Oldham East & Saddleworth, AV and political dialogue

[strange_complex]

I'm still busy campaigning away for Yes to AV, and will probably write more about the local campaign soon. But today I'd like to take a little time to point out why the Alternative Vote can deliver clearer messages to politicians about what voters actually want than First-Past-The-Post, using the example of this week's by-election in Oldham East & Saddleworth.

This by-election came about because the man who won the seat at the General Election, Phil Woolas (Labour), was found guilty of knowingly telling lies about his closest opponent, Elwyn Watkins (LibDem) in his election literature. The judges ordered a re-run of the election, and Woolas was suspended from the Labour party.

The resulting by-election attracted a great deal of media scrutiny, and prompted very sustained campaigning efforts from both the Liberal Democrats and Labour: the latter now headed by a replacement candidate, Debbie Abrahams. And the basic reason for all of this was that by-election results are normally seen as diagnostic - a sort of mid-term barometer reading on how all of the political parties involved are doing in the eyes of the electorate as a whole.

So what data has the result yielded, and what can it tell politicians about what voters are thinking? Let's compare the detailed results in OE&S from May 2010 and January 2011:

May 2010 - Turnout: 44,520 (61.2%) +4.4

Candidate	Party	Total votes	Percentage	% change
Phil Woolas	Labour	14,186	31.9	-10.7
Elwyn Watkins	Liberal Democrat	14,083	31.6	-0.5
Kashif Ali	Conservative	11,773	26.4	+8.7
Alwyn Stott	British National Party	2,546	5.7	+0.6
David Bentley	UK Independence Party	1,720	3.9	+1.8
Gulzar Nazir	Christian Party	212	0.5	+0.5

January 2011 - Turnout: 34,930 (48.0%) −13.1

Candidate	Party	Total votes	Percentage	% change
Debbie Abrahams	Labour	14,718	42.1	+10.2
Elwyn Watkins	Liberal Democrat	11,160	31.9	+0.3
Kashif Ali	Conservative	4,481	12.8	-13.6
Paul Nuttall	UK Independence Party	2,029	5.8	+1.9
Derek Adams	British National Party	1,560	4.5	-1.2
Peter Allen	Green	530	1.5	N/A
Nick "The Flying Brick" Delves	Monster Raving Loony	145	0.4	N/A
Stephen Morris	English Democrats	144	0.4	N/A
Loz Kaye	Pirate	96	0.3	N/A
David Bishop	Bus-Pass Elvis	67	0.1	N/A

Firstly, overall turnout was lower than at the General Election, as is typical for by-elections. So it's no good comparing overall numbers of votes - we have to focus on the percentage won by each party. On this basis, the obvious story is a collapse in the Conservative vote (-13.6) and a boost in the Labour vote (+10.2), with the Liberal Democrats holding more or less steady in the middle (+0.3).

But how do we explain this, and what does it mean? One problem with first-past-the-post is that we can't really tell. Clearly, it's out of line with how the three main parties stand in nationwide opinion polls, which would have conditioned us to expect a collapse in the LibDem vote rather than the Tories. But why didn't this happen? Is it because Elwyn Watkins had such a good relationship with the local electorate that his own vote has held up in spite of nationwide discontent with the Liberal Democrats as a party? Is it because Tory voters in the constituency deliberately decided to vote tactically for the LibDem candidate in an attempt to keep Labour out, as numerous commentators have suggested? Or is it some combination of both?

Under FPTP, we can only speculate about what happened, and what it all 'means'. At best, opinion pollsters can ask a sub-set of the voters about the reasons behind their decisions. But wouldn't it be even better if all of the voters in this constituency could have expressed their preferences clearly and unambiguously at the ballot box in the first place?

This is one of the things that AV offers. Imagine that Tory voters in particular - the main group suspected of tactical voting in OE&S - hadn't had to do so. Imagine that they could have expressed their true preferences by ranking candidates instead. Then we would be able to see more clearly what the real picture was.

The suggestion is that a third of people who voted Tory in the general election actually switched their votes to the LibDems in the by-election - not because they positively supported the LibDem candidate, but because they negatively opposed the Labour candidate, and saw Elwyn Watkins as best placed to defeat her under an FPTP system. If this is true, AV would make that shift, and those feelings, transparent. Under AV, we would be likely to see those same voters, now freed of the need to vote tactically, being able to put the Tory candidate whom they really preferred first, in the knowledge that if he were eliminated, their second-preference votes for the Liberal Democrat candidate would still be carried forward to help decide the final outcome.

