The detailled legislation for the boundary review and the AV referendum is now out – if you are not a hardcore political anorak, you may wish to skip this post! For those who are left, here are the details.

Alternative Vote is the more straightforward section. The referendum is May 5th and the question is “Do you want the United Kingdom to adopt the “alternative vote” system instead of the current “first past the post” system for electing Members of Parliament to the House of Commons?”. There is no minimum turnout or anything on the referendum and it is binding – if the referendum is won, the minister must bring the provisions introducing AV (which are all in the Bill) into effect.

Rules for boundary changes are much more complicated. First, the legislation proposes boundary reviews every five years, significantly speeding up the current timetable where they occur between 8 and 12 years apart (normally at the latter end). With fixed term Parliaments of 5 years, that means seats would change every Parliament (though it would also mean that the changes were normally quite small).

There will one national UK quota, rather than seperate quotas for England, Scotland, Wales and Northern Ireland. There is, however, an exception to this for Northern Ireland, where if the number of seats Northern Ireland is entitled to is more than a third away from a whole number (as it probably will be this time round), they will have their own quota to aim at, which prevents the boundary commission in Northern Ireland being left with an impossible task.

The quota will indeed be based around the electorate of the UK, minus the two protected seats, divided by 598. All seats must be within 5% of the quota, with three exceptions: the two preserved seats (the Western Isles, Orkney and Shetland), Northern Ireland under the circumstances discussed above, and any seat with an area above 12,000 sq. kilometers where the commission is satisfied it is not possible to get it within quota. As proposed before, there is a cap of 13,000 sq km on the size of consistuencies.

The Commissions may still take in account special geographical considerations, like size and shape, local government boundaries, local ties and minimal change (except, that is, for the coming boundary change, when they should not pay heed to minimising disruption). They may now also take into account the European electoral regions (making some of the possiblities I wrote about in this post less likely). All these considerations are subject to the rule on seats being within 5% of quota, which means that the boundary commissions will have to cross county boundaries if it is necessary in order to get within 5% of quota.

There is no mention of splitting council wards, but then, the old legislation doesn’t refer to them outside Northern Ireland either. In practice, the rule about seats being within 5% of quota may compel Boundary Commissions to split wards in places like Birmingham with very large wards.

In terms of speeding up the review process, the Bill is pretty brutal. Local inquiries on boundary changes are abolished – though it is slightly balanced out by the period for written representations being extended from 1 month to 12 weeks.

All of this is, of course, subject to whatever amendments get made as it trundles through the Commons. I expect some bits may have a tricky passage.

361 Responses to “AV referendum and 600 seats”

1 6 7 8
  1. I noticed back on continuation page 5 a post has appeared from “One Wight Campaign” (July 24th, 2010 at 12:00 pm) that presumably was in moderation as we moved over from page to page.

    That prompted me to have a look back at what would happen if they got their way. They would be 44.7% over the quota (or two at 27.7% under if it was split into two whole constituencies). The implication for Hants, Southampton and Portsmouth would would be that they fit 17 constituencies almost precisely without IoW. So, IoW having a single constituency and nothing crossing the Solent would mean that the combined Hants (inc Soton,P’mouth) and IoW would have one less constituency than the combination would have.
    If that happened on my calculation the beneficiary would be the combined Northants, Leics & Rutland; or London if NLR already had 17 constits.

  2. Amber, Anthony,
    I think I have discovered why my calculation of Sainte Laguë seat allocation only gave 49+2 to Scotland, whereas you thought it would be 50+2: I was using Pippa Norris’ spreadsheet, and it had wrongly classified Berwickshire, Roxburgh and Selkirk as being in North East, not Scotland (its electorate for the seat looks oddly low, too). While that isn’t the only difference between the PN numbers and Electoral Calculus, putting Berwickshire, Roxburgh and Selkirk in the correct region alone is enough to indicate that constituencies are likely to be England 503, Wales 30, Scotland 50+2, NI 15.
    Meanwhile, I have been wondering how the Boundary Commission will work to a fixed number of constituencies: the way they have worked previously doesn’t really fit this. In one way it would be logical if they used Sainte-Laguë to allocate seats to areas. But there are no obvious pre-set areas to get Saint-Laguë to allocate seats to: we know that they will have to cross some county boundaries, but not others. We know that some possible groupings (eg the former county of Avon, in my earlier post) will work for some numbers of seats, but not others. Perhaps they will use an iterative process, or perhaps they will use the method they have used in the past, and hit a problem when it gives the wrong number!
    As I pointed out July 24th, 2010 at 3:45 pm (continuation page5), Using Pippa Norris’ figure for the 2010 London Electorate, rather than BCEs would give 69.48 constituencies. If they were faced with that electorate, and treated London as a whole in deciding how many MPs to allocate, on recent practice they would allocate 69, and I suspect thus be one MP short nationally (or two if the treated Merseyside and Cheshire, including Halton as two separate areas).
    I would also be interested in anyone’s thoughts about whether the electorate of London when BCE do their calculation will be more like the early 2010 figures that are on the BCE website, or the ‘after the late registration frenzy’ figures that I think are the Pippa Norris electorate figures. Will high population mobility in London mean the extra electorate drop back off the register, now the GE is over?

  3. Ben – I’ve been pondering that too. My guess is that they will start the process by allocating numbers of seats to each county, then making necessary decisions about pairing counties, then comparing the resultant number of seats with their target for the country. If it doesn’t match, then they’ll reconsider their decisions on pairing to try and find a solution that does match.

    In terms of allocating numbers of seats, it looks to me as though decisions on pairing is where they have the most discretion and, therefore, where it will be easiest to tweak to get the right target number. Frankly though, my view is they’ve been rather too strict on the 600 and should have just introduced a fixed quota at 1/598 of the electorate of the UK minus the two protected seats, and just accepted that it would sometimes produce 599 or 601 seats. It wouldn’t be the end of the world. Still – unless it gets amended that’s what we have.

    Whether the new electorate drop off the register depends how strict individual boroughs are at cleaning their registers of non-respondents each year, I’d expect most of those new registrations to remain.

  4. Trying to figure out boundaries within Bedfordshire has led me to an interesting paradox. If you try to work out what the groupings of ‘counties’ would be first, it suggests one combination of ‘counties’, but if you then try to draw sensible boundaries within those groupings, the premise you were working on may fall apart. I suggested that given that Bedford Borough and Central Bedfordshire together justify 4 MPs, the BC would go for that, and treat Luton with Hertfordshire (Bedford Borough, Central Bedfordshire and Luton are all unitaries, and thus count as separate ‘counties’ for these purposes).

    But when I try drawing boundaries, the sensible cross-constituency Luton/Hertfordshire constituency is Luton South and Harpenden. But a sensible shape Luton South and Harpenden constituency would include Hyde, thus meaning that in fact the boundary between Central Bedfordshire and Luton would then be crossed, as well as the boundary between Central Bedfordshire and Herts, contrary to the reasoning for crossing the Luton/Hertfordshire rather than the Bedfordshire/Cambridge one (StNeots and Little Paxton have the right sort of numbers of voters, and go quite well with Sandy in a constituency, btw).

    Mind you, if you cross the Bedfordshire/Cambridge border, it looks like Central Beds has to cross its border with Bedford Borough too, and also Luton. To make things worse, I can’t figure out a sensible set of constituencies for Luton and SW Beds (taking in the extra territory/voters that the combination of the three would have to, to get constituencies close enough to quota), without having multiple crossings of the Luton/Central Beds border.

  5. Ben,

    Whilst Bedfordshire no longer has a County Council, I think that like the other Counties which now have Unitaries, it remains a “County” for ceremonial purposes.

    So I imagine that when looking at potential pairings, Beds & Hets is a possible pairing, but Beds is more likely to be paired with Cambs. Eastern Region could then be split “nicely” into Norfolk/Suffolk; Beds /Cambs; Herts/Essex.

    As for Luton, local logic dictates that the Dunstable area gets put into a “Luton West” seat rather than pairing Harpenden with Luton South – that would be a chalk and cheese match if ever there were one.

  6. Anthony,
    Thinking more about how the BC will allocate numbers of seats to areas to result in the right number for England, I think they will use the regions, which the Parliamentary Voting System and Constituencies Bill will introduce into their thinking for the first time. I think they will decide how many seats in each region, and then allocate them within the region. So, I did a Saint-Lague on the regions, and that suggests NE26, NW69 (I had been thinking 68), Y&H50 (I had been thinking 51), EM44 (I had been thinking 43), WM54 (not 53), SW53, SE82 (not83), East 56 and London 69 (not 70, even on the PippaNorris electorate figure that includes late registrations, I think). I don’t know whether BCE will do a Saint-Lague on the regions, but it does seem plausible, given that MEPs are allocated to regions using Saint-Lague, and Saint-Lague is already going to be used to allocated seats between England, Wales, Scot and NI, according to the Bill.
    I may have to rethink within regions on the basis of my suspicion that BCE will allocate seats to regions first.
    Except for Berkshire, it appears that ceremonial counties are irrelevant for the Boundary Commission/ Parliamentary Voting System and Constituencies Bill (at least unless amended). I also am pretty sure from their previous practice the BC will not pair/combine counties/unitaries unless they have to. And I am pretty sure that once Essex has been joined to Southend-on-Sea and Thurrock Unitaries, it will not be paired with another county. I agree Norfolk and Suffolk will be paired (since Norfolk has too many voters for 8 constits), but I think Cambridge won’t be (except with Peterborough) since the combined Cambs & P’boro fits 7 constituencies (just) within 105% UK-wide quota. Hertfordshire may just fit 11 and be above 95% UK-wide quota. I don’t think you can ignore that Bedford Borough and Central Beds combine to 4 constituencies comfortably within UK-wide quota. But it is not at all clear that the BC will prefer to treat Luton with Central Beds rather than Hertfordshire.
    As far as I can work out, they decide on pairings/groupings of council areas first, and then see what logical constituencies fit those pairings/groupings (and if none do close enough to the quota, they propose something illogical, but won’t revisit the pairings/groupings even when the illogicality is pointed out).
    If they did Luton & Central Beds can just about fit 4 constituencies (but a difficult 4.3% over UK-wide quota), leaving Bedford Borough to combine with Cambridgeshire (&P’boro), which would fit within quota (9 constits combined).
    If the BC did combine Luton with Central Beds, it could be done with a Luton S constit that takes Barnfield, Limbury, Saints from the current North, and loses the part of South East Beds to it (but on the BC website figures that results in a constituency 104.9% UK-wide quota size). The LutonN/W and Dunstable constit would include the rest of the current North, the rest of the current SE Beds ward and Dunstable, but not Houghton Regis. There could then be a West Beds constituency of LeightonB, Houghton R and Ampthill, and an East Beds of Flitwick, Maulden & Houghton conquest, Shefford, Silsoe and Shillington and all of the NEBeds Constit currently in Central Beds. How much sense does it make for Dunstable to be in one constituency and Houghton Regis in another?

