Friday 18 February 2011

How AV would have kept William Hague out of Parliament

William Hague is leading the Tory push against the Alternative Vote. Perhaps he remembers that it would have stopped him from getting into Parliament in the first place.

In most constituencies, an Alternative Vote election and a first-past-the-post election will very likely give the same winning candidate.

Whether you prefer AV or first-past-the-post depend on what you think of those cases where the winning candidate does change. This happens when most voters would prefer one of the other candidates to be the MP, rather than the candidate who is ahead on the first round. Supporters of the status quo complain that this is like giving the gold medal to the sprinter who was second. Supporters of AV respond that the candidate with most support from the voters overall is, as David Aaronovitch puts it (£) "the real winner".

The curious case of the close shave of Mr William Hague allows us to test these intuitions.

Hague came into Parliament at the February 1989 Richmond by-election, where he was defending a majority of 19,576 after the Conservatives led the Liberals by 61% to 27%.

This is one of the safest Tory seats in the country. (Half of all seats have not changed since 1970, and almost 30% since 1945). The Tories have held Richmond uninterrupted since 1910.

Yet Hague nearly lost. He squeezed home in large part because both the Social and Liberal Democrats (the 'Democrats'), who were later to rebrand as the LibDems, and the anti-merger SDP both ran candidates, campaigning hard and claiming a good chance of victory. It was towards the end of the hard-fought campaign that popular local farmer Mike Potter seemed to pull away in second. With a new field of parties, there would certainly have been voters unsure about which candidate was best placed to attempt a famous election upset.

William Hague (Conservative) 19,543 37.2 -24.0
Mike Potter (SDP) 16,909 32.2
Barbara Pearce (SLD) 11,589 22.1 -4.9
Frank Robson (Labour) 2,591 4.9 -6.9
Dr. Robert Upshall (Green) 1,473 2.8
David "Lord" Sutch (Loony) 167 0.3
Anthony Millns (Ind) 113 0.2
Lindi St. Claire (Corrective) 106 0.2
Nicholas Watkins (Liberal) 70 0.1

Hague's majority was 2,634 (5%).
Turnout 52,561 64.4

Under AV, Hague might well have had a bigger lead on the first round - probably closer to 10% than 5%.

But that is simply because first-past-the-post elections never reveal what the voters' real first preferences are. (This is why the LibDems always poll a lower vote share whenever we have PR elections - for Europe, Scotland, Wales or London). About 10-15% of voters routinely adopt a patchy "do it yourself AV" by voting for somebody other than their preferred candidate or party, while the majority of voters say they would be prepared to vote tactically where it made sense to do so.

In Richmond, Mike Potter's tactical Labour voters could have cast a real first preference, so that Labour would have been more likely to be closer to its real support around 12%.

But those voters who stayed loyal with a sincere vote for Labour rather than voting tactically to help cause an SDP upset would, under AV, also have had an equal opportunity to influence the outcome - with the chance to wipe out Hague's majority, and put the SDP and Tories close to neck-and-neck. If Potter could secure more SLD votes than Hague, he would be the choice of most voters.

It is almost certain that more Richmond voters wanted Potter to win the by-election than Hague. If so, AV would have delivered that outcome. (If not, it wouldn't).

Of course William Hague's defeat might not have been what they wanted in the SLD's Cowley Street headquarters. They would have had mixed feeling about a David Owen triumph, since the focus for the newly merged party was on seeing off the Owenite continuity SDP. Nevertheless, that intense battle over whether Britain needed three or four party politics would have been much less important to the Richmond voters. AV would have found out who most of them wanted to send to Westminster.

This does create the intriguing possibility that AV in Richmond 1989 could have both kept William Hague out of Parliament and strangled the Liberal Democrats at birth! (What's not to like, John Prescott?!). Perhaps not. However, the newly merged party - which was to trail the Greens by 15% to 6% in the European Elections that June - was extremely vulnerable, as Lord Rennard recounted. Had that shot in the arm enabled Owen's SDP to seem viable for even 18 months longer, who knows what might have happened? (Might it have been the Greens who faced a coalition dilemma last May?)

The SDP candidate Mike Potter commented on UK polling report last year that:

Had the old Alliance vote not been split, Hague would have lost that election by 10,000 votes, but would no doubt have been elected in 1992. He would not have been in a position to take the leadership when John Major stood down, but would now be the Tory leader. I suspect with Hague as leader, a Tory win in May this year would have been a certainty. Instead we are heading for a hung parliament.
(February 3rd, 2010 at 11:10 am)

Potter's AV majority may well have been under 10,000. He is right that the Tories would have been strong favourites to regain the seat in 1992, even against a popular local incumbent, though it may not have been Hague who stood again after a famous shock defeat.

