Long Read – Ballot Box Britain: GE 2019 Under PR

Just before last week’s General Election, I went back to 2017’s results to work out what they might have looked like under a form of Proportional Representation. The resulting Long Read post covered each part of the UK, giving a relatively detailed look at a possible alternative to Westminster’s antiquated, unfair and fundamentally undemocratic voting system. Having gone to all the bother of building that model, it was (relatively) easy to then apply it to this year’s results as well.

What distinguishes this from the flurry of “if this election was PR” seat totals circulating at the moment is the detail. Firstly, I’ve actually designed a proper system of PR rather than simply going “well X party got Y%, so they get Z seats”, so it has sensible rules such as % thresholds to be elected, common in most PR systems in Europe. Secondly, I’m going down to a much more local level, having built example districts to elect MPs from. Rather than just give top-line seat numbers, I think it’s useful to have that little bit more detail, especially in a country where “but what about the local link?” is the common response to calls for electoral reform.

I’m not going to go back through all the detail of how the system works or how I built the districts since that’s already in the original post. Make sure if you’re really curious you check that out as well! But just as a refresher;

  • Districts each elect between 4-9 MPs (with three exceptions being smaller)
  • All but one of those MPs is elected “directly” with local votes using the Sainte-Laguë method
  • The votes for every District in each Nation/English Region are totalled up
  • Remaining seat in each District is a “levelling seat” filled on the basis of the total vote for parties with more than 3% in that Nation/Region
    • This is how, occasionally, you’ll see party come second in sets for a district they came first in. Just levelling seats doing their job

For each Nation/Region I’ll not only be doing a comparison between the real result under FPTP, but also change versus the notional 2017 result outlined in the Long Read.

Note: Given I’ve input all this data manually, there may be very small errors in vote shares for a couple of places just because I’m only human, and it’s too early for other major sources like the Commons Library to have produced authoritative data to compare with.


Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • SNP – 45.0% (+8.1), 27 Seats (-21 / +5)
  • Conservative – 25.1% (-3.5), 15 seats (+9 / -2)
  • Labour – 18.6% (-8.5), 11 seats (+10 / -5)
  • Liberal Democrat – 9.5% (+2.8), 6 seats (+2 / +2)
  • Green – 1.0% (+0.8), 0 seats
  • Brexit – 0.5% (+0.3 vs UKIP), 0 seats
  • Gallagher Index – 1.1 (30.3 / 0.4)

The story of PR hypotheticals in Scotland at the moment is always that a big chunk of the SNP’s seats would go to other parties. They’d be down 21 here, with 9 to the Conservatives, 10 to Labour, and 2 to the Lib Dems. Compared to the 2017 PR estimate they’d be up 5 seats, gaining a seat in each of Clyde, Glasgow and Forth & Fife, plus two gains in Tayside. They’d also gain a vote lead in the two districts the Conservatives led in 2017. They also lead in 11 of the districts, up 2 from 2017.

The Conservatives would mostly have held on under PR, with a net loss of only two seats – dropping one in Tayside, Fife & Forth and Lothian & Borders, but gaining the Highlands levelling seat. They do however lose their vote share lead they had in 2 districts in 2017.

For Labour, although they’d lose almost a third of their seats versus 2017, it’s still 10 more than they actually won. Their losses would have been the three districts north of the Tay, plus one apiece in Glasgow, Clyde and Forth & Fife. Gaining the levelling seat in Lothian and Borders would give them a total loss of 5.

Finally, the Lib Dems would not have suffered the indignity of losing their leader under this system, holding all their seats versus 2017 and gaining two seats, one each from Aberdeen & Shire and Fife & Forth. That covers all the places they did win MPs in reality. They lead in the single member Orkney and Shetland district, the only one not to have an SNP lead.


Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Labour – 40.9% (-8.0), 17 seats (-5 / -3)
  • Conservative – 36.1% (+2.5), 15 seats (+1 / +1)
  • Plaid Cymru – 9.9% (-0.5), 4 seats (nc / nc)
  • Liberal Democrat – 6.0% (+1.5), 2 seats (+2 / nc)
  • Brexit – 5.4% (+3.4 vs UKIP), 2 seats (+2 / +2)
  • Gallagher Index – 1.9 (11.5 / 2.0)

The Conservatives making gains in Wales was one of the big stories of this election, going up 6 seats in reality. As will be a theme in other historically Labour areas (bar Scotland) that’s less to do with the Conservatives gaining massive amounts of votes and more to do with Labour losing a lot. That said, if you count up the bubbles on the image you might notice what seems like a bit of an oddity. Although the Conservatives come second in vote share, they actually win more of the directly elected seats than Labour – 15 versus 13. Given the Conservatives lead in 5 (+4) of the 9 districts, however, that’s not so surprising, and it’s what the levelling seats are there to help correct, which they duly do.

