The Boundary Commission for Northern Ireland has published its revised recommendations for Northern Ireland’s 18 parliamentary constituencies. My take in five lines:

• This is much more modest than the original set of changes in their provisional recommendations.
  
• I don’t see any significant impact on the next Westminster elections as a result of the new boundaries.
  
• At Assembly level, the SDLP will lose their current seat in Lagan Valley; but there will now be now a safe Nationalist seat in East Antrim (where the SDLP won an unexpected seat in the 1998 elections, but lost it in 2003).
  
• The changes to South Belfast, plus last week’s demise of the Women’s Coalition, will put Alliance in a stronger position to win a seat there; though the party still has quite some way to go.
  
• The entrenchement of six Assembly seats per constituency in the Good Friday Agreement results in the under-representation of voters in Newry and Armagh, North Antrim and Upper Bann, which are all large enough to warrant a seventh seat at Assembly elections.

(Thanks, as ever, to Conal Kelly for the map.)

I’ll look at this in more detail over the weekend. This is what I put together before breakfast:

• Belfast East constituency to include the Castlereagh LGD wards of Ballyhanwood, Carrowreagh, Dundonald, Enler and Graham’s Bridge. These wards are at present in the Strangford constituency;
  Does it make sense? Yes. The DUP successfully managed to keep Cregagh in East Belfast – the previous recommendations would have moved it to South Belfast, but the Commission has now swapped it for the Hillfoot ward instead..
  Demographic shift: The new East Belfast is roughly 3% more Protestant than the current constituency.
  Westminster consequences: no serious threat to Peter Robinson.
  Assembly consequences: DUP slightly strengthened at expense of Nationalists and PUP, but probably no change.
• Belfast North constituency to include the Newtownabbey LGD wards of Ballyhenry, Collinbridge, Glebe, Glengormley and Hightown, at present in the South Antrim constituency, and Cloughfern, at present in the East Antrim constituency;
  Does it make sense? Broadly yes. Specifically no. The weird division of the Shankill Road between North and West Belfast remains, and the new boundary weaves through the streets of Glengormley. Better to have left Cloughfern in East Antrim and extended North Belfast a bit farther northwest. (Query Shankill?)
  Demographic shift: The new North Belfast is 0.9% more Catholic edited to correct: Protestant than the current constituency.
  The new South Antrim is 1.6% more Protestant than the current constituency.
  Westminster consequences: In North Belfast this slightly accelerates Nationalist growth, slightly diminishes DUP dominance. In South Antrim the DUP are strengthened, probably not enough to make a difference.
  Assembly consequences: In North Belfast, DUP will slip slightly to UUP, but are so far ahead that it hardly matters.
  In South Antrim, second Nationalist quota is a little further away, so SF starting from a lower base next time (but probably would still take SDLP seat).
• Belfast South constituency to include the Castlereagh LGD wards of Carryduff East and Carryduff West at present in the Strangford constituency, and Wynchurch and Hillfoot, at present in the Belfast East constituency;
  Does it make sense? Yes; boundary with East Belfast is a bit inelegant but makes sense on the ground.
  Demographic shift: The new South Belfast is 0.5% more Protestant than the current constituency.
  Westminster consequences: none.
  Assembly consequences: Alliance slightly strengthened (from very weak position) at expense of UUP, and third Nationalist seat (currently second SDLP seat) looks weaker (but probably will be helped by demographic growth).
• Belfast West constituency to include the Lisburn LGD wards of Dunmurry and part of Derryaghy, at present in the Lagan Valley constituency.
  Does it make sense? Yes, apart from remarks about continued split of Shankill above under North Belfast; and the split of Derryaghy ward is unprecedented.
  Demographic shift: The new West Belfast is 0.2% more Catholic than the current constituency.
  Westminster consequences: none.
  Assembly consequences: DUP (or at least Unionist) seat is safer.
• Strangford constituency loses to South and West Belfast as noted above; it gets the Down LGD wards of Ballymaglave, Ballynahinch East, and Kilmore. The wards at present form part of the South Down constituency.
  Does it make sense? The most dubious of the new recommendations, in my view. It is news to me that Ballynahinch looks to Newtownards as a political centre.
  Demographic shift: The new Strangford is 1.3% more Catholic than the current constituency.
