Do manifestos make a difference? Almost no one cares what political parties have to say on actual policy

Jeremy Corbyn and Labour are going to get creamed in the election in two and a half weeks’ time. That, at least, is the consensus: the polls show around a 15-point lead for the Tories; one seat projection suggests that the Conservatives are on course for a 48-seat majority; you can get about 27 to one odds on a Labour majority at Betfair (and even a hung parliament is close to three to one).

Of course, that was the consensus just before the last election, too. And that time Labour did not get creamed. They merely lost, and by doing so outperformed expectations so much that some of their own ministers have since convinced themselves that it counted as a win.

Over the last few days the party manifestos for the 2019 election have been published, Labour’s having been designed with what looks suspiciously like Microsoft WordArt. Since they came out — especially since Labour’s — people have been eyeing the polls, looking for any movement; lo, there has been none. The release of the polls back in 2017 coincided with a collapse of the Tory predicted vote share and a surge in Labour’s, but there has been no such “manifesto bounce” this time around.

I’m a bit intrigued by the idea of the manifesto bounce. I am not a political journalist, but I’m inside the bubble of people who pay a wildly disproportionate amount of attention to political news. There’s a good chance, because you’re reading this, that you are too. But most people are not. According to polling, only about 5% of the general public were aware of, for instance, Jacob Rees-Mogg’s ill-judged comments about Grenfell; only 2% were aware of Tom Watson’s resignation or Ian Austin telling people to vote Tory.

More recently, a Labour manifesto pledge that was heavily trailed ahead of publication, the promise to provide nationalised broadband, reached the consciousness of a whopping 12% of people. Around half of people haven’t heard of any of the most-talked-about stories; in fact, about half haven’t heard of John McDonnell, as Danny Finkelstein pointed out in the Times last week. Finkelstein also says, rightly, that this is not because people are stupid or lazy: it’s because their lives are busy and this stuff is not vital to know. But still, it is the case.

If these numbers are even close to being right, it seems prima facie implausible to me that most of these stories have the faintest impact on polling. Let’s be generous and say that 20% of people are aware not simply that Labour have a manifesto out but that it pledges, say, increased investment in rail infrastructure and spending 3% of GDP on green research and development.

The margin of error on a YouGov poll of 1,000 people is 3% – that is, the creators are 95% confident that the true answer lies within 3% of the answer. So for the manifesto to have made a difference that could be confidently distinguished from random noise, nearly one person in every six who were even aware of what was in it must have changed their minds and decided to vote Labour.

And that 20% of people will, disproportionately, be the people who were already most politically informed, and so will hardly have been surprised to learn that Corbyn’s Labour want to combat climate change or spend money on public services. It just doesn’t seem like there’s room for manifestos to make much difference.

I did go and check this with a pollster and a political scientist. They agreed. “Generally speaking [manifestos] have bugger all impact,” the pollster told me. But there is an interesting outlier: 2017.

“Both manifestos were unusually impactful last time — or, at least, were perceived to be,” Rob Ford, a political scientist at Manchester and occasional UnHerd contributor, told me. “So everyone is very focused on them this time.”

The 2017 election was unusual. Looking at the poll trackers from the time, Labour’s two-week rolling polling average jumped from a nadir of 25% on 6 April on 13 April to 36% by polling day on 8 June. At the same time the Tories slumped, albeit less starkly, from an average of 47% to 44% between early May and early June. The manifestos were released in mid-May.

The received wisdom is that those manifestos drove the turnaround in the polls. But it’s very hard to be sure that this really happened, or if it’s a classic case of post hoc ergo propter hoc. If it did, it was probably driven by the terribleness of the Tory manifesto rather than anything particularly good about the Labour one. “I suspect there is a negativity bias with manifestos,” says Ford. “Not many opportunities to win support, but plenty of opportunities to lose it, by saying things that annoy people, like the ‘dementia tax’, or that reinforce negative perceptions of your party,” as in the decision to un-ban foxhunting.

(That may, he says, be why the Tory manifesto is so thin this time. You can’t piss people off with your policies if you don’t have any policies.)

Labour’s own manifesto, of course, is filled with extravagant spending pledges, and one negative perception of Labour is that they offer to spend lots of money without worrying about where it comes from. The things they promise are mainly popular, but, as John Curtice points out, people just don’t trust Corbyn to successfully deliver them. If the “you can’t win an election with a manifesto, you can only lose it” hypothesis is correct, then it might do more harm than good.

But more likely, I think, is that it will have negligible impact. Very few people will hear about it, and of those who do, most of them have already made up their minds; the numbers just don’t look big enough for it to work any other way. My suspicion is that it had very little impact last time, but that because they happened around the same time, and because we are creatures who need to tell simple stories with clear causes, the manifestos were seized upon as the trigger.

“All campaigns end up with symbols that we fixate upon after the fact,” says Ford. “The Conservative manifesto was one last time, but it was just one element in a much broader omnishambles; it’s not clear which elements were really decisive.”

It’s less than three weeks to the election. By this time in 2017, the polls had been narrowing for about a month, although the consensus was still that the Tories would win comfortably. There’s no sign of such a change now and the smart money is definitely on Labour getting a drubbing. But if they don’t, and people say the manifestos were the turning point, don’t believe them.

---

Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers