Politics: swings and roundabouts

By Richard North - May 8, 2021

One of the metrics used to chart election changes is the “swing”, between one party and another, first used in the form of a “swingometer” by the BBC in Bristol during the 1955 general election and then generally in every election since, including the current rounds.

This measure works best in a two party contest, and relies on the assumption that the collective vote share of those two parties is roughly constant between successive elections. Then, the proportion of the votes from one party moving to another – representing the “swing” – to produce a winner, gives a rough indication of the shifts in voting sentiment which have brought about the change.

This “yo-yo” portrayal, though, is an extremely crude measure and barely works, if at all, in a three-corner contest (and especially where there is tactical voting), where there are significant variations in turnout, or where there are asymmetric changes in voter behaviour – such as the decision of the supporters of one party to boycott their candidate, and refrain from voting.

In a complex electoral scenario, one might experience elements of all three phenomena, in which case the “swing” calculation becomes valueless, either for predicting the outcome of an election, or for explaining results, especially in respect on one particular seat (or groups of seats).

Nevertheless, when the result of the Hartlepool by-election was announced yesterday, most of the pundits – including the BBC – trotted out the usual mantra, noting that Labour had suffered a 16 percent “swing” to the Tories which had brought about the Conservative victory.

Yet, all the application of this simplistic metric does is obscure rather than enlighten, almost completely misrepresenting the situation as it has developed in Hartlepool over the years.

To get an inkling of what has been going on, one must go back to the very start, to the 1974 general election when Hartlepool was a new seat. Then, the contest was strictly a two-horse race, where Labour’s Edward Leadbitter polled 26,988 votes, against Conservative challenger Nicholas Freeman, who took 22,700 votes, with the turnout standing at 79.8 percent.

If we now fast-forward to the 2017 general election, the first in the post-referendum period – where Mrs May made the dubious tactical decision to go to the country – we see a new Labour candidate, Mike Hill, taking 21,969 votes to win the seat against the Conservative challenger, who gets 14,319 votes, beating the Ukip candidate (who polls 4,801 votes) into second place. By comparison with 1974, the turnout was substantially down, at 59.2 percent.

The 2019 general, however, is a bit of an oddity, where there is some concern about the progress of Brexit and the newly-formed Brexit Party fronts Richard Tice as its candidate in Hartlepool.

Tice makes a strong showing, with 10,603 votes, but the Conservative Stefan Houghton stays ahead of him with 11,869 votes. But the winner is Labour’s Mike Hill, whose vote drops to 15,464 on a turnout which has dropped to 57.9 percent.

That brings us to the current by-election, where one very obvious change is the poor showing of Reform UK, the successor to the Brexit Party. From 10,603 votes, we see a spectacular collapse to a mere 368 votes.

Received wisdom is that the Ukip and then Brexit Party votes in the past have come mainly from the Tories. However, Farage has long argued that, in the northern seats, he was taking votes from Labour, especially in the stronger “leave” seats such as Hartlepool.

With the collapse of the Reform UK vote, therefore, one might have expected some of the votes to have “returned” to Labour. But, in 2021, this is not the case. Mike Hill, the previous incumbent, has resigned over allegations of “sexual harassment and victimisation”, and is due to face an employment tribunal.

Hill’s successor, Canterbury-born Paul Williams, is parachuted in from nearby Stockton where he worked as a GP partner. And, although he campaigns vigorously, his reward is to suffer the lowest Labour vote in the history of the constituency, at a mere 8,589. If the Reform UK votes have been recast, they certainly did not go “back” to Labour.

Now here’s the interesting thing. If these votes didn’t go to Labour, only a fraction of them can have gone to the Tory challenger, Jill Mortimer, who only polled a mere 15,529 votes to win the seat, up only just over a thousand votes on Stefan Houghton’s showing in 2017 – representing only 68 percent of the vote polled by the Tories in 1974.

Thus, the real culprit here is the turnout. From 1974, when 49,688 voters passed through the polling stations, representing 76.9 percent of the electorate, this had dropped to 41,835 voters in 2017, giving a turnout of 59.2 percent (and not very much different in 2019, when the turnout was 57.9 percent). But, in the by-election just past, turnout plummeted to 29,933, calling in at 42.7 percent. Between 2017 and 2021, nearly 12 thousand voters stayed at home.

Given that the Tory vote largely held up, even if it was significantly down on historic levels, it is reasonable to postulate that most of the stay-at-home voters were former Labour supporters.

Looking at the results from this perspective, it is fair to say that the Tories have barely moved from their 2017 voting figure, and polled only 68 percent of their 1974 vote. Thus, one can conclude that the seat went to the Tories because the Labour vote had collapsed to an historic low, with the Labour candidate unable to attract Brexit Party (formerly Ukip voters) back into the fold.

In essence, the Tory showing is actually quite mediocre, on which basis, rather than suggest that there had been a swing to the Tories, it would be more appropriate to point to that collapse of the Labour vote: the Tories didn’t win the seat in any real sense. Labour lost it.

This cannot be said to be any great endorsement for the Tories and, in many respects it paints a picture of party politics in decline. In round figures, just four out of ten bothered to cast a vote and, of those, only two voted for Johnson’s party – a mere 22 percent of the electorate.

Yet, for all that, this result is being painted as a great victory for the Conservatives. In her analysis, the incomparably stupid Laura Kuenssberg gushes that the “rickety folding tables” [in the counting centre] “looked like they could hardly cope with the weight of votes for the Tory candidate, and now elected MP, Jill Mortimer, in Hartlepool”. With only just over 15K votes to record, one wonders how the clerks would have coped in 1974.

But, in Kuenssberg’s foetid, London-centric world, she sees the lacklustre Tory performance as “more evidence for the Conservatives that they are digging further and further into territory where once they were total outsiders”. In her terms, “They didn’t just win here, they romped home”.

Despite the over-cooking, though, nothing can disguise the fact that this is a bad result for Labour. Already, the wibbling from the Left is in full flow, with talk of Hartlepool voting “by a landslide for a Conservative”.

Thus, the indications are that Labour has no more idea why they lost the seat than the Tories have for winning it, and are completely unaware of the electoral dynamics which shaped the result.

More to the point, Labour is far from coming to terms with some basic truths, which means that there will be plenty more “Hartlepools” on the horizon.

With Labour in such disarray, Johnson doesn’t have to win any seats as long as Starmer is so obligingly losing them. On the other hand, if Starmer reads the runes correctly (a very big if), his task is not as daunting as would appear. All he has to do is stop losing.