Anyway, Labour still have the lead in votes and seats here, but it’s much narrower than in the 2017 version. They’d lose their seat in Ceredigion & Powys, plus one in each of Swansea & Neath Port Talbot and Bridgend & Rhondda Cynon Taf. Although they lead in votes overall, they only win 4 of the districts, down 4 from last time.

Conservative gains are much less dramatic in PR, up just one seat on the 2017 model, coming in Bridgend & Rhondda Cynon Taf.

For Plaid, a very modest drop in votes means they’d hold as many seats as last time. They do redistribute slightly however, picking up the levelling seat in Ceredigion & Powys thanks to their much more comfortable victory in the actual Ceredigion seat this time. That means they drop the one in Caephilly, Blaenau Gwent & Merthyr Tydfil. They also win the “no change” award, having won as many seats under FPTP as is fair in both 2017 and 2019 and not changing that number between those elections.

As there’s another party to spread seats to, the Lib Dems vote share increase wouldn’t have translated to any gains. They hold the same two seats as in 2017 – though that’s at least more than their actual 0, and would mean their local leader not losing her seat (only gained in a by-election earlier this year).

Brexit would have broken through in Wales, picking up the vacated levelling seats in Swansea & Neath Port Talbot and Caerphilly, Blaenau Gwent & Merythyr Tydfil.

Northern Ireland

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • DUP – 30.6% (-5.4), 6 seats (-2 / -1)
  • Sinn Féin – 22.8% (-6.7), 4 seats (-3 / -2)
  • Alliance – 16.8% (+8.8), 3 seats (+2 / +2)
  • SDLP – 14.9% (+3.1), 3 seats (+1 / +1)
  • UUP – 11.7% (+1.4), 2 seats (+2 / nc)
  • Gallagher Index – 3.3 (19.2 / 4.3)

One of the biggest stories out of Northern Ireland last week was that Nationalist MPs outnumbered Unionist MPs for the first time. Under PR that wouldn’t have come to pass, but they would have slipped to minority status.

The DUP only drop a single seat overall, but that’s because a gain in the Foyle etc district partly offsets a loss in each of Belfast and Antrim & Mid Ulster. That effectively entails a straight trade with the UUP. They are static on 2 seats, but pick up a directly elected seat in Antrim and Mid Ulster which is why the DUP then get that levelling seat in Foyle etc. Overall, Unionists therefore drop a seat.

Nationalist parties also drop a seat overall. Sinn Féin lose two seats, one in Armagh & Down and the other in Antrim and Mid Ulster. They hold all of their directly elected seats, so basically it’s the fact they lost too many votes to qualify for the levelling seats. The SDLP hold what they had in 2017 plus gain the Belfast levelling seat, which captures the growth in vote share that saw them win Belfast South under FPTP.

That means the cross-community Alliance Party pick up two seats. They go from no seats to the second directly elected seat in Armagh and Down, showing how strong the vote that handed them North Down under FPTP was, plus Antrim and Mid Ulster’s levelling seat. They’d also hold the balance between the Unionist and Nationalist sides, perhaps a fair indication of how Northern Ireland has changed over the past few years.

Compared to the 2017 PR version, there are no changes to which party leads in each district, remaining 3 DUP and 1 Sinn Féin.

North East England

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Labour – 42.6% (-12.9), 13 seats (-6 / -4)
  • Conservative – 38.3% (+3.8), 12 seats (+2 / +2)
  • Brexit – 8.1% (+4.2 vs UKIP), 2 seats (+2 / +1)
  • Liberal Democrat – 6.8% (+2.3), 2 seats (+2 / +1)
  • Green – 2.4% (+1.1), 0 seats
  • Gallagher Index – 3.5 (18.2 / 2.6)

There was a rare focus on the North East of England when the results rolled in on Thursday, as Labour lost a rake of seats here they had held for decades, including Tony Blair’s former seat of Sedgefield. Those losses came less from a surge in Conservative support and more from bleeding a large chunk of their vote to the other parties, most notably Brexit and the Lib Dems.

That means Labour take a big hit under PR as well, losing a total of 4 seats versus the 2017 model. They lose a whopping 3 in Durham alone, plus another in each of Cleveland and Sunderland & South Tyneside. That’s balanced out by picking up the Northumberland levelling seat. They also only end up leading in 2 districts, down 2 from last time.

Remember how in Wales Labour came first in vote share, but second in terms of the directly elected seats because the Conservatives had a lead in more districts? That happens again here, with the Conservatives leading in 3 districts versus Labour’s 2. The levelling seats do their job again here, and everything comes out neatly correct. Their 2 gains come from Durham and Cleveland, where they also take a lead in vote share in for a total of 3 leads in the region.

This is the only English region where the Lib Dems don’t come third, with Brexit taking that position instead. They effectively hold the Sunderland & South Tyneside seat UKIP would have had in 2017, and gain one in Durham. Both of these are directly elected, showing the strength of their vote.

For the Lib Dems, things are slightly weird as although they gain a seat overall, they do so by picking up the levelling seats in Durham and Sunderland & South Tyneside – the same districts as Brexit – and losing the Northumberland one. Effectively, although Northumberland is their best district by vote share, they improved their result in the big cities enough they are much closer to seats there due to there being more to go around.

North West England

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Labour – 46.5% (-8.4), 36 seats (-6 / -6)
  • Conservative – 36.1% (+1.3), 29 seats (-2 / +1)
  • Liberal Democrat – 7.9% (+2.5), 6 seats (+5 / +2)
  • Brexit – 3.9% (+1.6 vs UKIP), 3 seats (+3 / +2)
  • Green – 2.5% (+1.4), 0 seats
  • Speaker – 0.8% (+0.8), 1 seat (nc / +1)
  • Gallagher Index – 2.4 (14.3 /1.7)

The North West was similarly dramatic last week, as a slate of Labour seats turned blue. As a fun quirk of FPTP, the actual seats were better balanced between Labour and the Conservatives here, whereas in the North East the votes were more balanced but seats less so. Ridiculous system.

Anyway, Labour still retain a lead overall here, though they drop a lead in one district to go down to 5. They drop 6 seats as well, one in each of Merseyside N & E, Lancashire N, Greater Manchester W and Greater Manchester SE, and down 2 in Greater Manchester NE.

The Conservatives are up a single seat overall, though that’s a net figure from gaining one in Merseyside N & E and GMNE, but losing the Lancashire S levelling seat. Despite this they also pick up the lead in that last district as well, giving them 4 here.

A modest gain of 2 seats for the Liberal Democrats, dropping the Merseyside N & E levelling seat but picking up the ones in Lancashire N, GMW and GMSE, giving them 2 seats in the latter. That’s not a weird side effect of the levelling process, that’s actually their best area, coming quite close to actually winning two seats under FPTP here.

Brexit are much weaker here than in the North East, but largely because in the actual election there were many more Conservative seats here which they stood down in en bloc. They effectively hold UKIP’s GMW seat, gain a directly elected in GMNE, plus the levelling seat in Merseyside N & E.

Finally, the Speaker would be re-elected even under PR, though see the original post for musings about whether that’s what’d actually happen if we genuinely used PR. That’s effectively a gain vs 2017 as it’s a new Speaker.

Yorkshire and the Humber

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative – 43.1% (+2.6), 24 seats (-2 / +1)
  • Labour – 38.9% (-10.1), 22 seats (-6 / -6)
  • Liberal Democrat, 8.1% (+3.1), 5 seats (+5 / +2)
  • Brexit – 5.9% (+3.3% vs UKIP), 3 seats (+3 / +3)
  • Green – 2.3% (+1.0), 0 seats
  • Yorkshire Party – 1.1% (+0.3), 0 seats
  • Gallagher Index – 2.6 (12.3 / 3.5)

Yorkshire too was part of Labour’s northern woes this election, as they slipped behind the Conservatives to come second in terms of votes here as well as losing seats to them. However, they did narrowly win more seats under FPTP, which PR obviously corrects.

The Conservatives have a net gain of a single seat, but there’s a fair bit of shuffling. They gain a seat in North Yorkshire, Leeds West Yorkshire and Humberside South, but drop a seat in each of Leeds West Yorkshire and Doncaster South Yorkshire. They also pick up another district in terms of vote lead, for 4 total.

Labour’s substantial losses come one apiece from every district except Bradford West Yorkshire where they gain a the levelling seat to come out a net loss of 6. They’re also evens with the Conservatives in terms of district vote leads, with 4.

The Lib Dems two new seats come from the East Riding and Doncaster South Yorkshire wards, but they demonstrate their reinforced vote share by their Leeds seat coming as a directly elected this time.

Brexit are particularly strong in Doncaster where they pick up a directly elected seat, rounding out their haul with the levelling seats in Sheffield and West Yorkshire South.

No seats for them, but I’m highlighting the Yorkshire Party this time as a notable local party. They pulled 1.1% this election, which in the 28 seats they contested was a static 2.1%, the growth overall coming from standing in 7 more seats. I’m absolutely convinced that if we had PR these guys would have seats in parliament, bearing in mind they got nearly 4% for the Europeans earlier this year.

East Midlands

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative – 54.8% (+4.0), 27 seats (-9 / +3)
  • Labour – 31.7% (-8.8%), 15 seats (+7 / -5)
  • Liberal Democrat – 7.8% (+3.5), 4 seats (+4 / +2)
  • Green – 2.6% (+1.1), 0 seats
  • Brexit – 1.5% (-0.9 vs Brexit), 0 seats
  • Gallager Index – 3.8 (22.9 / 3.1)

Although the focus has been on the North and Wales, Labour’s losses weren’t confined to those three regions. They also dropped substantially in the Midlands. The East Midlands are the blandest region in this post though as no one except the big three picks up seats here.

A few gains for the Conservatives, no churn this time, just straight up additional seats in Lincolnshire, Nottinghamshire South and Leicester & Loughborough.

Labour’s decline is similarly neat, five straight losses, one in each of Lincolnshire, Nottinghamshire South, Northamptonshire, Leicester & Loughborough and Derbyshire North.

That leaves the Lib Dems to pick up the balance with a couple of extra seats. Again note that strengthening of their vote, which moves the Lincolnshire and Leicestershire & Rutland seats they won as levelling seats last time to directly elected. Their additional seats are then levellers in Northamptonshire and Derbyshire North.

Compared to the 2017 PR version, there are no changes to which party leads in each district, remaining at 6 Conservative and 2 Labour.

West Midlands

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative – 53.4% (+4.0), 32 seats (-12 / +2)
  • Labour – 33.9% (-8.6), 20 seats (+5 / -6)
  • Liberal Democrat – 7.9% (+3.5), 5 seats (+5 / +2)
  • Green – 3.0% (+1.3), 2 seats (+2 / +2)
  • Brexit – 1.4% (-0.4 vs UKIP), 0 seats
  • Gallagher Index – 1.3 (17.2 / 2.5)

Similar story in the West Midlands for Labour, which last week saw them shed a bunch of seats.

The Conservatives take the lead in a further 2 districts here compared to 2017, giving them a lead in 8 of them. As in some other regions, PR mixes seats up a bit versus the last election, dropping a seat in Herefordshire & Worcestershire, but net gains thanks to Staffordshire North, Warwickshire and Wolverhampton & Walsall.

Labour would do better under PR than they did under FPTP, but compared to last election’s PR version they’re down substantially. They are down a seat in every district except Birmingham, Dudley & West Bromwich, and Coventry, Meridan & Solihull. They are also down to leading in a single district (-2).

As is becoming a pattern, the Lib Dems three 2017 seats go from being levellers to directly elected, their two new ones being the levelling seats for Shropshire and Coventry, Meriden & Solihull.

The Greens strictly come very slightly below the threshold (2.97%), but since I’m rounding to one decimal place and that’s close enough that I think we can very safely say they’d definitely have met it under PR, I’m letting them pass as it’s a useful example. Anyway, they’d pick up two seats, one in each of Staffordshire South and Herefordshire & Worcestershire, which were their two best districts.

