It was reported today in The Times that the planned review of constituency boundaries and cut in MPs may be scrapped due to opposition in Parliament. If the report is to be believed, the Boundary Commission will now start to design new constituencies for a parliament of 650 MPs, the current number, rather than the planned 600.
The ERS has long taken the view that a cut to 600 MPs would be harmful without a corresponding cut in the number of ministers. Without a reduction in the size of the government, the ratio of ‘payroll’ MPs to backbenchers will tip in the government’s favour, strengthening the executive’s hold over parliament – just as each MP’s case-load increases as they represent more people.
That’s not the only problem with the changed rules from 2011, however.
As well as reducing the number of constituencies, the Parliamentary Voting and Constituencies Act 2011 also attempts to equalise them, by requiring that seats’ electorates do not differ more than 5%.
This was aimed at reducing first past the post’s electoral bias that was in favour of the Labour Party at the 2010 election. But, unequal seat sizes are just one of the reasons the first past the post biases parties over each other. Abstention, third parties and the geographical locations of voters all cause first past the post to be biased.
Strictly equalising seats in this manner is risky as constituencies change from representing natural communities with shared interests, to being mathematical agglomerations of voters.
Take Cornwall, a county with an exceptionally strong identity. The county’s population gives it, with 600 seats, enough voters for 5.5 constituencies, necessitating a ‘Devonwall’ seat that crosses into Devon. With an electoral quota of 5% there is no mathematical way in which seats can be created that are wholly in Cornwall.
An excellent 2014 study by the leading electoral geographers and boundary experts Ron Johnstone, David Rossiter and Charles Pattie simulated how many potential constituencies could be created in a UK divided into 75 principle areas.
With a 5% threshold, 650 seats and allowances for ward-splitting, 54 areas had a large number of potential constituencies wholly within their borders, compared to just 34 with 600 seats and no ward splitting. With more options, the boundary commission can concentrate on making borders match local communities.
But even with 650 seats knock-on effects can happen. If, for instance, Glasgow cannot be kept whole then at least one constituency must reach into the neighbouring areas, and this could create knock-on effects throughout Scotland.
What’s more, Johnstone, Rossiter and Pattie find that the 5% threshold increases the number of people who find their constituency changed from election to election. Relaxing the quota from 5% to 8% or 10%, reduces this disruption and makes more natural constituencies to be possible.
A 10% threshold, 650 constituencies and ward-splitting, gives 69 out of 75 of their principle areas lots of options for constituency design. And in terms of disruption, the majority would see no change at each election.
Updating electoral boundaries is a vital part of the functioning of our electoral system. If it is to work, our system should represent genuine constituencies of interest and give people the stability to not have their representatives constantly shift. First Past the Post is a bad system, but there is no reason to make it even worse for the voter.
Rather than simply increasing the number of constituencies back to its current figure of 650, the government should look instead to rethinking the way we draw up boundaries in this country to make sure that the next review uses an up to date register to produce the best, most representative constituencies it can.
 Generally council wards are used as the building blocks for constituencies. Urban councils though, often have wards that are so large that without them a seat would be below the threshold, and adding them will cause seats to exceed the threshold. Ward-splitting can avoid this.