This means that AV would have allowed local Tory voters to express their preferences in more detail, transmitting more information about what they actually thought about the candidates running for office via their choices at the ballot box than they were able to under FPTP. Indeed, this is of course true for all of the voters in the constituency, not just the Tory ones. I'm simply picking on local Tory voters because they are the most obvious example here of voters whose preferences we can't fully understand when expressed only through the FPTP system.

To me, what all this demonstrates is that AV is a superior electoral system to FPTP because it conveys a clearer message to politicians about what the people who are voting for them actually think and want. It enhances the political dialogue by making it easier for voters to indicate to politicians when they approve or disapprove of their actions. And if politicians want to do well under an AV system, they will need to listen and respond to the extra data which voters are providing to them. Indeed, they will have to, because of the requirement which is also part of the AV system that a winning candidate must secure the support of at least 50% of their voters. In other words, AV should mean an electoral system which is literally more democratic, because politicians become more responsive to the will of the voters.

I'm also quite aware that in this particular constituency, the likely message from voters to politicians which AV would have transmitted clearly, and which FPTP did not, is disenchantment with the Liberal Democrat party. If the LibDem share of the vote really did hold up partly thanks to Tory tactical voters, then that is perhaps not such great grounds for self-congratulation as Tim Farron tried to claim afterwards. And you know, much as I am still steadfast in my support for the party, I think it would actually be better for us to hear the clearer message that AV would have delivered in OE&S. It would have been a more accurate barometer of what we're getting right and what we're getting wrong, and a more helpful guide as to how to do something about it.

Of course, ultimately no by-election result can ever be a really accurate reflection of national political success or failure. Obviously local issues and local personalities play a huge role; as does the scope which voters tend to feel for registering a 'protest vote' in a situation where they know that it will have no effect on the party of national government.

But I would like all voters to be able to express their opinions more clearly in both by-elections and general elections. I want a better dialogue between voters and politicians - one in which our voices are stronger, and our candidates are forced to listen more attentively.

And that is (just one reason) why I am saying Yes to AV.

Click here if you would like view this entry in light text on a dark background.

Comments

[diffrentcolours]

On the doorstep, I heard just about every form of tactical and philosophical change in vote that you can imagine.

I saw Tories tactically voting Labour to break up the coalition. Tories tactically voting Lib Dem to keep Labour out. Lib Dems voting Labour over tuition fees. Former Lib Dems voting Tory because they couldn't stomach voting Labour or Lib Dem. Labour voting Lib Dem over Woolas. Tories and Labour switching to Lib Dem because they've "proved themselves" somehow.

Obviously the philosophical changes wouldn't show up as easily under AV. But there was a huge amount of odd voting during that by-election...

[strange-complex.livejournal.com]

Thanks - yes, that is a lot of oddness. I can't wait until AV brings an end to tactical bloody voting. It's such a ludicrous distortion of what people actually think.

[glitzfrau.livejournal.com]

Oh, people still vote tactically under PR. They just use different tactics.

[strange-complex.livejournal.com]

Although AV isn't PR, of course. It's preferential, not proportional.

This very lovely flowchart (http://www.anthonysmith.me.uk/2011/01/17/how-complicated-is-the-alternative-vote/) did the rounds yesterday, which (although obviously over-stating the case slightly for comic effect) captures beautifully how AV frees people from the need to vote tactically by comparison with FPTP. That's basically where I'm coming from.

[fearmeforiampink]

I saw that at a phone bank then too, I thought it was rather awesome, though I did raise the whole "People will still do some tactical voting once they're used to the system".

Essentially the tactical voting comes in if you think your second choice party will be eliminated before your first choice party, and your first choice party doesn't have much chance of getting in.

As an example; parties A, B and C. Parties A and B will always put each other as 2nd choice, C will put B as their second choice. C gets 40% of the vote, A gets 30%, B gets 20% (and we'll ignore the last 10% as it makes the example easier, say they voted for smaller parties with no second preferences).

Party B gets eliminated, their 20% goes to party A, party A gets 50% and wins.

However, if C/the C voters realises that they won't win, then they can move some of their first choice votes from C to B, so that B isn't eliminated first, A is. So, 11% move from first choice C to first choice B. At which point the numbers are A = 30%, B = 31%, C = 29%. C is eliminated (in another set of numbers, it might be B that was eliminated), and B jumps to over 50%.

In the first (non-tactical) situation, the representative is the first choice of 30% of the voters, and is acceptable/supported by/second choice for 50% of them. In the second (tactical) situation, the representative is the true first choice of 20% of the voters, but is acceptable/supported by/second choice for 90% of them.

