,
10 tweets,
2 min read
Read on Twitter
Twitter thread coming based on my modelling of the Leave and Remain 2016 vote by postcode (1/10)
Commentators have struggled to find what the “average” UK political place to visit is on Brexit–the places that voted in line with the country are more likely to strongly Labour or strongly Conservative - creating a confusing narrative (2/10)
Commentators have also struggled to understand what an “average” area is in England and Wales – is it a place that just voted out by 53.3% or something else? (3/10)
The distribution of voters by how “leave” an area is highly irregular and not a normal distribution – there are great more number of voters that live in Extremely remain areas (>90% Remain vote) than voters that live in Extremely leave areas (>90% leave vote) (4/10)
There is a gap between the average mean vote for Leave in England and Wales (53.3%) and the average median vote which is massively higher Leave vote (59.5%) (5/10) 
When commentators are talking about an “average” place which voted in the EU Referendum – they are better to go deeper intro Brexit country, whilst acknowledging that there are pockets of London/Bristol/Cambridge and Oxford that supply almost 0.5m blanket votes for Remain (6/10)
This pattern also explains why you are more likely to get Remain voters who don’t know leave voters than the other way round – it’s not a personal failing, it’s a function of maths and geography (7/10)
But it highlights a point – group-think is more likely to occur in closed metropolitan environments – more Remain voters will go about their daily lives with their views unchallenged than Leaver ones (8/10)
This asymmetric distribution also explains why metropolitan movements can create massive ideological bubbles which fail to penetrate as effectively throughout a country than more rural/populist movements (9/10)
It also explains that the problem with Remain campaign #1 is not that it didn’t engage the core enough – the opposite, its penetration throughout middle/suburban/rural England and Wales was v.bad (10/10)
