"Fair seats"

We are all familar with the LibDems' campaign for "fair votes" (a populist name for proportional representation - the case for which was demolished by William Norton here and here).  Perhaps the Conservatives need to be campaigning for "fair seats" or "fair constituency sizes" or something that rolls off the tongue a little more easily?

The unfair electoral system was the core theme of an article by Michael Brown in this morning's Independent.  He notes the huge gap between Brown losing his majority (he only needs to lose 32 seats) and of Cameron winning a majority (the Tory leader needs to win another 130 seats).  Mr Brown, a former Conservative MP, concludes that "a hung Parliament is now becoming the most likely outcome in 2010" and that it is in the interests of the Tories to move in the direction of PR because of the "bias and unfairness in the electoral system".

Michael Brown jumps to the wrong conclusion.  What we need isn't a change to our electoral system but a revolution in the process by which constituency boundaries are updated.  It currently takes, William Norton has noted, 26,908 votes to elect a Labour MP but 44,373 votes to elect a Conservative.  A large part of the explanation for this is the fact that southern and rural seats (which tend to be Tory) are decidedly larger than northern and urban seats (which are more likely to be Labour).  This difference in size reflects an often yawning gap between census reviews and boundary commission reviews - a yawning gap that Labour has little incentive to remedy.  Changes to constituency boundaries can take place many years after a population has shifted towards the south and the countryside.  There is also a problem of Scottish and Welsh constituencies being smaller although this is less of an issue than it was.

Perhaps Ken Clarke's Democracy Taskforce could investigate ways of accelerating the boundary review process?

Comments

[Moral minority]

"William Norton has noted, 26,908 votes to elect a Labour MP but 44,373 votes to elect a Conservative."

That is largely due to lower turnout in Labour-held seats. I agree with the proposal for regular boundary reviews and cutting back the number of Welsh and Scottish seats.

[Editor]

Thanks MM, differential turnout is certainly a factor here. It won't be possible to eliminate the whole problem.

[Justin Hinchcliffe]

I agree with Michael Brown.

[Christopher Mahon]

I entirely agree with this, and seems a decent (albeit only partial) solution to the West Lothian Question.

If we reduce the size of the Scottish seats to match their English counterparts then we too will reduce the importance of the West Lothian anomaly (Scottish votes on English only issues)

Seems a much simpler way round this question than say English parliaments or Grand Committees.

[Buckinghamshire Tory]

Does anybody know if this is a new situation (that traditional Tory seats are larger than traditional Labour seats)?

Or was this a problem for previous Conservative Prime Ministers (like Major, Thatcher, Heath and McMillan) as well?

[Richard]

I thought Ken Clarke was looking at this!?
I thought there were proposals along the lines of a statutory +/-5% difference in constituency size.

[Editor]

Thanks Richard; I'll call Mr Clarke's office and try and find out....

[Boy Blue]

"If we reduce the size of the Scottish seats to match their English counterparts then we too will reduce the importance of the West Lothian anomaly (Scottish votes on English only issues)"

This has already happened. Scottish seats were reduced from 72 to 59 before the last election. One consequence was to reduce the prospects of us winning more than one.

[Alan S]

Scottish seats are still smaller though Boy Blue even after that adjustment for 2005?

[Sean Fear]

Michael Brown's article is wrong. The Conservatives are not disadvantaged by the current electoral system, although Labour are clearly advantaged by it. The real losers are the Lib Dems and minor parties.

In 2005, we won 31% of seats on 33% of the vote. The Lib Dems won 10% on 23% of the vote. Minor parties (outside Northern Ireland) won 2% of seats on 8% of the vote.

But, once we go above 33%, the system starts to work in our favour. So, if we and Labour each poll 34.5% (and the Lib Dems and Others are unchanged), we win 232 seats (36% of the new total). If we go to 36%, and Labour fall to 33% (and the rest are unchanged) then we win 260 seats, or 40% of the total.

If the Lib Dem vote drops, then the system starts to work in our favour even more strongly. If Yougov's figures were repeated in an election, we would win 371 seats, 57% of the total, from 45% of the vote.

Why should we wish to swap a system which benefits us (if less than Labour) and gives us the chance of winning an outright majority, for one which won't do either?

[Iain Lindley]

The main reason for the disparities are differential turnout and differential swing.

Differential turnout makes up most of the disparity, and short of compulsory voting or replacing constituencies with an open list (a horrible idea) there's absolutely nothing you can do about it.

The "we need an x% swing to win a majority", based on the uniform swing, is also largely nonsense. There are more swing voters in the marginal seats and (as we've seen in previous elections) if there is a change in government the marginal seats will swing disproportionately.

