Verdict: my reluctance was justified, and I should have held firm.
It vastly exaggerates what CA did, takes several of their garbage, baseless claims at face value and ultimately makes a pantomime villain out of a minor character.
I've spoken to mathematicians who are adamant that "psychographics" simply doesn't work.
CA were selling snake oil. Nobody has ever been able to prove that any of it worked.
The Great Hack laps it all up.
Not to mention infuriating snippets like:
"CA would say you've gotta target this state, that state"
"How would they know that?"
"That's their secret sauce!"
Erm, any political science graduate could build a model to determine which states to target.
Ultimately I'm with @OwenGleiberman: it's understandable that people were shaken by Trump's win and sought desperately to find a bogeyman, to reassure themselves that without this one evil data firm, there's no Trump. But it's not true variety.com/2019/film/revi…
I'd also strongly suggest that anyone who still buys the vastly exaggerated version of CA's role and capabilities reads @Soccermatics's excellent book Outnumbered, which delves into the details of CA's claims and other cases of overhyped algorithms goodreads.com/book/show/3676…
Two final points:
• When Obama campaign used FB users' friend lists & behavioural nudges to persuade voters in 2012, it was written up as smart use of data docs.house.gov/meetings/IF/IF…
• Read this thread by an Obama campaign staffer explaining what they did
I'm not making value judgments on how either the Obama or Trump campaigns used FB data to try to persuade voters, but you cannot say one was smart and good, and the other was evil.
Right or wrong, they're both industry-standard ways of using personal data to attempt to persuade.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Essential chart from the new mega report on the general election by @Moreincommon_
The vast majority of people — including Reform voters — said the Tories lost because they were incompetent, not because they were too left or right wing.
And to the extent that people thought they were either too left or right wing, equal shares gave each answer.
There’s one very clear message and anything else is a distraction.
When asked what were the biggest mistakes the Conservatives made in government, the common themes are not left or right, but:
• Mismanagement
• Lack of integrity
• Incompetence
• Dishonesty
• Corruption
• “They are chaotic”
The lack of enthusiasm for Labour at this election really is striking.
Among those who plan to vote Labour tomorrow, the party is much less well-liked than in 2019, 2017 or 2015 (no data before that).
Quite a flimsy voter coalition that could unravel in the absence of results.
The Conservatives weren’t especially popular with their backers in 2019 (mainly a vote for Brexit and against Corbyn), and this was a big part of why they fell so far since then, but Lab voters this time are even less enthusiastic about their party than Tory voters were in 2019.
Of course, all that matters tomorrow is winning more seats than the opponents, and Starmer’s Labour will manage that very easily.
But if they don’t start delivering tangible results, it wouldn’t be at all surprising to see Labour start bleeding votes in all directions.
Some people have responded to that chart with "That can’t be right", or "We can’t be worse than America".
I’m afraid the chart is right. 15 years ago the UK’s record on homelessness *was* not too dissimilar to other developed countries, but things have rapidly deteriorated.
There has long been a gap between people’s views of crime locally (not a big issue) vs nationally (it’s terrible out there!), but there are signs this is now happening to economic perceptions too.
My finances? Going okay. The economy? Awful.
What’s going on?
My column this week asks whether the media (both mainstream and social) and its incentives to maximise engagement could be playing a key role ft.com/content/8cd76c…
With crime, it’s widely accepted that the main reason for this decoupling is media coverage.
People’s sense of crime levels is based mainly on what they see on TV and read in newspapers, and much less on what they or the people they know actually experience.
NEW: my column this week is about the coming vibe shift, from Boomers vs Millennials to huge wealth inequality *between* Millennials.
Current discourse centres on how the average Millennial is worse-off than the average Boomer was, but the richest millennials are loaded 💸🚀
That data was for the UK, but it’s a similar story in the US. The gap between the richest and poorest Millennials is far wider than it was for Boomers. More debt at the bottom, and much more wealth at the top.
In both countries, inequality is overwhelmingly *within* generations, not between them.
And how have the richest millennials got so rich?
Mainly this: enormous wealth transfers from their parents, typically to help with buying their first home.
In the UK, among those who get parental help, the top 10% got *£170,000* towards their house (the average Millennial got zero).