# Talk:First-past-the-post voting

WikiProject Elections and Referendums (Rated C-class)
C  This article has been rated as C-Class on the quality scale.
WikiProject Politics (Rated C-class, High-importance)
This article is within the scope of WikiProject Politics, a collaborative effort to improve the coverage of politics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
C  This article has been rated as C-Class on the project's quality scale.
High  This article has been rated as High-importance on the project's importance scale.

## Evaluation section

The "evaluation" section seems to contain only criticisms. It would be good if someone who knows enough about the topic to site some referecences could add some points in favour.86.160.156.109 (talk) 20:49, 14 May 2010 (UTC)

In the meantime I've changed the section heading to "criticisms." However, I believe the whole section should be deleted as per the tags that hang over it. Any thoughts on that?--Heyitspeter (talk) 20:00, 15 May 2010 (UTC)
I'd like to see citations for the elements of the section. It's well known that the criticisms exist (their validity is open to debate), but without citations the section just isn't credible. Also, as someone decidedly against FPTP, I'm very interested in hearing pro-FPTP reasoning and would like to see cited evaluations for this as well. Erik Carson (talk) 21:12, 29 October 2010 (UTC) (resigned after login)

## Disproportionate influence of smaller parties

This section contains the statement "However in PR systems, small parties can become decisive in Parliament so gaining a power of blackmail against the Government, a problem which is generally reduced by the FPTP system.[3][4]". The first citation doesn't seem to substantiate the statement except in the lone case of the article's subject, in which PR is combined with the relative ease of dissolving the government, creating the conditions for the mentioned blackmail; in essence, the dissolution option appears to be the more direct cause rather than the PR itself.

I only skimmed the second citation, but it isn't terribly long, and it doesn't appear to mention this sentiment at all, much less substantiate it. Erik Carson (talk) 21:12, 29 October 2010 (UTC)

## Confusing Introduction

The introduction seems to say two different things, and even if you disagree, I think you might agree that it is far too long. —Preceding unsigned comment added by 128.112.16.224 (talk) 19:49, 11 November 2010 (UTC)

## Grammar/Spelling

"an majority"? Frognsausage (talk) 08:02, 30 April 2011 (UTC)

## 2005 GE results

Before anybody complains - I've fixed the results given for the 2005 British general election. The results committed the old BBC fault of not realizing that the Speaker is non-partisan, as well as confusing Great Britain with the United Kingdom. Wereon (talk) 12:57, 3 May 2011 (UTC)

## Majority criterion

I just undid an edit by JMFriedman which suggested that FPTP does not meet the majority criterion. This is incorrect - it is one of the few citreria it does meet. See http://en.wikipedia.org/wiki/Voting_system#Summary_table which is correct.

The confusion may be that JMF has overinterpreted the criterion. The criterion only requires that if a majority of voters pick a candidate that candidate must win. It does not require the converse (i.e. that if only a plurality support a candidate, then the candidate must not win). Therefore, FPTP satisfies the criterion (indeed you could argue that it over-satisfies it). —Preceding unsigned comment added by 2.123.56.49 (talk) 14:17, 4 May 2011 (UTC)

See also Majority criterion, or for example http://www.fairvote.org/single-winner-voting-method-comparison-chart 2.123.56.49 (talk) 14:24, 4 May 2011 (UTC)

The chart and explanations at Fairvote.org is quite specific that Plurality Voting does not need to meet the plurality criterion, that the winning candidate does not have to have a majority over all other voters

Majority Favorite Criterion: If a majority (more than 50%) of voters consider candidate A to be the best choice, then A should win.[1]

—Fairvote.org
Looks pretty unambiguous to me. I will undo the revert and add this citation.
I suspect that Majority criterion has been hacked, so will go there to correct it.
No. The criteria at Fairvote is correct, but I think you still midunderstanding it.

Majority Favorite Criterion: If a majority (more than 50%) of voters consider candidate A to be the best choice, then A should win.

. If, in a plurality system, more than 50% of voters support A, then A does in fact always win. Thus plurality obeys the majority criterion. The majority criterion simply doesn't say anything about what happens when a plurality but not a majority support A. I'm going to flag this for independent review.