Where there is a problem is "population lag", whereby between boundary reviews Labour-leaning seats have seen population falls whereas Conservative seats have seen population increases. This is still a problem and more regular boundary reviews would solve it. However, with inner-city living becoming more fashionable, it is probably not as big a problem for the Conservative Party as it has been in years past. Here in Salford, for the first time in decades, the population is actually increasing, and none of the 20 wards have seen any notable decline in population.

[Iain Lindley]

Good post Sean.

[Sean Fear]

True enough Iain. A big pro-Conservative swing would see the unravelling of a lot of the anti-Conservative tactical voting that built up from 1992-2001.

In all likelihood, an outcome along the lines of Yougov's lates poll would see the Conservatives winning c.385 seats rather than the 371 that uniform national swing suggests.

[London Tory]

Does this problem not stem back to boundary changes that were made in the early to mid 90s? My understanding is that John Smith appointed Margaret Beckett to personally oversee Labour's submissions to the Boundary Commission at that time, and to lobby and object remorselessly on a seat-by-seat basis, especially when it looked as if the outcome would not be in Labour's favour. My understanding is that we (or more precisely the leadership under John Major) did next to nothing in response.

The fact is that when we had clear and visionary leadership- I am thinking of 1979, 1983 and 1987, we won easily under this system. We can do so again under David Cameron.

Michael Brown won Cleethorpes under this electoral system.

[Michael Veitch]

FPTP has 3 crucial advantages:
1) it is easy to understand for the voter
2) it provides a very clear line of accountability between the electorate and the MP
3) by making coalitions unlikely, it is much easier for the electorate to vote out an unpopular Government
PR (in any form) does none of these things, and having experienced both the AMS and STV variants first hand here in Scotland, PR is nothing short of a disaster.

[Dale]

Once again Mr Hinchcliffe has shown himself to be at odds with every other conservative.

[Robert McIlveen]

Even if all seats were of equal size there would still a pro-Labour turnout bias. Even under most PR systems there would be (unless the regions were enormous: Liverpool would still turnout less than Surrey).

The interesting way to redress the bias in the system is to slice a number of MPs off after each election to reduce the number of MPs. We are absurdly over-governed, and should aim for about 450 MPs. Taking a seat off the 10-15 most over-represented counties after each election would reduce the bias significantly, as most of them would be Labour...

[Steve]

London Tory: Agree entirely. We benefitted during the 1980s from this system. Labour is today. Swings and roundabouts. Our attempts at influencing the 1997 review were truly hopeless. This time around, under Roger Pratt, we were very good.

Sure, we review boundaries not nearly often enough, so speed them up (as Iain L suggests).

[Tim Roll-Pickering]

I've comment about the malapportionment problems before on this site but as well as the differential turnout and the time lag in the boundary changes there is another big problem.

When the Boundary Commission do their work they do not sit down to neatly divide the country up so that every constituency has exactly the same number of voters. There are several factors that are also taken into consideration:

* Using local government wards as the basic building blocks for designing seats.

* Not crossing county, unitary authority or London Borough boundaries except where absolutely necessary to remove a numerical imbalance and even then there's a limit on the number of areas they'll combine.

* Not creating unnecessary disruption so keeping existing boundaries if the numbers aren't that far apart.

* The crucial one for any justification of FPTP: respecting natural ties and not creating geographically absurd seats. It is no coincidence that both the largest and smallest seats are based on islands.

Now these rules & practices, have a tendency towards undersized urban seats. I don't think this is deliberate but rather that it's easier to get ten seats that are all close to the target number in a rural county where the wards are small enough as to make the task easy, whereas in a smaller (review wise) urban area that qualifies for 4.5 seats and has huge wards it's much harder to create seats that get the numbers on target. A county with ten seats with declining population only needs to decline by 6% to lose one. An urban area with four seats needs to drop near to 20% to lose one.

[Tim Roll-Pickering]

I've comment about the malapportionment problems before on this site but as well as the differential turnout and the time lag in the boundary changes there is another big problem.

When the Boundary Commission do their work they do not sit down to neatly divide the country up so that every constituency has exactly the same number of voters. There are several factors that are also taken into consideration:

* Using local government wards as the basic building blocks for designing seats.

* Not crossing county, unitary authority or London Borough boundaries except where absolutely necessary to remove a numerical imbalance and even then there's a limit on the number of areas they'll combine.

* Not creating unnecessary disruption so keeping existing boundaries if the numbers aren't that far apart.

* The crucial one for any justification of FPTP: respecting natural ties and not creating geographically absurd seats. It is no coincidence that both the largest and smallest seats are based on islands.

