"First and Second Past the Post" electoral system

Thande

Donor
In Britain prior to the Great Reform Act, some parliamentary seats (especially county seats) elected two rather than one MPs. The second seat was simply awarded to the runner-up in the election. The only other case I recall this applying was in the early USA, where the vice-presidency was originally given to the runner-up in the presidential election rather than a president running with a VP team-mate.

I'm wondering what a country would look like if a "FASPTP" system like this was formalised and institutionalised. Halve the number of constituencies by merging them, and then put both the candidates winning the first and second highest number of votes into parliament.

This would be particularly interesting because when the practice was used in Britain, it was before formalised political parties (Whig and Tory being as much a tribal identity as anything). As such, it was often the case that parties would not stick to a single candidate - say, six people would run in a constituency competing for two seats, 3 saying they were Whigs and 3 Tories, but it would be decided chiefly by personality (and connections, of course). Now things would change if we're talking a more modern setting with more organised parties and universal suffrage. Parties would probably stick to a single candidate per constituency, unless it's an ultrasafe seat in which case they might risk splitting the vote by running two candidates to try and pick up both seats.

Possible disadvantage:
If a country already has a very strong two-party system, like the USA, then pretty much every set of two seat will be held either by the Republican winner and Democratic runner-up or vice-versa, forming a situation where Congress is almost statically 50-50 Democratic:Republican. This would either deadlock things constantly or lead to formalised power-sharing 'one-party state', either of which is bad.

Possible advantage:
Stops parties in a three or more party (FPTP-type) system from saying "Vote for us in party 1 - you may like party 2 but if you vote for them they'll split the vote and then party 3 will get in, and none of us want that!" So parties like the Liberal Democrats in modern Britain would have a better chance of picking up seats. Also makes it easier for Independents, who often get as far as second place but can't quite overturn the core vote.

Thoughts? I think I'll run the numbers from the last British election and see what comes out as a (very flawed) test study.
 
In Britain prior to the Great Reform Act, some parliamentary seats (especially county seats) elected two rather than one MPs. The second seat was simply awarded to the runner-up in the election. The only other case I recall this applying was in the early USA, where the vice-presidency was originally given to the runner-up in the presidential election rather than a president running with a VP team-mate.

I'm wondering what a country would look like if a "FASPTP" system like this was formalised and institutionalised. Halve the number of constituencies by merging them, and then put both the candidates winning the first and second highest number of votes into parliament.

This would be particularly interesting because when the practice was used in Britain, it was before formalised political parties (Whig and Tory being as much a tribal identity as anything). As such, it was often the case that parties would not stick to a single candidate - say, six people would run in a constituency competing for two seats, 3 saying they were Whigs and 3 Tories, but it would be decided chiefly by personality (and connections, of course). Now things would change if we're talking a more modern setting with more organised parties and universal suffrage. Parties would probably stick to a single candidate per constituency, unless it's an ultrasafe seat in which case they might risk splitting the vote by running two candidates to try and pick up both seats.

Possible disadvantage:
If a country already has a very strong two-party system, like the USA, then pretty much every set of two seat will be held either by the Republican winner and Democratic runner-up or vice-versa, forming a situation where Congress is almost statically 50-50 Democratic:Republican. This would either deadlock things constantly or lead to formalised power-sharing 'one-party state', either of which is bad.

Possible advantage:
Stops parties in a three or more party (FPTP-type) system from saying "Vote for us in party 1 - you may like party 2 but if you vote for them they'll split the vote and then party 3 will get in, and none of us want that!" So parties like the Liberal Democrats in modern Britain would have a better chance of picking up seats. Also makes it easier for Independents, who often get as far as second place but can't quite overturn the core vote.

Thoughts? I think I'll run the numbers from the last British election and see what comes out as a (very flawed) test study.
I like this idea:)

I don't think it would lead to deadlock, I think it would lead to far less polarization within the parties- as you have large numbers of Republicans from places like Massachussets and large numbers of Dems from places like Utah scraping through on second preference. Thus their would be a lot of ideological overlap. It also means third parties would have a real chance at developing a presence.
 

Susano

Banned
Eh. Seems like a rather unfair system, though. The first guy or gal who could get double as many votes as the next one, and yet still end up with the same number of seats (1, obviously)... not very democratic, IMO...
 
