A quick Cambridge Analytica thought. Especially since the C4 broadcast, they look a lot like traditional American election ratfuckers, doing the sort of thing that Karl Rove used to get up to. I would enter a caveat here. As there is a ratfucker playbook, so there is a counter-playbook. You have squads of lawyers standing by at the capitol and volunteer poll-watchers at the voting booths. You challenge changes to the election law at the statehouse or in the courts. So on or so forth.
The problem, though, is that CA promised to deliver some of the effects by different means. You can’t send lawyers to a polling station to see off Facebook ads, and you certainly can’t do that to stop your own supporters’ and indeed your own response to the barrage of robot trolling from putting people off.
Also, there’s a question of scale. Classic dirty campaigning involved local knowledge, physical presence, and expertise in the laws of 51 states. A digital demobilisation campaign, though, can be delivered from a distance, in a semi-automated manner, at industrial scale. One in four US citizens seems to have been reached by at least one item of fake news. Not all of them can vote, so the percentage of electors was probably even higher.
As a general rule, changing the scale of something is always important and often has qualitative consequences.
“ratfucking, from a distance, at scale” is a good description, and also why the world weary types who trot out “it’s just the same that Obama did back in 2008” have got it wrong – that campaign still used secondary identifiers and had standard GOTV tactics in mind.
A number of commentators have noted the tendency for governments to apply top domestic audiences what is first tried out on insurgent populations elsewhere, and in that connection the stories from Kenya, Nigeria et al are interesting as is:
https://theprint.in/politics/exclusive-inside-story-cambridge-analytica-actually-india/44012/
On a separate note, for whatever reason it appears that such people tend to keep the control of companies operating in this way well within their grasp, looking at the records of Emerdata – thats two mercers and Erik Prince’s business partner:
https://beta.companieshouse.gov.uk/company/10911848/officers
That together with their frequent ex-military connections makes me wonder if you can trawl for such things by work backwards from names and graphing searches from places like companies house (Bout airline/flight plan style).
Worth adding in that there is some evidence about ability to affect turnout thru FB:
https://www.nature.com/news/facebook-experiment-boosts-us-voter-turnout-1.11401
I’ve been thinking a lot recently about how key the change of scale might have been.
A lot of people seem to concentrate on “CA weren’t doing that much different” without asking about scale (partially just the historical accident of the scale of social media usage reaching new highs) changed the way we need to think about the influence models.