Scottish Parliament 2019 'notional' results

[clyde1998]

Using the Electoral Calculus ward estimates (which I maintain my concerns about using out of date local election results as their base, notwithstanding the demographic adjustments that they do), I've estimated the 2019 general election results on Scottish Parliament boundaries.



Seat wise, compared to the Scottish Parliament election in 2016, the changes are:

• Con to SNP (4): Aberdeenshire West, Ayr, Eastwood, Edinburgh Central.
• SNP to Con (3): Aberdeenshire East, Banffshire and Buchan Coast, Moray.
• Lab to SNP (3): Dumbarton, East Lothian, Edinburgh Southern.

Obviously, the result in some areas may have been very different due to tactical voting, incumbency and different issues. A seat like Edinburgh Central is divided between four Edinburgh Westminster constituencies, two of which being SNP-Labour contests, one SNP-Lib Dem and one SNP-Conservative.

The spreadsheet is here: link

There are about 3,000 votes in error compared to the actual results (ie. going to the wrong party). The Conservative have 1,900 too few votes, the Lib Dems 380, the SNP 360 and others 1; Labour have 240 too many votes, Brexit Party and Greens 90 too many each. I don't know if that's an error on the Electoral Calculus' part, whether I've accidentally copied the 'predicted' ward results or anything else. An error of more than 100 is unlikely to be because of rounding. I've tried to check every seat again, especially the ones with results I thought were strange; that did reduce the error from about 20,000 votes but hasn't removed it entirely. I don't think the vote errors will flip any seats.

[Deleted]

I suspect Aberdeenshire West and Eastwood would've gone Conservative. Ayr is a bit more of a coin toss but I would be inclined to give it to the Conservatives as well.

Also pretty sure Edinburgh Southern would've gone Labour.

[clyde1998]

Dec 24, 2019 19:18:44 GMT Sandy said:

I suspect Aberdeenshire West and Eastwood would've gone Conservative. Ayr is a bit more of a coin toss but I would be inclined to give it to the Conservatives as well.

Also pretty sure Edinburgh Southern would've gone Labour.

Ayr only went SNP by about 100 votes on these figures, so probably would've gone Conservative in the context of a Scottish Parliament election. Aberdeenshire West was SNP by 500 votes, so same statement. Eastwood probably only went SNP due to the Brexit issue and would invariably would've stayed Conservative without it.

The Edinburgh seats are a mess in the sense that you're converting five Westminster seats into six Holyrood seats. Southern would almost certainly be Labour; it has territory that's within the East and South West Westminster constituencies, where Labour were a lot weaker relative to the SNP than South - despite that, they are only 300 votes behind on these figures.

I can't see the SNP having got as many seats as this shows had the election actually been held on these boundaries.

[Deleted]

Dec 24, 2019 19:51:55 GMT clyde1998 said:

Dec 24, 2019 19:18:44 GMT Sandy said:

I suspect Aberdeenshire West and Eastwood would've gone Conservative. Ayr is a bit more of a coin toss but I would be inclined to give it to the Conservatives as well.

Also pretty sure Edinburgh Southern would've gone Labour.

Ayr only went SNP by about 100 votes on these figures, so probably would've gone Conservative in the context of a Scottish Parliament election. Aberdeenshire West was SNP by 500 votes, so same statement. Eastwood probably only went SNP due to the Brexit issue and would invariably would've stayed Conservative without it.

The Edinburgh seats are a mess in the sense that you're converting five Westminster seats into six Holyrood seats. Southern would almost certainly be Labour; it has territory that's within the East and South West Westminster constituencies, where Labour were a lot weaker relative to the SNP than South - despite that, they are only 300 votes behind on these figures.

I can't see the SNP having got as many seats as this shows had the election actually been held on these boundaries.

That seems like a very reasonable analysis to me.  It is interesting that most of the notionally Conservative held North East seats are those which are fairly far down their target list, while seats which are (at least on paper) lower hanging fruit such as Angus North & Mearns remain SNP!

[lancastrian]

Dec 24, 2019 16:34:55 GMT clyde1998 said:

Using the Electoral Calculus ward estimates (which I maintain my concerns about using out of date local election results as their base, notwithstanding the demographic adjustments that they do), I've estimated the 2019 general election results on Scottish Parliament boundaries.



Seat wise, compared to the Scottish Parliament election in 2016, the changes are:

• Con to SNP (4): Aberdeenshire West, Ayr, Eastwood, Edinburgh Central.
• SNP to Con (3): Aberdeenshire East, Banffshire and Buchan Coast, Moray.
• Lab to SNP (3): Dumbarton, East Lothian, Edinburgh Southern.

Obviously, the result in some areas may have been very different due to tactical voting, incumbency and different issues. A seat like Edinburgh Central is divided between four Edinburgh Westminster constituencies, two of which being SNP-Labour contests, one SNP-Lib Dem and one SNP-Conservative.

The spreadsheet is here: link

There are about 3,000 votes in error compared to the actual results (ie. going to the wrong party). The Conservative have 1,900 too few votes, the Lib Dems 380, the SNP 360 and others 1; Labour have 240 too many votes, Brexit Party and Greens 90 too many each. I don't know if that's an error on the Electoral Calculus' part, whether I've accidentally copied the 'predicted' ward results or anything else. An error of more than 100 is unlikely to be because of rounding. I've tried to check every seat again, especially the ones with results I thought were strange; that did reduce the error from about 20,000 votes but hasn't removed it entirely. I don't think the vote errors will flip any seats.

I think a lot of the missing Conservatives are in Aberdeenshire West, because my rough adding up of Electoral Calculus' estimates has them over 1000 ahead there rather than 500 behind.