  Westminster consequences: depends on the ability of the UUP to mobilise tactical voting if the DUP have a bad year.
  Assembly consequences: Nationalist seat, barely missed in 2003, is slightly more possible now. Alliance looks most vulnerable – good areas in Castlereagh lost to East and South Belfast, in return for parts of Down where the party has no recent record.
• South Down unchanged except for losses to Strangford (major changes in provisional recommendations withdrawn).
  Does it make sense? See above remarks re Ballynahinch. The proposed division of Newry Town would only have restored the constituency boundary to where it was until 1983, but I guess the locals put up a strong case for no change.
  Demographic shift: The new South Down is 3.2% more Catholic than the current constituency.
  Westminster consequences: Brings the seat within reach for Sinn Féin.
  Assembly consequences: Second Unionist seat is now marginal; currently UUP (who held this constituency at Westminster until 1987) still ahead of DUP. On 2003 results SDLP better placed to pick up – for now.
• Upper Bann unchanged.
  Does it make sense? The minor alterations originally proposed were pretty silly.
• Lagan Valley loses one and a half Dunmurry wards to West Belfast, and Glenavy to South Antrim.
  Does it make sense? Recognises the demographic shifts, though the splitting of Derryaghy is inelegant.
  Demographic shift: The new Lagan Valley is 5.3% more Protestant than the current constituency.
  Westminster consequences: None.
  Assembly consequences: The SDLP seat held by Patricia Lewsley looks doomed; the Catholic percentage of the population is now only just over a quota, and this in a constituency where many Catholics vote Alliance. On the Unionist side the votes will be affected much more by the move of several local personalities from the UUP to the DUP than by the constituency boundaries.
• East Antrim constituency to be extended to include the Moyle LGD wards of Glenaan, Glenariff, and Glendun at present in the North Antrim constituency (the shift of Ballycastle in the original recommendations is withdrawn, as is the silly proposal to rename East Antrim “Antrim Coast and Glens”).
  Does it make sense? No. At the southern end, it would have been better to shift a couple more Glengormley wards into North Belfast rather than Cloughfern. At the northern end, it’s simply absurd to suggest that the Glens look to Jordanstown rather than Ballymena as a regional centre.
  Demographic shift: The new East Antrim is 4.0% more Catholic than the current East Antrim constituency.  The new North Antrim is 2.6% more Protestant than the current constituency.
  Westminster consequences: UUP have a larger pool of potential Nationalist tactical votes to regain East Antrim with, if their fortunes ever revive.
  Assembly consequences: One very safe Nationalist seat in East Antrim, rather than the marginal one lost in 2003. Probably enough spare votes to keep Alliance in play, so likely loser would be third DUP seat.
  In North Antrim, however, the second Nationalist seat gained in 2003 is vulnerable.
• It is proposed to alter the boundaries of the East Londonderry and Foyle constituencies by transferring the two Derry LGD wards of Banagher and Claudy from the Foyle constituency to the East Londonderry constituency.
  Does it make sense? Yes.
  Demographic shift: The new East Londonderry is 2.1% more Catholic than the current constituency. The new Foyle is 0.3% more Catholic than the current constituency.
  Westminster consequences: None.
  Assembly consequences: In East Londonderry, Nationalists closer to a third seat but still some way off. In Foyle, Unionists slightly further from a second seat that was never very likely.
• No changes are proposed to the Fermanagh and South Tyrone, Mid Ulster, North Down and West Tyrone constituencies.
  Does it make sense? Yes. Although the Mid-Ulster and West Tyrone constituencies are both very new in their current form, they are reasonably sound natural units and like Fermanagh-South Tyrone close enough to the quota that no change is necessary. Tinkering with North Down could have given a bit more flexibility for the rearrangement of Strangford.

Edited to add: West Belfast/Lagan Valley – I have rerun the figures making the most sense I can from the census. On my current estimate the new West Belfast is 0.2% more Catholic, not (as I had first thought) 0.7% less, and Lagan Valley 6.3% more Protestant rather than 5.3%. I think the DUP seat in West Belfast remains marginal but salvageable – the total Unionist vote share actually increased there in 2003 from 1998 – but the SDLP seat in Lagan Valley does not.

NB also minor (cough) correction to North Belfast as originally described.