THREAD ON BOUNDARY CHANGES: Guido Fawkes is reporting that boundary changes will not now involve cutting the number of seats to 600. This is unsurprising given past rebellions, but if correct it will have a number of consequences. 1/
order-order.com/2020/01/02/tor…
order-order.com/2020/01/02/tor…
Firstly, the reviews completed in 2013 and 2018 will have to be junked and the whole process started again with, presumably, 650 seats. That means changes to the legislation. Whether other aspects (the tolerances, protected seats, nation allocations etc) will change is unclear 2/
Secondly, the next review was due to start in 2021 with December 2020 electorates. But since it sounds like the legislation is going to be reopened anyway, it would make sense to start it immediately, given that a spring 2024 election would mean less time than usual 3/
Re-snapping the electorate would also remove the controversy over the use of the 2015/2016 electorates. As I wrote at the time, this row was mostly a lot of hot air anyway, as the impact of the difference was trivial, but it's one less row 4/
It's also worth pointing out that we can't be sure about the impact of the changes until the boundaries are drawn. Electoral Calculus estimates that the "old new" boundaries would have given a Con majority of 104, but if those get junked, that figure isn't particularly helpful 5/
What we can say is that despite the realignment at the general election (Labour losing heartland seats with falling electorates) Labour still has a substantial advantage from the current boundaries, as its seats have electorates around 3,800 smaller than Conservative seats. 6/
Therefore, bringing the two into line would – as has always been the case in the past – add Conservative seats, so the notional majority that the Tories would be defending would be bigger than 80. 7/
Nope. That's still not what gerrymandering means. The UK has independent boundary commissioners (usually senior judges). This is not the US.
Voter ID – depends on the details. Can be done well or badly (if that latter, probably hurts Labour).
Postal votes – depends where the fraud is. Most of the accusations tend to be in very safe seats. Impact probably limited.
FTPA repeal – helps incumbent.
Postal votes – depends where the fraud is. Most of the accusations tend to be in very safe seats. Impact probably limited.
FTPA repeal – helps incumbent.
Electoral bias/skew/proportionality are functions of many things, of which boundaries are only one factor (usually a small factor). Distributional efficiency, differential turnout, contests with smaller parties, tactical voting etc usually matter far more
Assuming 650 seats, Dec '19 electorates, rest as per the last review, I get (with changes vs current boundaries):
Scot 56 (-3)
Wales 32 (-8)
N Ireland 18 (=)
E 61 (+3)
E Mids 47 (+1)
London 76 (+3)
NE 27 (-2)
NW 73 (-2)
SE 91 (+7)
SW 58 (+3)
W Mids 57 (-2)
Yorks & Humber 54 (=)
Scot 56 (-3)
Wales 32 (-8)
N Ireland 18 (=)
E 61 (+3)
E Mids 47 (+1)
London 76 (+3)
NE 27 (-2)
NW 73 (-2)
SE 91 (+7)
SW 58 (+3)
W Mids 57 (-2)
Yorks & Humber 54 (=)
Reason for the big drop in Wales, as last time, is that moving to a UK-wide quote removes the over-representation it had since before devolution (which Scotland also had until 2005). Everything else reflects changes in electorates
Within the South West region, it looks like Cornwall would have a theoretical allocation of 5.96 seats, well within the 5% tolerance of 6 seats, so in this scenario that the controversial “Devonwall” constituency wouldn’t be needed
