In this election, it is not surprising that there has been a huge controversy about turnout. Jeremy Corbyn's leadership pitch was, from the start, that he would bring back lost voters, people who have given up and felt alienated from the system. He said, from the start, that Labour needed to win over people who had stopped voting. Part of the surge in Labour's support during this campaign has been driven by groups of voters who ordinarily don't vote, or who haven't voted before. That, to an extent, reflects success on Corbyn's part. But it is one thing for people to be politically won over by Corbyn; it is another thing for them to be persuaded they must vote.
There is always the risk that the bubbles of political life, encouraged by social media -- what they call the "pluralization of life-worlds" -- has created a false sense among some of us of just how widespread and how deep the excitement is. That risk is tempered by the very obvious successes of Labour's campaign, by the tremendous reception Corbyn gets in diverse settings, by his surprisingly strong media performances, and by the sheer calamity that is the Tory campaign. Nonetheless, I think we all -- all of us who will campaign for Labour tomorrow -- worry that we're missing some hidden structural resources of conservatism, or confusing wish-fulfilment for analysis. Even when we try to check our biases, passionate longing has a way of making one dream.
And we are dreaming, and not just that Labour will do well. Let's be honest, we want there to be a Spring surprise, a shock Labour win, a rupture, something that radically confounds received wisdom. We want the entire set of assumptions by which neoliberalism rules, above all the internalised defeat that goes under the mantra 'There Is No Alternative' to be irradiated. We want the cultural predicates, by which everything should have a market price, and by which refugees and those on benefits are sadistically punished, while banks and large corporations endlessly get away with it, to be overturned. We want, on Friday morning, to wake up to a different country.
That, though we all know the odds against it, though we all know how much power is mobilised to stop precisely this from happening, exerts a gravitational pull on all our analysis. And we have to be endlessly careful not to give in to it.
In this election, because of the unusual degree of volatility and uncertainty, polls have been wildly divergent in their assumptions. Some polling companies have consistently predicted a Labour share of the vote around 33-35%. Those would include ICM, Opinium, ComRes, Panelbase and (until tonight) TNS. YouGov (until tonight), Survation and ORB have shown much higher Labour shares, in the region of 38-40%. The ironic role reversal tonight was that it was TNS which put Labour up 5 points to 38%, and YouGov which, with some methodological adjustments, brought Labour down three points to 35%.
Choosing between these is impossible, and reflects a mix of desire and gut instinct. One of the problems with polling, one of the reasons why it can get things wrong, and why weighting is such an issue, is that its method of inquiry starts from the assumption -- which then has to be corrected, post-hoc -- that political action is based on individualised units of opinion, in which every opinion is of equal weight and competency. In fact, ideologies are usually formed in collective, group contexts, based on shared experiences: your class, your race, your gender, your locality, your religion, your nationality, your street, your occupation, and so on.
Posing the question in an individualised way sidesteps the collective way in which people form ideas and make decisions. And it has the effect of forcing people to rely on impressions gained from recent media campaigns. That renders polls particularly susceptible to media-driven buzz. Recall that before the 2015 problems which produced such reflection from pollsters, there was also the 2010 "Cleggasm". I don't remember too much panic about that, but polling companies seriously overstated the Liberal vote (with the effect of uncharacteristically understating the Labour vote) in that election year.
This is, of course, not to dismiss polling. It is a highly technically advanced way of gaining a snapshot of ideological forces in motion -- raw material which politicians and public relations professionals are able to work with, and work on very effectively. It is often wrong, but not usually that wrong. And with the huge differences among polling organisations, they all have a degree of rationality in their predictions, informed by past practice as well as (frankly) a degree of unconscious ideology manifesting as 'gut instinct'. They have a huge amount of future business riding on their predictions, so they have every interest in getting it right. YouGov probably has a mixture of professional concerns in adjusting its estimate and business concerns -- the costs of underestimating Labour are a lot lower than those of overestimating labour.
But it is to say that there is nothing authoritative about polling. It is always just raw material to act on. We read the polls looking for points of weakness in our opponents, moments of opportunity, areas where intervention can make a difference, ambiguities and ambivalences in what appears to be a consensus.
And as we approach polling day, we read them with only one answer. Let's get as many people out to vote as possible. Let's make all the pollsters wrong. Let's confound their expectations. And even if we can't, even if what we're doing now is building a basis for the future, even if we know that we were never going to turn Britain into a left-wing country within just two years, or just one election campaign, in the wake of Brexit and after years of reaction... Still, let's bring on the flood.