« | Home | »

Episode Five: Flatliners

Tags:

The BBC has a handy electoral poll tracker on its website. Here’s where we are:

bbc poll of polls

I’ve never seen polls like that. The thing which really stands out is that nothing stands out. They are as flat as the flattest of flatnesses. Five months ago, on 7 October, Labour was on 34 per cent and the Tories on 32. In the latest polls, Labour are on 33 per cent and the Tories on 34. That small advantage is not stable, though. The parties are passing the tiny lead back and forth between them like dope-smokers conscientiously sharing a joint.

Most people don’t think much about party politics. To get their attention you need, in the language of the pros, to ‘cut through’. For all the noise and kerfuffle, this election hasn’t. It isn’t self-evident why, and no doubt this very fact will be the subject of study for some time to come, but there are a number of possible reasons. One of them might be that the parties and their leaders are so boring and so small, and arguing within such narrow margins of neoliberal consensus, that the electorate just can’t be bothered to think about them. The problem with that line of argument is that something very similar could have been said about many other elections, without there being this same picture of non-engagement and stasis.

The fixed-term parliament might be playing a role. When the government can choose the timing of an election, the electoral game turns on picking a moment of waxing popularity. If the right moment doesn’t come, the government hangs on as long as it can, in the hope of a convenient accident (small war, big scandal). When it clings on like that, recent history suggests, it tends to lose. Here are the last eight elections, arranged in order of how long the government waited before calling it:

1997 5 years 1 month lost
2010 5 years lost
1992 4 years 10 months won
1979 4 years 7 months lost
2001 4 years 1 month won
1983 4 years 1 month won
1987 4 years won
2005 3 years 11 months won

 

The general rule seems to be that the longer a government is in office, the more likely it is to lose. The exception is 1992, when everyone thought Labour would win, until the first exit polls came through. (I remember that vividly because my wife and I went over to a friend’s house to watch it. We set out at 9.50 expecting to arrive at a party, and walked through the door at 10.05 to an atmosphere of recent bereavement.)

The causation doesn’t run from longevity to unpopularity, but the other way round: governments which were worried about losing stayed in office as long as they could. The election always came as an event whose timing had been determined by a trajectory of popularity and public feeling. Not this time, though. The date has been set for five years, and the polls look set too: stuck down, glued in place. We’ve seen this one coming a long way off. The parties haven’t had to scramble and cobble together a pitch for our attention. Maybe that’s a problem: our opinion of the contestants is all too fixed and definite and well-known.

I wonder if another contributing factor is social media, especially Twitter. Twitter is the speediest and most powerful news source that there has ever been. But it’s very very easy – too easy – to arrange your supply news and information and opinion so that you only hear from people who already agree with you. You can’t be on Twitter without screening out trolls and morons, and it’s a simple matter to cross over from that to screening out all differing viewpoints. You can confirm a viewpoint or perform an allegiance in 140 characters, but you can’t change somebody’s mind. If we can read one thing from that poll tracker, it’s that people’s minds aren’t being changed. It’s almost as if, to adapt Jean Baudrillard, the general election is not taking place.

Comments

  1. Amateur Emigrant says:

    I’m not sure that social media is making that much of a difference in the way you suggest. In the past you would read a certain newspaper – you were a Telegraph reader, or a Guardian reader, or a Sun reader, and your impressions of the political landscape would be shaped by the editorial stance of your paper of choice, along with the broadcast coverage. I don’t see that the self-selecting ‘talking amongst yourselves’ aspect of social media is very much different in that way from the pre-Twitter era.

    In fact I think these days I see more comment via Facebook and Twitter from news sources I wouldn’t previously have looked at. Often they tend to be the more swivel-eyed rantings offered up by my network for ridicule and outrage, but not always. In any case, unless you are genuinely undecided in your voting intention, and interested enough to want guidance, then even 140 pages of comment won’t change your mind, let alone 140 characters.

    Personally I think the lack of ‘cut through’ and the narrowness of policy variation between the two main parties is more significant. Such cut through as has happened took place some time ago. UKIP snared the support of the 15% of the population who blame foreigners for everything. Even in Scotland the SNP’s cut-through was achieved in the referendum and their domination over the collapsed Labour vote has also been a flat line on the graph since September.

    The true picture of non-engagement will come from the turnout. Will more than two-thirds of the electorate actually give enough of a toss to cast a vote? In such a close run election if there is no significant increase over the 2010 turnout then there will be cause for concern.

  2. Geoff Roberts says:

    I have the impression that the effects of social media are more insidious than you have suggested. It’s possible – to ‘vote’ (and bet) on anything you like and to punch in ‘like’ or yes or whatever every thirty seconds, no matter what the issue may be – ‘Was Milliband more effective than Cameron?’ Or ‘Will Manchester City beat United?’ There is absolutely no consequence, which is why people put up the most absurd comments in the media. This is probably why flat is flat – the question has become a proposal which I find vaguely amusing. If you were to suggest in Twitter ‘Ms Thatcher will win the election,’ I bet that you would get 50% + agreeing, just for laughs. So political conflicts are reduced to a set of vague statements – to identify the core differences between Labour and Conservative you need some though and some research, not simply an ‘off the top of my head’ feeling.

  3. streetsj says:

    UKIP’s “appeal” was I think more than just xenophobic. At the time of the European elections, when they polled more votes than any other party, it was much more of a general protest vote/ anti-politician vote.
    I was expecting its support to drain away back to its traditional homes but I’m less confident of that now. It has held up much better in the opinion polls than I expected but, who knows, maybe the reversion will come in the voting booth.


Advertisement
Advertisement