Digital Politics 1: First, they break you down
A short series on UK political messaging, social media, and the Culture Wars
The first (this one) is about the recent past, and the power of segmentation.
The second (which I’ll publish next week) is about the present, and what happens when complaint gets weaponised.
And the third (which might come the week after, or sooner if I get my act together) is about what progressives can learn from fandoms.
Back to Brexit
It seems to me that digital politics in Britain got off to a bad start.
While there had been some online political campaigns in the UK before 2016, digital politics really broke through as a mainstream concern in the run up to the referendum. Whatever you think of the tactics used by Vote Leave, their approaches to digital campaigning and data analysis have shaped the Tory party’s campaign management and policy development over the last seven years and their legacy can be felt in the set of rather gloomy, contrarian policies currently being rolled out by Sunak ahead of the 2023 party conference.
To understand how we got to where we are, this post is going to rewind and reflect on some of ways data and digital technologies were being used in 2016, at home and in the US.
An incomplete set of recollections
What follows is not a forensic examination of the digital and data landscape in 2016 but a retrospective view of some the activities that shaped what came next. And because it’s as good a place as any, I’ll start with what I was doing.
One the early projects we ran at Doteveryone involved putting digital mentors into the offices of four MPs. This meant that from May to July 2016 we had four mentors moving between the parliamentary and constituency offices of Yvette Cooper, Calum Kerr, Matt Warman and Norman Lamb. The mentors were chiefly there to do some problem solving and see what could be gained by having someone on hand who would “think like a technologist”. In several cases this meant doing pretty basic things like setting up a rule that put the 38 Degrees emails into a single folder and drafting useful copy for “contact us” pages on the MP’s websites.
This was also a fact-finding mission. When our chair Martha Lane Fox announced Doteveryone on BBC One, the fact that that understanding technology was an important part of political life that most MPs had no training for was a core theme of her speech. Even now, MPs without ministerial posts are supported by very small teams, and if the same person has run your constituency office for 20 years it is entirely possible they will have no experience of how “digital transformation” has changed the working environment. This was also in the very early days of the Parliamentary Digital Service and the roll-out of core IT was still a work-in-progress.
For reasons to do with funding and political fairness, we couldn’t publish a great deal of our actual findings. We had hoped to roll the programme out throughout parliament, but issues of governance and impartiality got in the way, however there was still plenty to learn from our sample of four. The two things that stand out now include the fact that some of the case management systems being used were held together with digital string and external good will, while digital processes — where they existed — tended to be ad hoc.
Meanwhile, in another bit of the political forest, Dominic Cummings had discovered CRM.
I’ve written elsewhere about the fact that Cummings seemed to be a bit blinded by the science of this, but whatever his team were doing, it almost certainly wasn’t held together with string. According to Cummings’ blog, the Vote Leave team hired physicists to run voter segmentation and do A/B testing; even at the time the physicists seemed like overkill – it would have been significantly cheaper to bring in a solid marketing team – but whatever they did and whoever they used, it appears to have worked.
While Cummings doesn’t use those terms, the clever thing his team appears to have done is to bring sales and marketing tactics to political campaigning: they scraped available data to create a pretty good voter segmentation and then identified promising leads.
This meant they weren’t speaking to potential Leave voters as a singular mass, but tailoring communications. (If Cummings’ blog post is to be believed, this was specific enough to enable them to knock on potential Leave voters’ front doors to speak with them. This raises a few data ethics points, but I’m going to save those for another day.) From what I can tell, the thing that made the difference wasn’t cutting-edge data science or physicists or machine learning, it was running the show like a solid marketing campaign from c.2010.
The run up to the 2016 referendum also sticks in my mind as the time another shift occurred: political news in the UK started to move at the speed of social media. In the summer of 2016 I’d still habitually pick up the evening edition of the Evening Standard at 6pm, but for the first time I was finding that the headlines printed at 4pm were wildly out of date. While the specifics of whatever Farage was up to during those long afternoons have now slipped my mind, it was certainly the case that political events had stopped being staged in time for the World at One or the Six O’Clock News and were running on Twitter Time instead.