  7. Ben,

    No less sensible than dividing Luton itself. Still preferable than straying acoss the boundary in to Herts.

    If I understand you correctly, you are saying that the Luton / Dunstable conurbation is too big for two seats but too small for three.

    That is presumably a common problem and requires hinterland to be added or else parts of the conurbation to be subtracted.

    Unless one has a strong locus for arguing the point one way or another in a specific area, I see little point in worrying over potential boundary changes until the BC has announced what areas it will be using for its dleiberations. After all, amendments to the Bill could throw a spanner in the works.

    For example, the target quota could be amended to deliver fewer seats, or the BC could be asked to indicate groupings of seats so that they could be combined into 200 3-member seats to enable STV to be used instead of AV.

  8. Paul,
    Yes, the Luton / Dunstable conurbation is too big for two seats but way too small for three, if it includes Houghton Regis (but it may be OK for 2 if it doesn’t), and even if you find a way of getting 4 constits out of Luton and Central Beds, Bedford Borough is too large for 1 and too small for 2.

    Certainly amendments to the Bill could throw a spanner in the works: I don’t think those you mention are likely, but Cornwall and the Isle of White are the two most likely, IMO (in each case they don’t want border crossing constits). I wonder what the impacts would be of each of them. It is only by trying the sort of calculations Anthony and I have been that you discover some of these implications, though.

    “Unless one has a strong locus for arguing the point one way or another in a specific area, I see little point in worrying over potential boundary changes until the BC has announced what areas it will be using for its dleiberations.”

    Anthony warned “if you are not a hardcore political anorak, you may wish to skip this post”. Once you have got past 350 responses and several newer threads, I think you have to be a _very_ hardcore political anorak.

    However, it may be that by suggesting likely impacts (eg a Luton S and Harpenden constituency) that people don’t like, that momentum grows for amendments to the Bill (eg allowing quota + or – more than 5%).

  9. If, as I suspect BC Sainte-Lague the regions to allocate seats to regions, I think it will come out as NE 26, NW 69 (not 68), YH 50 (not 51), WM 54 (not 53), EM 44 (not 43), East 56, SE 82 (not 83), SW 53 and London 69 (not 70) (comparison with adding up the counties and groupings of counties/councils I had been working on, and Saint-Lague between them, ignoring regions).

    I then had a look to see which counties/groupings would be affected if they did it this way – again using Sainte-Lague within regions.

    Lancashire, Blackburn & Blackpool would get 15 constituencies, not 14
    West Yorkshire would only get 20 constituencies, not 21,
    Hants, Southampton, Portsmouth and IoW would only get 18 constituencies, not 19
    Northants, Leics (and Leic), Rutland would get their 17th MP
    Combined Warks and Staffs (incl SoT) would get their 17th

    So, even using the same electorate figures, how many constituencies are allocated to a county not only depends on groupings, but also in some cases on whether or not they decide to go for fairness between regions before fairness between groupings of counties, or whether they go straight to fairness between groupings of counties (if they feel able).

  10. James

    “Indeed, we have always had AMS for Hollyrood! Though the reason Labour agreed to this was largely to make it harder for the nationalists.”

    It is true that some in the Labour party were influenced by he expectation that devolution would, in George Robertson’s words, “see off the nationalists.”

    I never regarded GR as the sharpest knife in the box, and a dozen years later when it clearly hasn’t worked I’m still waiting fo an expanation for how it was supposed to work, never mind an excuse for why it didn’t.

  11. Paul,
    In answer to my “How much sense does it make for Dunstable to be in one constituency and Houghton Regis in another?” you said “No less sensible than dividing Luton itself. Still preferable than straying acoss the boundary in to Herts.”

    The problem is, Bedford Borough + Central Beds + Luton have at least 6000 too few voters for 6 constituencies of 95% of the UK quota (and 28,000 less than 6 times the quota). So even having a constituency boundary between Dunstable and Houghton Regis, you haven’t avoided the need to cross the 1973-1997 county boundary.

    If you don’t like pairing Harpenden with Luton South, you need to suggest where else the 1973-1997 county boundary should be crossed (and to make it more difficult, the BC are being asked to take account of regional boundaries, which suggests not crossing the boundaries with Bucks, MK or Northants).

1 6 7 8