[Tim Roll-Pickering]

(continued)
Now these rules & practices, have a tendency towards undersized urban seats. I don't think this is deliberate but rather that it's easier to get ten seats that are all close to the target number in a rural county where the wards are small enough as to make the task easy, whereas in a smaller (review wise) urban area that qualifies for 4.5 seats and has huge wards it's much harder to create seats that get the numbers on target. A county with ten seats with declining population only needs to decline by 6% to lose one. An urban area with four seats needs to drop near to 20% to lose one.

[Andy Stidwill]

One of the main factors is that turnout is much lower in Labour seats than Tory ones, which means Labour needs fewer votes to win a particular seat. There's nothing that can be done about this, unfortunately.

[Editor]

Fascinating thread. I'm learning lots. Thanks everyone.

I've spoken to Ken Clarke's office, Richard, and I'm told he's not looking at the size of constituencies. I'm not giving up and am calling one or two other people on the subject...

[Andrew Woodman]

Are we not committed to reducing the number of MP's? I would presume that means making seats that represent 100000 instead of 80000. Would that not be a solution to this problem.

[Joe James Broughton]

That is the main reason.

Although, I think some of the seats are suspiciously drawn - York Outer, and the
abolished Vale of York.
(Although am hopeful Tories will win York Outer on the basis of the May 2007 result +14 points since May 2003).
It still looks like Labour have been better at arguing for doughnutting to protect or offset seats that are lost, and there still seem to be a few very small seats that have survived, and will probably decline further.

As I understand it, the Boundary Commission are not allowed to use projected electorates/populations, just a fixed point in time.

The safe Labour seats generally have low turnouts. Many Labour MPs are returned
with just 17,000 or 18,000 votes even on shares of well above 50 per cent. But it's 50+ WITHIN THOSE SEATS, and still counts for less when added to a national vote.

One idea - but probably unworkable in a democracy - and not thought through - is
to merge seats where turnout is below 55%, but that is probably unacceptable an
d would have lots of unintended consequences.

But if the Conservative vote rises several points to around 40+ per cent, the electoral bias should - in part - unwind - but not completely (1992).

[Andrew Woodman]

I'm behind you Editor!!

[Steve]

We've definitely argued better this time around. Labour will lose somewhere between 10 and 20 seats next time automatically. And quite right too!! Lol.

[Andrew Woodman]

Not helped us in some areas though Steve. In Derby we've gained 1 extra safe seat but made 3 others a lot more difficult.

[Sean Fear]

Broadly speaking, the boundary changes give the Conservatives another 12 seats, and reduce the number of Labour seats by 7. That is better than 1997, by far.

[Dave H]

Based on the 2006 electorate figures the following English counties are due extra or less seats even after the new boundary changes:
Cambridgeshire +1
Greater Manchester -1
Oxfordshire +1
Somerset +1
South Yorshire -1
West Midlands -1.

Also if seats in the other parts of the UK were the same size as in England, then

Wales would be down 7 or 8
Scotland down 3 or 4
Northern Ireland down 2.

Surely post devolution there can be no argument why Wales and NI require smaller seats than England?

All parts of the UK should use the same electoral quota and ways of speeding up the boundary reviews or using more upto date/forecast data introduced.

[rightsideforum]

"One idea - but probably unworkable in a democracy - and not thought through - is
to merge seats where turnout is below 55%, but that is probably unacceptable an
d would have lots of unintended consequences."

LOL I like the thinking.

I think the BEST thing we can do as political parties is to drive the turnout up.

It really is a disgrace that so many people don't get out to vote. It is your DUTY to vote. People often talk about how we should 'give them a reason' to vote which is of course nonsense.

People have every reason to vote, and near enough no excuses not to.

[eugene]

I have to say that sometimes the Tories are the problem. The Isle of Wight is easily large enough for two seats.....and the Tory MP wrote to me saying that they like it that way...''do not want to split the island''. Britain is an island for crying out loud. The next election will probably be so close....lets hope by refusing to even think about sliting the Isle of Wight does not rob us of a majority.

[Tim Roll-Pickering]

eugene: The Isle of Wight is frankly without easy solution because it isn't quite large enough for two seats. You either have one seat with is the largest in the UK (as per the status quo), two seats which are the smallest in England (and drawing a satisfactory boundary is harder than many think - indeed the mid 1990s review found this a near insurmountable obstacle) or one seat on the island and a Solent spanning "Southampton Central & Cowes" constituency that has no natural ties at all.

Dave H: Northern Ireland is guarenteed "between 16 & 18 seats" in the current legislation with 17 the optimum target, though the legislation creating the Assembly would get in a mess if there are anything other than 18 constituencies (as it assumes 6 members per constituency & 108 overall). Also it's hard to come up with a workable 17 seat solution, let alone a 16 seat one - you'd have a real mess in cutting Belfast to three seats for a start as you'd need a "Belfast South West" seat combining the mid & upper Falls with the Malone Road - not exactly the most natural ties going!

Scotland does use the same quota as England, it's just that because of islands and existing local authority boundaries it's not possible to get the exact number of seats that a strict application of the formula demands (indeed most of your "3 or 4" oversize disappears if we accept the principle of at least one seat for each island group). Wales similarly has some awkward geography. The problem is compounded by the way quotas are calculated on the basis of the *existing* number of seats, thus ratcheting up the growth.

[Old Hack]

Another more conservative Conservative idea may be to reduce the number of Parliamentary seats. There is no reason why an MP could not handle around 100,000 constituents (Andrew Turner MP already does this in the Isle of Wight).

One might also redistribute seats from urban constituencies with smaller electorates to more suburban and rural areas which tend to be under represented.

The way to do this is simple - tighten up Boundary Commission guidelines and set some tougher boundaries which a review recommendation must meet.

If one wanted to go one step further one might specify that where the case for an additional is marginal that one should round down for boroughs and round up for county seats.

[Gordy D]

Tim Roll-Pickering, I totally agree.
I would not dare go down the Malone Road late at night, it is much too rough. But I dont see why the likes of inner Belfast could not be cut down to three seats. There has ben a move out to the suburbs just like other cities.

[Tim Roll-Pickering]

Gordy D: Belfast has overspilled beyond its council boundaries and indeed the 1994/1995 review (in response to an initial 17 seat proposal containing a three seat Belfast) came to the conclusion that the declining electorate in Belfast is best handled through expanding the seats into the overspill suburbs (many of which have more in common with their neighbouring parts of Belfast proper than two very disparate parts of the city do) rather than removing one seat and trying to put the remaining three together. From recollection very few submissions to the most recent review suggested a three seat solution, even when there was strong debate about which precise wards to add to Belfast seats.

Old Hack: To be honest the precise number chosen for the target sized seat isn't going to make much difference. The problems of getting the numbers right with the building blocks and areas not having an exact entitlement (e.g. in the current review Surrey is entitled to 11.43 seats) would remain.

What would need to happen would be an abandonment of the use of authority boundaries as rigid and maybe even of using the local government ward as the base unit - perhaps the polling district. This may sound like a simple solution but imagine the chaos that will ensue from trying to work to more precise numbers. The potential number of combinations will rocket, the review areas would grow in size, there'd be massive debate about just which direction should be gone in to make up the numbers and so forth.

And most crucially for any single-member constituency system, the seats produced would often not reflect "natural ties" and easy geography. If FPTP is defended for anything beyond it normally delivering a single party majority in the Commons then it's defended for the natural ties. "The member for constituency 482" or the Australian style "The member for Baldwin" don't have the same effect as "The member for East Surrey".

[Tim Roll-Pickering]

Gordy D: Belfast has overspilled beyond its council boundaries and indeed the 1994/1995 review (in response to an initial 17 seat proposal containing a three seat Belfast) came to the conclusion that the declining electorate in Belfast is best handled through expanding the seats into the overspill suburbs (many of which have more in common with their neighbouring parts of Belfast proper than two very disparate parts of the city do) rather than removing one seat and trying to put the remaining three together. From recollection very few submissions to the most recent review suggested a three seat solution, even when there was strong debate about which precise wards to add to Belfast seats.

[Tim Roll-Pickering]

Old Hack: To be honest the precise number chosen for the target sized seat isn't going to make much difference. The problems of getting the numbers right with the building blocks and areas not having an exact entitlement (e.g. in the current review Surrey is entitled to 11.43 seats) would remain.

What would need to happen would be an abandonment of the use of authority boundaries as rigid and maybe even of using the local government ward as the base unit - perhaps the polling district. This may sound like a simple solution but imagine the chaos that will ensue from trying to work to more precise numbers. The potential number of combinations will rocket, the review areas would grow in size, there'd be massive debate about just which direction should be gone in to make up the numbers and so forth.

And most crucially for any single-member constituency system, the seats produced would often not reflect "natural ties" and easy geography. If FPTP is defended for anything beyond it normally delivering a single party majority in the Commons then it's defended for the natural ties. "The member for constituency 482" or the Australian style "The member for Baldwin" don't have the same effect as "The member for East Surrey".

[eugene]

Thanks Tim....that is a good argument shared by the MP for Isle of Wight. However, it is surely wrong that the 100 000 voters on the IoW have the same voting power than say a few over 10 000 Scots on Shetland have. OK- so the IoW may have to take in a bit of Southampton for the second seat....there are examples of that. Sittingbourne and Sheppey in Kent immediately springs to mind. Sheppey and Kent mainland people joke about each other in a light-hearted manner and have little in common...but it is a united seat in Westminster with no negative consequences.

[Chris A]

Absolute rubbish. Southern seats and northern seats will be almost identically sized after the boundary review. Michael Brown can't even get the numbers right as the Tories would need about 110 seats to win an outright majority and Labour would need to lose 23. Again it is a myth to say that Scottish seats are smaller - since the reduction to 59 for Scotland they are almost identical to English ones.

[Newmania]

All most enlightening ,I am going to have start popping in regularly again. Thanks .

Does it boil down to this ? FPTP obviously benefits larger Parties if you measure benefit in total votes to total seats. Sean Fear is therefore confusing two subjects. In this sense the current system might benefit any major Party. The question is whether one Party in particular ie Labour benefits unfairly. So the article is right.
The number of people voting is irrelevant if startling .In a FPTP system votes will be cast or not cast in the knowledge of their effect. The actual number in the Constituencies is the only point that matters and this will be against Conservatives as the population becomes wealthier and moves away from Labour areas . This long term drift is why the devolved settlement is of such crucial importance to Labour . Of course the boundary Commissions should be quicker reflecting the faster rate of internal migration especially for work.

I like the idea of reducing the Number of seats in Westminster for Scotland and Wales. Apart from anything else with the massive reduction of the competence in their own country they have bugger all to do and we need a lot less MP`s anyway who act as glorified social workers, another over represented fellowship

Finally reading some of the Heath Robinson variations in PR it occurs to me that people doing weak degrees in Comparative Politics have nothing better to do and the endeavour may well be something like the attempt to find the “Philosophers Stone”The points made against PR are in my view crushing .( Michael Portillo predicted that Gordon would offer the Lib Dems PR if he lost Scotland …eek)

[Richard]

Apparently I am not alone in believing Ken Clarke would be looking at this, Lord Baker was clearly under the same impression. Here is a link to his PMB from earlier this year that sought to introduce a +/-5% statutory limit.

http://www.publications.parliament.uk/pa/pabills/200607/parliamentary_constituencies_amendment.htm

[Editor]

Thanks Richard. There's a lot for us to look at here. I'll now extend my consultations to Ken Baker's office!

[Editor]

Chris A: "Southern seats and northern seats will be almost identically sized after the boundary review."

Are you sure? Genuine question.

Isn't there still a serious time lag problem between boundary reviews and population changes?

[Joe James Broughton]

" Isn't there still a serious time lag problem between boundary reviews and population changes? "

I'd have thought so yes - by definition.

As I understand it, the Boundary Commission almost certainly works them out on the basis of a fixed point in time, and can't make a projection on population or electorate.

[Tim Roll-Pickering]

eugene: It's very easy to focus on the most extreme cases (incidentally Shetland doesn't have a seat of its own at Westminster but is combined with Orkney - indeed this and the number of Northern Irish seats are, IIC, the only malapportionment guaranteed by legislation - but Na h-Eileanan an Iar aka The Constituency Formerly Known As The Western Isles does and that's the smallest) but both cases come down to the especial problems of island geography and indeed alternatives have been opposed in previous reviews. I don't think Sheppey quite comes into the same sphere (remember Sheppey is not a local government area of its own).

Newmania: It's not *just* the voting system that has produced the past results of *more than one* parties being able to secure a majority in the Commons but also the coincidence of the voter distribution and parties developing effective techniques. Other countries that use FPTP have not had such a happy outcome - indeed using FPTP contributed to the spectacular collapse of the Candian Progressive Conservative Party in the mid 1990s.

[Tim Roll-Pickering]

Editor: The answer is yes. The boundary review uses the figures at the start of the review, which does mean that the figures for both allocating the number of quotas and carving up the seats are the same. However the boundary review can often take several years to complete and since the changes are not legislated until they've all been done this widens the gap further, as does the number of elections. The current boundaries were drawn up on the 1991 figures, the new boundaries on something like the 2000 figures.

The only minor concession to subsequent shifts that I'm aware of is that the Boundary Commission can include it as a factor of consideration when discussing local ties and the homogenuity of an area - for instance a growing town might argue that in this review it should be kept as one oversized seat instead of having bits trimmed off for its neighbours, so that next time round the town can be cut in two. There's some discussion of this in the Kent review, specifically for Ashford. However this factor can only really affect keeping seats with the existing outliers and doesn't allow the BC to create seats that are on paper numerically substantially larger or smaller than the quota.

[William Norton]

The problem of seat distribution follows from the rules that (put simply) constituencies should not cross local authority boundaries, or should only do so as a last resort as combinations of whole local authority wards, and should represent discernible "communities".

The most likely outcome is to have a county where the towns are slightly under the population quota for a whole number of constituencies and the non-town parts of the shire (everything else) is slightly over. The choice for the Boundary Commission is either to draw funny shapes all over the map which don't really correspond to real whole communities or to accept that the urban seats will tend to be underweight (and having fewer voters, it takes fewer votes to elect an urban MP) and rural seats will tend to be overweight. Which outcome wins will largely depend on the history of constituency boundaries and the evidence presented to each enquiry [query: how many sitting rural MPs I wonder have volunteered to forfeit wards to urban towns to even out the electorates?]

This problem is an inescapable consequence of having single-MP constituencies which are supposed to have roughly equal number of voters. On the whole, the country gains more than we lose. You certainly don't solve the "problems" by tinkering about with PR.

[William Norton]

The problem of seat distribution follows from the rules that (put simply) constituencies should not cross local authority boundaries, or should only do so as a last resort as combinations of whole local authority wards, and should represent discernible "communities".

The most likely outcome is to have a county where the towns are slightly under the population quota for a whole number of constituencies and the non-town parts of the shire (everything else) is slightly over. The choice for the Boundary Commission is either to draw funny shapes all over the map which don't really correspond to real whole communities or to accept that the urban seats will tend to be underweight (and having fewer voters, it takes fewer votes to elect an urban MP) and rural seats will tend to be overweight. Which outcome wins will largely depend on the history of constituency boundaries and the evidence presented to each enquiry [query: how many sitting rural MPs I wonder have volunteered to forfeit wards to urban towns to even out the electorates?]

This problem is an inescapable consequence of having single-MP constituencies which are supposed to have roughly equal number of voters. On the whole, the country gains more than we lose. You certainly don't solve the "problems" by tinkering about with PR.

[Philip]

I totally agree with the Editor. As soon as we get to power, we should implement English MPs only votes on England-only matters, and initiate redrawing constituency boundaries so that each constituency has the same number of electors.

(Of course we should never entertain any thought whatsoever on such an undemocratic idea as PR, which gives parties with least votes a share in Government)

[Ian McKellar]

The last Boundary Commission proposed a cross-river seat in Merseyside comprising four wards of Wallasey with two wards of Liverpool (with no physical contact between them). Actually this was a cross-estuary rather than a cross river seat. As it is, it is looking like Wirral West will now be the smallest seat in England with out 57,000 electors.

Wirral Borough on the formulae merits approx 3.55 seats and with a falling population, it is going to be difficult in justifying four seats. So rather than continue the link with the Merseyside Boroughs,in the case of the peninsula, the sensible option would be to link it with Cheshire rather than with Merseyside

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster with a Yes\No option if there had been no amendments by parliament, and with the 2 options of the Boundary Commissions and a no option put to the vote with a 2nd preference vote?

Another possibility would either be multi-member constituencies in more populus areas, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV - 100 candidates constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a £100,000 deposit payable for it repayable if the list entry got 5% plus of the Popular Vote then parties could stand seperately and in coalition giving people the choice of whether to vote for an absolute majority executive or a lesser majority, or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster with a Yes\No option if there had been no amendments by parliament, and with the 2 options of the Boundary Commissions and a no option put to the vote with a 2nd preference vote?

Another possibility would either be multi-member constituencies in more populus areas, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV - 100 candidates constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a £100,000 deposit payable for it repayable if the list entry got 5% plus of the Popular Vote then parties could stand seperately and in coalition giving people the choice of whether to vote for an absolute majority executive or a lesser majority, or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster with a Yes\No option if there had been no amendments by parliament, and with the 2 options of the Boundary Commissions and a no option put to the vote with a 2nd preference vote?

Another possibility would either be multi-member constituencies in more populus areas, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV - 100 candidates constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a £100,000 deposit payable for it repayable if the list entry got 5% plus of the Popular Vote then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster with a Yes\No option if there had been no amendments to the Review by parliament, and with the 2 options of the Boundary Commissions and a no option put to the vote with a 2nd preference vote?

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV - 100 candidates constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a £100,000 deposit payable for it repayable if the list entry got 5% plus of the Popular Vote then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster with a Yes\No option if there had been no amendments to the Review by parliament, and with the 2 options of the Boundary Commissions and a no option put to the vote with a 2nd preference vote?

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV - 100 candidates constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a large deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster - the number of constituencies each nation could get could be based on their proportion of the UK electorate.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV - 100 candidates constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a large deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster - the number of constituencies each nation could get could be based on their proportion of the UK electorate.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV on a winner takes all basis - constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a large deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster - the number of constituencies each nation could get could be based on their proportion of the UK electorate.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities - perhaps MPs with less than a third of the Popular Vote could have 1 vote in parliamentary debates, those with a third to a half could get 2 votes, those with 50% to 2/3 of the popular vote could get 3 votes and those with more than 2/3 of the popular vote could get 4 votes.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster - the number of constituencies each nation could get could be based on their proportion of the UK electorate.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities, or MPs could each be given different numbers of votes according to the total numbers of votes they got.

A compromise between First Past the Post and STV might be that in a seat if no candidate had won 50%+ of the Popular Vote, to allow candidates from the bottom of the list to donate the votes they had received to other parties with each candidate at the bottom having their vote either donated to another candidate or eliminated. Might be interesting to see who might have won the 2005 General Election under such a system.

Another alternative would be to have a single constituency executive on a list system elected by STV on a winner takes all basis - constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a large deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

[Yet Another Anon]

Isn't there still a serious time lag problem between boundary reviews and population changes?
Maybe there should be a boundary review for Parliamentary constituencies immediately after every General Election and referendums on each of the seperate reviews in England, Scotland, Wales and Ulster - the number of constituencies each nation could get could be based on their proportion of the UK electorate.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities, or MPs could each be given different numbers of votes according to the total numbers of votes they got.

If candidates were allowed to donate their votes to other candidates then in the event of no one getting 50%+ of the Popular Vote in a seat there could be a sort of local convention, a it like STV but giving preferences to the candidates to use rather than voters. Might be interesting to see who might have won the 2005 General Election under such a system and it could even lead to coalition arrangements with parties winning no House of Commons seats.

Another alternative would be to have a single constituency executive on a list system elected by STV on a winner takes all basis - constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a large deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities, or MPs could each be given different numbers of votes according to the total numbers of votes they got.

An alternative would be to have a single constituency executive on a list system elected by STV on a winner takes all basis - constituting an executive with similar powers in the UK to those of the French President, they would choose the PM and cabinet (If names were allowed to apear multiple times in different lists, but with each 100 name entry having to have a large deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities, or MPs could each be given different numbers of votes according to the total numbers of votes they got.

An alternative would be to have a single UK wide 100 seat constituency executive elected using STV on a list system on a winner takes all basis, having similar powers in the UK to those of the French President, they would choose the PM from amongst their number. If the same people were allowed to appear in different lists, but with each list entry having to have a £100,000 deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, then parties could stand seperately and in coalition giving people the choice of whether to vote for an majority executive (absolute or otherwise) or some kind of hung executive. It would require 2/3 of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments.

Multi-member constituencies in more populus areas would help balance out situations in which one party had a vote more spread out, or granting extra votes to MPs who won with large majorities, or MPs could each be given different numbers of votes according to the total numbers of votes they got.

An alternative would be to have a single UK wide 100 seat constituency executive elected using STV on a list system on a winner takes all basis, having similar powers in the UK to those of the French President, they would choose the PM from amongst their number. If the same people were allowed to appear in different lists, but with each list entry having to have a £100,000 deposit payable for it repayable if the list entry got 5%+ of the Popular Vote, and a fixed percentage of parliamentarians (in a parliament with about a quarter the current number) to vote against to vote down the Executive's legislation or reject their appointments - in that case no one could argue against the mandate of the winning list entry.

[Yet Another Anon]

Sorry about all those posts, I'd assumed the spam filter had blocked them out and was altering them not realising they would be posted.

Far from stopping spam, the spam filter seems to have actually ended up causing it - beginning to wonder why I bothered signing up on Typepad in the first place.

[Tim Roll-Pickering]

Philip: Leaving aside the "English MPs only votes on England-only matters" for the many other threads there have been on this issue, I don't think your proposal is workable. It's all very well to make a grand statement of principle that it should be the case "that each constituency has the same number of electors" but there are many, many problems with this. As I and others have raised, a combination of geography and local authority boundaries means that it's hard to get the seats exactly on size within existing parameters and keeping some notion of "natural seats". Throwing this to the wind, drawing whatever set of boundaries one likes and aiming for statistical exactitude would be a gerrymanderer's paradise - gerrymandering doesn't necessarily involve different seat sizes (indeed that's called "malapportionment").

[Deputy Editor]

The folk at Make My Vote Count have responded to this thread:

fascinating discussion is underway over on ConservativeHome on the impact of boundary sizes, differential turnout and the like. A lot of number-crunching, some good points, some slightly more dubious ones and some contributions which show just how ignorant and ideologically driven (brainwashed by the cult of FPTP) a number of people are.

The electoral reform movement is going to be doing some myth-busting exercises in the new year, to try and impart a little bit of reality and commonsense into the debate. If you can't wait, see here and here. But to see the scale of what we are up against, step forward William Nortonm who writes in one of his this previous piece and this one on the ConservativeHome site:

"Like some great stomach the juices of PR will dissolve each of us into our component parts ready for digestion."

We welcome lively and open debate. And there are some good thinkers and people up for a reasoned, intelligent debate within the Conservative party - the likes of Ken Clarke, Andrew Tyrie, Oliver Heald, Douglas Carswell and (though not a parliamentarian) Ferdy Mount. So more of them and less of the mindless bile please.

[Scottish Conservative]

I agree this is a good discussion. I think the point about Scottish overrepresentation is a bit petty but rather predictable the way things are going with u guys. Following the application of the English electoral quota at the last review we had a theoretical entitlement to 57 seats but ended up getting 59 due to the application of 1. the statutory provision relating to Orkney and Shetland and 2. the application of the special geographical considerations rule to the seat formerly known as the Western Isles - sounds a bit like the artist formerly known as Prince:-)

The reason Scotland has not ended up with 57 is that there is due to the ratchet effect. Thus when smaller seats are created it does not mean that the other seats need to be larger to make up for it. Thus the subsidiary rules which allow for the creation of smaller seats have a ratchet effect on the overall number of seats and so instead of our theoretical entitlement to 57 we end up with 59 seats in Scotland.

There are 2 keys points to make here. Firstly, these rules are applicable throughout the UK and so Scotland doesnt get some special advantage - the number of seats is merely the result of the application of UK wide rules. Secondly the ratchet effect occurs in other regions as well for instance in London which I believe is also overrepresented. Once again you have a subsidiary rule which leads to the creation of smaller seats which ratchets up the overall number of seats. In London the rule which does this is not special geographical considerations as in Scotland but the need to respect local authority/borough boundaries.

Being an anorak I did a dissertation on comparative redistricting laws and almost all countries have something like the rule on special geographical considerations allowing for smaller seats in rural areas. I mean even a seat like Argyll and Bute has 28 inhabited islands and a coatline longer than France - is it really such a scandal if it has 65000 instead of 69000? even the smaller Scottish Parliamentary seat of Inverness East, Nairn and Lochaber covers an area almost 6 times bigger than Greater London. And at the end of the day is it worth the fuss of doing away with the specail geogrphical considerations rule to get rid of 1 (non-Labour) seat?

From a partisan political point of view I wouldnt get too hung up on the 1 extra Highland seat that comes out of the special geogrphical considerations rule. If I were a tory gerrymanderer id want something like the following.

1. a review every parliamentary term - hard to legislate for this as we dont have fixed terms but could say 2 years after every GE.

2. A strict permissible level of deviation say 2-3% either side of the quota in cases where subsidiary rules do not allow for a greater devaition.

3. do not allow the need to respect local authority boundaries to allow for devaition from the quota - restrict that right to special geographical considerations.

4. something like the Australian system whereby if changes in a seat means it devaites by over a stipulated level ie -5% you have a mini review in region before the next major review.

5. the sub-allocations are a problem ie if Merseyside is entitled to 8.51 seats it gets 9 whereas another region with a theoretical entitlement to 8.49 will get 8. if english regional identity is as weak as claimed there should be more flexibility here to cross reional lines.

6. id clarify the provision relating to local ties. this can sometimes be interpreted as administartive ties as it was in my home review in Renfrewshire. it should be redefined so that local ties are to do with the demographic nature of the community concerned rather than administrative ties. this will mean putting villages together and towns together wherever possible.

So there u go Elbrideg Gerry eat ur heart out :-)

[Richard]

As 'Make My Vote Count' have mentioned Oliver Heald, I have heard that fair seats is a bit of an obsession of his.

[Tim Roll-Pickering]

Whilst the issues of geography are always going to be a problem, can I propose one change to the rules that might be a step in the right direction:

* Modify the rule that no more than two London Boroughs can be combined to one that no individual seat in London can be in more than two London Boroughs.

Because London Boroughs tend to be small, overall London winds up with extra seats because it's hard to create 0.6 of a constituency within these constraints. In the current review a counter proposal to tie Tower Hamlets, Hackney and Islington together with five seats was thrown out precisely because of this rule, and instead each borough is standing alone with two seats each.

This would reduce the unneccessary overrepresentation of London (and I am a Londoner), and on a wider basis help get closer to equal sized seats et al. I don't think London has the problems of rural scattered voters that other ares with smaller constituencies do and it's not that hard to get round, even in the outer suburbs.