South West England

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative – 52.8% (+1.4), 30 seats (-18 / +1)
  • Labour – 23.4% (-5.8), 13 seats (+7 / -4)
  • Liberal Democrat – 18.2% (+3.2), 10 seats (+9 / +1)
  • Green – 3.8% (+1.5), 2 seats (+2 / +2)
  • Brexit – 0.4% (-0.7 vs UKIP), 0 seats
  • Gallagher Index – 1.6 (28.5 / 2.7)

We’re now into the Southern English regions which are heavily Conservative. Labour had a lot less to lose in these areas seat wise, and although they still drop a fair few seats in the PR models, they would wind up much better represented than they are under FPTP. Brexit’s vote share is also massively underrepresented as they contested few seats.

A modest gain versus 2017 PR for the Conservatives here, losing a seat in Bristol and Avon in exchange for one apiece in Devon North and Gloucestershire.

Labour’s losses come one from Goucestershire, Dorset, Somerset and Devon North.

In 2017 this was notable as a region the Lib Dems retain such strength they’d have won all their seats as directly elected. That almost remains the case here, with a few of them moving earlier in the process. However their vote share increase only translates to one extra seat, being the Somerset Leveller (see what I did there?)

One of the Green’s historically stronger regions, with an MEP here since 2014, they’d have picked up two seats under PR, one in the Bristol and Avon district where they do typically well, plus another in Dorset which is a relatively strong district.

For those up on their politics for this area, I did check whether Independent Claire Wright who has recorded a succession of strong results in the Devon East constituency, would manage to get in under PR. Even assuming no threshold for Independent candidates, she doesn’t have enough votes to be either directly elected or pick up a levelling seat.

Compared to the 2017 PR version, there are no changes to which party leads in each district, remaining at 7 Conservative and 1 Labour.

East England

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative – 57.2% (+2.5), 34 seats (-18 / +1)
  • Labour – 24.4% (-8.3), 14 seats (+9 / -5)
  • Liberal Democrat – 13.4% (+5.5), 8 seats (+7 / +3)
  • Green – 3.0% (+1.1), 2 seats (+2 / +2)
  • Brexit – 0.4% (-2.1 vs UKIP), 0 seats (nc / -1)
  • Gallagher Index – 1.6 (26.9 / 2.3)

Again, South of England so very few Labour seats under FPTP meaning very few Brexit candidates meaning this hypothetical is even less representative of a PR reality.

Whilst the Conservatives would actually share out over a third of their seats to other parties compared to FPTP, in our PR world they pick up one more seat than 2017. That’s a net gain as they’re down one in Suffolk but up in Norfolk and in Essex South.

Labour would be on nearly three times as many seats as under FPTP, but that’s down 5 versus the PR 2017. That’d be a loss in Bedfordshire, Essex South, Essex North & West, Suffolk and Norfolk.

A very solid percentage increase for the Lib Dems here would translate to a gain of 3 seats under PR, direct seats in each of Suffolk, Essex South and the levelling seat for Bedfordshire.

Like the West Midlands the Greens are actually a tiny fraction below the threshold (2.96%) but are let through on rounding for illustrative purposes. Those seats would be in Suffolk which is by far their best district, with Essex North and West a distant second.

No seats for Brexit here, but worth noting that means they’re effectively losing the UKIP seat won here on the 2017 model. I reckon they’d absolutely have crossed the threshold if they actually stood candidates in Conservative held seats.

Compared to the 2017 PR version, there are no changes to which party leads in each district, all 8 remaining Conservative.

South East England

Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative – 54.0% (+0.2), 46 seats (-28 / nc)
  • Labour – 22.1% (-6.5), 19 seats (+11 / -6)
  • Liberal Democrat – 18.2% (+7.7), 16 seats (+15 / +7)
  • Green – 4.0% (+0.9%), 3 seats (+2 / nc)
  • Brexit – 0.3% (-2.0 vs UKIP), 0 seats
  • Speaker no longer in this region (nc / -1)
  • Gallagher Index – 1.4 (28.5 / 2.1)

Last region for the Many Conservatives, Few Labour, Almost Zero Brexit caveat about the actual FPTP results!

The Conservatives are static in terms of overall seats, with just a few trades around the region. Additional seats in Kent East, Kent West and East Sussex are balanced by losses in West Sussex, Surrey West and Hampshire North.

Labour’s hefty six seat loss comes in at one in each of Oxfordshire, Berkshire, the horrifyingly named Southampton, New Forst and Isles of Wight, East Sussex, Kent West and Kent East.

Under FPTP the Lib Dems really were robbed in this region, 15 seats short of their fair share. Even in PR terms they almost double their number versus 2017. Then they would have been 3 districts short of a clean sweep, in SNFW, Kent East and Buckinghamshire. In addition to those seats, they also pick up a second seat in each of Berkshire, Oxfordshire, Hampshire North and Surrey West.

Although the Greens are up a bit since 2017, it’s not enough to add to their total of 3 seats. One of their levelling seats is bumped from Kent East over to West Sussex instead, which is a better reflection of their support spread this time.

One more seat shared out amongst the parties here thanks to the previous Speaker, John Bercow, standing down.

Compared to the 2017 PR version, there are no changes to which party leads in each district, all 8 remaining Conservative.


Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Labour – 48.1% (-6.4), 36 seats (-13 / -5)
  • Conservative – 32.0% (-1.1), 24 seats (+3 / -1)
  • Liberal Democrat – 14.9% (+6.1), 11 seats (+8 / +4)
  • Green – 3.1% (+1.3), 2 seats (+2 / +2)
  • Brexit – 1.4% (+0.1 vs UKIP), 0 seats
  • Gallagher Index – 1.5 (15.8 / 2.2)

In recent elections, London has basically been “What if the North, but with Lib Dems and maybe Greens?”, and that shows in results here. It also means it was a better reflection of Brexit support, with plenty of candidates.

Labour keep their overall lead here, but they do lose a vote share lead in 2 districts which still leaves them on a commanding 9. The handful of seats they lose here are from Hounslow Hillingdon & Harrow, Hackney Haringey & Islington, Ealing Barnet & Camden, Greenwich & Bexley and Lambeth & Southwark.

Although they increased their vote share nationally, the Conservatives went slightly backwards in London, so they’re down a seat in PR terms. It’s the Enfield and Barnet levelling seat that gets the chop, though they do pick up vote leads in 2 districts, a net gain of one.

The Lib Dems would pick up the bulk of the seats the other parties dropped, gaining representation in Hounslow Hillingdon & Harrow, Enfield & Barnet, Ealing Brent & Camden and Greenwich & Bexley. They also narrowly take the lead in one of the districts from the Conservatives.

Greens more comfortably cross the threshold here, giving them two seats. Those come in their two strongest districts of Hackney Haringey & Islington and Lambeth & Southwark.

The Whole Picture


Votes and seats in total (vs actual 2019 seats / 2017 PR seats);

  • Conservative, 43.6% (+1.2), 288 seats (-77 / +9)
  • Labour, 32.2% (-7.8), 216 seats (+14 / – 55)
  • Liberal Democrat, 11.5% (+4.2), 75 seats (+64 / +26)
  • SNP – 3.9% (+0.8), 27 seats (-21 / + 5)
  • Green – 2.7% (+1.1), 11 seats (+10 / +8)
  • Brexit – 2.0% (+0.2 vs UKIP), 10 seats (+10 / +7)
  • DUP – 0.8% (-0.1), 6 seats (-2 / -1)
  • Sinn Féin – 0.6% (-0.2), 4 seats (-3 / -2)
  • Plaid Cymru – 0.5% (nc), 4 seats (nc / nc)
  • Alliance – 0.4% (+0.2), 3 seats (+2 / +2)
  • SDLP – 0.4% (+0.1), 3 seats (+1 / +1)
  • UUP – 0.3% (nc), 2 seats (+2 / nc)
  • Speaker – 0.1% (nc), 1 seat (nc / nc)
  • Gallagher Index – 1.4 (11.8 / 1.9)

As you’d expect, the big difference between a proportional result and the actual one is the lack of a Conservative majority. The fact Labour took a big hit in terms of votes opens up a much wider gap than the 2017 PR estimate as well, though they would have fared better than under FPTP. The other big difference is that Labour losses go overwhelmingly to the Lib Dems, who increase their number of seats by over half, plus the Greens and Brexit, all three of which were thoroughly scuppered by First Past the Post.

Staying on the Greens and Brexit, apart from the certainty they’d have a much higher vote share under PR (indeed, even under FPTP in Brexit’s case had they not stood down in the 317 Conservative seats), they also lose out slightly in this example compared to the simple versions you’ll have seen on social media. That’s because the regionalised nature of this example system effectively meant Brexit only picked up seats in the North plus Wales, versus Greens in the South of England. Whilst still quite far from what the UK is likely to implement, it’s definitely closer to reality than those simple versions.

