The problems with FPTP– and with AV

It has been an interesting week for election-watchers, especially those of us interested in the dynamics of competition in single-seat districts. Canada had its election, with historic shifts in voting patterns, on Monday. Tomorrow the UK votes on whether to retain FPTP or move to the Alternative Vote (AV). And, just to make things even more interesting, voters in parts of the UK–Scotland and Wales–will be voting in MMP elections tomorrow as well.1 Quite a week–and tomorrow is quite a day–for electoral systems!

Here I will offer some observations about why I do not like either FPTP or AV (except from a researcher’s standpoint, for which they are terrific!)

The problem with FPTP is that it is fundamentally a system to elect a local representative in a world in which–at least for Canada and the UK–a general election is mostly a contest among national parties. That’s fine if there are just two parties of any significance. You still get the tension between hundreds of local contests and the clash of national parties. But if most districts are two-party contests, notwithstanding some number of “safe” seats for one party or the other, the system works, on its own terms: A series of local playings of the national contest between government and alternative government.

However, decades ago in Canada and the UK, the voters (and the party elites) largely stopped playing this game. Third parties have become more and more significant, and not only regionally. There seems to be a widespread view, even in the academy, that national multipartism masks local two-partism–that most districts feature two “serious” candidates, just not necessarily the same two in all parts of the country. That may have been true at one time, but it ceased being so some time ago. Now many British and Canadian districts feature a strong third party, and can be won with barely a third of the vote. Or even less. Canada’s election could be a step back to a more two-party pattern, given the collapse of the Bloc and the poor performance of the Liberals, but the latter may well be back. So it is far too early to say.

Sometimes voters in a given district even “tacitly” coordinate to send a minor party to parliament, not because it is best positioned to represent the specifically local interests of the district’s voters, but because the small party has invested in winning this one district that happens to have a demographic base consisting of the type of voters the party appeals to. I am thinking especially of Elizabeth May’s move across Canada to Saanich and Gulf Islands, which she won for a Green party that invested everything there. Caroline Lucas and the UK Greens last year are another such case, although Lucas at least had represented the same locale in other offices previously. It is not that there is anything inherently wrong with this strategy and outcome. Not at all! It just is another piece of evidence that voters and elites not playing the FPTP game.2 The contest in such a district becomes not about local representation, per se, nor about voting for the current or a potential governing party, but about voting for a fringe national party.

Then there is the whole micro-targeting strategy. To the extent that a party tailors its message to ever-smaller subsets of its constituency in swing districts, it, too, is not playing the FPTP game as we (used to) know it. It ceases to be a national campaign, speaking to broad swaths of citizens collectively, and becomes instead a disaggregated message to relatively small blocs of voters who just happen to live in swing districts. Again, not necessarily about local concerns, per se, but about ever-narrower demographic slices.

OK, so British voters can put a stop to all of this by voting for AV, right? Not so fast.

The best argument that the pro-AV camp in this referendum seems to have come up with is that your MP will “work harder” and will have to earn a majority of the district’s voters. I assume MPs tend to work pretty hard as it is, and to the extent that many of them already are pretty close to the median voter in their district (even when winning 40% or less), it is not clear that they have to work any harder under AV. Moreover, given that the proposed version of AV for the UK would allow voters to give only one or as few preferences as they wish,3 it is simply not true that the system will guarantee endorsement of every MP by a majority of voters.

Fundamentally, it seems that the argument for AV in an existing FPTP system where two-party competition is no longer the norm is a reactionary one.4 It puts the emphasis back on who wins the district and by what share of the vote. Yet FPTP parliamentary democracies have mostly gone well beyond that, as I started out with in my overview of the problem with FPTP.5 If significant percentages of voters are routinely voting for parties that have little hope of winning their district, but instead will be a clear third or fourth place finisher, it says they don’t really care about who represents the district. They care about national politics. And the two Green examples mentioned above suggest voters are capable of coordinating when what they care about national politics is electing a nationally small party with what are perceived to be fresh ideas. In neither case is AV necessary, and in the main, it’s not helpful if it is trying to put the genie back in the bottle and return to the good old days of majority winners in each district (as a presumed ideal).

And I would think that AV would be a micro-targeters dream. (Is there evidence for that in Australia, or am I out of line here?)

