Charles Kennedy attacked the electoral system in Parliament today, showing that the Liberal Democrats always retreat to their self-interest whenever they get the chance. The BBC also reported rumours that Liberal Democrat members of the House of Lords are planning to break the "Salisbury Convention" - the longstanding agreement that the House of Lords will not block proposals that were in the manifesto of the party that won the general election.
It's not heard to see that these two points are clearly linked. Many have made a lot of the fact that at the last general election the Labour Party only won 35.2% of the vote (there's a slightly higher figure, but only using figures for Great Britain) and also of the fact that in England alone the Conservatives received more votes than Labour.
Neither of these is a novel situation. In the last 100 years ago the emergent government did not have the largest share of the vote in both 1910 elections, 1923, 1929, 1951 or February 1974. In most general elections no party or formal coalition has won over 50% of the votes cast. Figures for individual parts of the UK are harder to comeby, but in 1964-1966 and again in 1974-1979 Scottish MPs made the difference at Westminster. When the West Lothian Question was first coined amidst the 1970s proposals for devolution it was no academic theory but a very real possibility.
But that was on the basis of MPs at Westminster, not votes cast in the ballot box. Currently the way our political system works it is reliant on receiving the support of a majority in the Commons and the total numbers of votes cast across the country do not play a direct role in that. It is very easy to throw some figures together and claim that the majority of the country has voted against a particular policy, but very few people ever seem to produce some number crunching that creats a majority for any package of proposals. Under proportional representation, with the series of hung parliaments that are likely to ensue (and the almost perpetual holding of the balance of power by the Liberal Democrats - no wonder they're so keen for it) there would be many, many policies that would fail the very test that the PR lobby uses to discredit the current system.
The disparity on England is another fluke of the system. It is forgotten that Labour MPs tend to get elected with fewer votes than Conservative MPs, for a variety of reasons (including the continuing decline of the number of voters in inner cities, rotten turnouts in the same constituencies and traditional Labour voters in rock solid safe seats being so disillusioned with Not So New Labour that they decline to turn out and vote for it). Redistribution should reduce the problem, though even then there will be anomalies due to the way the rules work (in London in particular the Boundary Commission rules on not tying together any more than two boroughs means that parts of the capital will have more MPs than on a strict overall allocation). But also it's a built in safeguard against any one part of the country becoming too dominant. A parliament needs to represent the nation as a whole and a system that requires multiple individual victories is one way to do it. The comparison is with many sports where one side can score more goals/runs/points but still lose overall - yet no-one seems to object to such a state of affairs.
And the Liberal Democrat proposals for PR contain some holes as well. They advocate the Single Transferable Vote. But it's telling that they propose a few exemptions for this. In the Highlands and Islands of Scotland they would retain single member constituencies. Currently most of the Highlands and Islands have Liberal Democrat MPs, including Charles Kennedy himself.
STV would also carry a lot of technical problems in allocating. The idea of constituencies with five or six MPs could be hard to implement on the ground. One example is my home county of Surrey.
Surrey has eleven local government districts. They are generally of equal size, and the current Westminster constituencies are all based on them, but with sufficient variations that make carving them up for even single member constituencies tricky. Trying to cut the county up into a five and a six member constituency would be a nightmare.
Basing the seats on the local government districts is a natural starting point, and breaching the boundaries of any would stir up a hornets nest, especially as on the numbers it isn't really necessary. (In theory STV would not require many boundary changes, but rather a change in the number of MPs each seat is allocated.)
Starting out is fairly straightforward. Because three of the districts in the east stretch from the London border to the Sussex border, whilst a fourth is completely contained within them, it's clear that one seat has to have as its nucleus Tandridge, Reigate & Banstead, Epsom & Ewell and Mole Valley. But after this things get complicated. Either one or two districts will have to be added. The only possible additions are Waverley, Waverley & Guildford, Elmbridge or Elmbridge and Spelthorne.
Adding Waverley and Guildford would result in a severe numerical imbalance in the county since Guilford especially has more than 1/11th of the electorate in the country. Furthermore it would detach Guildford from many of the north west districts that it has strong ties to. Adding just Waverley makes sense on the numbers game, but results in a seat that stretches from Farnham to Oxted, detaching a lot of the links between rural parts of Waverley and Guildford. (Indeed the village of Bramley has fought very hard to avoid being transferred from the Guildford constituency to South West Surrey in the current boundary review. The prospect of a greater detachment would severe many links in the area.) The alternative route could only feasibly go as far as Elmbridge - Spelthorne's connections are mainly with other districts, particularly Runnymede. Even then many of the ties would be weak and the seat would be undersized, again repeating the numerical imbalance. And this is just one county.
Another problem area would be Northern Ireland. The current legislation recommends the province has seventeen seats, though the Boundary Commission presently recommends eighteen as the only way to prevent a major restructuring of the province's electoral geography. But either way it becomes difficult to cut the province up into five or six member constituencies. The west of the province (the three counties of Londonderry, Tyrone and Fermanagh, or the nine districts of Fermanagh, Dungannon & South Tyrone, Cookstown, Omagh, Strabane, Magherafelt, Derry, Limavady and Coleraine) in the west make a natural unit for a five member constituency, but in the east the districts are not well spaced - Belfast City requires three MPs (the present four seats encorporate a lot of the suburbs). But then the rest of County Antrim (districts Moyle, Ballymoney, Ballymena, Larne, Carrickfergus, Antrim, Newtownabbey and Lisburn) requires four. Counties Armagh and Down (districts North Down, Ards, Castlereagh, Down, Banbridge, Newry & Mourne, Craigavon and Armagh) require five. Alternatively it would be natural to attach Belfast City to surrounding suburban districts such as Castlereagh, Lisburn, Newtonabbey and (perhaps) North Down - but this would leave the rump of County Antrim as a detached block of three MPs and Counties Armagh and Down would also be left floating. Trying to compensate by cutting up the west just doesn't work and in no way is a "natural" arrangement.
(Of course if the political process in Northern Ireland ever gets anywhere or the government ever decides to get on with thinks we could finally see the very long awaited review of local government in the province that will severely reduce the number of districts. How this will affect a notional set of STV constituencies is unclear at the moment.)
There is a lot in principle to recommend STV as a system for Westminster elections. But it falls down both on the difficulty of producing viable constituencies and also because it fails on some of the very points that current campaigners for proportional representation are making against the present system. All electoral systems have their deficits, but to make a sudden change to a system that doesn't seem to answer many of the problems and just generates some new ones is not the answer.