.     You vote for Candidate C.

This has the effect of decreasing A to 50%, decreasing B to 48%, and increasing C to 2%. Your actual impact on the election is minimal, given that A wins whether you vote for C or refrain from voting at all.

Here, then, is the upshot: if you cannot in good conscience vote for A or B, then in terms of sheer outcome (winner), it doesn’t matter, given the parameters, whether you vote at all. But if you want to vote your conscience, voting for C has the effect of registering your disapproval of A and B (by decreasing their share of the vote) while also registering your approval of C [edited for coherence] (by increasing C’s share). So, kids: in this type of a case, math supports you.[3]






[1] I have clearly rounded each of these numbers.

[2] Here, I am using “actual impact” to differentiate between only two options: voting for a particular candidate or not at all. It’s just a way of keeping track between the two strategies, and I wanted to show A’s and B’s significance in the whole thing. I also recognize that the relative percentages A and B may have in the real world can vary: I’m really just investigating this type of situation (and sufficiently similar ones).

[3] I realize, of course, that there are other, principled reasons to vote for a third-party candidate or not at all. I just thought this was a neat way, in relevant cases, to choose between the two options.

2 comments:

  1. Assuming, of course, that your choices/opinions don't have any affect on other voters (and they on even more potential voters.)

    ReplyDelete
    Replies
    1. Yes, that's right. Even in my fake scenario, I assumed that other votes had an impact, which really they won't be nearly so accessible (though still may have an impact). I had to stipulate that we knew something about it or other, so we could get on with looking at the effects of the vote on the vote share given the one particular vote (I also implicitly assumed there would be no further votes, again for simplicity's sake). We can always abstract out the time factor and run the same analysis on any completed set counterfactually, and the math holds, anyway. :)

      Delete

Please remember to see the comment guidelines if you are unfamiliar with them. God bless and thanks for dropping by!