What are the disadvantages of first-past-the-post electoral systems?
We all know the situation could arise in the U.S. where one candidate wins the popular vote but another one the electoral college. Given that the same could arise in the United Kingdom and other countries with a first-past-the-post or "winner takes it all" electoral system, what are the genuine disadvantages of first-past-the-post electoral systems? Besides the given examples?
Is it fair to inquire about the disadvantages without taking into account the advantages?
I could have reformulated the question, but at least for me the disadvantages were more important as I did already look up the advantages (the very few that exist)
Simple plurality voting has very little in its favor in any election with more than two candidates, but the top disadvantage is vote splitting. FPTP only allows voters to vote for a single person, and since the vote can't be transferred, if that person doesn't win that means the voter might as well have not voted at all.
Imagine a situation like the following:
- Person A is liked by 60% of the electorate
- Person B is liked by the same 60% of the electorate
- Person C is liked by the remaining 40% of the electorate
Assume that the 60% don't like C, and the 40% don't like A and B (it doesn't really matter either way, since FPTP doesn't care who you like after #1). It's unclear if A or B should win, but clearly C shouldn't -- a majority of people don't like him. However, if half the 60% vote for person A and the other half vote for person B, the results are:
- Person A gets 30% of the vote
- Person B gets 30% of the vote
- Person C gets 40% of the vote and wins
Because person A and person B were so similar, they split the vote and neither got enough to win; they would've been better served if one of them had dropped out.
Very similar to (but probably more ridiculous than) vote splitting is the spoiler effect. This happens when a similar minor candidate gets very little of the vote, but it's enough to tip the election the other way. For example:
- Person A gets 47% of the vote
- Person B gets 4% of the vote entirely from voters who would've voted for person A
- Person C gets 49% of the vote
Person C wins by 2 points, but it's person B's fault -- if he hadn't run, person A would've won. Worst of all, everyone who voted for person B is kicking themselves because the person they liked least is now the winner. One of the desirable qualities of a voting system (according to Kenneth Arrow) is independence of irrelevant alternatives -- the introduction of a new candidate shouldn't change the winner unless it's to that candidate. Plurality voting is particularly bad at this.
The result of all this is that FPTP bodies tend towards dual party systems after a while; the similar candidates fight it out in primaries so that in the real election, there's only two major choices and the effect of vote splitting is minimized. Nonetheless, spoilers can still cause problems. FPTP doesn't gather sufficient information to be certain when spoilers have changed a result, but in Gaming the Vote Poundstone estimates that at least five U.S. presidential elections have been decided by spoilers
And of course the American primaries still use a FPTP system, too, so these effects are apparent there. One could argue that the reason Trump won the Republican nomination in 2016 is because there were so many other candidates splitting the votes amongst themselves for so long that Trump established dominance with just the remaining minority; though one could argue against it as well, as there's not much data on who, say, Rubio supporters would have gravitated towards if he'd dropped out earlier. One might wonder if a similar effect will happen with the Democrats now.