Looking at the proportionality via the Gallagher Index, under FPTP the result was even more disproportional than it had been in 2017. The reverse is true with this PR example, as the larger number of regions where smaller parties crossed the threshold in helps smooth things out.

Looking now at the districts by party, the Conservatives have a vote lead in 59 (+9), Labour in 29 (-12), SNP in 11 (+2), DUP in 3 (nc), Lib Dems in 2 (+1) and SF in 1 (nc). All the districts Labour lost their lead in went to the Conservatives, whilst they themselves lost a lead in two districts to the SNP and one to the Lib Dems.

Finally, let’s talk about prospective governments. As always, the 5 seats in total held by the impartial speaker and abstentionist Sinn Féin are ignored for these purposes.

  • Conservative – DUP (294 vs 351)
    • There’s a net gain of 8 seats for this option, but it’s still far below the 323 for a majority. If our imaginary UK was going through Brexit still, the DUP might also not be as keen on supporting the Conservatives given their opposition to aspects of the Brexit deal.
  • Conservative – DUP – Brexit (304 vs 341)
    • A net gain of 15 seats for the most Pro-Brexit parties, but still far short of a majority.
  • Conservative – DUP – Brexit – UUP (306 vs 339)
    • We can bunch all the parties formally on the Right of the political spectrum together, but still come behind the combined Centre and Left parties.
  • Conservative – Lib Dem (363 vs 282)
    • With a net gain of 35 seats for this option, it would provide almost as large a majority as the actual Conservative one possesses. Again the likelihood of this depends on the timing of PR in this alternate universe. If the Lib Dems did their 2010 coalition and Brexit is an issue, likely an unattractive prospect.
  • Labour – Lib Dem (291 vs 354)
    • Goes from narrowly short in 2017 to massively short in 2019 thanks to a net loss of 29 seats.
  • Labour – Lib Dem – Green (302 vs 343)
    • A smaller net loss of 21 seats for this option, but still not enough.
  • Labour – Lib Dem – Green – Alliance – SDLP (308 vs 337)
    • Even if you bring Labour and the Lib Dems’ Northern Irish counterparts, still comes below what they need.
  • Labour – Lib Dem – Green – SNP (329 vs 316)
    • Of the options for a left-leaning government, this would probably be the most likely as whilst still broad (though not beyond what is common in many other European countries) it needs to satisfy the least number of parties.
  • Labour – Lib Dem – Green – SNP – Plaid – Alliance – SDLP (339 vs 306)
    • Less a governing arrangement and more a look at parties likely to have backed a second EU Referendum, should that have been on the table in our imaginary PR UK. A net loss of 13 seats though still a lead. I’ve not included the UUP here as I’m not entirely sure what their position is at the moment.

Perhaps more difficult to govern with compared to the 2017 PR options, which had the ability to make relatively comfortable majorities behind either a right or a left led government, with the latter even having the potential for slender majorities without any nationalist parties. This time any Labour led government would have to get the SNP on side. The breadth of parties required to form certain governments may once again take the British breath away, but again I’d emphasise that’s quite common in many European countries and they have yet to collapse into the sea.

Remember, under FPTP the “backroom wrangling” goes on within the big broad-church parties of the Conservatives and Labour (see the European Research Group and Momentum, for example), and there’s no option for voters to back specific factions within those big parties. At least under PR, the combination of open lists allowing voters to pick out factions within the big parties and proportionality in general boosting smaller parties prospects, voters get their say over how much weight a given group should have in government.

The reality is of course that a single party majority government will not be introducing PR. But now more than ever, the case needs to keep being made for fair votes. This isn’t about which parties are currently in or should be in power – it’s about accurately reflecting the diversity of opinion that exists amongst voters. So long as Urban Conservative and Rural Labour supporters are being denied representation, our electoral system isn’t fit for purpose. So long as the Lib Dems go under-represented, our electoral system isn’t fit for purpose. So long as people aren’t voting Green, Brexit, Yorkshire Party, or any other smaller parties for fear of wasting or splitting the vote, our electoral system isn’t fit for purpose. It remains my firm view that any party, indeed any individual, that is not committed to the cause of Proportional Representation cannot call themselves a democrat. It isn’t the 1860’s anymore. Do as most of our neighbours have and give us an electoral system fit for the 21st century.