Here's some more standard texts confirming this very point: http://books.google.co.uk/books?id=tGsQl-wxbKAC&pg=PA174#v=onepage&q&f=false (see table 7) or Democracy defended
(Gerry Mackie) 144.32.196.109 (talk) 10:11, 16 May 2011 (UTC)
In the plurality system, it is certainly true that a candidate with more than 50% of the valid poll will win. That is also what your citation says - but no more. But the reason why Plurality fails to pass the majority criterion is that it is not necessary that A should have an overall majority to win. There are many instances, certainly in UK politics, where candidates have won having secured as little as 35% of the valid votes. The majority criterion simply says that a majority must support A for A to win: any system [such as plurality] in which A wins without acheiving majority does not pass the test. It fails.--John Maynard Friedman (talk) 11:57, 16 May 2011 (UTC)
OK, so we agree exactly on the ananlysis. We disagree on what the criterion actually says or means.
The version at fairvote only makes a positive statement, it doesn't make the additional negative statement required by your definition. Similarly the statement on the wikipedia page: Majority Criterion. Similarly all of the external sources I've previously cited.
I've given links to the Voting System page in wikipedia (the table has not been recently changed) and three external sources all of which explicitly classify plurality as MC compliant. I'm still waiting for any citation which explicitly classify it as non compliant.

144.32.196.109 (talk) 15:42, 16 May 2011 (UTC)

OK, I've had an insight. Maybe the problem is the language used? Voting system theory is a branch of game theory which is a branch of mathematics. Voting systems criteria therefore use the formal language of mathematical proofs. If you're not aware of this, then I can see that it might lead to misinterpretations.
My interpretation is that the MC is of the form 'A implies B'. You are arguing for a stronger reading: 'A implies B and not A implies not B'. These are differenct: see for exampe Inverse (logic) quote 'the inverse of a conditional is not inferable from the conditional'.
The formal language of mathematics provides unambiguous ways of stating either of these. In the first case, we say 'If A, then B'. In the second, we would say 'If and only if A, then B'. See if and only if, necessary and sufficient condition. The MC is stated in the first form. Does that clear it up?
I guess if we adopted your stonger form, several other voting systems would also change wtheir compliance status. But the table in Voting systems would then have multiple inconsistencies with those in Nurmi or Fairvote, whereas at present they are consistent.

144.32.196.109 (talk) 06:45, 17 May 2011 (UTC)

Ok yes, I understand the mathematics and accept your logic. So to put it succinctly: "If A gets most votes, he must win. But if A does not get most votes, he may still win." I suggest a footnote to clarify this because I suspect that I won't be the first to fall for this mathematical fallacy. --John Maynard Friedman (talk) 11:07, 19 May 2011 (UTC)
Following this change, I added in language here that will hopefully make the issue more clear. As I see it, FPTP can sometimes meet majority criterion even when it falls short other times. --Sgt. R.K. Blue (talk) 05:53, 27 May 2011 (UTC)

As a new reader to this article I'd like to reopen this discussion - I don't fully understand the last few steps in the logic as to why this does not meet the Majority Criterion - I agree with the simple interpretation of MC where it only requires 'A implies B' and I don't see where in the article about MC, or the references, where that is disputed. I note that on Voting System that FPTP is marked as complying with MC... So this inconsistency does need to be addressed somehow. --DannoNZ

Hello Everyone, the section is currently blatently wrong, as the footnote referenced for this section completely contradicts the wikipedia article. I made an edit to correct it but John Maynard Friedman reverted it because he claims that there is consensus to the contrary (see my talk page for his comments). Is that the case? How can there be consensus if the wikipedia article is in complete contradiction with its own footnote on this point? If supposedly people think that the wiki article is right on this point why can't someone find a footnote to support the proposition rather than to contradict the proposition? — Preceding unsigned comment added by Encycloknow (talkcontribs) 06:13, 9 October 2011 (UTC)

#### Criticism moved from article:

This is mathematically incorrect. For example 9 people vote (5 blue, 4 red). A majority (more than 50%) of voters consider candidate A to be the best choice, then A should win - right? No, if there are 3 seats red could win two (getting two votes in each leaving blue with one in each) and blue win one (with all three votes). Blue loses the election even though they had an absolute majority of votes!
+
+ 1st seat (red,red,blue); 2nd seat (red,red,blue); 3rd seat (blue,blue,blue). This doesn't need to be sourced wiki is claiming something that is simply wrong, incorrect, does not follow the rules of mathematics - wiki is wrong here - simple.
Unsigned, but added by IP 109.158.138.156. The article needs to make clearer what it is saying to avoid misunderstandings. Dbfirs 22:38, 5 January 2012 (UTC)

Hi the problem is not thinking that A implies B is the same as not A implies not B it is that in this case A does not imply B. Getting more than half the votes does not imply victory in FPTP.

For example the following with 9 voters, three seats and two options (A&B).

(AAB) (AAB) (BBB)

B wins 5 out of 9 votes (therefore over 50% so under majority criterion must imply victory)

However B secures only 1 out of 3 seats (therefore loses under FPTP)

Simple as that FPTP fails the majority criterion. — Preceding unsigned comment added by 109.158.138.156 (talk) 22:41, 5 January 2012 (UTC)

Yes, obviously, in a parliament, one party can be elected to form a government having received fewer votes overall than the losing party, and I think this fact needs to be mentioned in the article. Nevertheless, in each constituency, the first past the post system guarantees (rather trivially) the majority criterion. Dbfirs 22:50, 5 January 2012 (UTC)
How has this come back in? We agreed many moons ago that FPTP does NOT meet the majority criterion because it is possible for a candidate to be elected with achieving an overall majority. For example, in most UK constituencies, the winning candidate barely scrapes 40% and in a few it is as little as 30% when the opposition is hopelessly split. Conversely of course there are 'pin a red/blue rosette on a donkey' constituencies where the winning candidate gets 80%. For FPTP, an overall majority is sufficient but not neccessary. The majority criterion says that it must be necessary as well as sufficient. Thus FPTP fails this criterion. --John Maynard Friedman (talk) 20:13, 9 January 2012 (UTC)
Where in Majority criterion does it say that 50% is necessary for a win? I read it as just a sufficiency criterion. Have you forgotten that you understood this last time? Please reconsider. Dbfirs 08:20, 10 January 2012 (UTC)

My apologies for putting my criticism in the article in the first place

By the inclusion of "Although the criterion is met for each constituency vote, it is not met when adding up the total votes for a winning party in a parliament." the article is no longer highly misleading. It is, however, I think still incorrect.

FPTP is about the election of the whole parliament not the individual seat - do others agree. Therefore it would be better to say that FPTP fails MC but insert "Although the criterion fails when looking at the results of a parliament as a whole, it is met when looking at each constituency individually." Can someone else phrase this better?

FPTP is made up of the "winner takes all" system (at a constituency level) combined with "PR" at the parliament level (PR of the seats not of votes ie take 57.3% of seats get 57.3% of parliament). By only looking at a constituency level in judging MC we have only checked if the "winner takes all" system passes MC (which it does) not if FPTP passes MC (which it does not). What are your thoughts? — Preceding unsigned comment added by 109.152.233.103 (talk) 23:23, 11 January 2012 (UTC)

ps it is not about "one party being elected to form a government having received fewer votes overall than the losing party" it is about a party not being elected after achieving, not just more votes than any other, but after achieving more than half the total votes (ie with FPTP you can lose with 50%+ of the vote therefore MC fails).

Well, no, not really. I am aware that some people in many countries vote for a party regardless of individual members, but this is not how the usual voting system is intended to work. Those of us who vote for individuals are happy that FPTP satisfies MC in our constituency. I agree that the party system distorts results at government level, but the article is about the voting system, not about the subsequent formation of governments, coalition or otherwise. In the UK, the government is often elected by a minority of voters, but this is a criticism of the constituency system, not of the FPTP voting system. Dbfirs 09:10, 12 January 2012 (UTC)

Hi, I think I am guilty of not reading the article properly. It clearly states that FPTP is the same as winner takes all. FPTP being the system we use to elect individual MPs and not the Government as a whole (I might be wrong but I think this might be what has caused confussion for a number of people). In which case the article is right, though I still think it has been improved by the clarification that "Although the criterion is met for each constituency vote, it is not met when adding up the total votes for a winning party in a parliament."

I also agree that under the Westminster model we theoretically elect MPs and not governments (thus MPs being able to cross the house without calling a bi-election).

It is very confusing to me that if we are talking about FPTP being the system that elects the individual MPs (rather than parliament as a whole) that gerrymandering is talked about in such detail. Gerrymandering only effects the broader election of the parliament - not the FPTP system as described in this article (ie a system of winner takes all that have nothing to do with the effects of segmenting the votes into different constituencies). If FPTP is more complicated, and in fact the system that electes the parliament as a whole (with division into constituencies being integral to FPTP), then such a system would not comply with MC. — Preceding unsigned comment added by 109.145.38.155 (talk) 23:27, 12 January 2012 (UTC)

## Merge

So this merge template to merge with Plurality voting system has been up there since the beginning of this year. I'm not very familiar with the subjects but is there a major difference between Plurality voting system and First-past-the-post voting. Devourer09 (t·c) 01:19, 24 October 2011 (UTC)

First past the post systems have a predetermined maximum number of votes (i.e. the electoral college) whereas a plurality has no preset maximum (i.e. electing a representative from one district). They are substantively different.

I don't understand your distinction because "first past the post" is just another name for "single winner plurality voting". In a two-round "first past the post" system, there is no "predetermined maximum number of votes". I would support a merge, but not strongly. Dbfirs 08:20, 13 January 2012 (UTC)

Plurality has nothing to do with FPTP system. This system in fact reduce plurality (i.e. numbers of parties in parlaiment), for examole US and Canada. Definitly opose mixing plurality with FPTS. — Preceding unsigned comment added by 99.224.228.119 (talk) 19:01, 13 June 2012 (UTC)

## Terrible intoduction

I read the intro and the first half of the article and still have no idea what FPTP voting is. I mean, even the introduction says that it's both "the person who gets the most votes wins" and simultaneously "the person with a majority of the votes doesn't necessarily win". WTF?

Perhaps this is a clear distinction to a polisci major, but it's awful for everyone else. Worse still, nothing in the article actually clarifies this. No examples, nothing. This article is less a description of what FPTP voting is, and more a discussion of its value as a system of voting, which sucks. Monolith2 (talk) 16:14, 23 August 2012 (UTC)

I agree, and revised a bit. The absolute majority part was problematic anyway, given where the link to that redirects. Sgelbman (talk) 00:30, 24 August 2012 (UTC)
@Monolith2: Where did you find "the person with a majority of the votes doesn't necessarily win"? This sentence was not part of the introduction. The introduction said "The winning candidate does not necessarily receive an absolute majority of all votes cast.". That's logically something totally different. It is the same difference like reading "Not every running world champion runs 100m in 5s." ($A \not\Rightarrow B$) and understanding "Running 100m in 5s does not necessarily is good enough to become running world champion." ($B \not\Rightarrow A$) Therefore, your problem with the introduction was not that you are not a "polisci major", but that you misread the semantic logic of the sentence.
In the new version, the whole information was just erased. In my opinion, the old version was more informative and therefore better. Maybe the old version can be modified in a way that the logical misinterpretation Monolith2 showed us, can be avoided without completely deleting the statement. --Arno Nymus (talk) 20:53, 25 August 2012 (UTC)
The old version's use of "absolute majority" was very misleading, and also not particularly informative (even if it had been accurate) given the extensive discussion of what distinguishes FTPT from other plurality systems in the body of the article. Sgelbman (talk) 16:12, 26 August 2012 (UTC)
That's a huge improvement, thanks! Monolith2 (talk) 22:38, 29 August 2012 (UTC)