The Stoke Central and Copeland by-elections
By-elections are a strange beast. They are an ‘event’ for parties and the media and fun for political anoraks, but their impact on the narrative usually far outweighs their electoral significance or predictive value. In turn, they are themselves difficult to predict, and national polling is not a good guide.
When it comes to individual constituencies, many people look at them through the lens of history. Both Stoke-on-Trent Central and Copeland, taken together with their predecessor seats, have been Labour for 82 years. But very long term electoral trends also reveal that the margin by which these seats are more Labour than the country has been declining.
This is not unlike the story in many English and Welsh Labour heartlands. In overly simple terms these seats contain fewer of the demographic groups that have switched to Labour from the Liberal Democrats than the national average, and more of those that have switched away from Labour to Ukip (or to the Conservatives, or indeed abstention).
And like other such areas, both defied their Labour MPs to vote Leave in significant numbers, Copeland by 60 per cent and Stoke Central 66 per cent, according to constituency estimates from Number Cruncher Politics.
Stoke is complicated both politically and demographically. Ukip came second at the last election, 17 points behind Tristram Hunt. Ukip doesn’t get a lot of help from the age structure or ethnic composition – the proportion of people over 65 in the 2011 census was similar to the national average, as is the proportion who are white.
But Stoke Central’s relatively diverse communities also show signs of poor integration. A number of postcodes have double-digit percentages of households where English is spoken poorly or not at all, putting them in the bottom 2 per cent in the UK for that measure. Elsewhere in the constituency are numerous deprived, majority-white neighbourhoods, where identification with Englishness is strong – classic Labour-Ukip battleground territory.
Additionally, Eurosceptic local independent Mark Breeze, who came fourth with just under 7 per cent of the vote in 2015, is not standing this time, which should help Ukip.
Copeland is an altogether different constituency with lower deprivation, less diversity and a substantial rural element. So substantial in fact, that Copeland has the lowest population density of any currently Labour-held seat. Think empty nesters and semi-detached houses. Labour’s challenge there will be the older age profile, with 21 per cent of residents aged over 65.
The geographically dispersed population creates a problem for the Tories, whose planned downgrading of West Cumberland hospital services means a relatively long prospective trip to Carlisle (an hour from Whitehaven by road) for A&E and maternity services. But on the other hand, the largest employer is Sellafield, and the nuclear card is relentlessly being played against Labour and its leadership.
Local factors are often discussed in relation to by-elections, but difficult to measure, unlike in general elections where the entire country goes to the polls on the same day. And as John Woodcock points out, it isn’t up to the parties to decide what the central issue is. Since it could plausibly be the NHS, or Sellafield, or something else altogether, it’s hard to say who the issues might favour in the Cumbrian seat.
The same could also be said of the improvement in Lib Dem fortunes. While no-one seriously anticipates Lib Dem victories, the party has made vote share gains in all of its recent contests, so where those votes come from will be critical.
Despite all of the challenges however, Labour has a reasonable chance to hold on in each case, not least because its principal opponents have challenges of their own.
In Copeland, the Conservatives’ problem is that they are in power. Governing parties have lost vote share in 96 per cent of by-elections since 1983, as voters take the opportunity to land a free hit, and turnout holds up better among opposition supporters.
Even when – as now – a governing party has large mid-term poll leads, such at the peak of New Labour or the Tories during the Falklands war, swings towards the government in by-elections are still not normal. The Tories would be doing very well simply to stand still, and on top of that they have Jamie Reed’s 6.5 per cent majority to overturn.
Ukip’s problem isn’t specific to Stoke, but by-elections in general. Apart from the two occasions in 2014 when it had defecting incumbents (and therefore local knowledge, some local infrastructure and in one case a personal vote) and its near-miss against Labour at Heywood and Middleton the same year, the anti-EU party typically hasn’t done much better in by-elections than the swings in national polling would suggest.
Any serious Ukip challenge would need Conservative tactical votes, but the absence of opinion polling (to convince Tories that UKIP would have a better chance of toppling Labour) makes this harder.
So while both Stoke Central and Copeland look highly uncertain, Labour supporters probably have a bit more reason to be hopeful for these by-elections than some of the commentary would suggest.