Obviously, that's a very notional situation, but it does demonstrate that tactical voting can exist under AV.

[strange-complex.livejournal.com]

Thanks for this - it's a good example. But I guess it also demonstrates that things have to be quite finely balanced for tactical voting to become a worthwhile choice under AV. And it would definitely take a fair while for people to get used to this sort of thinking. I'm kind of hoping we'll have moved onwards from AV by then.

[fearmeforiampink]

I'd expect a small amount at the first election and more at later elections of newspapers or similar giving people voting memos, which tell them which order to vote for people in. But you're right in that it will take a while for that to come in on any large scale, and we should've moved on by then.

On the other hand, I'd also argue that it's not neccessarily a bad thing – in the above example, the tactical voting results in a representative who may have 10% less true first preference backing, but has 40% more first-plus-second backing – everyone that we're tracking finds them acceptable as an MP, even if they're less people's first choice.

[po8crg.livejournal.com]

Tactical voting under AV involves switching to a candidate that is less popular than your first preference; how could such a candidate run an effective enough campaign to achieve their desire?

Under FPTP, you can run that campaign to get people to switch from a less popular party to a more popular party.

[primitivepeople.livejournal.com]

Be careful what you wish for.

The Scottish elections in May are under full PR - the Lib Dems will be wiped out, whatever system is used.

[primitivepeople.livejournal.com]

I think that's part of the problem - most PR systems don't really do a much better job of representing people than FPTP does, and I don't think it will eliminate tactical voting.

For example, I'm a Labour supporter who wants to keep the Conservatives and the Lib Dems out of power at all costs, so all my 2nd, 3rd, 4th etc. choices under AV won't be for people I support, they'll just be for the least-bad alternatives.

If the Lib Dems are willing to share power with Labour in Scotland while propping up the Tories in Westminster, it goes to show what an unprincipled, opportunistic bunch they are.

[primitivepeople.livejournal.com]

No, it means the electorate has no idea what it's voting for and that policy is dictated entirely by whichever other party is wooing them. The "Coalition" has been a massive disaster for the Lib Dens and it's no surprise their support has almost vanished. People don't want "pluralism", they want the policies they were promised. Tuition fees? VAT? NHS reform? Speed and scale of cuts? Lib Dem voters feel massively betrayed in these areas, and all of this matters far more to the average person than voting reform.

[fearmeforiampink]

The problem comes from the fact that some people voted for tution fees to be scrapped by voting Lib Dem, some voted for them to be kept or raised by voting Tory or Labour. Since the Lib Dems didn't get a majority, it's unsurprising they can't put in all their intentions.

I'd imagine that if (and it's a big if) coalitions become more common in this country, you'd have party manifestos being split into two categories – things which are utter red lines that the party will demand and not back down on, and things that it will fight for, but which are not guaranteed if the party forms a coalition with another party.

[primitivepeople.livejournal.com]

It's got nothing to do with the Lib Dems getting a majority. They pledged to vote against tuition fee rises, which has nothing to do with how many MPs they have. Every single one of them promised to vote a particular way, then loads of them didn't. You can justify that however you want, but people won't forget in a hurry.

[viscount-s.livejournal.com]

Personally, I’m less concerned about ‘giving politicians messages’ – I don’t think that byelections can be used as any sort of barometer as to the effectiveness of a government as people treat them as referendums on the incumbent party and – as you rightly point out – turnout tends to be lower. As such, using elections as a tool to communicate with party hierarchies is merely a byproduct of AV and not necessarily a useful one – it could conceivable lead to ‘government by focus group’ which simply leads to political parties having no core values, save the desire to say anything to win power. *

The main thing to my mind is the legitimacy issue – people who are elected via the FPTP system, and yet have less than 50% of the total vote in a constituency have, to my mind, a ‘smaller mandate’ than someone elected by the clear majority of the people. This is particularly important when major & potentially painful changes have to be enacted so having a better mandate than the largest fraction of those people who voted should be required.

S

*Yes, I accept that most political parties are this way already, but even so.

[strange-complex.livejournal.com]

Oh, I entirely agree that giving greater weight to legitimacy is a more important feature of AV than communicating views, and also that by-elections will always be a very imperfect barometer. I'm focusing on the dialogue issue in this post because I felt that the OE&S result demonstrated the room for improvement there particularly clearly. But of course the two things come together - politicians who are forced to think harder about how to appeal positively to a wider proportion of the electorate will be helped to do so by having access to better data about that electorate's preferences.