Politics: an “overwhelming victory”?

By Richard North - October 7, 2023

What makes Thursday’s Rutherglen and Hamilton West by-election interesting – albeit only marginally so – is not the result, per se, but the reaction to it.

With the Labour candidate, Michael Shanks, polling 17,845 votes, a triumphant Keir Starmer has told jubilant activists that they “blew the doors off” with what is being hailed as the party’s “overwhelming victory”.

The Times has John Curtice endorse the euphoria, with the pundit noting that parties that win a by-election often succumb to the temptation to exaggerate the significance of their success.

But in truth, he goes on to declare, “it is difficult to dispute the claim of Scottish Labour’s leader, Anas Sarwar, that his party’s success in Rutherglen was ‘seismic’”. The scale of his party’s victory, Curtice opines, “suggests the SNP’s decade-long dominance of Scotland’s electoral politics is now under serious threat for the first time”.

Citing election data to support his view, he tells us that Labour’s performance was its best in any Scottish contest since the party’s Westminster representation was decimated in the 2015 Westminster election.

Its share of the vote at 59 percent, he says, was only a little below its tally of 61 percent in the constituency in 2010. With the SNP trailing on 28 percent, “the result was redolent of what once was regarded as normal for a west of Scotland by-election: Labour success with the SNP trailing”.

In an egregious example of arguing from the particular to the general, we are then invited to consider that, if the 20 percent swing from the SNP to Labour were to be replicated across Scotland as a whole, Labour could find itself with 40 seats, while the SNP would be reduced to half a dozen. Such an outcome, Curtice says, “would almost exactly match the outcome of the 2010 election”.

The BBC’s political editor, Chris Mason, is no less fulsome, grandly declaring that the result “transforms the Scottish political weather”. In so doing, he says, it “changes the forecasts some will make about the next general election”.

Politico has it that: “Scottish Labour dares to dream again after thumping Rutherglen by-election win”, quoting victor Michael Shanks declaring: “There is no part of this country where Labour can’t win”.

Standing back from the hyperbole, though, we have to remind ourselves that this was a by-election and that such contests have their own rules. Results can be treated as a pointer to a forthcoming general election, but it is also the case that they are not necessarily replicated in a full-blown electoral contest.

What the pundits – and most of the legacy media – so often fail to do is to highlight the turnout, in this case a mere 37.2 percent compared with 66.5 percent in the 2019 general election and the 61.5 percent in 2010 when Labour held the seat before losing it to the SNP in the 2015 election.

If we now look at the actual votes cast for Labour in this by-election just past, we see Michael Shanks taking 17,845 votes, representing 22 percent of the total electorate (as opposed to votes cast), compared with the 18,545 votes taken by the losing Labour candidate in 2019, when he took 23 percent of the total electorate.

By comparison, therefore, in this by-election cast as an “overwhelming victory”, the winning Labour candidate actually polled fewer votes than his unsuccessful predecessor in the 2019 election.

One difference, of course, is the turnout, dropping from 53,794 in 2019 against the 30,477 who could be bothered to vote at the by-election. But the other difference is the way the SNP vote has collapsed, from 23,775 (29 percent of the electorate) in 2019 to a mere 8,399 (10 percent of the electorate) last Thursday.

While the pundits parade the 24.1 point swing for Michael Shanks, in strict terms the actual number of people voting Labour this time round has dropped slightly. What gave Shanks his victory was not any great surge of support for his party, but a collapse in the SNP vote.

For sure, there may have been some SNP voters who moved to Labour, and indeed some Conservative voters may have changed allegiances, as the Tory vote was down, but the evidence is entirely compatible with this being another “stay-at-home” election, on the part of the incumbent’s supporters.

This one might have expected in any events, as the incumbent, Margaret Ferrier, was driven out of her seat over a breach of Covid rules, after being arrested in in January 2021 and charged with “culpable and reckless conduct”, for which she pleaded guilty and was later sentenced to community service.

As with Owen Paterson and the North Shropshire by-election in December 2021, voters will sometimes punish the transgressor’s party, even if they have since departed the scene.

But if one is trying to read the runes, one place to look is the Selby and Ainsty by-election in July of this year when Keir Mather of the Labour Party took the seat from the Conservatives after the resignation of the incumbent.

Mather gained his seat on the largest swing in a by-election since the 1994 Dudley West by-election (and the second largest ever swing to Labour in a by-election), with a turnout of 44.8 percent compared with the 71.7 percent in 2019.

There, with a posted swing of 21.4 points, we see a similar dynamic where the Labour vote stayed roughly the same (increasing from 13,868 in 2019 to 14,456 in 2023), while the Conservative vote collapsed, dropping from 33,995 to 12,295.

We see this same dynamic again and again, even when there is no change in the winning party, as with the City of Chester in December 2022. Then I wrote under the title “staying at home”, noting that the Labour winner, Samantha Dixon cast 17,309 votes, against here predecessor’s 27,082.

The delivery of that the Guardian headlined “the best result since 2010” rested entirely on the collapse of the Conservative vote, dropping to 6,335 from 20,082 in 2019.

To give him his due, Curtice – described by the Telegraph as the UK’s “most eminent psephologist” – does concede that the situation is more nuanced than the results would indicate, although not with any great clarity.

Voters often use by-elections to cast a protest vote against the government, he says, leading sometimes to very high swings. He interprets the 17-point fall in SNP and the 11-point fall in Conservative support as “strongly suggesting” that voters have lost faith in the governments in London and Edinburgh. But, he warns, in a general election some of their lost voters may return, and Labour will not necessarily do as well as they did in Rutherglen this week.

Despite this, he can’t help but be transfixed by the notional “large swing Labour recorded in Rutherglen, together with the 24-point swing from the Conservatives the party recorded in the Selby by-election in July. These are just the kind of performances that oppositions record if they are on course to win a general election, he concludes, pointing to the Mid Bedfordshire and Tamworth contests which are to be held in a fortnight.

The only serious note of caution seems to come from the Financial Times, which has as its headline: “Labour buoyed by Rutherglen win but pollsters sound note of caution”. The by-election victory in Scotland has energised activists, the paper says, but pollsters warn against over-interpreting the result.

Here we have Chris Hopkins, political research director at pollster Savanta, telling us that “the margin of victory is impressive, but that can be chalked up to by-elections being a bit weird, and easier to squeeze the Tory vote”.

Patrick English, associate director at pollster YouGov, thinks it is realistic to expect Labour to get around 25 Scottish seats at the general election thanks to falling support for the SNP and the coalescing of the pro-unionist vote around Labour. But he adds: “Does this mean Scottish politics has gone back to pre-referendum format? Does this mean Labour could get 40 constituencies in Scotland? No, let’s not be silly about it. It’s a by-election”.

Even then, these pundits miss the point. Although there is no means of knowing exactly what is going on, the evidence supports the thesis that we are not so much seeing a swing towards Labour, as voters deserting the Tories (and, in this case, the SNP) – staying at home rather than changing their votes.

Quite possibly, both phenomena are at play but, if we are seeing primarily stay-at-home votes, this suggests that support for Starmer’s Labour is far more fragile than the election results would indicate.

To an extent, this is supported by Emily Gray, Scotland managing director of Ipsos. She says her polling suggests that while the public wanted change they were still not “deeply in love” with Labour. When it comes to Starmer and his party, she says, voters are “still fairly lukewarm”.

Whether we have an “outstanding victory” or not, therefore, there might still be everything to play for.