A collaborative blog for Current Affairs and Policy Debate

MR, DR and PR

In Uncategorized on September 25, 2009 at 9:35 pm

For the first time in a while I have decided to not just begin, but finish a note about voting systems. I have not done this because it is a breathtakingly exciting topic. Indeed, it is academic and dry. Rather, I wish to address what I regard to be some misapprehensions about the subject.

When approaching the issue, most people divide it into a debate between two broad categories of system: Majoritarian, and Proportional Representation. One of these categories is too vague, the other too specific; for the roles most people cast them in.

What people normally mean by majoritarian systems is systems which often grant a majority to the “winning” party, rather than a system of majority rule. These are usually single-member systems such as First Past the Post (FPTP, used in Britain), Alternative Vote (AV, used in Australia) or a two-round system (France). Other systems exist, such as Block Voting, but these are rarely used and for the sake of simplicity we will not focus on them.

Single-member systems do not always grant a majority to the party with the most votes. Indeed, they occasionally grant a majority to the 2nd party (cf. United Kingdom general election, 1951), and extra seats won by a party depend very much on the votes won being in the right areas. For example, in 1992, John Major’s Conservatives lost 40 seats despite only losing a net total of 0.3% of the vote, from the previous election.

This is because the basis for single-member constituencies is Direct Representation (DR) – it is not actually to provide a majority for the winning party. However, these systems most often provide excessive swings for any party that secures a significant lead over its rivals, so the description is fair, if not entirely accurate.

However, when people use the term Proportional Representation, the problem is the opposite: that they are using a specific term to describe a broad variety of systems, some of which are not actually PR in the slightest. One of these is a favourite bugbear of mine: the Single Transferable Vote (unfortunately abbreviated to STV).

Those of you who have heard me on this subject know that I do not consider STV to be PR. I am increasingly of the opinion that it is unhelpful to regard it so.

This is because STV elects individuals, not parties. And the moment the electorate decides to stop voting on party lines, any semblance of party proportionality will be lost.

And if there is one thing common amongst all true PR systems, it is that they must elect parties. As for a system to deliver truly proportional outcomes for parties, then parties as institutions must drive the process. It is for this reason that analyses have found AMS to be far more consistently proportional than either STV, or even any List system which utilises smaller multi-member constituencies. For any form of local or individual basis for a voting system is liable to frustrate the noble ideals of proportionality, because PR is an inherently anti-individualist theory of representation.

What STV actually sets out to achieve is to elect representatives that reflect the broadest possible range of opinion in a community, and minimise so-called “wasted votes”. It must be said that certainly from a mathematical point of view, it does this very successfully. But this theory of representation is certainly nothing to do with parties as institutions, and would be more usefully termed “Moderate Representation”, or MR.

To those who think that this is only an academic point, a technicality, I ask you to take a look at this survey: http://www.democraticaudit.com/download
. In 1997, Labour would not only have still gained a majority on a minority of the vote (admittedly reduced to 44) under STV, but the Conservatives would have won even less seats – compared to the current system, which they were already under-represented by! This struck many people as a surprise, because they had been led to believe STV to be PR – but as STV is actually nothing of the sort, the widespread unpopularity of the Conservatives party and the widespread consensus that Labour was the preferable alternative, even amongst non-Labour voters, conspired to deliver something that was not even vaguely describable as PR.

Thus we have at least three schools of representation that lie at the basis of voting systems, DR, which seeks to elect representatives that directly represent localities, MR, which seeks to elect representatives that broadly reflect the range of opinion in a community, and PR, which seeks to elect party representatives.

Some systems fall into more than one of these categories, and some frustrate all of them. Borda Count is a Direct Representation system that seeks to elect the most consensual candidate, to the extent that a candidate with a majority of first preference votes can be defeated – the only system I know of which can do that. Such a system could be said to be both Direct Representation and Moderate Representation. On the other hand, Block voting is a multi-member, winner-takes-all system which is certainly not MR, and is arguably an example of a more true majoritarian (or more accurately pluralitarian) system.

So it is clear that these categories are inadequate, but this method of analysis already serves us better than the old binary, divisive method, as it looks at the basis of each voting system as well as its outcomes. It is clear that for the sake of accuracy, it is in need of qualification; but I believe that this is a good place to start from when assessing the most common modern voting systems, as it shows somewhat more clearly the variety and range in need of consideration.

  1. Very informative post. I personally do not like the term PR because so many people use it to describe any system that isn#t FPTP. The nearest to a pure PR system I suppose is the open list nationally, but even that requires some adjustment with respect to who gets the one or two seats that are slightly unproportional (if that makes any sense).

    I prefer to take each system individually, and given a free choice I would pick AMS for the reasons stated on my blog (shameless plug I know).


    I wonder which system you would pick to replace FPTP if indeed you wish to replace it.

  2. I am not arguing that STV is not PR from a particularly negative point of view — I’m not a fan of party-based voting systems in general.

    AMS is party-based, although to varying degree on how many “top-up” MPs there are to single-member ones. This creates an interesting conundrum: what to set the ratio at. Generally speaking, whatever criteria is used to set it will be arbitrary, for instance the Jenkins Commission used one that would have most likely delivered a hung Parliament in 1992, but not in 1987 — despite the fact that the change in the net popular vote for the Conservatives was very minor, around 0.3%.

    Incidentally, it is not true that AMS is always “not directly a proportional system”. The survey I linked to, by Democratic Audit, indicates that where enough “corrective” top-up members are elected, it can be more proportional than list PR or STV (the latter not being a particular surprise, admittedly!). In Germany, where the top-up ratio is about 50%, the results are very proportional if I recall correctly.

    I also disagree that the AV part of AV+ is pointless — the whole point of AV is that it makes the single-member election part more fair, by requiring a majority of voters (or as near as you can get in an election with more than 2 candidates) to support the candidate who reprsents the constituency. By the way, it isn’t only 2nd preferences that count — it’s third, fourth and fifth — the BBC stuffed up the definitions when they did their FAQ about voting systems this Summer. The system which uses 2nd preferences only is Supplementary Vote.

    I am also suspicious of the idea that AV is inherently less representative. From the local level, it is almost certainly *more* representative, and from the national level, the only way we can tell is by running polls consecutively with *FPTP* elections, where people will be looking at the election from the perspective of the current system. Evidence from Australia, where AV is used, indicates that it is not generally more disproportional than FPTP.

    Generally speaking, I tend to think that more important is to increase democratic responsibility in both the Lords and Local government, so to balance out what people vote for (the amount of issues people have to weigh up when voting in elections to the House of Commons is ridiculous considering the nature of any system of elections) before considering what voting system is best for the Commons. In a democratically reformed House of Lords, I would probably support the use of Single Transferable Vote to elect Peers for one term only, for about twice the length of a Commons parliamentary session. Depending on what the balance of power is between the two chambers, I would probably prefer to see AV in the lower chamber to First Past the Post, but would be open to the idea of AMS or AV+.

    I enjoyed reading your blog post on the subject, btw.



    • I am for House of Lord’s reform as well. The first steps were taken under this Labour government that I support wholeheartedly, it was billed as a first step and not the complete package, I would be keen to kick out the hereditaries and elect the House of Lord’s. Although to go into this further is off topic.

      I do not like AV because the results can often be less proportional than FPTP, I saw the theoretical figures in a text book (I no longer have it and cannot quote a title so inadeuate I know) for the 2005 election using predicted preferences and it was worse, and I think if we can drag the nation to electoral reform we should be aiming higher than the same again. I saw Jack Straw today at the Progress rally at the Labour fringe I was not convinced by his arguments at all (pro AV + FPTP). We cannot have any system that can elect a governing party that was got less votes than another party. I did not see the BBC definitions so I cannot comment, but that was not the cause of my mistake.

      I would like to see the House of Lords have an open list nationally (so not regional) which as I said should be easily the most proportional, mathmatically it can only be out by a handfull of seats whichever formulae are used.

      Regionality is not the only kind of representation that should be in our system. Eradicating the regional representation and personal reelection motives would help enable the Lords to maintain a certain amount of non-partisan independence.

      However for the House of Commons which currently dominates the legislative and executive functions regionality and reelection are the only constraints on our politicians so is essential. This is why I would like to keep the single member constituencies, which are still fairly small (only twice as big on a 50/50 ratio with same size parliament) but create a larger amount of party proportionality.

      Hopefully the lack of reelection for the Lords would not give it a large enough mandate to create conflict between the two houses. I explained why I didnt like STV and the reasons hold for the Lords.

      Apologies for the essay, but I find it really interesting.

      • Yeah, I agree re: the hereditaries and largely when it comes to electing the Lords (though I would keep appointment of Crossbenchers via the House of Lords Appointments Committee).

        I’ve seen theoretical figures for the 1997 election, though admittedly not for the 2005 one. The problem with these type of figures is that they’re based upon polls taken consecutively with FPTP the elections, which means people are answering questions on an alternate electoral system through the context of the current one. Because each voting system influences the way the political culture develops, these theoretical analyses can’t really be taken as much — the best thing to do is analyse countries which already use a system, such as Australia — and I personally remain unconvinced that they have a less proportional system to our own, if admittedly no more proportional.

        I must admit that I prefer AV over FPTP, simply because on the Constituency level it is far more fair. Third parties no longer have to worry about splitting the vote — as too candidates with even less momentum. The preferential element means that smaller parties can also exercise an indirect effect upon elections, even if their direct effect is minimal.

        “We cannot have any system that can elect a governing party that was got less votes than another party.”

        Unfortunately, that can happen under most voting systems. That’s why Malta (at least, I think so, though it may have been Tasmania) introduced a law that would grant top-up seats to any party if one were to win an election despite getting fewer votes, despite the fact they use STV. Under List PR and mixed AMS it is a slimmer chance, but it still remains, and there are other structural problems with such voting systems, such as with AMS ways of exploiting the corrective top-up means of election.

        Sorry, the reference to the BBC definitions was merely a nitpick, and possibly a mistaken one. It’s a pet peeve of mine, when voting systems were on the news, they completely stuffed up the way they reported it. Anyone with five minutes browsing the web should be able to find out the difference between Alternate and Supplementary vote, but for some reason they failed to do the necessary research.

        I have a motto when it comes to Lords reform, that you may disagree with; that any reform should build from the existing strengths and be careful not to risk jeopardising them. I think one of the existing strengths is a greater independence of central party control compared to MPs, so I would not personally prefer to see elections through any list system, even open list. I’d prefer STV because generally speaking, it is a relatively moderate system of election whilst retaining the basis of individual means of election — even with open list, the chances of independents are slimmer, and parties can determine the exposure a candidate gets.

        When it comes to STV for the Lords, I don’t agree that the issues hold as much sway. Because of the nature of the Lords (a revising chamber, rather than a controlling one), the difficulties re: transparancy and ease of understanding do not have the same ramifications, and ideally, the Lords wouldn’t work in a way that made one-party control possible (by which I include alliances, by the way), which I regard as one of the greatest strengths of the chamber at the moment.

        I agree on personal re-election motives, but not on regionality — simply because the House of Commons being the chamber which a government is (more or less) formed from, there will inevitably be elements of regional representation work that it does not satisfy.

        Re: constituencies, I’m not sure that I agree there. For the same reason that I disagree with Cameron on reducing the number of MPs, incidentally. Constituencies are already big enough to mean that there are discrepancies between the size of constituencies in England and those in rural Wales and Scotland. Such a state of affairs is a problem already, enough so to mean that for a rural Scottish seat, John Reid still had to hire a ferry (and put it on expenses, I remember the story in the news) to do constituency work, so I remain unconvinced of the merits of either enlargening constituency size or reducing the number of MPs.

        Don’t worry about the length of replies — I enjoy reading them, and I’m more than guilty of that myself. If you wanted to, you’d be welcome to write for this blog, btw — I was very impressed to find an article I’d read on Labour List on your blog, you’re clearly a talented writer.