Vote Leave’s engagement with data and the shift away from traditional media cycles represented a paradigm shift in how digital politics worked in the UK — but it wasn’t a shift that took everyone along.
It seems weird to say so now, but in 2016 the Internet was still pretty novel in some parts of Establishment life. Until the Cambridge Analytica scandal broke (and, indeed, for some time afterwards), things that happened online were quite frequently not taken seriously by the kinds of people who relied on their secretaries to print out their emails.
Of course, hindsight is a wonderful thing, but it’s also worth remembering that while a lot of progressive people in the UK were Extremely Online, this overall level of digital engagement didn’t necessarily map to how the Remain campaign was run.
While the operational Vote Leave team were taking advantage of the kind of tools a decent corporate would have used to shift FMCG goods or summer holidays, I don’t have the sense that Remain were doing the same. I have no special insight into how the digital parts of the Remain campaign were run (as far as I know, no one has written thousands of words of blog posts detailing their technology approach), but I’m going to take a bet here and say that staying in Europe was treated as being an objective public good rather than as a really tactical campaign. Losing the Referendum was a shock to the Extremely Online people who’d been group hugging on Facebook, but the data-driven legwork wasn’t necessarily being done to convert people who felt sceptical, left out or otherwise unconcerned.
And it turned out that Vote Leave weren’t the only ones who were good at segmentation.
Later that year, just after Trump had won the US election, I found myself at the kind of event I often went to when I ran a think tank – a closed-door policy convening in a stately home – at which I started chatting to someone who worked in security monitoring. While there’s been no shortage of analysis of how Trump won the election, this snippet of conversation — which describes how a particular kind of digital-first populism is born — has stuck in my head ever since.
The analyst observed that one of the unusual things about Trump had been that he had an opinion about almost everything.
While presidential candidates usually stick to a few well-chosen promises, Trump was offering lines on the kinds of esoteric topics that slick political campaigns usually avoid. As a result he was building tiny coalitions everywhere, among people whose most strongly held opinion or frequent source of irritation was suddenly being voiced by someone of significance. This led to the formation of micro-alliances that spanned demographics and political allegiance.
Just like the folks who had joined the Facebook group about sleeping on the cold side of the pillow in 2009 or the people who follow feline influencers on TikTok, people with nothing in common beyond hating it when bananas stay green were reassured to suddenly find other people who hated it too. This diffusion of opinions also meant there was not a single argument to rebut, but hundreds, if not thousands, of sometimes trivial and sometimes nonsensical perspectives that were just not on the radar of the opposition.
Whether this was an organic or an intentional strategy, the net effect of this was similar to the segmentation run by Vote Leave, which built a broad church by bringing otherwise disconnected people together in support of the same cause.
Digital Social Capital
Of course, none of this will be news to anyone who has studied social media, but this repatterning of our allegiance and opinions is still something that politicians, pollsters and policymakers are coming to terms with.
A constituency-based, first-past-the-post system needs people in the same neighbourhood to have some things in common in order to gain critical mass, but for some demographics (particularly younger people and everyone without secure housing) your social network is more likely to live on your phone than on your street. You don’t have to be a genius political strategist to realise that this is more of a problem for Labour, which doesn’t tend to attract so many older homeowners, than it is for the Conservatives.
My assessment is that the puzzle for Labour messaging is that it needs a much more complex set of messages than it has deployed so far, so it can:
(a) build a broad church that galvanises many people around change
(b) make the specific concerns of swing voters in key constituencies feel included in that broad church without overly sweating the details of how it all fits together; after all, politics is complicated, and lots of things can and do happen at once
(c) convert the digital social capital of younger progressives into actually getting out the vote
The next post in this series is about how the Conservatives have achieved (a) and (b) through a process of Other-ing and Complaint; the third will be about what Labour can learn from fans and fandoms about creating places of inclusion and celebration.
Thanks for reading Just enough Internet ! Subscribe for free to receive new posts and support my work.