My take on AV would be different if the system could make a large difference in the way national politics works. And in style maybe it would do so, although I suspect that claims about reducing negative campaigning are exaggerated. (Candidates still have an incentive to see that certain contenders are eliminated from the count before others.) Fundamentally, most UK elections would have had the same basic shape of partisan forces in parliament with AV as they had under FPTP. So you get a reactionary effect at the district level without a clear corresponding progressive effect at the national level.

If FPTP is broken, as I believe it is in the UK (and arguably Canada, even if less this week than it seemed before), the only solution worth the effort is MMP or STV or another proportional system. If only the voters could have the chance to plump for PR…

1. Plus voting for many English local councls; are these all plurality of some form (often, I believe, in multi-seat districts)? There could have also been STV races on tap, in Scottish municipalities, but these are no longer concurrent with the Scottish Parliament elections.

2. If enough of this sort of thing happens to subvert FPTP, it’s fine by me!

3. Which is fine; I do not like the Australian requirement to rank every candidate.

4. But not in the horrifically specious way that William Hague and Margaret Beckett claim in a cross-party no-on-AV article in The Telegraph: that it would take Britain back to the days of the rotten borough by undermining one person, one vote.

5. India, the largest FPTP parliamentary democracy by far, is at least partially an exception to this point. More to come on that, as Indian district patterns are an ongoing research topic of mine.

55 thoughts on “The problems with FPTP– and with AV

  1. There is a significant advantage to AV that you have neglected, You can win an FPTP election by mobilising your own electors or demobilising your opponents’ electors. In FPTP elections the attack ad, the third party ad and the soundbite are king. AV abolishes that option because even the most benighted campaign operative will not try to drive away potential second preferences.

    You also understate, I think, the extent to which major parties need to at least keep their bridges open to minor parties when minor party second preferences matter. Admittedly the ALP seems to have forgotten this recently and their devastating losses in Victoria and New South Wales bear witness to that.

    I agree completely that preferences should be optional, as they are in New South Wales, Queensland, Tasmania and the ACT, if for no other reason than to force the major parties not to assume they can always cruise in on the second preferences of minor parties.

    AV also has a large impact after the election. All but one of the various independents and Greens owed their success at the last federal election to Labor preferences. The side they would choose to back in government was more or less dictated by the electorate, as it should be.

    I cannot speak for Julia Gillard’s apparent deathwish in her repeated attacks on the Greens.


  2. > “The Chartists, the Suffragettes, the great reformers – none of them campaigned for AV or anything like it”

    No, Mr Hague, of course not. Instead they were campaigning for open primaries, a 55% majority to dissolve the House of Commons, a poll tax, and all the other loony electoral brain-farts emitted by Brtish Conservative Party MPs since 1980…

    Mr Hague might also find a regrettable lack of soundbites from Disraeli or Lord Salisbury urging Britain to withdraw from the European Union, too.

    I meant my “Simon de Montfort considered MMP and rejected it” comment as a parody but it is rather alarming to think that someone once considered a suitable alternative Prime Minister of the United Kingdom is taking it seriously.


  3. Antony Green is triple-dumbfounded.

    My personal favourite remains the claim that Fiji is abandoning AV. Somehow the anti-AV types neglect to say that Fiji is an entrenched military dictatorship which shows no signs of holding elections under any system in the near future.

    Tom Lundberg’s case is deeply unpersuasive. If a referendum as conservative as this one is defeated the chance of a more radical referendum such as STV or MMP is zero, zip, nada.


    • My guess is that reform to PR will remain in deep slumber regardless of the outcome. But I am not as certain as Alan seems to be that the less auspicious outcome for a future consideration of PR is the ‘no’ vote.

      Green’s posts are fascinating! Thanks for the link.


  4. Hague is evidently unaware that AV has twice been passed by the democratic branch of the UK parliament and was rejected on both occasions by all those hereditary Chartists and Suffragettes in the House of Lords.

    Somehow I find it impossible to think of Hague ever being an alternative prime minister. Alternate, yes. Hague in paisley and mullet, no.


  5. “And the two Green examples mentioned above suggest voters are capable of coordinating when what they care about national politics is electing a nationally small party with what are perceived to be fresh ideas.”

    Hold it there. It’s not an either/or issue whether the obstacles of FPTP for participation can be overcome or not. They can be easy to overcome, and possible to overcome by a lot, or they can be immensely hard to overcome even when “overcome” means getting a token vote.

    Saying that the Greens don’t need AV, because they can just focus all their attention on one district and sell themselves completely out to that small group in order to get a token representative…

    AV will make a lot of votes meaningful that are currently cast even though the voter realizes they are totally meaningless. That is a small, but pretty unquestionable measure of improvement in my eyes.


  6. Wilf, Brendan O’Leary’s piece is behind the FT’s imprenetrable force-shield. Could you give an even more succinct summary?

    Also, can anyone illuminate what on earth Frank Furedi is talking about in this particular paragraph?

    “… the AV system would give party managers an even greater say in who can run for parliament. At least at the moment, now and again a relatively free spirit who has not been parachuted into a constituency by party bosses can become a serious contender for a seat. With AV, however, party managers would have even greater scope to handpick on-message candidates…”

    FF doesn’t offer any examples or even theories as to why this is so. He does not – needless to say – deal with obvious counter-examples going back to Alex Coles such as, oh, I dunno, Jim Killen, John Hatton, Rob ‘n’ Bob (Oakeshott and Katter Jr), Clover Moore, and… oh, why bother. This is the UK AV referendum after all. Real-life examples don’t come into it.

    Allowance must be made for the fact that he’s writing in spiked, a journal whose in-your-face bovver-boy contrarianism started to stale around 2007. There’s probably a market for an iPhone app that can auto-generate your own spiked op-eds simply by taking a position held by 80% of more of the people who have an opinion on it, and claiming that that position is a manifestation of top-down elitism in a post[-]ideological society characterised by managerialism.


  7. Harald, for whatever it may be worth, my remark about the Greens was in the spirit of my critique of FPTP no longer working the way its supporters claim. It was not meant to be taken as an argument about what the Greens’ position vis-a-vis AV should be.

    In fact, the same logic I was making could apply to Greens and similar parties under AV: find one district and concentrate your energies there, to break open the otherwise quite tightly shut door to parties of that sort. A case in point here is the Green breakthrough last year in an Australian House seat in Melbourne.


  8. Tyler Cowen has an AV thread on his blog, where I commented.

    AV does seem to make it harder for minor parties to gain representation in Parliament, though this may be more of an issue for minor party politicians than for minor party voters, who would get courted for their second choice votes.

    Most of the other effects of AV seem to be small but positive, several of which have been listed here or on the linked thread.

    I have to admit that the “oppose AV because it is not proportional representation”, which I have also seen on the Crooked Timber blog, makes no sense to me because proportional representation is not an option in the referendum, nor is it likely to be an option in any referendum any time soon. The chance of moving just from FPTP to AV itself seems to be something that comes around only once every ninety years or so. Elections in at least England will be conducted through single member districts for the forseeable future, so the question for advocates of multi-member districts is whether they want to make the single member district elections work as well as possible, or to sit out the process.

    The proposed change towards making constituencies more equal in population may well turn out to have a bigger effect, if the referendum proposals pass.


    • My understanding is that the changes in boundary delimitation take effect regardless of the outcome of the referendum. So, assuming the polls are correct, the Tories get the best possible outcome for their party: new boundaries and FPTP.


  9. I was able to retrieve Brendan O’Leary’s letter to the FT via Lexis Nexis. I can’t legally quote from it, but the succinct summary would be:

    1. AV makes it hard for minorities to elect their own representatives.

    2. It makes a two-bloc system more likely–and this is why Milliband supports it.

    3. Supporters of PR should vote no.


  10. What is the evidence for each of the O’Leary propositions? Hopefully we will get beyond:

    Australia has MPV
    Australia has two parties
    Ergo MPV causes two parties

    Otherwise we might be forced to apply an equally rigorous argument to FPTP:

    The US has two parties
    The US has FPTP
    Ergo two parties cause FPTP

    Or indeed
    Australia has two parties
    Australia has MPV
    Ergo two parties causes MPV

    Indeed it would be possible to apply this level of logical analysis to prove that there only two political parties in India and that the change to MMP in New Zealand was caused by FPTP.

    Post hoc propter hoc is not always a pretty sight.


  11. Mr Hague might also note that the French revolutionary Constitution of 1791 – which was influenced by Tom Paine and other British radical democrats, whom Mr Hague’s political forefathers were busily repressing at the time – required that elections be by “absolute majority of votes”.

    He has a hide trying to claim the historical mantle of the Chartists and the Suffragettes.


  12. Matthew, I think your argument treats politics only through the lens of representation. But from a voter’s point of view as they experience elections, I can’t believe you wouldn’t see AV as an improvement. No, it’s not proportional voting, but as a winner-take-all system, it’s way ahead of “top of the heap” plurality voting (we must stop this “first past the post” nonsense, as there is no post).

    With plurality voting, you have one shot to make a difference. If you know your real favorite can’t win, you can’t make a difference and vote for your real favorite. With AV, you don’t have to worry about that .

    As to the working hard point, I’d say it’s true based on US elections. The candidates who do more grassroots campaigning, who hustle more and get to know more people, are better at getting second and third choices. That’s why campaign spending hasn’t meant as much in several big AV-IRV races in the USA.

    Finally, opposing AV after the horrifically unfair campaign by the AV opponents is rubbing salt in the wound — sometimes you can learn something about the value of something by who fights against it.


  13. I’m honored to know that Andrew Sullivan is reading. And honored as well that Rob has commented.

    If it helps reduce the salty wounds, Rob, let me make clear that I still favor AV for US single offices (executive and Senate, if we can’t do more fundamental Senate reform). And I am rather agnostic at this point about AV for US House and state legislatures. It is quite likely the best we can hope to do. But I feel the UK can do better–someday. I will admit that this miserable campaign has shaken my belief in that, but I wouldn’t rule out a more wide-open process at some point in the future. The problems of FPTP (sorry) won’t go away, and forms of PR are used in so many other UK elections that it won’t just fall off the agenda forever, no matter how bad the prospects are in the short run as a result of today’s referendum.


  14. The strategy of voting no to a conservative proposal on the basis that it somehow gets you to a less conservative proposal is perhaps the silliest of all the anti-AV arguments. There is just no obvious path to a PR referendum or even a second AV referendum.

    This is the best chance of any electoral reform in the house of commons for the last century. Defeating it is not the most obvious strategy for encouraging deeper electoral reform. See you in 2111.


  15. Top of the heap — TOTH — unlike FPTP it’s an acronym you can actually pronounce. It even has an appropriately icky sound (perhaps I’m biased). Best of all, it is scrupulously accurate.

    I’m with MSS on the general point that whether AV for legislative elections is a step forward or step backward depends on the institutional and political context. And I agree with his assessment of the context in the U.K.

    If I were a citizen of the U.K., I would have been planning to vote No right up to the moment they handed me my ballot paper. Then, I don’t know. I might have been persuaded by the truly dreadful arguments of the No side — persuaded, that is, to vote Yes as protest against the campaign itself.

    Right this minute, I’m glad I’m not a U.K. citizen and didn’t have to make that choice.


  16. Here’s an idea: introduce a trinominal system.

    Each party fields 3 candidates. In the first round, 2 seats are allocated to the top 2 parties. 1 seat is allocated to compensate any results.

    So, if party A has 50%, B has 30% and C 20%, in the first round parties A and B would win a seat. In the second round, C would get the top-up seat.


  17. Something of a case study in the fate of the supporting partner in a coalition, as well, these elections. Projections are suggesting that the SNP will win an overall majority in Scotland, which uses an actual PR system (MMP) for itzs regional assembly. Thats a nice rebuke to the whole “government stability” argument against any change in FPTP, in what will most likely be a fairly depressing day for electoral reformers in the UK. Even if one is a sceptic on AV and the whole “evolution to PR” argument, its clear now the scale of the poltiical challenge facing reformers.


  18. Derek, your example is just SNTV, or 3-seat plurality with one vote per voter. If you mean PR, then what formula did you have in mind? One might expect party A, with more than double the votes of party C, to win two seats under PR, although with simple quota and largest remainders–not by coincidence, the “PR” system most similar to SNTV–the outcome would indeed be as you suggest. Formula matters a great deal at low M. (Terminology: what you sketch is not a two-round system, and “top up” implies a two-tier system.)

    3-seat districts have been proposed before; I recall such a proposal in Israel about 20 years ago, although I think it was within a two-tier (i.e. “top-up”) system.

    El Salvador has a very large number of 3-seat districts with simple quota/LR, which is why its “party C” (the PCN) has long enjoyed substantial over-representation.

    I’ve always thought 3-seat districts (with either open-list D’Hondt PR or STV) would be a good system for California, where we currently have two assembly districts nested in each senate district. Go unicameral and keep the senate districting arrangment only with M=3. I have never thought about it in the context of Canada or the UK.


  19. In the first paragraph of the main entry, I should not have forgotten to mention the Northern Ireland assembly was also being elected. It uses STV, adding to the diversity of electoral systems in use on Thursday across the UK.


  20. From one arriving from Sullivan’s blog…

    I’ve surfed sideways to read your thoughts on “top two” voting, hoping for some analysis in comparison to all of these other systems, and I have to say I’m befuddled by the argument you (and others) use to argue against a “top two” system.

    I live in a county that went 75% for Obama in 2008. There hasn’t been a Republican elected in a partisan election here in my memory. While I love the politics of the area, it makes for very warped elections, wherein the primary is the general, and local Republicans are forced to either vote in a primary for who’s going to get killed in the next general, or register as Democratic just to have any say. It means they tend to be extremely angry and frustrated, while the Democratic politics largely conform to who various power players get behind. This isn’t democratic, and it doesn’t work very well.

    On the other hand, we do have “non-partisan” local elections for certain offices, which largely come down to endorsements by a three well-organized PAC/committee combos that carefully study candidates and issue endorsements, which tend to sway a total of about 45% of the voters, but almost always determined the outcomes. They effectively serve as de facto parties, and these elections are generally well debated, thoughtful, and responsive to well informed public opinion.

    The knock against the “top two” primary systems is that somehow it would keep the Republicans off the ticket where I live. So what? They never win anyway! The local third parties seem to be scared of a “top two” system too, largely because it would frequently keep them from having a Green or a Libertarian to vote for on the ballot. My response is, you can’t even win second place in a system with no gates on it, what makes you deserve to be on the final ballot? *Elections are about picking a winner, not providing a venue for personal expression.*

    I want a top two system locally because it the two major parties would have to give up their role as sclerotic gatekeepers and actually have to get into the business of political organizing. I want a top two system because a non-omphaloskeptic third party that actually decides it wants to win and not just whine has a chance to overtake a weak second party, get into a runoff, and really go after the front runner, and maybe actually win an election or two. And I want these over IRV systems, because I think the IRV systems are still open to back room gamesmanship as to who you ask your voters to put where, whereas in a top-two system, the opportunities for vote switching and massing are much lower.


    • Michael, thanks for surfing on over from Sullivan’s!

      We had a brief discussion of California’s new “top two” system a while back. I don’t know if that’s the thread you are referring to or not.

      The best recent example of Instant Runoff in action in the USA might be the recent Oakland mayoral election.

      Really, I’d appreciate keeping this thread to FPTP vs. AV, because the issues of top-two systems in California and elsewhere are rather different. The previous threads should be open for comments.


  21. I can’t help to be a bit surprised that the results from the referendum have not been counted and released yet. It should be faster to count ballots with a binary yes-no question than for the different multi-candidate elections. Or they could have done both at the same time, but the austerity has perhaps reached the election authorities, so that they might not have the budgets and staff to do the counting quicker. And it’s strange that there have been no exit polls – in a much smaller country like Norway there are usually two at every election.


  22. The priority was counting the local and regional elections. But referendum counting is now well underway and there should be results soon, given that these ballots can’t take too long to count.

    I’m watching the returns at The Guardian‘s dashboard.

    It is a shame there are not exit polls, if that is correct.


  23. Rob, I made that choice. And I did not change my mind when I was handed my ballot paper yesterday. What would have been the purpose of that act?

    Yes, the nature of the campaign was appalling often descending into silly boys’ games of ‘look who is on my team!’ No wonder, voters were turned off.

    The British media seemed less interested in generating a balanced and detached debate about the strengths and weaknesses of AV and other voting systems and more interested in the point-scoring of partisan politics – especially the BBC who unfortunately set the agenda for British news coverage on a daily basis.

    Meanwhile the debate in the academy felt introvert although I share the blame having declined to appear on the BBC News channel a few days ago (bad timing). This morning, I tried to redress this on Radio Canada. Hopefully, I was successful in dispelling the notion that AV is a proportional system but I wouldn’t bet on it.

    As an optimist I’d like to believe that the outcome of the AV referendum will not put electoral reform on the shelf in the UK for a generation. I do feel very sorry for Nick Clegg though. A year ago, he seemed to be scoring all the points.


  24. Thanks for all the interesting comments. I voted ‘no’ without hesitation. I’ve seen all the arguments on both sides, and understand (and respect) how lots of people voted ‘yes’, but could not get over the possibility that AV would raise the threshold on small parties, making it harder for them to win any seats at all. There is, of course, the fact that Australia has had AV for nearly a century, and some Canadian provinces went from AV to SMP, not PR. Being risk averse, I see no point in ditching a system that does allow small parties to win at least some seats in favour of another majoritarian system that could actually make things worse. I agree with everything Professor O’Leary said in his Financial Times letter (you can actually get free access to a limited number of articles if you register, by the way).

    The Scottish result was a surprise (and a nice one for SNP supporters!). It will be interesting to see how the parties interact in the parliament under these conditions. Plenty of material for political scientists!

    Tom Lundberg, University of Glasgow


  25. No part of Australia has had AV for anything like a century. NSW has had AV since 1980 and Queensland since 1992. It is truly amazing how even after the referendum, we still keep getting these counterfactual pronouncements delivered with utter certainty.


  26. Alan, I know that the terminology in some Australian circles is somewhat different, but the clear consensus among political scientists is to call your system for the House “AV,” and consider the question of compulsory or optional expression of second (etc.) preferences to be a variable within AV.

    As far as I can tell, even Australian political scientists (e.g. Ben Reilly and Ian McAllister) agree with calling each of the variants on ranked-preference voting in single-seat districts “the Alternative Vote.”


  27. I think Alan is following Antony Green’s usage. AG has been reserving “AV” for the version with optional preferences, to avoid confusion in the UK referendum debate arising from claims that the optional-preferential version being offered there would lead to 5% of ballots being invalid, or candidates winning from third place on the first count, both of which do occur with the federal full-preferential version, but are 99% unknown under OPV as used in NSW and Queensland.

    Unfortunately, calling the Australian federal version “compulsory preferential voting” leads to confusion also, since it is mixed up with a legal duty to cast a ballot. Whereas other countries with compulsory voting (eg Belgium) do not use preferential ballots at all.


    • Yes, the Antony Green blog about AV in Queensland and NSW indeed confused me at first, especially when he wrote about “compulsory preference voting,” and for exactly the reason Tom notes.


  28. Yes. Try explaining to someone that all three jurisdictions I’ve lived in (Qld, NSW, ACT) have “compulsory optional-preferential voting” and you see why 68% of British voters opted for the devil they knew…

    In Australia, people usually say simply “preferential voting” for the federal (and Vict, SA, WA) method, although this sometimes leads into terminological cul-de-sacs (“Australia uses preferential voting for the House of Reps but not for the Senate” – err…) and is over-broad since many non-STV/ AV systems invite voters to use numbers to rank candidates. (Eg: the D’Hondt list system used in ther ACT 1989-95; MNTV as used for many Qld local elections, and formerly for Italian intra-list votes; Borda or other points systems as used for Slovakian minority seats and Austrian/ Norwegian intra-list votes).

    I have tried to think of alternatives and narrowed the field down to “mandatory-preferences voting” (MPV), “full-preferential voting” (FPV), and “exhaustive preferential voting” (XPV), but none of these is really satisfactory.

    Ideally, we would have started with “Alternative Vote” (AV) for the federal version and “Optional Alternative Vote” (OAV) for the Qld/NSW variant (and “Limited Alternative Vote” [LAV] for the London/ Colombo plan), but electoral terminology is so heavily path-dependent that it’s probably easier to get OAV itself adopted than a new term for it.


  29. Tom, the problem is that the exhaustive system is not the default and was not the plan originally proposed by people like Droop, Hare and Clark. Indeed Lakeman commented several times that she had no idea why Australia required exhaustive preferences.

    She was obviously unaware of the enthusiasm of MPs for playing the system at the margins. Broadly the Coalition has pursued exhaustive preferences and strict definitions of formal voting, Labor has pursued optional preferences and looser definitions of a formal vote. The Coalition, just quietly, is also convinced that there is a vast undetected level of voter fraud at least as immense as that claimed by the Republicans in the US.

    The default name should probably belong to OPV, the system as originally proposed by its designers.

    The introduction of optional preferential voting, known as AV elsewhere, made considerable changes in both states. The level of informal voting declined markedly. In Queensland the National and LIberal parties merged into a single party, despite their history of bitter conflict, including times when the Liberals sat in Opposition to a National state government.

    And the inability of small parties to get elected under OPV somehow failed to prevent One Nation electing 11 of the 89 MLAs in 1998 in Queensland.


  30. I could think of a two-tier system for the U.S. Electoral College.

    – “House Electoral Votes” will be allocated in each of the Congressional Districts, going to the candidate with an absolute majority using IRV
    – “Senate Electoral Votes” will be allocated proportionally.

    The PR formula is a modified version of

    Say California 2000

    Total seats 54 (52 CD+2 at-large)

    Gore 33 53.45%
    Balance 21 2.21
    Bush 19 41.65%/2.21= 18-19= -1
    New balance 2 4.9/2= 2.45%
    Nader 0 3.82%/2.45%= 2
    Other 0 1.08%

    Gore 33 53.45%
    Bush 19 41.65%
    Nader 2 3.82%


  31. The problem with electing members of the electoral college with Single Member Districts is mainly because many states in the U.S have gerrymandered districts.

    The present allocation of electoral college votes by who ever wins the most votes in a state by plurality is better than electing electoral college members in Congressional districts.

    It would be better to abolish the electoral college, and use a two round system if Preferential Voting can’t be used by all the states. I would prefer preferential voting over a two round system.

    The First Round requiring 50% plus one, to win outright, and the second round, the top two candidates.

    The electoral college does lead to reverse pluralities once in a blue moon like in 2000, but if it happened twice in a row. This system would be abolished. Look at what happen in NZ with FPTP in the late 70’s and early 80’s. It eventually lead to MMP being adopted in 1993 by referendum.


  32. The US already has the most partisan redistricting system in the world. Derek’s proposal, formerly known as the District Plan, would empower redistricting authorities to decide presidential elections.

    The District Plan would not prevent reversed majorities. It would favour concentrated parties over dispersed parties. It would privilege rural areas over urban areas to a significant degree.


  33. > “open primaries, a 55% majority to dissolve the House of Commons, a poll tax, and all the other loony electoral brain-farts emitted by Brtish Conservative Party MPs since 1980…”

    Add to that: directly-elected local police chiefs. (Not chosen by FPTP, of course; are you nuts? Someone might win with only 36% support!).

    An electoral system used for 90+ years in the world’s 6th-oldest continuous democracy (AV), or one that has been exhaustively examined by political scientists (STV) = “weird”, “foreign to our ways and customs’, “un-British”.

    Populist sound-bites generated by Thatcherite think-tanks or borrowed from the American Republic = “reforms essential to decentralising power to the people in the Big Society”.

    Is there any consistent principle behind this other than “attract votes to, and maximise seats won by, the British Conservative Party”?


  34. Tom, I remind you we having an atheist prime minister who is living in sin in the Lodge while opposing gay marriage because she is a cultural traditionalist. And the US has a president who is deeply committed to human rights and believes he has power to issue lettres de cachet. The UK conservatives sound like models of reason to me.


  35. I see the chances for any large scale change to a better voting system as being dead.

    It’s bad enough that the opponents of electoral reform can so effectively lie to and confuse voters.

    With the proponents misrepresenting an improvement as simple and obvious as AV then telling voters they would rather keep FPTP, we have no hope at all. The chances of electoral reform have gotten worse rather than better over the last few years.

    As for AV, its problems are those of any single seat district system and the only potential down side is a physically larger ballot paper or slower count.

    Several advantages of AV are being ignored.

    AV eliminates strategic voting and the need for pre-election polls. Regardless of who wins any seat, the first choice count is a valid and comprehensive indication of voter preference, in contrast to the ballots cast under FPTP where the results are distorted by strategic voting. This gives smaller parties and independants the recognition, if not the seats, they deserve. It thus allows smaller parties and independants far better acces to the system than FPTP.

    Fair Vote Canada has complained that AV can reduce proportionality compared to FPTP. They cite an election some years back where AV would have given more seats to an over represented Liberal party. The example may be valid but the complaint is not. AV will favor the large centrist party candidates. Sometimes this will make proportionality worse, and sometimes it will make proportionality better.

    An example of the latter is the recent Canadian election where a right wing party won a majority with 40% of the vote. AV would have favoured the parties with the other 60% of the vote, one centrist and the other left wing.

    The smaller Green party would not have achieved fair proportional representation but its supporters outside the leader’s riding would have been able to effectively express their left/centre/right preference.

    In addition, with AV the Green party leader, Elizabeth May, would have won with a majority rather than a plurality of the votes under AV. Under FPTP she could have lost to a less popular candidate.

    Alternative Vote is a simple step forward and anyone who wants electoral reform should support it.