Eh. Seems like a rather unfair system, though. The first guy or gal who could get double as many votes as the next one, and yet still end up with the same number of seats (1, obviously)... not very democratic, IMO...

A compromise could be achieved:everyone gets a second preference. If your guy wins, then the second preferences of people who voted for the winner get counted in determining the second representative. However those second preferences are counted as 0.5 while those who's first preference didn't win get counted as 1, thereby ensuring the minority has a shot.
 
Double member constituencies in the United Kingdom continued until they were abolished by the Representation of the People Act 1948. From 1922 there were 17: ten cities in England, one city in Scotland (Dundee), three counties in Northern Ireland, and three university constituencies. The single transferable vote was used in the university constituencies.

Sometimes from 1900-1929 Liberals and Labour ran in harness, as did Liberals and Conservatives sometimes from 1918-1931.
 

Thande

Donor
Thoughts? I think I'll run the numbers from the last British election and see what comes out as a (very flawed) test study.

OK, here's what happen if you apply it to the last election result. Obviously this is rather unscientific because if everyone knew this system was being used, voting patterns might be different and the parties could run two candidates in ultrasafe seats. But anyway.

Counting both first and second highest votes per constituency you get:
Conservative: 501
Labour: 420
Lib Dem: 297
SNP: 35
Plaid Cymru: 9
Independents: 6
DUP: 12
Sinn Fein: 7
SDLP: 6
Alliance: 1
Respect: 1
Green: 1
TUV: 1
Total: 1297

Note: Conservatives include UUP, Labour include Labour Co-op. Speaker counted as independent

Obviously too much for a parliament, so just for the sake of argument, halve and round all the numbers.

Conservative: 251
Labour: 210
Lib Dem: 149
SNP: 18
Plaid Cymru: 5
Independents: 3
DUP: 6
Sinn Fein: 4
SDLP: 3
Alliance: 1
Respect: 1
Green: 1
TUV: 1
Total: 653

As in OTL, no-one has a majority and the probable result is a coalition government. However, unlike OTL both a Con-Lib and a Lib-Lab coalition is feasible.

Comparing to the OTL FPTP results below, the major beneficiaries are the Lib Dems and the SNP. Note also that the BNP and UKIP don't get anywhere and the Greens and Alliance don't win any more than the one seat each they did in OTL: none of them came second anywhere.

OTL:

Conservative: 304
Labour: 255
Lib Dem: 57
SNP: 6
Plaid Cymru: 3
Independents: 3
DUP: 8
Sinn Fein: 5
SDLP: 3
Alliance: 1
Respect: 0
Green: 1
TUV: 0
Total: 650
 
Counting both first and second highest votes per constituency you get:
Conservative: 501
Labour: 420
Lib Dem: 297
SNP: 35
Plaid Cymru: 9
Independents: 6
DUP: 12
Sinn Fein: 7
SDLP: 6
Alliance: 1
Respect: 1
Green: 1
TUV: 1
Total: 1297
How did you end up with an odd number? Were there three uncontested seats?
 
I think people did get two votes for the old multi-member constituencies - when Churchill sat for Dundee between 1908 and 1922 he usually ran in harness with a Labour member - at the 1922 election, which he lost, the winning members were a Labourite and a Prohibitionist. The Prohibitionist won because he and the Labourite supported each other.
 
OT This is roughly how elections in Chile work and they do result in very closely fought elections. Since it takes a big landslide in any district to win both seats, nearly every district has one member from each coalition.
 

Thande

Donor
OT This is roughly how elections in Chile work and they do result in very closely fought elections. Since it takes a big landslide in any district to win both seats, nearly every district has one member from each coalition.

I'm starting to get paranoid about Chile, every time I have an idea it turns out Chile had it first...I think Chile may be following me around.
 
Didn't the VP undermine the P and vice-versa in the early US when they were of different? It didn't happen to often as the running-mate thing popped up early to deal with it.
 
Probably weakens party primaries - there's much more incentive to run as an independent if you lose the primary, since the odds of winning the second seat are higher.

