StatCounter

Tuesday 12 January 2010

The inequality of the UK's electoral system

I have blogged before about the unfairness of the UK's electoral system that is incredibly biased in favour of the Labour party. So I was interested to read this piece on Conservative Home:
"ConservativeHome Contributing Editor Paul Goodman considers the need to redraw constituency boundaries to level a playing field which is currently weighted in Labour's favour.

Let’s start with a simple comparison. Imagine that at the coming election we win 40% of the vote, Labour 30%, the Liberals 18% and others 12% [The results in the overnight ICM poll]. All else being equal, we’d have a Commons majority of eight.

Now reverse those first two figures. Imagine that Labour gain 40%, we take 30%, and the other two figures stay the same. Labour would have a majority of 138 – an 130 seat difference on the same share of the vote, according to UK Polling Report.

...

Take Liverpool Wavertree, on its 2009 boundaries. These were drawn up from data obtained as long ago as 2000 – almost a decade ago. In 2000, it had 72,256 voters on these boundaries, and was thus roughly the Boundary Commission for England’s preferred size. This year, it will have 62, 952 – some 10,000 below that ideal.

On the other hand, consider South Northants. In 2000, it had 68,773 voters. At the next election, it will have 79,563 – a growth of some 15 per cent in nine years – and will be well over the Commission’s norm.
I’m told by an academic who cited these examples to me that, unlike some others, they’re illustrative of a general trend. Voters are moving out of Labour-dominated city areas into Conservative-leaning suburbs and shires, especially in a wide belt that stretches north-east across England from Devon to Lincolnshire.
And under present rules, the Boundary Commissions are always several years behind change. As we’ve seen, the Commission for England’s latest review is based on findings ten years or so out of date. The only-just-abolished boundaries – under which Sheffield Brightside, for example, had only 50,801 electors – were drawn from data gathered as long ago as 1992.

Furthermore, our four Commissions don’t work from the premise that equal constituency size trumps everything else. New Zealand allows a deviation of 5% from either side of the preferred size, or quota. America permits less than 1%. Our Commissions balk only at variations of more than 20%.

...

Fortunately, we’re already committed to the separate proposal of reducing the number of seats by some10% – thereby cutting the cost of politics. This is reasonable in its own terms: in a Commons in which some legislators double up as Ministers, there’s no reason why MPs shouldn’t take another 10,000 or so voters. The measure should be in the first Queen’s Speech of a Conservative Government.

As part of that reduction, the Boundary Commissions should work from the principle that constituencies must be roughly the same size, with a much smaller variation from the quota, of New Zealand margins or thereabouts. County and ward boundaries should be adapted to meet this norm, rather than the reverse, as is often the case at present. Reviews should be more frequent – perhaps taking place every five years rather than ten, or being, in effect, continuous rolling assessments. The appeals procedure must remain independent, but be speeded up. In particular, the present rights of objection must be scrutinised closely."

1 comment:

John M Ward said...

Good piece (as usual)!

Of course, exact electorate size-matching isn't the only criterion: keeping complete communities within a constituency is important too. For an area of that size it shouldn't be too difficult to do without a large variation in numbers.

For local council wards it is much tighter, though I have recently conducted my own exercise on my local council's urban wards, which I think would work out quite well in that regard.

One interesting difference between local and national boundaries is that council wards are based on projected population ny the middle of the expected period the new boundaries would apply. Thus, if new boundaries were to be introduced somewhere this year, and were expected to remain in place for (say) twelve years, the population/electorate figures would be based on what is expected in 2016.

With parliamentary boundaries, it all works on current/historic data (like census data, though not quite as bad as that!) so is already out of date when the boundaries take effect. Daft and inconsistent — but what's new there, eh?