  36. Alas, Michael, why go for simple steps when one can hold onto distant dreams of a grand marble staircase with gold statues of Hare and Clark?


  37. The NSW Parliamentary Library has a preliminary analysis of the recent state election. It is worth quoting the report on Balmain, the seat won by the Greens:

    This tight, three-way contest was the only one at the 2011 election in which the second placed candidate on primary votes was elected. In this instance, the Liberal candidate led the Greens candidate by 1.9% on the primary vote only to be defeated by a 7% margin in the final count, following the distribution of preferences from Labor and Independent candidates. This marks the first time a Greens candidate has been elected to the NSW Legislative Assembly.

    Needless to say the result defeats several arguments against AV.

    (1) The electorate achieved a decisive result.

    (2) AV operated to elect a minor party candidate and several independents.

    (3) Comparing the radically different results in the legislative council and the legislative assembly shows that STV is much better than AV.


  38. I’ve always been in favor of a runoff system. I like the idea of runoffs having 3 or 4 candidates.

    I favor the idea that voters should rank the top n candidates they want to see in the runoff, while in the runoff they should rank the candidates they want to see win.

    Of course, the first round could use approval voting, where the limit of the number of candidates voters could approve is equal to the number of slots up for the runoff. The runoff then would allow voters to vote for the candidate they want to see win.


  39. Mark Thompson, “The Tories’ UKIP problem shows why they were wrong to oppose AV: Rather than appealing for tactical votes from UKIP supporters in the Eastleigh by-election, the Tories should have supported a voting system that ends this dilemma.”
    New Statesman “The Staggers” blog (12 February 2013),


  40. “Plus voting for many English local councls; are these all FPTP?”

    They are all pluralitarian (as is true in Wales as well, but Wales did not have council elections in 2011). Most councils use either single-member wards, or else multi-member wards that elect a single member at each election (generally, electing a total of three each, with one a year for three years in a row, followed by a fourth ‘fallow’ year with no elections).

    Some wards elect multiple members at the same election (most common after redistributions, but for some councils this is standard), using plurality at-large/bloc vote (each voter may cast as many votes as their are seats, but no more than one vote per candidate).


  41. Toronto is going to AV or something similar to elect its mayor:

    As you might expect, this happened due to something of a perfect storm of a bad first past the post outcome. A candidate, whose political philosophy most Toronto voters disagreed with, won the election for mayor with a plurality in the low 30s due to vote splitting among candidates who were positioned closer to the median voter. And the candidate turned out to be fairly unqualified for the job as well. You would expect the City Council to take steps to make sure something like this doesn’t happen again.


  42. “Preference count: the fine art of shafting one’s coalition partner

    Labor and the Liberals are considering a preference deal to shut out their junior coalition partners. The parties need to decide: who is the real enemy here?”

    – Charles Richardson, (5 July 2013),


  43. “… [Lynton] Crosby clearly has a formidable record: steering John Howard’s Liberal Party to four election wins in Australia and running Boris Johnson’s two successful campaigns for Mayor of London. But his wizardry can be overstated: he oversaw the shambolic Libertas campaign in the 2009 European elections (they won one seat in Europe). And then there was the 2005 Conservative general election campaign. It increased the Tory vote share by a miserly 0.7 per cent of the vote. The cost of the “Are you thinking what we’re thinking?” campaign can still be felt in the state of the Tory brand: 40 per cent of voters say they would never ever vote blue.

    “The 2005 election showed the limits of importing successful electioneering from Australian to Britain. Australia’s use of the Alternative Vote forces every voter into a straight choice, between the (conservative) Liberal Party and the Australian Labour party. Crucially, voting is also compulsory in Australia, which lends itself to negative campaigning: offering a compelling reason why the electorate should not plump for the alternative is enough.

    “Britain’s electoral dynamics are very different. We live in a multi-party world; even if the Tories are successful in attacking Labour’s electoral weaknesses on welfare and immigration, voters may plump for Ukip or the Lib Dems instead. 35 per cent of the electorate did not vote for anyone in 2010: they need a positive reason to bother. Relentless negativity is less effective as a campaigning technique when voters can choose whether or not to vote….”

    – Tim Wigmore, “Is the wizardry of Conservative strategist Lynton Crosby a myth? Lady Warsi reminds the Tories of their problem with demographics. It is unclear whether the Conservatives have a long-term plan to adjust to the changing face of the British electorate.” New Statesman (11 August, 2014 – 13:12),


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.