Richmond is again the safest Conservative seat in the country after the 2010 election - with Hague polling over 60% of the vote when holding the seat since. But the difference between the real by-election and the hypothetical AV one does perhaps demonstrate how AV can help to bring more safe seats into competition.


But even William Hague might find a case for having mixed feelings about AV. In another parallel universe, where we still had first-past-the-post in 1989, AV might have helped William Hague later on.

Had Tony Blair introduced AV or AV+ in the 1997-2001 Parliament, there can be little doubt that the party would have had to pursue the early "Kitchen Table Conservatives" modernising strategy promoted then by new number 10 strategy chief Andrew Cooper and his colleague Daniel Finkelstein, instead of returning to a core vote strategy to try to limit the scale of their defeat in 2001, so that the Tory party's somewhat faltering attempt to reconnect to the electorate did not begin until David Cameron's election in 2005.



Mark Pack makes the good point in the comments that William Hague was only Conservative Party leader because of a version of AV being used, in successive ballots, so that MPs could transfer their votes. He would have lost a first-past-the-post leadership election.

Hague trailed Ken Clarke on both the first and second rounds, before winning on the third (when Clarke and Redwood proved with their failed attempt that candidates' pacts can not dictate how voters will transfer, with Hague winning 18 transfers to 8).

Presumably the Conservative Party used this system deliberately, so that it would find the leader with the support of most of the Parliamentary party, rather than just electing the person with most first preferences if most MPs preferred somebody else. (The case for AV is simply that a Parliamentary constituency would be better served at Westminster by the same courtesy being extended to the voters).

So perhaps it is also the knowledge that he would never have endured being Tory party leader with first-past-the-post that explains why Mr Hague is now so keen to keep it.


Mark Pack said...

Under first past the post William Hague would have also lost the Conservative Party leadership election in 1997. It was only the transfers from others that took him to victory.

Though whether him losing would have been good or bad for him and his party is a fun counter-factual...

no longer anonymous said...

What's your opinion on the Jenkins Report's view on AV?

Sunder Katwala said...

I wrote about this in a Fabian Review essay in 2007, noting that Jenkins looks at the Commons in isolation, and arguing that combining AV (a more plural majoritarianism than FPTP) with the check of a legitimate, PR-elected second chamber, could achieve broadly similar goals behind Jenkins' AV+ hybrid.


"The main argument against AV is that it can sometimes be more disproportional than FPTP. Labour would have won an even larger landslide in 1997. The Tory party was toxically unpopular: the electorate as a whole preferred a Labour to Tory government by 58 per cent to 32 per cent. Almost everybody who had voted for neither main party preferred Labour. AV will only boost a party which is generally popular in this way. FPTP is different: a party feared by the majority could triumph by organising its vote efficiently. That is because geography (an arbitrary factor) decides FPTP elections, while AV uses (politically relevant) voter preferences.

The Jenkins Commission proposed using top-up seats to mitigate AV's tendency to exaggerate decisive victories. There is a good argument in principle, but the practical barriers are formidable. The 'two classes of MP' argument returns. Under what is called 'AV plus' every constituency on the electoral map would need to be redrawn, making it difficult to hold a referendum and implement reform in one parliament (while straight AV would use the current constituencies). It would be bizarre to hold a final general election under a system rejected by voters, while holding a one-off election under AV while preparing for 'AV plus' would be unnecessarily confusing.

A much better approach than seeking to achieve every objective through a complex, hybrid voting system would be to link electoral reform with a broader democracy agenda.

*** Democracy package ***

British politics used to be 'winner takes all'. But much has changed since Lord Hailsham warned of 'elective dictatorship' and Margaret Thatcher demonstrated its practical potential when combining a majoritarian voting system and unwritten constitution in western Europe's most centralised state. We have since seen devolution, the Human Rights Act and unprecedented judicial activism. There is a strong case for codifying a new constitutional settlement.

The electoral reform debate has remained strangely isolated from these changes. But linking voting reform to the unfinished business of Lords reform can combine effective government with greater parliamentary proportionality.

Gordon Brown should propose, in Labour's election manifesto, a referendum to introduce the Alternative Vote for Commons elections, alongside a second chamber which would be 80 per cent elected, by proportional representation. This democracy package could also include a review of local government, including voting systems (where there is a strong case for PR, perhaps by STV given the existing tradition of multi-member constituencies).

Unknown said...

All of which makes some sense, if you treat it on a case-by-case basis. Then generalise it to a Parliament, and see what happens.

Here, the combined centrist parties had 54% of the vote. Giving the current "centre party" 40% of the national vote, and assuming an AV ballot, Electoral Calculus predicts a Lib Dem majority of 392.

And there's the problem.