I caught up with David Runciman, Professor of Politics at Cambridge University, after he delivered The Political Quarterly's annual lecture 'Nobody knows anything: Why is democracy so surprising?' to discuss the volatility of electoral politics, fake news and the state of democracy.
In your talk, you mention the “five P’s” who have been wrong about recent elections: politicians, pundits, pollsters, political scientists and prediction markets. You argue that pundits, pollsters, political scientists have been getting their predictions wrong for a long time, and don’t have strong incentives to change this necessarily. But it seems to me that politicians and prediction markets do have strong incentives to get predictions right. What do you think has derailed them in recent years?
My argument was that for [pundits, pollsters, and political scientists], being interesting is often better than the business of being right, whereas for politicians and prediction markets, being right is their livelihood. And it’s one of the big puzzles, because they should have sufficient diversity that they pick up on the signals that other people miss. My feeling is that those other two groups are also suffering from groupthink. They’re listening to each other rather than listening to the signals that are coming from the wider public.
But then the questions is: why is that? Because the whole point of them is that they’re not closed off, they’re meant to be open towards lots of different opinions.
I suggested a couple of reasons. One of which is the feedback loop. The people in politics listen to what the City thinks, and the people in the City are listening to see what the betting markets think.
And then the other thing that I genuinely think is a big problem for contemporary politics is the divide between the people in all of these groups, who tend to be professionals in one form or another, and so have gone to university, and the people who left education earlier.
I don’t think it’s surprising that people who comment on politics in newspapers get it wrong, because you might expect that, and they’ve done that for two hundred years. But with these other groups there seems to be a bigger disconnect in the past two years.
I think your idea about groupthink and the feedback loops is a really useful one. In your talk, you also mention how dramatic UK elections in recent years have thrown us all “off kilter”. And your basic claim there is that the digital revolution has increased the supply of information, and this makes politics more unpredictable. Could you comment on some of the evidence you find is the strongest in linking increased information supply to increased voter volatility?
I think there is clearly evidence for increased volatility. One of the things that political parties have got wrong in recent years is that election campaigns are not meant to matter. There is meant to be a sort of settled set of opinions out there, and then election campaigns churn it up a bit, but when it comes to the vote, the settled opinion reasserts itself. And people revert back to the things that they knew before the campaign started.
And now, not just in UK elections but around the world you’re seeing that the four six weeks of a campaign actually changes peoples’ minds. And that does seem to be evidence that short-term cascades of information actually alters their [voting] behaviour. That does seem to be relatively new, since the last three to four years.
We can see these things happening, [although] it’s much harder to provide the evidence that shows this volume of information over here creates this volatility over there. But these two things do seem to be connected.
People inhabit different information universes which don’t overlap. And then, within those universes, there does seem to be a lot more movement as people change their political behaviour.
But it’s so early. People have been studying politics and elections for ten, twenty, thirty years are having to adjust to a reality which is two, three, four years old. It can’t be a coincidence that this period of time in which people sharing news on Facebook has become, for many, the dominant source of information. It’s almost certainly part of the explanation.
That’s interesting, and made me think: what about the issue of digital literacy? 12.6 million adults in the UK lack basic skills in this area and five or so million adults have never used the internet. I know these are relatively small numbers compared to the country as a whole but how do you see this fitting in to your theory about voting behaviour?
That’s a good question. My understanding is that the vast majority of those people are older people – though not all.
Again, one of the things that has been overturned recently is this idea that ‘old people vote and young people, particularly students, don’t vote’. But in the 2017 election, the fact that young people and students turned out to vote took everyone by surprise. Some of that churn and volatility seems to be disproportionately affecting certain parts of the population. And, of course, there are still people voting in traditional ways according to traditional patterns.
I heard a presentation by YouGov where they said one of the things that they missed was that there was this view that you can segment populations by age group. There’s an age group that are 65+ and you treat that as a homogenous group. They’ve decided that that was a mistake. There are now so many people who are 75 – 80 + and those people are really hard to get to. They’re often living in care homes, they’re not online, but they vote. And they tend to vote Tory.
It’s a complicated picture. There’s churn among young people, there’s stability among older people, and there’s a group of older people who maybe do behave in traditional voting ways but aren’t being picked up.
We heard a bit about your views on the functioning of democracy and how surprise was potentially an important part of it. How should we measure whether or not democracy is functioning better than it was, say, twenty years ago? I mean, what kind of criteria could we use – if that’s a possible thing to do?
That’s a big question. One of the really useful functions of these surprising elections is just to make people aware of the country that they live in. A lot of people said after Brexit, after Trump, especially after the two [UK] general elections, is that they found something out about their democracy that wasn’t visible before. And that’s got to be, I think, a positive thing.
But there is that surprising-ness that reveals itself within democratic elections. The thing that I’m warier of is you get randomness and surprise in the functioning of the institutions themselves. They actually become destabilising for the framing of democratic politics.
It’s kind of amazing the extent to which Americans accept that Trump is their president. Despite the resistance. It’s still the case that his legitimacy, though it’s questioned, is not really in doubt. So surprise within that context is probably a good thing. But then when that context itself starts to get a bit unpredictable so people aren’t really sure what the shared values and norms are, they aren’t really sure about which institutions are legitimate…
I think one of the wider consequences of Trump and Brexit, not so much the [2015 and 2017] general elections is that it would make the functioning of the institutions themselves unpredictable. Maybe not elections, elections are probably the one thing that we’d probably still cling onto, but some of the other basic functioning like the value of parliamentary government, role of referendums versus representative democracy – some of the basic frameworks of politics – if you can’t be sure from one year to the next which foundational rules will apply, I think that’s more hazardous.
The two are related. You get more of the second because of the first. Because if you’re going to elect Trump, if you’re going to vote for Brexit, you don’t just get information about the country you live in, you get consequences, which are, particularly in Trump’s case, potentially the destabilising of the institutions.
So there are surprises within democracy and there are surprises about democracy. And I think on the balance sheet, I’m inclined to the pessimist side. I think democracy is really struggling to cope at the institutional level. You do need predictability in the way institutions function.
In terms of Trump’s legitimacy – and I’m not questioning the fact that most Americans are not pushing back too much on the fact that he should be in the White House at all, but the issue of ‘fake news’ is in some ways unsettling his legitimacy. I felt like in your talk you implied that fake news is a misleading term: ‘My news is different to your news’. Can fake news be so readily dismissed?
There definitely is quite a lot of fake news out there and there’s unquestionably some manipulation going on. The main thing I was trying to say was that I’m uncomfortable with the idea that the answer to the question ‘Why is politics so surprising?’ is: ‘Fake news’.
It’s just too simple.
If you take that as the answer, you’re missing the bigger picture, yes. And there’s something uncomfortable about saying that it’s because “Bad people deceived gullible people” because it takes away the extent to which voters are actually finding out things for themselves. There’s a danger that we end up calling one person’s preference for a news source over another news source ‘fake’ or ‘illegitimate’.
It’s not clear to me that the people who calling some things ‘fake news’ are themselves on the side of the angels when it comes to objective reporting. It’s a variant of this thing where people say: ‘The people who voted for Brexit were lied to, the £250m [for the NHS on the bus]… they didn’t understand the issue’. I’m not sure the people who voted to remain understood the issue.
I don’t think anyone did. And it smacks of class bias as well; people being very judgemental about Brexiters.
Yes, and particularly, I have to say, highly educated people. [Remain voters] were often just as tribal as the other side.
Absolutely. Fake news or not, there is of course the problem of the echo chamber, of filter bubbles, and people not interacting and engaging with dialogue. So, my last question is how do you view the role of mainstream media in the future. Does it still have a place where people with differing views can interact?
I’m not sure whether people ever really interacted through mainstream media. There’s a nostalgia in this country for a time when everyone watched the BBC and ITN. I’m not sure that there was a huge amount of interaction going on when the only way you could get your news at nine o’clock was the nine o’clock news, but there was clearly a shared water cooler thing, there was a shared grounding for what people were then thinking about.
There was a synchronicity in time at least, and then potentially conversations happening in spaces that weren’t filtered.
That has gone. I don’t think there’s going to be a time where the BBC will manage to capture all the different things that people think. The same with newspapers.
And there is a dystopian version of this which is that people are self-curating their news sources but it’s all from one giant corporate entity. I don’t think Facebook is a malevolent organisation, but just that kind of monopoly power. You don’t predetermine what happens on these networks but you are the provider of the network.
Facebook would say, well unlike the old days of the BBC and ITN, on our network people really are interacting, they really are exchanging ideas, they really are communicating. Which is sort of true. But there is something chilling about the thought that all of this is happening on a framework that the people who are exchangers of information don’t control and don’t understand and don’t see.
I mean, the thing about old-fashioned media is that there was at least some transparency in what was going on.
And they change the algorithms and the way that they operate their platforms without revealing what those changes are.
There’s that famous thing you can do where you go on the Guardian website and you can use an ad revealer to show you at any given moment how many advertisers are tracking you. Machines are watching you. And so much of it is hidden.
And, of course, we know that people act differently when they know they are being observed, as well.
It’s so new that we don’t really deeply understand what it means and it’s changing all the time. I mean people’s behaviour is unquestionably changing all the time as they react autonomously and independently to their understanding of what’s happening to them. I’m a bit nostalgic for the times when people at least understood what reading the news meant. It meant you read it and they supplied it. Now, reading the news means they’re watching you, that’s different.
I think there’s a level of trust which is lost as well, of people who are the gatekeepers of knowledge.
Yes, and to repeat I don’t think it’s because there are bad people who are exploiting the system to engineer election results. In a way, it’s more serious than that. Good people as well as bad people – I think that most people who work at Facebook are decent people, they’re not proto-fascists. They’re probably well-meaning liberals. And the fact that those people engineer the system is in some ways scarier.
Anya Pearson is Social Media and Events Editor at The Political Quarterly.
A recording of David Runciman's lecture is available below.