Back in March of 2024, I wrote a piece about the ambiguity of the term "wasted votes." In it I reproduced three or four inconsistent meanings of the expression that can be found around the web and are used by advocates for various reforms. But after reading Steven Hill's useful 2023 Substack essay, "Which Proportional Representation Method is Best for America?" I see that I left out an important conception of the term, one that Hill relies on extensively as a reason for preferring Ranked Choice Voting ["RCV"] to several popular Proportional Representation ["PR"] alternatives.
While Hill doesn't actually define "wasted vote" in that piece, it is fairly easy to infer that he takes a vote to be wasted just in case it produces zero representation or voice for the voter who cast it. So, for example in an election for President or Senator in our winner-take-all system, if you voted for any losing candidate, Hill considers that to be a wasted vote. Where there is PR on the other hand, it is possible to vote for a second- (or third- or fourth-) place candidate and still get representation or voice. So, according to this conception, when that happens the votes are not wasted.
That particular understanding of waste is somewhat different from each of the publicly available explanations I considered in my blog last year, including the definition that FairVote, an organization with which I believe Hill is closely associated, provides (or at least at one time provided) on its website; but it can't be denied that Hill's version has an intuitive, even somewhat comforting, ring to it. I mean, one really might sensibly think I guess I could just as well have stayed home if it turns out that, subsequent to the next swearing-in ceremony, nobody will be representing one's views in the legislature, or council, or whatever.
I think there are a couple of problems, however. First, it seems to me that it actually does make a difference whether a candidate wins by 30% or .02% of the vote. That is, the votes for losing candidates do matter in some sense, so long as they are counted and publicly published. Since close elections may constrain winners, it may be better to vote for a loser than stay home. So it could be reasonably concluded that such votes aren't really wasted.
Nevertheless, votes for losing candidates do seem wasted in SOME sense: what does the voter get for it? The problem is that if we consider RCV, Hill's favored way to reduce wastage, it seems wrong to infer that, so long as Smith has ranked Crawford at all, and Crawford wins a seat, that Crawford should be thought to be providing a "voice" for Smith. For example, I may prefer Vance to Trump--and may have strategic reasons for ranking Vance at all--but one can't reasonably infer from this that I want Vance, or that he will be speaking for me should he get elected.
The idea that every ranking is a wanting is simply a problem for all ordinal voting methods: not even a top-ranking should automatically be thought to be a wanting. This is one reason that I believe Approval Voting ["AV"] is superior to every ordinal method in the search for proportionality--or at least would be if we could depend on voters to faithfully follow AV rules and vote for all and only those candidates of whom they actually approve. (Please note that I acknowledge with regret that voters may very well not follow those rules. See this article.)
In fact, Hill's conception basically makes the determination of whether or not a vote is wasted depend on something that might be entirely extraneous to who one takes to be an appropriate representer of one's views.
Take this cuckoo example. Suppose there is a PR voting rule that says "Put a mark next to all those candidates you have heard of, making the size of the mark bigger based on how frequently you have heard of him/her. If you have heard of this candidate a LOT, make a very large mark, and if you have only heard of him/her very infrequently (or you think you've heard of him/her but aren't entirely sure whether you've heard of him/her at all), make your mark very small. The votes are to be tallied as follows: the size of all marks made are to be "added up" and the five candidates producing the largest aggregated mark are to be declared winners.
Under this rule and Hill's definition of "waste," a ballot will be wasted only if it includes no vote cast for a candidate among those who amass one of the five biggest splotches. In other words, in this scenario, those who voted only for candidates that most people hadn't heard much about would be considered to have wasted their votes, but those who cast ballots that included a smudge for any of the five best known candidates would NOT be wasting their votes. Obviously, that produces a very odd and unacceptable conception of voice.
Now, I am not suggesting that under RCV the rule requiring the ranking of candidates by how one compares them to others running for the same office should be considered tantamount to the crazy voting rule specified above. But I do think that if wasting a vote is thought to be a matter of not getting ones "voice" heard in the future chorus of representatives, a vote must be a matter of APPROVING a candidate as being someone one believes to be fit for the job, someone who is better than nothing; not just a matter of whether one likes him/her a bit more than some alternative considered to be almost equally awful.