Thursday 23 June 2011

The Future for Electoral Reform is AMS

After the AV referendum a quick consensus formed that Electoral Reform is off the political table for a generation.  This was a consensus between opponents of reform and, bizarrely, most supporters of Reform, who seemed to suffer a massive collective loss of nerve.

I, personally, could not disagree more.

Over the medium and long-term the time is ripe for change, regardless of the result of that referendum. Reform failed in 2011 due to a combination of temporarily awful political circumstances, the presentation of a weak alternative and gross incompetence on behalf of the Yes Campaign.  None of these circumstances need recur, and it is highly likely that the long-term trends will continue to strengthen the argument for change, as they have since the 1960's.  The decline of the two party vote, the rise of the minor parties, the increasing inability of FPTP to properly represent the democratic wishes of the people of Britain.  None of these things are going away.

What is needed is for the Reform movement to pick itself up off the floor, knock itself hard on the head and learn the lessons of 2011.  Only honestly admitting that it got things horribly wrong and committing to change can give hope of success in the future. The Electoral Reform movement needs a dramatic modernisation, like Tony Blair's refounding of New Labour or David Cameron's modernisation of the Conservative Party, to achieve its aims in an age where politics and campaigning are professional and serious businesses.  It needs a thorough reconsideration of both Aims and Methods.

In this article I consider the aim for reformers by suggesting what I consider to be the best achievable alternative to FPTP.  And a superior alternative to AV.  In a following article I will suggest some ideas about a change in tactics and strategy that I think reformers need if they are to actually achieve their goals within a generation, and avoid repeating the disaster of 2011.
The massive 2011 vote against AV doesn't have to kill hope of reform for a generation. But it quite probably has put paid to any hope for change to AV itself for at least that long.  Or to put that another way, any hope for change within the next two decades can only exist on the basis of abandoning AV.  Good, I say. AV was adopted mostly because it was what was on offer, and it only became what was on offer for reasons of Labour Party convenience.  AV was capable of solving at most one of the numerous problems with the current system, and in a manner that had the potential of making other problems worse.

It did have one particular advantage though that should not be forgotten in its tidal wave of defeat. It was quite similar to the current system.  This made it an achievable reform. And this is my first criterion for a candidate for replacing FPTP. A further attempt at change should be focused on a similarly achievable reform, sufficiently similar to the current system to be recognisable as operating on similar principles, and sufficiently different to AV to seek distance from its calamitous defeat. Regardless of the problems with FPTP the massive No vote shows there is considerable public sympathy or at least overwhelming familiarity with its principles.  Any proposed alternative must work with this familiarity rather than against it.

It should also not be based on the same principles as AV i.e. preferential voting. This means not only AV, but also the other alternatives to FPTP that have been seriously proposed by reformers, namely STV and AV+.  AV+ was the system recommended by the Jenkins commission on reform in the late 90's. It is AV with an additional top-up of PR apportioned seats. It is a remarkably complicated change, as one would perhaps expect from a committee, and should be rejected for that reason and for being largely reliant on AV.

STV is the long-time preferred alternative of the Electoral Reform Society, Lib Dems and most other UK reform groups, and is currently used in Ireland. It is AV in multi-member constituencies, which unlike AV gives largely proportional results. STV is the preferred system of a majority of reformers. However, regardless of this, it should be abandoned, at least as a medium term aim. The staggering defeat of AV means that its central mechanism is politically discredited for the foreseeable future and because it requires voters to accept change to preferential voting and much larger multi-member constituencies, in reality, like AV+, it is too large a change to be sellable at once.

Both the reform movement's concentration on STV for decades and the strength of its conversion to AV in the previous year can be explained by its obsession with preferential voting. Most organised reformers are just convinced of its superiority to simple majority voting, regardless of other considerations. However, it has been rejected in the form of AV for now. It would appear to be a change and complication too far and, quite frankly, it is not worth sacrificing the chance of achieving real improvement by other means, merely out of a quixotic attachment to the wonders of preferential voting.

Where does this leave us if we've already rejected FPTP, AV, AV+ and STV? Except in Acronym hell. Another option worth mentioning is Closed List PR.  This would be a very simple system where you just vote for a party, and then the votes are counted and seats portioned out to the parties equal to its percentage of the vote.  This is the only true PR system.  However its side effects are so awful that it is generally rejected even by hard-core PR enthusiasts. Basically the problem is that voters have no control over who is actually elected, and there is no geographical connection between voters and representatives or sense that representatives represent everyone, rather than merely those who voted for them.  It is hence a massive leap from the current system, though it does bear the award of being the joint simplest system with FPTP.  Though from the opposite side of the spectrum.

So, ignoring Closed-list PR, AV, AV+, STV and FPTP, what is possibly left?

The answer to that is very simple. It's more proportional than FPTP, maintains constituency links, is a modest change from FPTP, is widely used by some European countries and within the UK itself, makes every vote count and is relatively simple compared to AV or STV but would still have given single-party government from our more decisive of recent electoral victories.

This system is the Additional Member System (or AMS).  In particular in a form I like to think of as FPTP+.

It is a combination of our current FPTP system used for UK General Elections and the Proportional Representation D'Hondt system we use for European Elections.  It would work like a combination of the two, producing a composite system that hopefully maintains the main advantages of both, while smoothing away their most stark problems.

The way it would work is simple.  Most MP's would be elected the same way as now, one per constituency under FPTP, with every bit of the country having a constituency MP.  In addition to these ordinary constituency MP's there would also be top-up list MP's.  Parties would gain a number of these MP's in proportion to their share of the vote, taking into account those MP's already elected in the constituencies.  The system works like our current FPTP system, but the top-up list MP's act to dampen the extremity of its results. Guaranteeing a degree of proportionality and ensuring that if you get enough votes you will get seats.

An AMS election would be simple.  Each voter gets a ballot paper with two sections.  One where they vote with a cross for whatever candidate they want to be their local constituency MP, exactly as now, the other they vote for the party they support, which goes towards deciding who gets the list seats.

In particular for the UK I would recommend the following arrangement.  I would suggest keeping a House of Commons at its current size of 650 MP's.  Of these 500 would be constituency MP's and 150 list MP's.  List MP's would be allocated by the D'Hondt system based on the list vote, taking account of the number of constituencies already won.  List MP's would not be based on the vote over the entire country.  Rather I would suggest multi-member list constituencies across the country based on the UK regions used for European Elections.  These could be subdivided to give list constituencies of an appropriate size of 4-8 MP's. I would also suggest allowing candidates to stand as both constituency and list candidates at the same time.  I think of this particular arrangement of AMS as FPTP+.      

That's phrasing it technically.  Basically it would be the same system currently used for Scottish and Welsh devolved elections.  Just with a higher proportion of constituency MP's to list MP's than they have.

I believe this system has a number of immediately apparent advantages.

Firstly, to go back to the point I mentioned above, it is familiar.  It's two 'parts' are systems already used UK wide for General and European elections, and together they are used in almost exactly this way in Scottish and Welsh elections.  It is also the same system used by Germany and a number of other European countries.  As a change from FPTP it would be modest, and I believe, achievable.  Indeed it would be down-right familiar in considerable parts of the country.  Everyone would still have an MP elected in the normal way, albeit in somewhat larger constituencies.  But they would merely also have further MP's elected to represent their area. On a practical note, it can be seen straight-away that this completely negates two of the major arguments used recently against AV.  Unlike with AV it would be impossible to argue that AMS was an unpopular, marginal system only used in one major country, as was argued, with some basis, against AV.

Secondly, it would also be much harder to argue that the system was complicated or alien, as it is used widely in Britain both in total and in its constituent parts without difficulty.  There is even the example of a English speaking Commonwealth Nation, which has already successfully switched from FPTP to AMS:  New Zealand.  Any argument that the system was complex could be answered by saying that the Scottish and Welsh have no problem with it.

Thirdly, unlike both AV and FPTP a degree of proportionality is guaranteed.  This system will always produce as or more proportional results than the pure FPTP we have now, unlike AV.  There will thus be no incentive for a No2AV, Yes2PR type vote as there was with AV.  Almost all supporters of change should have no problem of principle supporting this change, even if they would prefer an even more proportionate option still.

Fourth, it would make every vote count.  Whether your candidate wins or loses at the constituency level you still have an incentive to cast a vote at the list level, knowing it will go towards securing representation for your party of choice.  Equivalently on the political party side, it will give parties an incentive to campaign even in no-hope areas, knowing that they need to maximise their vote to gain vital list seats. I believe this is superior to FPTP and also AV.  AV made 'every vote count' by allowing people to put their 2nd, 3rd 4th choice towards influencing a result in a constituency, probably for a party they didn't want anyway though, but at least hated slightly less than some other party. FPTP+ gives every voter a chance to influence the over-all result in favour of the party they actually want to win seats.  This would also obviously solve the problem of tactical voting in a better way than AV, by allowing people to cast an effective list vote that will reveal the true relative level of party support and still vote in their constituency election.

