With the Conservative Party Conference set to take place in Birmingham from Sunday, Chris Game considers Tory reluctance over proposals for a revamp of the electoral system
The Conservatives conferencing here in Birmingham aren’t keen on AV – the Alternative Vote electoral system on which we’ll all have a referendum vote next May.
This year’s election was difficult enough. They finished seven per cent ahead of Labour, yet still ended up in a coalition. AV, they believe, will make it even harder to win a parliamentary majority. As we’ll see, they’re probably right. But it’s not entirely a one-way process.
The Dudley North Conservatives might have made their trip down the Wolverhampton Road to the ICC led by their own MP, had the 2010 election been fought using AV.
In fact, their candidate, Graeme Brown, was narrowly defeated by the sitting Labour MP, Ian Austin, which ruined completely their planned joke about there still being one Mr G Brown on the Government benches in the Commons.
You may recall Dudley North. It was a Conservative target seat, which earned its ambulance station a visit at the tail end of David Cameron’s all-night campaign tour during election week. It also staged surely the election’s longest count, with repeated recounts and an eventual declaration late on Friday.
The Alternative Vote is a system, which, as its name suggests, enables voters to list not only their most preferred candidate, but also their second, third and further preferences, which are then taken into account if their first preference is eliminated in the count.
It isn’t complicated though it can require several counts before one candidate emerges with over 50 per cent of the votes.
Simulating elections under different electoral systems is difficult, not least because sceptics can always argue that, if the system were different, people would cast their votes differently. True, but we can at least try.
The best attempt this time has been by an Essex University research team using their massive British Election Study database. A representative sample of over 13,000 voters was asked immediately after the election both how they had actually voted, and, using a simulated ballot paper, how they would have voted in a comparable AV election.
Respondents were invited to rank up to seven candidates, each standing for a political party. The response rate was slightly higher than the actual 65 per cent turnout, with three-quarters expressing a first preference, 70 per cent a second preference, and so on, with about 40 per cent prepared to rank all seven candidates.
From these rankings, you can calculate distribution ratios for second and subsequent preferences – for example, the proportion of voters whose first preference was Lib Dem who gave their second preference to the Conservatives (26 per cent in the West Midlands), rather than Labour (37 per cent).
After the first count, no candidate was near the majority required to be elected under AV, so in the next count the second preferences of the 173 National Front voters are distributed amongst the remaining candidates, but are too few to make much difference.
The next round redistributes second preferences of British National Party voters, plus some third preferences of NF voters whose second preferences had been for the BNP. 45 per cent of these votes go to the UK Independence Party candidate, which enables him very nearly to overtake the Lib Dem – but not quite.
When the accumulated UKIP votes are redistributed, almost half go to the Conservative – Graeme Brown in our example – which is sufficient to push him into first place, but still short of the necessary overall majority. A final run-off count is needed, and, while more of the redistributed Lib Dem preferences go to Labour, the Conservative just makes it over the line.
The theory is that, because far more voters see their second and maybe their further preferences contributing to the final result, they will feel AV is a more involving and less exclusive system than our present ‘first-past-the-post’. Moreover, the winning candidate is able to claim the support of a majority of those who voted, whereas Ian Austin became one of the 112 MPs elected with the support of less 40 per cent, and one of the 434 with less than 50 per cent.
However, if, as a Conservative, you view the world first through blue-tinted spectacles, then AV does indeed lose some of its appeal. Dudley North was very much the exception – in fact, the Essex University team reckon it was the only seat the party would have gained in 2010. Against that single gain, they could have lost 10 seats to Labour and 13 to the Lib Dems.
AV is not a system of proportional representation (PR), as the Lib Dems constantly, but correctly, point out. However, as the currently most under-represented party they would have been the chief beneficiaries from AV, both arithmetically and tactically.
Gaining seats also from Labour, they would have increased their representation from 57 to 89. It is way short of the 145 they might have won under a perfect PR system, but, with the Conservatives and Labour down to 283 and 248 respectively, Nick Clegg would have had the arithmetic to enable him to negotiate a majority coalition with either major party. Whether personally he would have welcomed that opportunity is another matter.
As for the Dudley North projection, it is even more meaningless than these simulations usually are, because the other part of the Coalition electoral reform deal – the Conservative bit – is the Reduce and Equalise scheme, which will have all of us voting next time in differently sized and shaped constituencies.
The Parliamentary Voting Systems and Constituencies Bill, which makes provision for the AV referendum, will also trigger a programme of boundary reviews. These reviews will reduce our present 650 MPs not to 585 – the 10 per cent cut promised in the Conservative manifesto – but to 600, and equalise the size of constituencies at around 76,000 electors, compared to England’s present average of around 72,000.
The Conservatives are convinced that the unequal size of constituencies is a major source of the electoral bias against them in the present system. They look at the average electorate of 72,375 in the seats they won in May and the average of 68,564 in Labour’s seats, and reckon there’s well over a million surplus votes to be saved there for a start.
It’s a fair point, in both senses. The extreme case, of the Isle of Wight having six times more electors than the Western Isles, seems a mockery of democratic fairness and equality, but even in Birmingham, Hall Green is more than 15 per cent larger than Erdington, and these are newly redrawn constituencies.
If the new electoral quota were to be 76,000, the West Midlands would probably lose five seats, giving it 54 instead of 59. Birmingham, with its 2010 electorate of 650,000, could keep its present allocation of 9, but, with all constituencies having to be within 5 per cent of the quota, it might be a close call.
The political impact would perhaps be less dramatic than the Conservatives hope. One projection, by the independent research organisation, Democratic Audit, calculates that, if the 2010 election had been fought in 600 constituencies, the Conservatives would have won 294 seats (13 fewer than their actual 307), Labour 233 (25 fewer), and the Lib Dems 50 (7 fewer).
Even without the introduction of AV, the Conservatives’ cherished and hugely disruptive reform would still have left them short of an overall majority. Arithmetically they might have contemplated a coalition with Ian Paisley’s Democratic Unionist Party, but would that really be so much more attractive than the Lib Dems?
Chris Game is a lecturer at the Institute of Local Government Studies at the University of Birmingham.
Our team of writers will be providing exclusive online content throughout the Conservative Party Conference in Birmingham at birminghampost.net/tories