Hard to overstate just how rapid the changes in US party coalitions have been.
The Democrats are now the party of high education *and* high income voters. Just 12 years ago, the inverse was true.
On these terms, the Trump coalition was closer to Obama 2008 than the Harris one!
Firstly, let's look at how views change with age.
Men's views are U-shaped: the youngest and oldest men are the most positive, dipping in middle age.
Meanwhile women's views vary little by age group. There is only a slight upwards slope in positivty from 18 to 75.
/ 1
Stats-based forecast for tonight’s Eurovision final (average overall ranks shown from 250,000 simulations)
My model got 20/20 qualifiers correct during the semi-finals — let’s see how it fares tonight!
Genuine question: Why can we not, at the very least, produce half of ballot papers in reverse alphabetical order?
Seems like changing one’s surname to Aarons by deed poll might actually be a better use of election resources than actually campaigning!
Now the dust has settled from the local elections, it's time to talk about vote-rigging. In particular, the way in which a sinister force has been manipulating British democracy in council after council, year after year. That force? The goddamn alphabet. (1/?)
Today we've launched a new Workforce AI Tracker in partnership with the @FT, including our first white paper for the project.
🧵 Subsequent reports will focus on particular topics in depth, but here’s a quick ‘state of play’ of where we’re at...
So what comes next? We outline three scenarios in the white paper, or the “ABCs”: Augmentation, Bifurcation, or Crash. The data we have collected indicates the middle option is the most likely. Find out more by reading the full white paper below.
Excited to reveal the first @focaldataHQ House of Representatives MRP poll of 2026.
🔵 Democrats: 228 seats (+13)
🔴 Republicans: 207 seats (-13)
Fieldwork conducted up to Monday this week, and the model accounts for last night's redistricting referendum in Virginia.