        All the best,


  3. I am not as strongly against appointment as many reformers are, but direct representation will always beat patronage (however that is derived) for such an institution.
    Your point about how to judge electoral systems is a fair one, but change for changes sake is wrong, because it only alienates the public from further change. AV is we can agree is no more proportional (with respect to parties a definition many forget) and I still fear it would be less proportional. It is more complicated to the electorate and would hamper the chances of any proper system to be in place. Your point about small party power is correct, but it is not quantifiable. All that can be said is that in some seats small party’s second choices could be sent to another party but this will still be more important in some areas than others, it will still place too much power with party hierarchies.
    I am intrigued why you believe an open list when used with one national constituency is not necessarily the most party proportional. You mention structural problems with list, what are they. I know lots of arguments against list, but I appear to be more ignorant than I believed on the structural problems (other than the divvying up of the final few seats which is contentious) of the system.
    I accept that the house of Commons does not always represent regions perfectly. But it is one function that is exercised better than most. I would like to see a second chamber remain a step back from the day to day grind we see politicians go through every day to win popularity, I think for this local representation would have to be removed from the upper house.
    I also think that it will be inevitable that giving the House of Lords a proper democratic mandate would increase the role over a period of time. They are more active now than when they were full to the brim of hereditaries, that trend would continue further. I would not be surprised to get a member of the Lords as Prime Minister (in a few decades), unless it was legislated against.
    As I say I think the loss of local representation is not a good thing but perhaps a necessary sacrifice. In Robert Brown’s perfect world we would see greater devolution (similar but not the same as Miliband’s second devolution) somewhat compensating.
    I will be happy to submit something, maybe next week after my fresher’s week at university.

  4. I think what I see as the advantage of appointment of crossbenchers via the commission is that it helps ensure that the balance of power remains between parties and alliances. Even with something like STV, it’s possible to get disproportionate results, so it strikes me as good for some of the crossbenchers to remain to hold the balance of power, as well as to bring expertise to the Lords.

    I don’t think AV is all that more complicated — it’s fairly easy to explain the concept of preferential voting, in that all it amounts to with AV is what is effectively a series of simulated run-off elections. I find it’s invariably more difficult to explain the way the top-up formula works with AMS, and the way the surplus votes are redistributed in Single Transferable Vote.

    AV doesn’t improve the proportionality of outcomes whatsoever — the reason I prefer it to FPTP is entirely on the constituency level, as I dislike the fact that, say, in Leominster, if someone decides to vote Labour it can increase the Conservative MP’s majority.

    I don’t think it would hamper chances of further reform — even weak electoral reform would set a precendent — at the moment it’s an entirely unprecedented option.

    I think it the extent to which smaller parties gain influence is quantifiable — you merely have to look at the constituencies MPs fail to get a majority of the support of the constituents. In these cases, the supporters of smaller parties would play a crucial role. Certain constituencies will still be ‘safe’, although to a lesser degree, and party heirarchies will still have quite a bit of power, granted, so it’s not perfect, but it is a step.

    Sorry, Open list on a national basis would be very proportional — but I don’t think that’s a realistic option, given that even in the EU elections, where closed-list is used, there are still regions. I don’t think that open list on a national basis would be a sensible idea, either, in that the lists would be huge, and thus individual accountability of candidates weakened further. When it comes to region, that study by democratic audit, if I remember correctly, demonstrates that AMS (mixed-member) is the more proportional system.

    The structural problems with AMS are the fact that a party can artificially ‘split’ itself, one of the ‘parties’ only contesting list seats, the other contesting constituency seats, thus creating the possibility of gaining more seats than proportionally should have been awarded to it. (I believe I’m correct in saying this has actually happened in Italy, though I’m not sure: the person to ask is Tim Roll Pickering — his blog is linked to on our blogroll). On the regional list level, we saw in the EU elections how it’s possible for a party to increase it’s share of the vote significantly, but fail to pick up seats, whilst one with a smaller increase can gain disproportionatly. And even on the national level, there can still be issues, depending on how the vote is spread between parties and on the formula used, where parties benefit dispropotionately to their share of the vote or are even punished despite gaining more votes.

    I’m not sure that regionality is a function exercised better than most in the Commons, at least in practise — due to the fact that the Commons is the chamber of government and thus the level of party control, MPs tend to operate better as agents for their constituencies outside of it rather than inside it. I don’t think that local representation need mean that Peers are dragged into popularity games, if they are only elected for single terms — because that removes the incentive of winning popularity once in office.

    I’m not sure when it comes to the Prime Minister — the conventional role of each chamber is so strong that I honestly doubt that the idea that elections to the Commons determine the government would die out, particularly as there are other countries with more powerful 2nd chambers where the PM must be a member of the lower house (such as Australia, although this may actually be constitutional, I’m not sure).

    I’m afraid I don’t know anything about Miliband’s second devolution, though it sounds interesting. Which Miliband?

    Lastly, sounds great! If you email me I can add you to the contributers here — it’ll need to be the same email you signed up to wordpress with, otherwise I won’t be able to add you. I’m at dingdongalistic (at) googlemail (dot) com.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: