Tuesday 2 February 2010

Five reasons to be cheerful about the Alternative Vote ...

Gordon Brown has today set out the government's plan to legislate for a referendum, to be held within 18 months of the General Election, on whether to change the electoral system to the Alternative Vote.

Next Left has tried to provide accurate information - see"What is the Alternative Vote?" for resources as well as to debate the issues. Peter Kellner believes AV best balances different goals of an electoral system while; Stuart White has consistently argued against this and for PR as producing the most egalitarian social outcomes.

I have supported electoral reform (and PR) all of my adult life, move often arguing for Labour to faster. But I think the (non-PR) Alternative Vote would help achieve a major pluralist change in British political culture, so don't agree with (many) fellow reformers who think it barely changes anything at all.

I argued that in a Fabian Review essay in Autumn 2007How to reform the electoral system that AV is a desirable pluralist change which would break the century-long reform deadlock. I have defended that argument more recently in debates between reformers for Progress and a Guardian correspondence with Neal Lawson.

So what are the key arguments for the Alternative Vote?

1. Every MP needs to seek 50% of the vote, strengthening the constituency link.

Next Left has shown why AV means voters can stop the 'Jedward' candidate, in a scenario where one X Factor act has most initial support among 10 acts in week one. We don't crown them the victor if the vast majority of the audience would prefer another act to beat them. Under first past the post, we do.

In choosing a single winner, in almost every institution in society, the system used usually tries to find a candidate with majority support. (Note that the Conservatives use transferable votes and additional ballots to select candidates with majority support among their own selectorate, yet argue that the voters should not get the same chance).

2. An end to tactical voting: every voter can use their heart and head.

Many voters - Labour voters in the South, Tories in the north, Greens and LibDems where their party isn't competitive, strong Eurosceptics worried about a 'wasted' vote letting a more pro-European candidate in face a tactical voting dilemma. So AV, unlike first-past-the-post, allows every party to poll its full support everywhere, and would for the first time see elections reveal the actual patterns of support across Britain. (Which could also help to depolarise the regional divide in current voting patterns, which might impact on policy too).

Which means a very important change ... No more dodgy campaign bar graphs ever again. This has an enormous impact on the campaign dynamics. At present, in many constituencies and by-elections, the parties seem most focused on producing misleading bar graphs about the horse race. 'Only we can beat X here' becomes the most important message. Now, it would have to be about why voters should support them: the content of their campaign and argument, not their place in the horse-race.

3. Picking the right winning party.

The Alternative Vote, like the current electoral system is a "majoritarian" system. But, because it is mostly challenged by supporters of PR, its major advantage over the current system is overlooked. It is much better at giving a majority to the party which is most popular at national level.

What few people have noticed is how far a shift in electoral geography means the first-past-the-post system is broken in this crucial 'what it does on the tin' claim of its own supporters. Its chances of electing a majority government at all are much lower over the last three decades: its ability to pick the right winner in a close contest depends on an accident of geography which has not been in place for two decades.

It has got away with it because we so rarely have close elections: the two parties only finished within 5% of each other last time for the first time in 30 years. Whenever they do, the electoral system struggles to pick the winner. This was the focus of my Fabian Review essay which set out why:

the hidden truth of our democracy is that the electoral system is broken, and no longer fit for purpose. If elections ever became close contests again, this would reveal that we are playing Russian roulette with British democracy every four or five years.

Peter Kellner's post refers to his analysis (in his submissiont to Jenkins) showing that AV would have given a majority to the major party with more support among the whole electorate, while FPTP would not. This is an argument I have never seen FPTP supporters respond to, but Kellner's historical analysis appears valid.

4. AV is the most extremist-proof electoral system

AV would make it harder than the current system for the BNP to win seats in the House of Commons, requiring them to appeal to half the electorate in a seat like Barking, removing the chances of coming through with 25-33% in a three or four way contest.

By the same token, it is difficult for small parties. However, the Greens' chances would probably be much better than those of the BNP, since one can imagine them winning second preferences from a range of parties in their top target seats in places like Brighton. Since the removal of the "wasted vote" argument allows smaller parties to poll their full support: it makes clear, where environmental candidates may attract say 10% of the vote, that candidates who want to win face more pressure to have something to say about their concerns and issues.

5. More pluralism and grown-up politics

The central argument for AV is less about proportionality than about pluralism. Our current electoral system creates incentives for much exaggerated hostilities between those who you are ideologically closest to. Try talk to LibDems and Greens about each other, especially if there is a by-election on! But politics, like society, is becoming more plural, but the electoral incentives are against this. Preferential voting changes this.

What I think is especially important about preferential systems (AV, AV+ or STV) is that they insist on pluralism on the campaign trail, rather than only making it part of a post-election negotiation. (Some PR systems, like those in Scotland and Wales, do not do this). If you want and need votes from most voters, not only your natural supports, it provides a positive reason for a politics of mutual respect while arguing differences more clearly too.


Questioning common arguments against AV.

1. AV would be a fix to keep the Conservatives out forever.

This is a popular myth, which we will hear from newspaper columnists and bloggers. But it is evidently nonsense. A preferential voting system did not prevent Boris Johnson being London Mayor.

There is no doubt that the Conservatives would have been able to form majority governments several times under AV, including in the 1980s. The reason is simple: many more voters preferred a Conservative-led to a Labour-led government.

The difference would have been subtler, but still significant: Conservative candidates would have had to appeal to a wider electorate than was the case in the 1980s.
And the lack of any checks and balances on the Thatcher government would now be different in a range of ways: devolution would prevent the poll tax being imposed on Scotland without consent; the human rights act and the loss of a built-in hereditary Tory majority in the Lords would check the 'elective dictatorship' which Lord Halisham had warned about under the previous Labour government.

Almost all claims about party advantage or disadvantage under AV are exaggerated, and any attempt to reform electoral systems on that basis is likely to be frustrated too. (Nor is it possible, when a referendum is held, given that a narrow party agenda would never win a broad public majority). Voters would react to a new system in ways that can not be predicted.

AV would have helped Labour in 1997, when everbody who did not vote Labour or Tory preferred Labour to win, but would have hurt them in 1983, when the opposite was true. (An AV election at the time of the last European elections would probably have been worse for Labour). AV could well be good for the Liberal Democrats. The Jenkins Report suggested it would halve the disproportional under-representation. But many LibDems believe their share of first preference votes could fall (as it has done in most PR contests), as voters have a full range of alternatives.

Nobody knows for sure would gain or lose as voters and parties adapt to a more open system. All that can be said confidently is that it would be better for candidates and parties with a broad popularity, worse for "pariah" parties which are widely detested, and that it would disadvantage candidates and parties with intense but narrow support if they are also strongly opposed by a majority of voters. Such candidates and parties can win under first-past-the-post.

2. AV "exaggerates majorities"

There is something in this common argument from PR supporters, but I don't think the projections from the highly unusual 1997 election are the knock-down argument that is often presented. (Especially as voters behave differently in different systems). The scale of the 1997 election under first-past-the-post was itself a kind of "do it yourself AV" among the voters against an almost unlikely "pariah" party of government. 31% voted Tory, and 32% preferred a Tory-led government to a Labour one, which was the preference of three-fifths of all voters.

What is overlooked is the important difference between AV and first-past-the-post in why a party wins a majority.

Under AV, only a party which is broadly popular across the whole electorate can win big. That isn't the case under FPTP, where an efficiently concerntrated vote could see a party win with a plurality even if most voters feared that party a great deal. (So extreme parties like the German Nazis or Communists in Weimar Germany would have done better under FPTP than a preferential system).

This is because AV picks a winning party for a relevant, not an irrelevant, reason.

The idea that FPTP favours large parties, not small parties, is a myth. It favours geographically concentrated parties, big and small, and discriminates against those with support around the country. Welsh Nationalists often do better than the big parties on a votes to seats ratio for Westminster; Scottish Nationalists do much worse! This is morally and politically arbitrary.

