I’m watching various people I follow on Twitter and Facebook who support the Labour Party (or occasionally other left-wing parties), and I’m getting the definite impression that Labour is currently imploding. Or exploding. Or shattering. Whatever, it seems less like a party and more like lots of angry people trying to blame each other for what went wrong.
My personal opinion of what went wrong varies wildly depending on what mood I’m in, so I won’t go into that here. But I do have a very definite opinion as to why no-one knew it was going wrong. Because this shattering of opinion after the election defeat is a sharp contrast to the unanimity of opinion before the election that Labour Were Right. (Or maybe, that Labour were the best party with a realistic chance of winning).
Now, apparently, the majority of people who voted disagreed with me and my friends, given that Labour lost. But I didn’t see that. I get most of my updates from Twitter and Facebook. And the big thing about Twitter and Facebook is that you only see updates from people you know* and have chosen to follow. And these are mostly people who are quite like you.
Now, of course, I see plenty of dissenting opinions. But these are usually retweeted by someone who agrees with me and with a comment along the lines of “look at this idiot”. I don’t even have to think of my own reason to dismiss them, it’s pre-supplied. Under those conditions, a different point of view is unlikely to persuade me to rethink. And of course I see other opinions on less-selected sources of information such as the BBC. But that’s a fairly small quantity of information compared to the torrent of updates on social media. It’s easy to assume that the ranting UKIP supporter on Newsnight doesn’t represent most people, and that most people agree with me, when most opinions that I see agree with me.
Only, as it turns out, those opinions were a very selected group.
Who are all now angry with each other for not seeing something that we all missed.
Which is that with social media, we see mostly opinions of people who agree with us, not people who don’t.
I’m not saying that the selectivity is new. People have always mixed mostly with others like us. But the selectivity is much more hidden: on social media you connect to total strangers and find that they agree with you. It’s easy to think that you’re seeing a random slice of the world when you’re actually seeing the slice of people who agree with you.
The problem with homogeneous groups is that without dissenting opinions, you can make progress much faster, but you don’t see obstacles. I’ve heard groupthink described as “the fastest, most efficient way to run off a cliff”. (I can’t find the source for that, but if you let me know I’ll credit them).
I’m not sure what the solution is. Follow people you hate on Twitter? Ban politics from Twitter and Facebook**? I’m fairly sure that it’s not something we can reliably correct for consciously – our brains heuristically form opinions based on the number of pieces of evidence we see; and post-hoc adjustments are generally of smaller magnitude than the effect from a biased selection of data. In other words, you might, when really concentrating, remember to rule out the selection effect, but it’s unlikely to change what you really believe.
I wonder if similar effects are causing a polarisation of opinions: left/right, pro/anti-Islam, pro/anti-Israel, pro/anti-immigration, and all the various other divides we have across the country. I wonder. I don’t know. Plenty of other things are causing divides too: increasing inequality, for example.
And, of course, all of this will only be affecting those of us who spend a lot of time looking at social media. There are a whole lot of other people out there who don’t use social media, and we have no way of ever knowing what they think.
Until we have an election.
* For a given value of “know”, depending on how quickly or reluctantly you add people on social media.
** Actually, that might be a good idea for other reasons, not least that I’d personally have to post less crap.