This isn’t a question into which I’ve got any personal insight. I went out canvassing in several seats, and I couldn’t swear to you that Bury South wasn’t one of them; the name of the candidate doesn’t ring any bells, though, so I’m guessing not. So I don’t think Bury South was the place where a mock-furious resident jokingly threatened to come and batter us – or rather, as I quickly realised, a genuinely furious resident seriously threatened to come and batter us, and would have done if he hadn’t had to go back inside for his outdoor shoes. Nor was it the place where a hailstorm began, apparently centred on me personally, in the (long) two minutes between my ringing a doorbell and the door opening; or the place where someone who wasn’t even there explained patiently through his Ring device that my party leader was in fact a terrorist, in case I hadn’t realised; or the place where an Asian man and his partner told me that yes, they were definitely going to vote Labour, but told me very quietly and closed the door as quickly as they could.
Ah, the memories.
But no, I don’t remember Bury South. So this is based purely on publicly available data (viz. Wikipedia) and one or two weird tricks in Excel.
Click to embiggen, probably (WordPress has been very weird lately).
What’s going on here? These are the vote shares of the main parties (red, blue, orange), plus UKIP (purple), independent Right-wing parties and individuals (navy) and the Greens and independent Left-wingers (green). Rather than ordering them from Right to Left, I’ve grouped the two major parties and all the minor parties together (ordered Right to Left in both cases). The purple block includes the Brexit Party (2019) and the Referendum Party (1997); the orange block includes the SDP (1983 and 1987). The navy block includes Ivan Lewis (2019) – unfair, perhaps, but he certainly wasn’t standing as an independent Left-winger. Percentage shares are given every time a party gets 3% of the vote or more.
Although the vote shares of all parties add up to 100% in each column (check the first couple of columns if you don’t believe me), the overall height of the column is scaled to turnout. To put it another way, the total turnout can be read off on the left-hand Y axis from the height of the composite column; the (complementary) height of the translucent grey column represents the proportion of the electorate who didn’t vote (less than 20% in 1992, more than 40% in 2001).
The other wrinkle is the red line. This, measured against the right-hand Y axis, gives you the Labour percentage majority over the Conservatives at each election: positive every year from 1997 to 2017, negative 1983-92 and 2019. (Which is another reason why it would be fatuous to call this a “Red Wall” seat; when people talk about places that have been safe Labour seats time out of mind, they’re usually going back a bit further than 1997. “Nay, lass, it’s all Labour round here – has been since Euan Blair were a lad…”)
So what do we see? First, in 1987 and 1992, we see mobilisation of non-voters, primarily to the benefit of the Tories. Labour are coming back from the 1983 low, but – in this seat at least – they’re mainly coming back by reabsorbing the SDP vote and driving the Lib Dems back down to single figures.
1997 looks different, and the two elections after that look the same only more so. Turnout is down in 1997, and it looks as if it’s Tories who are staying at home (although a few of them have gone over to the Referendum Party). There’s also been a substantial shift from the Tories directly across to Labour, who now take the seat. Turnout is through the floor in 2001 and 2005, and again the Tory vote is hitting historic lows; the Lib Dem vote is recovering, however, apparently mainly by taking votes back from Labour.
Then there are 2010 and 2015. The Tory vote is recovering, but only slowly; the real action is in the ‘minor party’ section, which – in this seat as in several others – appears to have been (a) a repository for anti-system, ‘sod the lot of them’ votes and (b) a playground for the far Right (in this case, BNP and English Democrat as well as UKIP). The Lib Dem vote collapses in 2015, as it did in most places; the beneficiaries, in ascending order, are the Greens, Labour and UKIP.
Now look at 2017. Turnout’s up a bit, but what really leaps out is the level of two-party polarisation: even with a Kipper, a Lib Dem and a right-wing independent (listed in descending vote share order), Labour and the Tories together take almost 95% of the vote. Even in the three-party days the two parties’ share never reached 91% – and it had been below 80% at the three(!) previous elections. Voter mobilisation and massive polarisation, greatly to the benefit of Ivan Lewis MP (and was he grateful?).
2019, finally, was… 2019: turnout falls; the minor-party area takes 13% of the vote instead of 5%, as separate fringes of pro- and anti-Brexit voters make their respective points; the Tory vote increases a little while the Labour vote declines quite a lot; and Ivan Lewis himself standing in person isn’t really in the race but does attract 1,366 votes, in a seat taken by the Tories with a majority of 402.
What happened in Bury South, then, was that the New Labour years drove down political participation, demoralised Tory voters in particular, and created a relatively small but significant group of voters whose main motivation was to protest against what they saw as a rotten system. The Coalition, austerity and the collapse of the Lib Dem vote hardened this group’s opposition to politics as usual. In 2017 voter mobilisation and polarisation saw most of those voters going to the Tories, but a minority of them – together with the Green and some of the surviving Lib Dem vote – went to Corbyn, seeing him (correctly) as an outsider planning to shake things up. Finally, in 2019 – just as the bad name that four years of negative campaigning had hung on Labour finally began to cut through – the party’s Brexit positioning brought it into the realm of “politics as usual”; the minor-party vote duly revived, along with the (quietly continuing) revival of the Tory vote; and Christian Wakeford took the seat for the Tories by a margin of 0.8%.
To put it another way, what happened in 2019 was a small-scale replay of what had happened in 2015 – which in turn was only possible because of what had happened in 2001 and 2005 – together with the unwinding (after much persuasion) of what had happened in 2017. Add unfavourable background conditions (the debasement of the national debate, a cynically effective Tory campaign) and unpredictable local factors (Ivan Lewis MP (ret’d)) and you’ve got a Tory win.
How Labour win back similar seats I’m not sure, although one answer would lie in the mobilisation and polarisation exemplified by the impressive 2017 result. (And 2019 didn’t just happen, let’s not forget; a lot of people put a lot of work into reversing that result.) That said, 2017 nationally was also a record year for the Tory vote (highest vote share since Thatcher, more votes than Labour took in 1997); a rising tide floats all boats if you’re not careful, as the 1987 and ’92 results here demonstrate. The reverse strategy – depolarisation, demobilisation and generally driving down the vote – seems to have worked rather well in the Blair years, but I would urge anyone planning a repeat of that particular strategy to remember that New Labour began by exclusively driving down the Tory vote; the attack on Labour’s own vote came later, and began from a high base. Also, of course, the chart rather strongly suggests that 2001 and 2005 led (through the medium of a lot of grumpily apathetic ex-Tory voters) to 2010, 2015 and 2016, which is very much where we came in.
One other thing to stress about Bury South, finally, is that it was a close and a flukey result, as several of Labour’s 2019 losses were. None of the above would have mattered if one in six of Ivan Lewis’s voters in 2019 had stayed with Labour – or if one in 50 of Christian Wakeford’s had stayed at home.
On the bright side, Wakeford’s our comrade now, so none of it does matter! Isn’t democracy great?
Workers’ Playtime 