Cameron’s plan to remove anti-Tory bias in electoral system ‘won’t work’

David Cameron wants to cut the number of MPs by 65. As he explained in September in a speech called Cutting the Cost of Politics, that's partly because he wants to save money. But he has also argued that this would help to remove the anti-Tory bias operating in the electoral system. In this respect, according to recent research published in two academic journals, the Conservative leader is wrong.

The system is deemed biased because the Conservatives need more votes to win a seat than Labour. In 2005, in England, Tory candidates were elected in seats where the average number of people voting was 47,793. Labour candidates were elected in seats where the average number of people voting was 38,964. The overall size of Tory constituencies was bigger too. The average size of the electorate in a Tory-won seat was 73,217. The average in a Labour-won seat was 67,684.

As a result, the Tories won one seat in England for every 41,983 Tory votes cast somewhere in the country. Labour won one seat for every 28,111 votes it picked up in England. (For the Liberal Democrats, it was even worse; they needed 110,591 votes per seat.)

Cameron has argued that cutting the size of the House of Commons, by making every constituency larger, would remove some of this bias. Earlier this year he said:

I believe every vote should carry the same weight, which means levelling up the size of constituencies. That would help reduce the number of MPs, save money and give people confidence that their vote really mattered.

But, in an article in the January issue of Parliamentary Affairs (which unfortunately is only available to subscribers) the academics Galina Borisyuk, Ron Johnston, Colin Rallings and Michael Thrasher argue that the Cameron strategy is unlikely to eliminate the anti-Tory bias in the system.

They have studied previous attempts by the various boundary commissions to make the electoral system fairer by changing constituency boundaries and they have concluded that trying to make "Tory" constituencies the same size as "Labour" constituencies won't make much difference.

That's because most of the "bias" in the system does not relate to the size of particular constituencies. They say that Labour just does better because its vote is more effectively distributed, and they conclude:

Hopes among Labour's rivals that revising constituency boundaries might level the playing field are very largely misplaced, therefore. Labour continues to benefit from electoral size but its real advantage currently stems largely from a better distributed vote – it acquires fewer surplus and wasted votes than its rivals. It is also benefitting more than other parties from the general decline in electoral turnout, requiring fewer votes for its victories.

The Parliamentary Affairs essay echoes the conclusions of another article on electoral bias in the current issue of Political Quarterly. It's by Ron Johnston, Iain McLean, Charles Pattie and David Rossiter and it's also only available to subscribers. But here's their conclusion:

Many observers of and commentators on the British electoral scene consider that the present system for translating votes into seats is unsatisfactory. Among them, a number associated with the Conservative party are convinced that they have been disadvantaged at recent general elections because of variations in constituency size that favour Labour. For them, modifications to the procedures for creating and revising parliamentary boundaries – associated with a reduction in the number of MPs, which in itself will not materially affect that bias – will be sufficient to eliminate their disadvantage.

Substantial research has shown that this is not the case. It may remove the Conservatives' feeling of unfairness if constituency electorates are equalised – more so than is currently the case – but major biases will remain; they are a function of the geography of voting across whatever map of constituencies is generated, but cannot be eliminated by slightly changing the rules by which those maps are produced.