Fifth, this would also solve the problem of safe seats. Unlike the rubbish the Yes campaign for AV was putting out, the problem with Safe Seats was never that they existed.  If voters in a seat consistently vote for the same party then good for them, that is their business. But that if you were in such a seat and didn't support the majority party then you may as well stay at home or vote for any other party, or even the party you hated or whatever. It would have no impact on the national election or result.  As far as having any impact went you were effectively disenfranchised.  FPTP+ gets rid of that problem, unlike AV, by allowing you to cast a list vote that will actually go to securing election for the party you support even if there is no point turning up to your constituency election.

Sixth. FPTP+ is more resilient to another of the core arguments against AV.  Anti-AV campaigners made a lot of the idea that AV violated 'one person, one vote', by taking account of some people's 2nd and further preferences but not others.  This issue does not exist under an AMS system where everyone has two votes, in effect, and everyone's two votes are taken into account in the same way.

These are all particular advantages to FPTP+.  However, I think there is a more general advantage to this system, that trumps anything offerable by pure FPTP or AV.  Electoral systems are fundamentally about representation.  Turning people's opinions and votes into the suitable representation in parliament and in our government and laws in the most effective manner possible.

This is the point where I have a confession to make.  I classified my opinion in the AV referendum as NO2AV, YES2PR.  Despite this, and the fact I classify myself as a reformer, I do not actually support true proportional representation. A true PR system would have distortions just as great as our pure FPTP system. I classified myself as as No2AV, Yes2PR supporter because I want more proportional representation than our current system provides or guarantees.  A pure PR system arguably provides fair representation in terms of numbers in parliament, but it does not in terms of decisions taken and rewarding co-operation and cohesion. Tiny minorities can have disproportionate power by holding the balance needed to gain a majority, which can be as distorting as disproportionate numbers in parliament.

Also, and there's no polite way to phrase this, marginal ideas should be marginalised. Any political cause, no matter how ridiculous, will find some people to vote for it, and no matter how excellent, find some people to oppose it. In a democracy more popular ideas should be privileged over less popular ones in terms of representation, with popularity acting as the only democratically legitimate proxy for the quality of those ideas. Majoritarianism does this. It also provides an incentive for groups to work and stick together, to attempt to appeal to as large a swathe of society as possible, and stretch over a broad range of ideological ground.  It discourages and punishes splits, extremism and focusing on appealing to narrow sectional interest.  This is as it should be.  1 idea held by 30% of the population should have more representation than ten ideas each held by 3% of the population.

However, I believe there must be some limit to this. A decent number of votes should lead to some representation, even if it is weighted down compared to its raw proportion of the vote compared to more popular parties.  This is because strict numerical proportionality is not in of itself the most important thing.  It is far more important that you have some representation and only roughly in terms of scale how much that is. For example, the Liberal Democrats are grossly under-represented compared to their share of the vote. But its voters are represented roughly because they do have a decent sized block of MP's and if that number changed by ten or twenty more or less it would not dramatically change those voters' representation. However, if that number of MP's were to fall dramatically or vastly increase then I imagine that would affect it noticeably. UKIP, on the other hand, is entirely unrepresented despite having the support of a million people, and thus these voters are arguably far more shortchanged than the mere numerical under-representation of the Liberal Democrats.

Majoritarianism for the sake of majoritarianism is just unsustainable. In terms of legitimacy the argument for it is non-existent. Just how small a percentage of the vote are majoritarians prepared to have a single party government elected on?  Is 35% not too low already?  Would they really rather see a majority government elected with 33% or 30% or 25% of the vote than see a coalition or minority government?  

Perhaps unusually for a Conservative I am worried neither by hung parliaments nor by Coalitions.  Under any system apart from one rigged to produce extreme majoritarianism they are an inevitable and natural feature, which can work in a situation populated by realistic adults, both as voters and politicians.  Both the current Coalition and the SNP government in Scotland decisively answer the case against coalitions and minority governments (while also conveniently demonstrating that FPTP does not guard against Coalitions, nor AMS make single-party government impossible).  Those who say Coalitions are un-British are historically short-sighted and ignorant and far too willing to ignore the fact that even when we have not been governed by formal Coalitions we have been governed by internal party coalitions that have often been just at fractious and difficult. The historical argument that Britain does not do hung parliaments is a symptom of our historical short-sightedness, and general belief that no events before 1939 ever actually occurred. From 1885 to 1945 Britain only had single party majority government for 9 out of 60 years, with the rest taken up by a mix of minority and coalition government.

My fundamental problem with FPTP is not that it is not strictly proportional, but really rather that it is arbitrary.  Its majoritarianism occurs on no consistent or rationally explainable basis. And it neither weights down less popular parties nor rewards more popular parties consistently. A consistent majoritarianism would always weight down parties compared to more popular ones and weight up parties compared to less, however popular or not they were.  This is fair to all parties because they all have the same strong incentive to secure more votes.  FPTP doesn't work like that though. It is totally inconsistent. It is a system where the Lib Dems may get 17% of the vote at two elections and 9 seats at one election and 48 at another. A party's vote may go down and its seats up, and then next election its vote go up and its seats down. A party may get 31% of the vote and 166 seats, another 27% of the vote and 209 seats, and another 25% of the vote and 23. A party may get 170,000 votes and receive 8 seats and another at the same election get 5 times as many and receive none. Or at one election the lead party receives 35% of the vote and a 3% lead over the next party and receives 360 seats, a clear majority, and the next the lead party has 36% of the vote and a 7% lead and gets 306 seats, a hung parliament. And I could go on.

The reason and traditional rationale for this is that our general elections are not just one national election, rather they are 650 or whatever individual constituency elections.  And our system is a relic of when this was in fact the case.  Now it is clearly not though.  A general election is largely a mechanism for determining the national composition of parliament.  This, in turn, is largely denominated in terms of party lines.  Our current system only extremely roughly reflects this reality though.  And in the fact it does as well as it does is largely coincidence.  At times it has done a better job and at times a worse one.

This leads to another fact about representation.  The problem with FPTP and also AV is that they are only capable of adequately conferring representation on an individual constituency basis.  And on this level arguably AV does a better job.  But this is largely irrelevant if both fail on the national stage.  I am a liberal Conservative who currently lives in a consistently Labour seat.  But I do not feel entirely unrepresented just because my MP is a Labour party drone for the simple reason that my views are partially represented by Conservative MP's elsewhere, even though I am not in their constituency.  Representation is both national and local.

In Britain, perhaps to a peculiar extent, representation really is also local. We value the independence of our MP's, and praise it where it is especially found, and we cling with pride to the notion that our MP's may owe their candidacy to a party, but their election is solely in the hands of the particular voters of that constituency.  We also rightly cherish the closeness of that connection, as well as the magnanimous notion that an MP must represent all his constituents rather than merely those who vote for him. A mark of this is the manner in which MP's are only allowed to be referred to in parliament as 'the honourable member for x'.  Their name is unimportant, all that is important is that they have been chosen to represent a particular community, and from this they gain their authority. This sense is so deeply ingrained in our political psyche that it would be wrong to derail it.  We also correctly take this sense of connection to be stronger the fewer people a representative is responsible for and with closeness of geography and culture.

On these clear principles the top-up, multi-member list constituencies would in fact bring representation far closer to most people, by giving relatively local representation to the 60% of voters largely shut-out in any one region.  A mostly unseen problem with the current system is that even most of the proportionality it does get comes through massive divergences in overwhelmingly disproportionate results in different regions. Most UK regions are, within themselves, massively dominated by a single party on a minority of the vote.  It is only the vast demographic differences between different regions that produce even vaguely proportional results nationwide.  FPTP+ addresses representation at this level.  It's top up seats will elected Conservatives in Scotland, Labour members in the South and Lib Dems almost everywhere, meaning Conservatives in Scotland will no longer have to look to MP's far to the South, nor Labour voters in the South far to the North, nor the millions of Lib Dem voters to a few distant and scattered islands of representation around the country that few of them have actually had a chance to influence or vote for.

What we need then is a system that contains both the key intuitions behind proportional representation and majoritarianism.  That can ensure representation that is both tied to individual voters as closely as possible, and tied to the wider national opinion.  In other words a semi-proportional system.  Only STV or AMS can deliver this, and AMS is both closer to the traditional British model and less discredited than STV (for the foreseeable future).  Sub-regional multi-member constituencies would be large enough to provide a close link to the national share of the vote, while also being close enough to have some sense of connection and identity to the areas they cover, in addition to the 500 constituency MP's who would oversee constituencies not unrecognisably larger than they are currently.

AMS in the form of FPTP+ would improve representation at every level, enfranchise voters across the country who are or would be excluded by FPTP or AV, and update our electoral system for a more pluralist and connected age.  It is a modest and simple addition and improvement on FPTP and remains true to its principles that, despite their shortcomings, are so familiar and built into our understanding of politics and democracy.  In many ways it is a thoroughly conservative proposal for reform.  And although that may not commend it to some people, I believe that places it in a long, successful and unequalled history of steady, peaceful, evolutionary improvement that has helped make Britain one of the most long-lasting and peaceful democracies and best governed countries in the world.


Post a Comment