Forget leaflets: this election may be decided by shifting populations

What is the point of political campaigning? It’s bordering on heresy even to ask this question, let alone to suggest that six manic weeks of battle buses and TV debates and leaflets destined only for the recycling bin might not be worth the millions they cost. So we political junkies carry on parsing every throwaway remark for meaning, obsessing over every stunt and gaffe, searching for that one mythical moment that will in retrospect seem to have changed everything. If nothing else, it’s comforting to think there must be some kind of logic to it.

Related: Labour has a week to turn this election into a winter war on austerity | Frances Ryan

But never has so much time and money been spent on the basis of so little evidence. It certainly feels right that TV debates should help the wavering make up their minds, but research suggests undecided voters are the least likely to tune in; the debates work more like football matches, with fans watching to cheer on their side. It seems obvious too that mud thrown during campaigns must stick, but when asked which election stories they remembered from the last week, 40% said “none” and 17% just said “lies”, according to polling by the former Conservative deputy chairman Michael Ashcroft.

Most startlingly of all, a recent review of dozens of campaign field experiments led by the US political scientists Joshua Kalla and David Broockman concluded that the average effect of campaign contacts and advertising on US voters was precisely … zero.

Campaigning isn’t completely pointless, of course. It’s great at reminding people there’s an election on, which pushes the committed out to vote. Candidates taking “unusually unpopular positions” have an impact too, according to Kalla and Broockman, which helps explain why in 2017 Theresa May’s lead collapsed after she published her social care proposals. But campaigning is bad at changing minds, with their study estimating that at best it may convert around one in 800 voters. Broadly speaking, we all put far too much weight on what happens in the six weeks before an election and not enough on the long-term trends shaping the previous five years.

Why are people in coalfield communities once devastated by the miners’ strike surprising even themselves by considering voting Tory? Why did Labour do unexpectedly well in Canterbury last time? And why is the Surrey commuter belt now flirting with the Liberal Democrats? Of course this topsy-turvy election can’t be reduced to any one single factor. But a surprisingly unexamined driver is demographics. It’s a well-worn truth both that Britain is ageing, and that older people are more likely than younger ones to vote Conservative. But it isn’t ageing uniformly, and in a general election that matters.

As the thinktank Centre for Towns pointed out in a report on ageing last year that deserved more attention than it got, the populations of many of the so-called “red wall” towns across the north and Midlands, where Labour is under siege from the Tories, are ageing faster than average.

Young people move away to work or study in the big city and don’t come back, because they see fewer prospects for them back in Grimsby or Ashfield or Bolsover. A vicious spiral can set in, where an exodus of bright young people leads to a declining skills base in their old home towns; companies become reluctant to invest there, creating even fewer reasons for the next generation to stay. Big cities are getting younger, and southern towns are ageing roughly in line with national trends. But many of the smaller towns that Boris Johnson must win to get a majority are quietly haemorrhaging the age group most likely to vote Labour, and becoming dominated by a more hostile one, while younger people’s votes pile up in the big cities where Labour was already strong.

This isn’t just about pensioners’ greater tendency to have backed Brexit, either. These are sweeping generalisations with many individual exceptions, but older people tend to hold socially conservative attitudes on issues such as immigration and defence; pensioners have been relatively shielded too from the long post-crash squeeze on wages and working-age benefits, at least compared to the young. And voters who remember the hard left in the 1980s may be more sceptical about Corbynomics than those to whom it’s new and exciting. Brexit has driven a new wedge between Labour and its former heartlands, but the truth is they’ve been growing apart for years, in ways not easily reversed by last-minute offers of free broadband.

In the south, meanwhile, human geography plays out very differently. Places such as Canterbury became richer pickings for Labour in 2017, just as Cheltenham and Guildford are now for the Lib Dems, at least partly because these towns are changing as 30-something professionals find themselves priced out of London. Young families seeking somewhere affordable to live have spread first to the outer suburbs and then beyond, across Surrey and Gloucestershire and Kent and Sussex, bringing socially liberal attitudes and frustration with the economic status quo into once sleepy market towns and villages. (If the Canterbury effect isn’t working for Labour as it did in 2017, that’s perhaps because the Lib Dems are more competitive this time around among these mainly remain-leaning voters.)

As the Centre for Towns report stresses, demographics can’t explain everything; they’re just part of a bewildering jumble of social and economic changes, some pulling the result one way and others another. But the pattern it identifies suggests the Tories should be closing the gap in northern towns while Labour should be piling up votes in cities such as Bristol, which is indeed what we saw in 2017. What drives elections may not always be individuals changing their minds, but places slowly changing over the years as a consequence of decisions both politicians and voters have often unwittingly made.

Related: Digital election: what demographics are the parties targeting?

Old habits die hard, so knowing all this won’t stop us political junkies obsessing over every moment of the last few days. Nor is it to denigrate the heroic efforts of people out knocking on doors in the freezing cold and dark, each time hoping this conversation is the one that makes the difference.

But elections are won in the years leading up to them, not the weeks. This time more than ever it’s hard to shake the feeling that the letter is already in the post, even if nobody yet knows what it says; that all we’re really doing, in these last frantic days, is keeping busy until we can open it.

• Gaby Hinsliff is a Guardian columnist