Quick thread response to this note by Matt. He's right that age divide is very large currently. And 2017 and 2019 had largest age gradient since 1960s. But... these elections don't look so different, when you adjust for income, home ownership and education 1/n
https://twitter.com/GoodwinMJ/status/1493728848248115204
@jrgingrich and I have a chapter in the forthcoming IFS Deaton Review where we look at this. Here we show the relationship between age and voting Conservative between 1964 & 2019. Left is bivariate (just age), right is adjusting for income, education, homeownership and gender 2/n 
The likely culprit is, you guessed it, education, which once drove people to the right but in the last couple of elections in particular has pushed people to the left. note though, this is a long run story. 3/n
What I think is surprising though, is that WITHIN age groups, education has similar pattern. In 2019, over 50s with degrees / postgrad behaved rather like under 50s. It's just there aren't so many over 50s who went to university. But as we move into Gen Xers that changes 4/n
So will the age gradient survive or is it an artefact of educational differences of the generations who came of age before and after the great expansions of the early 1990s? I guess we'll see n/n
Should quickly add, this is using the British Election Study and working to reconcile the education, income etc indicators across surveys since the 1960s. These are pretty sparse models that don't include region, marital status, etc. But descriptively useful I think. n+1/n
• • •
Missing some Tweet in this thread? You can try to
force a refresh
