This draft thread header was written in December 2019, but due to pressure of work I never had a chance to send it in. Alterations are shown in italics. 

In 1918, Couponed candidates took 20 seats, Asquithian Liberals three seats, and Labour or Independent Labour candidates twelve seats in Wales. I mention this not from any particular desire to talk about 1918 in Welsh politics, but because it marks the last occasion to date when Labour failed to top a nationwide election in Wales. Admittedly in the 1999, 2007 and 2016 assembly elections they failed to win a majority but they were still by far the largest party and it proved impossible to form a government without them. The only time since 1918 they have not won an outright majority of seats at Westminster was 1931, when they won 16 out of 35 seats, still ahead of the Liberals on 8 and the Tories on 11. Traditionally, it was said they did not count the Welsh Labour vote, they weighed it. It saved time and gave much the same results.

2019 was in fact no exception. Admittedly, it was Labour’s weakest performance at a Westminster election in Wales since 1983, when they won only 20 seats of 38. But they won 22 seats, almost 632,000 votes (40.9%) and had nine seats where they scored over 50% of the vote. By any standards that is a handsome performance. Had it been replicated UK wide, Corbyn would be leading a majority government right now. 

But this masks weaknesses, and where there are weaknesses, there is value. Let’s run the rule over them.

1) Labour were almost obliterated outside the South.

In 2017 Labour had 48.9% of the vote and held 28 seats in Wales, including five in the north – Delyn, Vale of Clwyd, Wrexham, Clwyd South and Ynys Mon – that fell to the Tories this time. Their vote share fell by eight points, to 40.9%. Outside the south, the sole seat that Labour held was Alyn and Deeside, by the wafer thin margin of 213 votes. This is suggestive of a party being driven back upon its heartlands.

2) Labour’s vote has become less efficient.

In 2005, Labour won 29 seats with 594,821 votes. In 2019, they won 22 seats with 631,855 votes. This is partly explained by population growth in Cardiff and partly by higher turnout. But it is also a striking indictment of Corbyn’s grossest error – his strategy piled up huge votes in selected pre-existing Labour seats, while ignoring long held seats that were becoming marginals and semi-marginals in other areas. Wales, with its tradition of voting Labour (only Montgomeryshire has never had a Labour MP) was foolishly neglected, and a price has been paid. 

3) The old heartlands are not as hearty as they once were. 

Even in the south, there were some close shaves. Both seats in Newport were held with majorities of less than 2000 votes, as was Gower, while Bridgend was lost. Even Torfaen was held by less than 4000 votes. Although nine Labour MPs achieved a voteshare of 50% or more, just five had a majority of over 10,000. In 2005, those figures were ten and eleven. That is even worse for Labour than it sounds, because of the seats with ten thousand plus majorities, three are in Cardiff, one (Rhondda Cynon Taff) is turning into Cardiff hinterland, and therefore only Merthyr Tydfil and Aberavon could be considered traditional Labour heartland. One of the Cardiff seats – Cardiff Central, now Labour’s safest seat in Wales by any measure – was a Liberal Democrat seat until just five years ago, indicating an increasingly volatile electorate.

4) Brexit is big in Wales.

In 2005, UKIP stood in 31 seats and got 20,297 votes – only in three seats did it get over 1,000 votes (Torfaen, Pontypridd and, perhaps surprisingly, Gower, where it got 1,264). In 2019 the Brexit Party stood in 32 seats and got 83,908 votes. Its worst result was in Arfon, where it got 1,159 votes. Clearly there are a lot of angry voters out there right now, far more than in 2005, who reject the EU and Labour’s cosmopolitanism. In Wales, in seats where Brexit did not stand, such voters seem to be shifting to the Tories. What should therefore disturb Labour is that in five of their seats the Brexit vote was bigger than their majority, and in six more if the vote transferred en bloc to the Tories (yes, yes, I know) Labour’s majority would be shaved to under 2000. Despite this, the main conclusion must be:

5) Wales is turning to two party politics 

In 2005, Labour won 29 seats, and were second in a further eight. In 2019, they won 22 seats and were second in a further 13. So far, so not very different. But burrow into the detail and a deeply disturbing trend for Labour emerges. In 2005 the Liberal Democrats won four seats and were second in eight, Plaid won three and were second in a further eight. The Tories won three and were second in sixteen, usually a very distant second. Now, Labour are second in 13 seats, while the Tories were second in 24 of the 26 seats they didn’t win.