What matters under AV is how popular the leading candidates and parties are with the whole electorate, including those who did not vote for them.

3. AV is not vulnerable to the anti-PR arguments which resonate most strongly

AV is not PR. Whether this is a strength or weakness is a matter of taste.

The point is missed often by supporters of the current system, who have rarely paid attention to what they are 'agin'.

AV is still quite frequently wrongly described by politicians and commentators as the "Alternative Vote system of proportional representation" (as in ConservativeHome's immediate reaction last year). This then allows a range of stock GCSE politics stock objections to PR to be voiced, even though these are entirely irrelevant to a comparison between the current system and AV.

The following anti-PR arguments are irrelevant to any argument against AV from supporters of the current system. I fail to see what of value in the common system is lost in a switch to AV, and FPTP supporters making anti-PR arguments do not provide one.

* "The tail wags the dog" and smoke-filled rooms.

This objection of "disproportionate" power to smaller parties under PR in coalition negotiations is irrelevant here, because AV does not make coalitions more likely. (It is also the case that AV+ would often deliver majority governments, in all but two post-war elections, but would require parties to reach a higher share to win a majority).

* "Two classes of MPs".

I am sceptical of the importance of this charge, though it is one often voiced by MPs, as an objection to the Scottish/Welsh German-style system. (Some MPs are also hostile to overlapping constituency responsibilities under STV: that seems to me a weaker argument still). In any event, it does not apply to AV at all. There is no change to the constituency boundaries at all. Similarly, any concern about larger constituencies, under a regional PR system or STV in rural areas, or the (marginally) larger AV+ constituencies falls too.

* Letting in the BNP.

AV would make it harder for the BNP to win representation than the current system does, though parties would have more incentive to address real issues which were driving support for protest candidates, beyond an extreme racist fringe. Several forms of PR (such as those in use in Scotland, Wales and London would make BNP representation more likely), though that can be argued as throwing up a democratic preference that should be voiced and addressed.

So why not PR?

This is a good challenge. A good argument which can be made for PR, as opposed to AV, or indeed for the Alternative Vote + which is halfway between the two, despite being labelled "PR" by most. I am sympathetic to it: I agree with Stuart White about the egalitarian outcomes delivered in more consensual democracies, and am against any 'theological' arguments between different systems. I will support any more pluralist system which can carry the day, support STV for local government and a PR elected Senate alongside an AV or AV+ Commons, and have supported calls for ideas like citizens' conventions to deliberate about the choice which should be put. (But I don't favour a multi-option referendum: it would be a tactical disaster, where supporters of the status quo could easily sow "systems confusion" to try to win).

But I also think it is much harder to win that argument for full PR than its supporters often assume.

There is broad support for PR but it is also quite shallow - and I do not think the broad cultural argument for a shift to full PR has been won.

Take these findings from the Joseph Rowntree Reform Trust polling last summer.

Put one way, the case for PR is broadly popular, and has been consistently for 20 years.

This country should adopt a new voting system that would give parties seats in Parliament in proportion to their share of the votes.

63% agree with that, and only 22% disagree, so what possible problem could there be for refomers.

And 56% agree that the electoral system produces governments which don't represent most ordinary people (25% disagree).

But here is the very next question, to the same 1000 people in the same poll.

The present system of voting is the only way that the country can get strong one-party governments which can get things done.

Well, 53% agree with that and only 29% disagree.

61% would like a referendum to sort the issue out, with 24% against.

That is where the case for PR is most vulnerable: coalitions and power-sharing remain dirty words in British political discourse and alien to the British political culture, even as our institutions change and support for doing politics differently grows.

This is an argument (along with 'lets in extremists' - which has much more resonance now than two years ago) where PR advocates have struggled to engage fully with how to win the argument with non-expert audiences).

I would argue that the case can be made better in practice than theory. So I argued in the electoral reform debate at Saturday's Progressive London conference that there was a "paradox of pluralism". Electoral reform (whether PR or AV) would help to unlock a more pluralist culture in British politics, which the current system rules out. But to get electoral reform, we need to win the argument that pluralism works in politics.

But this need not be a 'chicken and egg' problem. We see stronger support for PR in Scotland after a decade of devolution.

If we want to create a more pluralist politics, we have to get on and show that it can work. I think the Alternative Vote could help us to do just that.

The Alternative Vote would make British politics most engaging and more pluralist. It is difficult to see any strong argument, except force of habit, for opposing the change. Nor should the strongest advocates of pluralism and PR make what they see as the best the enemy of the good.


James Graham (Quaequam Blog!) said...

"Next Left has shown why AV means voters can stop the 'Jedward' candidate"

This might be true, but one word: Boris.

Sunder Katwala said...

If most viewers wanted Jedward in the end, of course they should win.

That Boris won fair and square should lead Tories to ? their kneejerk opposition to preferential voting as some sort of political stitch-up.

Sunder Katwala said...

Here is a comment from Mehdi Hasan of the New Statesman.


Good blog. “Extensive, detailed and authoritative”, to borrow an Iraq line from Tony Blair.

In fact, I’ve never seen the case for AV expressed so forcefully or convincingly. But you still haven’t quite convinced me. The biggest objection remains the disproportionality of AV and the evidence suggesting that the 1997 Labour landslide would have been even bigger under AV (200-plus?).

In fact, according to the Jenkins commission, even the Tories would have had a slightly bigger majority in 1992 under AV. It’s a fudge of a system and I, for one, am fed up of Labour fudges.

Here’s wise ol’ Woy Jenkins’s verdict on AV:

The Commission's conclusions from these and other pieces of evidence about the operation of AV are threefold.

First, it does not address one of our most important terms of reference. So far from doing much to relieve disproportionality, it is capable of substantially adding to it.

Second, its effects (on its own without any corrective mechanism) are disturbingly unpredictable.

Third, it would in the circumstances of the last election, which even if untypical is necessarily the one most vivid in the recollection of the public, and very likely in the circumstances of the next one too, be unacceptably unfair to the Conservatives.

Fairness in representation is a complex concept, as we have seen in paragraph 6, and one to which the upholders of FPTP do not appear to attach great importance. But it is one which, apart from anything else, inhibits a Commission appointed by a Labour government and presided over by a Liberal Democrat from recommending a solution which at the last election might have left the Conservatives with less than half of their proportional entitlement.

We therefore reject the AV as on its own a solution despite what many see as its very considerable advantage of ensuring that every constituency member gains majority acquiescence".

Sunder Katwala said...

I should have linked Mehdi Hasan's New Statesman blog with that last comment.

The Provisional BBC said...

Hi Sunder,

I've disagreed with you several times over at LC, but don't think I've done you the courtesy of laying out my objections to AV yet.

I'll confine my discussion to AV, rather than AV+ which I oppose more vigorously. My basic position is that I can live with AV, I acknowledge that it has some strengths over the current system but prefer the current system overall.

It should also be noted that I'm talking about parliamentary elections only here. In theory my ideal system would probably be district councils elected under AV, county councils abolished, regional assemblies elected under STV, parliament elected under FPTP, House of Lords abolished!

Yes, it is difficult for fascists to get elected under AV. It's difficult under FPTP, too. No BNP member has ever been elected to the House of Commons. It is slightly more difficult under AV than under FPTP, so if I had no other concerns about AV, I'd support the change. But the small additional difficulty is not reason enough to sway me.

No more dodgy bar charts is also an attractive proposition. I think you do overstate your point slightly here. We know from previous experience that some parties will use dodgy bar charts even in full-PR elections! Nonetheless, I accept this practice would be reduced once most people understood the new system. It won't necessarily stop tactical voting, though. We could get a different kind of tactical voting, where people try to send messages through the voting system by voting for single-issue candidates first, then the candidate they want second, on the assumption that the single-issue candidate will be knocked out before the candidate they want. Incentivising this kind of behaviour could produce anomalies - the French prsidential election of 2002 wasn't run under AV but suggests one way this could go wrong.

Ok, on to arguments against AV:

The first is the slippery slope argument. Not traditionally a strong argument, but your article shows how it could happen! AV may pave the way towards full PR, by introducing a more pluralist kind of politics. So for those of us who have objections to features of PR that are not present in FPTP or AV, there is an instrumental argument for opposing AV.

You assume that readers will be in favour of more plurality. I'm not. The thing I value the most about the British party political system is that the fundamental cleavage of our politics is class. The Labour Party, when it is doing its job properly and advancing the interests of the section of society it was set up to advance, should be fundamentally opposed to every other party, in reflection of class antagonisms. There should be no plurality, no politics of mutual respect.

The biggest problem is the uncertainty. You say yourself that "Nobody knows for sure would gain or lose as voters and parties adapt to a more open system." Would more independent/single issue candidates win parliamentary elections? We don't know. Could we end up with schisms that are more important than class in our political system? We don't know. When I am broadly happy with FPTP, why would I take that risk?

Hopefully you can now accept that it's possible to have reasons in support of FPTP and against AV that are logical, but display different preferences about what kind of political system and culture we want than yours?

tim f

Sunder Katwala said...

Tim F

I appreciate the thought out response. And it is interesting that you could live with AV. I can see the 'instrumental' objection, on which you don't place much weight, as well as the anti-pluralism one as a different conception of politics. (I might see that this as now trying to reverse something which has changed over the last 3 decades under FPTP, but you might counter that is more reason not to open the floodgates).

On "Labour as a class party", it would be interesting to see what you make of the Soskice/Iversen argument put by Stuart White in the next thread up.

If we were to count working majorities as being those in double figures, the 20th century 1918-1997 score between the Tory-led right and Labour-led left was something like 13-3 to them. So the Tories have mostly done the politics of class alliances under fptp rather better than Labour, it would seem.

Cantab83 said...

Sunder, I think you are being overly optimistic about the extent to which AV would increase plurality in British politics.

Yes AV has many of the positive aspects of FPTP:
Maintaining the constituency link.
Excluding extreme minority candidates.
Delivering a majority government.

And it has the additional benefits of eliminating the need for tactical voting while conferring greater democratic legitimacy on the elected MP.

But the idea that it would lead to greater cooperation between parties as you suggest in Point 5, is I believe misplaced. Just because the votes are transferable between candidates does not mean that the way those votes are transferred is within the power of the losing candidate or party. Any attempt by those parties to suggest otherwise would be portrayed as a pact or deal of some sort that would be damaging to the parties involved. History shows that the electorate resent being taken for granted or told how to use their vote. So parties would continue to keep their distance and prosecute their phoney policy wars.

Another point that many fail to acknowledge is that the projections for the election result that a future AV system might deliver assume that the constituency boundaries would remain the same as now. This is highly unlikely. If the electoral system were to change, then the Boundary Commission would surely attempt to mitigate any trend towards greater dis-proportionality in the result by redrawing the electoral map.

However, I do see AV as a positive move in the right direction, but it is not the be-all and end-all of electoral reform. Far from it! Without other reforms AV will be no more than a form of palliative care for our electoral system.

A majoritarian system like FPTP or AV is OK for choosing a government, so long as the power that the government can exercise is itself constrained by some form of democratic proportionality. That is why AV (or FPTP) can only work fairly when it is combined with a reformed Lords with real PR. And that means we need a reformed constitutional settlement between the two legislatures with a new Parliament Act that removes the supremacy of the Commons.

Finally, what about other measures to improve electoral legitimacy and accountability?
(1) Compulsory voting.
(2) Term limits for MPs to abolish jobs for life and increase the amount of new blood in Parliament.
(3) Mid-term elections for the Upper House.

If we want real change in the way politics works in Britain we need a lot more than just AV.