Beware geeks bearing gifts

Today I know MacNumpty will be in election geek heaven. The proposals of the Boundary Commission are out, redrawing the now a decade-and-a-half-old Scottish constituency maps.

Recent research by the Electoral Commission found that people generally did not know why this happens. It's because populations change, and if the size of MSP's constituencies vary greatly it makes one person's vote worth more than another. You could argue that the first past the post system and 'safe' seats does that as well, and I certainly wouldn't disagree.

As ever, they have made some giant howlers. Helensburgh & Lomond for example, where I spent a large part of my formative years, has now been sliced in two, with 30% going to Argyll & Bute (the local authority) and 70% going to a new 'Dumbarton & Helensburgh' constituency that stretches as far as Duntocher. I am sure Jim Mather will be as disturbed to find Faslane in his constituency as my colleague Gareth will be to find Jackie Baillie as his MSP. Clydebank & North Renfrewshire spans both sides of the Firth of Clyde. And if the Commission followed their rules and paid heed to local council ward boundaries when devising the map of Edinburgh please, please give me some of whatever they were drinking at that meeting.

Of course, the main event is always how this would impact on representation.

Step forward Jamie Hepburn (Cumbernauld & East Airdrie), Christine Grahame (South Midlothian, Tweeddale & Lauderdale) and Jackson Carlaw (East Renfrewshire) - now (notional) constituency MSPs. Goodbye Cathie Craigie, Ken Macintosh and Jeremy Purvis from Parliament. [Edit - hat doff to Tory Boy, who corrected me that Jeremy Purvis did not in fact stand on the regional list.) Also goodbye for Lewis Macdonald, whose seat is gone. But it's not all gloom for the Lib-Lab alliance. Welcome back Sylvia Jackson, whose Stirling swings from being an SNP marginal clung to by Minister for Parliamentary Business Bruce Crawford with 620 votes to a Labour marginal held by just 321-491 votes, depending on how you do the estimates. Bruce Crawford doubtless survives on the list.

Goodbye whoever loses the Glasgow Labour catfight. 9 constituencies in the City Council area (10 in the Parliamentary region because it includes Rutherglen at the moment) go down to 8. Where gains an extra one to compensate? Why, the North East. Yes, step forward Angus East & Mearns, a new and notionally SNP seat. (2,789 over the Tories, since you ask.) Again, all silver linings have a cloud, and Nicola Sturgeon's hold on Glasgow South Central looks shaky at best.

This throws up a problem. The method used for estimating seats notionally by Rallings & Thrasher in England employs local government voting figures. They're not ideal, but they're the best you can do, and there are one or two little ways you can adapt them in the mathematical model. That's the kind of trickery I'm employing here. Sometimes though you just can't come up with a viable number at all. Using any variation on R&T looking back at the last election would make both Glasgow Govan and Edinburgh East & Musselburgh Labour holds rather than SNP gains. They can't be notional holds when you know they were actual gains. For this reason, I feel I have to classify the successor seats to these two and Argyll & Bute as impossible to call with anything but subjective methods. That caveat declared, I'll list them all as SNP nonetheless.

The bottom line, by my highly contestable calculation:

Labour 34 (-3) [Edit from 33]
SNP 24 (+3) [Edit from 25]
Lib Dem 10 (-1)
Tory 5 (+1)

And the moral of it all? Twofold. Firstly, be glad this isn't the USA, and the commission who are recommending this is headed by a judge and not a partisan political appointee. Secondly, even with a fair distribution by population, a narrow but tangible SNP lead of 1% in May of last year would have resulted in a 14% deficit in seats. Be glad of having a system of proportional representation - even if we don't yet have the best one.

Come to think of it, I think some of the flaws of that system are addressed by these boundary changes. This cannot be in any way even a vaguely reliable projection, but based on the way the limited number of list seats in some situations fails to compensate for the entirety of the disproportionality of the constituency seats, and the fact that this came into play in three regions at the last election, I would hesitate to - but nonetheless feel I must - tentatively project the following as the result of the last election on these boundaries:

SNP 50
Lab 44 [Edit - from 43]
Con 17
Lib 15 [Edit - from 16]
Grn 2
Ind 1 [Edit - who could forget Margo...]

Put that in your pipe and smoke it.

(PS although this is dated earlier tonight I did observe the embargo and didn't publish until 00:01)