Amusing (and weird) idea: what if each legislator could cast a number of votes on bills proportional to the percentage of votes they won? So David Cameron would cast 59 votes (having won about 59% in his constituency), while Gordon Brown would cast 65 and Nick Clegg would cast 53; their runners-up in their constituencies would cast 19 votes (Dawn Barnes, LibDem-Witney), 14 votes (Douglas Chapman, SNP-Kirkcaldy and Cowdenbeath), and 24 votes (Nicola Bates, C-Sheffield Hallam) respectively. That way, they'd be measuring the popularity of the member within the district, rewarding the most popular members.
It also would have a weird impact on constituency voting patterns - note that in the example above, Kirkcaldy & Cowdenbeath, by splitting 65-14, has more votes in Parliament (79) than Witney (78) or Sheffield Hallam (77). For comparison, Bethnal Green and Bow (Galloway's old seat) would have just 63 votes (43 for Rushanara Ali (L) and 20 for Ajmal Mansour (LibDem)), while Wells would have 87 votes (44 for Tessa Munt (LibDem) and 43 for former paymaster general David Heathcoat-Amory (T)). Heathcoat-Amory, here, has as many votes as Ali won; in fact, he'd have more votes than several winning candidates.
 
Probably weakens party primaries - there's much more incentive to run as an independent if you lose the primary, since the odds of winning the second seat are higher.

Amusing (and weird) idea: what if each legislator could cast a number of votes on bills proportional to the percentage of votes they won? So David Cameron would cast 59 votes (having won about 59% in his constituency), while Gordon Brown would cast 65 and Nick Clegg would cast 53; their runners-up in their constituencies would cast 19 votes (Dawn Barnes, LibDem-Witney), 14 votes (Douglas Chapman, SNP-Kirkcaldy and Cowdenbeath), and 24 votes (Nicola Bates, C-Sheffield Hallam) respectively. That way, they'd be measuring the popularity of the member within the district, rewarding the most popular members.
It also would have a weird impact on constituency voting patterns - note that in the example above, Kirkcaldy & Cowdenbeath, by splitting 65-14, has more votes in Parliament (79) than Witney (78) or Sheffield Hallam (77). For comparison, Bethnal Green and Bow (Galloway's old seat) would have just 63 votes (43 for Rushanara Ali (L) and 20 for Ajmal Mansour (LibDem)), while Wells would have 87 votes (44 for Tessa Munt (LibDem) and 43 for former paymaster general David Heathcoat-Amory (T)). Heathcoat-Amory, here, has as many votes as Ali won; in fact, he'd have more votes than several winning candidates.

Intriguing system. Probably would encourage people to vote more, so that their candidate has a greater say.
 
Possible disadvantage:
If a country already has a very strong two-party system, like the USA, then pretty much every set of two seat will be held either by the Republican winner and Democratic runner-up or vice-versa, forming a situation where Congress is almost statically 50-50 Democratic:Republican. This would either deadlock things constantly or lead to formalised power-sharing 'one-party state', either of which is bad.

Are there perhaps areas where you might get say a Green candidate coming second though?

You might also get Tea Party candidates coming second in certain areas.
 
Intriguing system. Probably would encourage people to vote more, so that their candidate has a greater say.

Except they have to vote for the right candidate. If a race doesn't collapse to two contenders, the constituency is in danger of electing a pair of weak legislators - a 37-33-30 split is a spectacularly bad result for the constituency, regardless of party supported.

This modified-FASPTP does allow minor-party representation, but it does weaken minor-party strength relative to pure-FASPTP. It gives more power to members who dominate their constituencies, but also more power to constituencies that generally have two dominant candidates (e.g. 49-47) or one really dominant and one semi-dominant candidate (e.g. 63-33).
 
Bump. Sorry for being late, but I remembered: Japan used a voting system for some time where each constituency elected several MPs - WP doesn't say much about the old system (which was changed in 1993), but AFAIR each constituency elected three to six MPs, and parties were allowed to run multiple candidates.
 
two and four party political arrangements are possible with this system, but the most probable arrangement (aside from temporary aberrations) is three roughly equal parties. At any given time, there's going to be a first place party, a second place party and an Opposition party ready to step in if one of the two higher-ups chokes. (assuming that your polity has little or no need for regional parties.)

Opposition is a slot that can only hold one party; a "second opposition" party either moves up fast, displacing one of the others, or withers quickly.
 
Bump. Sorry for being late, but I remembered: Japan used a voting system for some time where each constituency elected several MPs - WP doesn't say much about the old system (which was changed in 1993), but AFAIR each constituency elected three to six MPs, and parties were allowed to run multiple candidates.
It was pure Single Non-Transferable Vote. Nowadays Japan uses a mixed system - SNTV + block seats by proportional voting, d'Hondt method.
 
Top