I'd like an explanation

[andrewducker]

So far as I understand it, the current division of voting districts in the UK is unfair.  Labour gets a boost such that it will always get more seats than the Conservatives would given the same number of votes.

There's a bill going through at the moment which will reduce the number of MPs to 600 and alter the size of distrcits to be more even.

The first bit I'm leaving to one side for the moment - it's the resizing bit that seems to be causing the most fuss.

Labour are decrying it as gerrymandering.  And I can see that it _could_ be used that way.  But as no statement has been made about what the new districts will be, I can't see how on earth they can come to the conclusion that they will be slanted against them.

I don't understand how there can be an argument against voting districts being of (roughly) equal size, in the interests of fairness.  Can anyone enlighten me?

Comments

[gonzo21.livejournal.com]

As far as I can tell Labour are just upset because they will have fewer seats. Though the results will be fairer.

I think I'm generally in favour of the plan, although I can see it being problematic in very rural areas, where one MP in the Highlands say, might wind up having to cover an impossibly large area of diverse communities.

[a-pawson.livejournal.com]

It has already been announced that the seats of Orkney & Shetland and Na h-Eileanan an Iar (essentially the outer Hebrides) will be excepted from this. They will continue as they are. Other than that, I can't see it being a problem. The Highland constituencies are already spread over a huge and often remote geographical area.

[gonzo21.livejournal.com]

Ah, fair enough. I can't see any objection to it then.

[a-pawson.livejournal.com]

The result will be slanted against Labour because Labour hold many more seats in large cities, which is where the majority of the seats with lower electorates are.

Since the actual boundaries are to be set by an independent body (the electoral commision), and the aim is to make each constituency cover roughly the same number of people (with two exceptions), it doesn't strike me as unfair in the slightest. Essentially there new arrangements are likely to be fairer. But Labour are likely to lose out overall terms, and so are complaining rather loudly.

[khbrown.livejournal.com]

When constituency boundaries are created is it on the basis of the registered electorate in an area or the population in that area? Are there sometimes statistically significant differences between the two?

[the-gardener.livejournal.com]

Answer to your first question: the registered electorate.

Answer to your second question: potentially. But it cannot be said what the differences might be, because (a) the population census is taken only once every ten years, and (b) by definition it will not be known what percentage of the electorate has failed to register to vote.

[drdoug.livejournal.com]

I suspect a lot of the force behind the protest is straight-up political self-interest.

But I think there is a potential genuine issue about the process by which the boundaries are to be drawn. It's frustratingly hard to get any actual details, but I understand that Labour have concerns that the new process won't have multi-party input and agreement. If you're worried that the process has insufficient scrutiny to prevent it being gerrymandered, it's silly to wait until it produces gerrymandered constituencies and object at that point.

[bohemiancoast.livejournal.com]

Although Labour tend to lose out because they have more of these small seats, the real issue is in how the boundaries are set -- this can make a huge difference.

Labour argue that there are more people in the urban areas than are on the electoral register -- and that's probably right, but the solution is to make the electoral register complete, not to over-represent voters in those areas.

[errolwi.livejournal.com]

Or you base seat size on population as recorded in the census, which will presumably have less variation from the truth than self-selected enrolment.

The Australian method is also interesting (e.g. they are only allowed a projected variance in electors of 3.5%)
http://www.aec.gov.au/FAQs/Redistributions.htm

[the-magician.livejournal.com]

The main problem with the census is that it only happens every ten years. I'm also not sure whether the information about the number of eligible voters is accurately captured or whether it is legally allowed to be released to the independent body in a format that allows the number of eligible voters per smaller area to be determined. Without that, you can't accurately estimate how many voters there are under your new parliamentary district.

[pseudomonas.livejournal.com]

Or you add a question (or questions) to the census that allow one to determine whether a respondent is permitted to vote.

[andrewhickey.livejournal.com]

It's two reasons:
1) "Wah! No fair! We might lose out!"
2) "This gives us an excuse to block the AV referendum we supported in our manifesto, meaning we can shaft the Lib Dems and hopefully split the coalition. Even though that'll mean an election and a Tory majority government, because there's no way the Lib Dems would ever work with us after that, it's still worth it because we'll be the only 'progressive' force left!"

[andrewducker]

To be fair, the quote I heard was that if the resizing and the referendum were split into two bills thy'd support the referendum one.

Personally, I think they'd be crazy to oppose AV - getting rid of FPTP pretty much removes any chance of a Tory government ever again. That's not why I support it - but I can see the appeal.

[andrewhickey.livejournal.com]

Yeah, but the whole reason they said that is because the coalition agreement specifies that they be one bill. And the whole reason for *that* is so that they could use the fact that the resizing was in the Tory manifesto to keep Tory rebels from blocking the AV referendum...

A switch to *STV* would probably block any future Tory gov't (without a coalition partner), but an AV switch would mean a much less dramatic change - hence the Tories being OKish with it... of course, AV makes it much, much easier to then transition to STV (as, actually, do equal-sized constituencies).

[matgb]

getting rid of FPTP pretty much removes any chance of a Tory government ever again

I don't think that's true. All else being equal, from what I've seen, under AV, 2010 would have given the Tories a majority, LD voters were splitting against Labour this time around. Well, specifically against Brown, but...

But it does reduce their chances of getting a thumping majority, would have stopped Thatcher in the 1980s, and really helps all smaller parties, although the LDs disproportionately so.

Watching Scottish politics under AV is going to be interesting. I predict the Scottish Conservatives will wrap up and there'll be a new Free Democrat style party and a smaller religious nutters party fairly smartish.

[zz]

I want to be a rotten borough. *sulk*

[cheekbones3.livejournal.com]

Depends how you define size. Currently they theoretically cover equal numbers of people, although for more extreme constituencies this isn't the case. I'm not sure whether the proposals are thinking of more rigidly having even numbers in every constituency, but I imagine that if that is the case, Labour's urban strongholds will be diluted a little, and the Tories rural strongholds strengthened a little.

The general problem with enforcing equally-populated constituencies is that some will be absurdly huge, and you end up lumping disparate areas together such as in the North of Scotland, or the Isle of Wight.

The best solution? Not sure actually!

[phillipalden.livejournal.com]

I don't think politicians like "equal" voting districts.

We had a huge gerrymandering scandal here in the U.S. and a very powerful Senator (Tom DeLay) lost his seat, (and political career.)

[missedith01.livejournal.com]

It's the way the size of the constituency will be counted. I think the ConDems are planning to discount unregistered voters, which disproportionately affects Labour constituencies because Labour supporters tend to be poor and poor people tend to be more likely to fail to register for a number of reasons.

[ciphergoth.livejournal.com]

Now that *is* interesting! Any links to more on this? Thanks!

[missedith01.livejournal.com]

There's a bunch of stuff out there but I offer a link to a recent Peter Hain article in the Guardian as a starting point as it uses the word spatchcocking. I never heard the word before but I now intend to introduce it in as many conversations as poss. I shall spatchcock it into this post, for example! ;-)

http://www.guardian.co.uk/commentisfree/2010/jul/28/tories-sandbagged-clegg-electoral-reform

[andrewducker]

Thanks for that!

[andrewducker]

I just found this:
http://markreckons.blogspot.com/2010/07/left-fisk-forward.html
which points out that it's the same method that was used under the previous Labour government.

Now, there may be an argument to say that it's not the method that should be used, but it doesn't half feel like hypocrisy.

[missedith01.livejournal.com]

Now, I think he's wrong and that the Boundary Commission don't currently do what is being proposed, but I must say I haven't checked their rules and I now can't find an authoritative statement of them online. (pah)

However I think even if this was the case (method proposed is same as method now used) I am going to disagree with MarkReckons for the following reason: the current method of setting boundaries allows for quite a bit of spread from the average number of voters in a constituency. This, as I recall, is because we like parliamentary constituencies which make sense on the ground, for example, match up reasonably well with local govt boundaries. In a system like this, if a few thousand voters fail to register (perhaps due to some pernicious policy like the poll tax) it's not likely to make any difference to the BC's thinking regarding what makes a sensible and tidy constituency in that area. The constituency persists, and hopefully later the voters reregister. If numbers of people is only one of several factors affecting the setting of the boundary, not the most important factor or one on which there is plenty of wiggle-room, the way people are counted is not likely to be crucial.

The system which is being proposed moves us away from this towards a constituency being about a parcel of the country containing a certain number of voters. In a system like that, the question of who is registered and who isn't at the time of the review becomes very important. If one makes numbers of people the single most important element in setting a boundary, the way people are counted becomes very important indeed.

[andrewducker]

Gotcha. I can understand that as an argument. Voting with the "natural" community you're a part of is, after all, rather traditional in this country. I can see there being a lot of places where it's just not going to be possible to line up the local towns/villages and a voting unit of the right size.