The Liberal Democrats are a pitiful rump of two second places, one of them very distant, while Plaid, despite winning four seats, were second in precisely none. (The balance is Blaenau Gwent, where Brexit were second – coincidentally, in 2005 that seat was won by an Independent.) The Tories even came second in Llanelli and the Rhondda where Plaid have previously won those seats at Assembly level. Admittedly, in the Rhondda it was by a huge margin, but in Llanelli they ran Labour much closer than I think anyone expected – just 4,670 votes. In 2005 Labour had a majority of 7,234 over Plaid in Llanelli with the Tories 4,514 votes further behind. Although the Labour vote share has barely shrunk (and indeed Plaid’s share is down but comparable) the Tories have found 7,000 votes from somewhere.

This is because, as vote shares show, the Tories are rallying opposition behind themselves. In 2005 three parties won over 250,000 votes, with Plaid winning a respectable 174,838.  In 2019, two parties won over 155,000 votes. The Tories went from 297,830 votes in 2005 (almost exactly half Labour’s tally of 594,821) to 557,231 this time around, less than 75,000 behind Labour. Plaid’s vote held up quite well, with 153,269. The Liberal Democrat vote however has moved almost en bloc to the Conservatives, falling from 256,249 and being a serious national player to a pathetic 92,171 – just 8,000 more than the Brexit Party got.

Minor party candidates have also been more or less obliterated, winning 24,355 votes in 2019 compared to 48,721 in 2005 – a cut of half. This is one reason why although Labour’s raw vote has increased its majorities have declined. One of the great advantages for Labour and the reason they have hung on to power for so long is because their opponents are badly split. This election marks the moment that changed.

6) All this presupposes the existing boundaries will be used in 2024.

This is of course a very bold assumption. Even though the boundary review has restarted it is most unlikely Wales will be allowed to keep 40 seats. And it is the Valleys that would get hammered. On the proposed boundaries, my rough expectation would be Labour 14 seats (-8) the Tories twelve (-2) and Plaid two (-2). There are a few new seats that look marginal Llanelli (potentially three way) and Newport. Both of those are held by Labour. Labour desperately need this boundary review to be stopped or they face a stiff challenge to keep their 101 year record going. Oh, damnit, they can’t, they face a government with a majority of 80.

7) So where is the value?

Not, in fact, in 2024, given point 6. In 2021 there will be elections to the Welsh assembly. For the first time, the Tories can realistically paint themselves as the main alternative to Labour. UKIP, who acted as spoilers in 2016, will be no more and Brexit are likely to have followed them. Mark Drakeford will have to explain that after 22 years of Labour government, economic stagnation, underfunded public services and rampant corruption scandals at all levels of government are not his responsibility. Good luck with that. His story on Covid-19 is at best decidedly mixed as well – how much he will be blamed for the catastrophic mishandling of the ‘circuit breaker’ and how far he will get credit for the rollout of vaccines is questionable. Admittedly, in the Assembly elections Plaid will have a relevance denied to them at Westminster. But given Plaid’s targets and positioning, that is more likely to damage Labour than the Tories.

Therefore, I would say there is value to be had in Tories – most seats at the next assembly elections. You can get around 6-1 on that at the moment, where it should be around 3-1. Bear in mind, 23 seats might well be enough for that if, as I expect them to, Plaid get about 14-15 in total. That looks feasible on these numbers. Polls showing the Tories just two points behind do not exactly suggest Labour are pulling clear either.

There is a lot of ruin in a political party (just ask those who, like me, could see a potential breakthrough by the Liberal Democrats in 2019). Welsh Labour may rejuvenate itself, somehow. A century is a long time. Labour have proven remarkably resilient in ways nobody expected. They have a huge pool of loyal voters. But there is still the chance here of a sea change in Wales. I have no doubt I speak for even many of Labour’s 631,855 voters in saying it is long, long overdue.

Y Doethur

Y Doethur teaches History and Politics in Staffordshire. He is of Welsh (Valleys) descent and lived in Wales from 2001 to 2008.