Bet a huge part of this is compositional? Surely women with degrees have been getting younger and men without degrees has been getting older? Just because of the expansion in HE.
Even when you're familiar with this trend, the data is absolutely eye-popping https://www.wsj.com/articles/the-yawning-divide-that-explains-american-politics-1540910719 …pic.twitter.com/qK5dx2f8Gj
-
-
-
The gap has opened from about 30 points to 75 in the last 4yrs. Would be generous to call 10pts of that compositional
-
Split of splits is a very deceptive measure. If they both go from 50-50 to 60-40 then the splits go to plus minus twenty and the split of splits goes from zero to forty.
-
Fair. But attributing even a 10pt change in the splits rather than the 2nd derivative is generous over 4yrs. US attendance hasn't been rising as sharply as UK. Especially when the smell test of the particularly sharp post-2016 move is much clearer.
-
The age related effects are huge though. Under 40 split is like +15 democratic, and over sixty is +10 republican. So if you shift the average age of those groups you can easily pick up 30 point change on split of splits.https://news.gallup.com/poll/172439/party-identification-varies-widely-across-age-spectrum.aspx …
-
(Maybe ‘easily’ is overselling it - but compositional effects can definitely be significant here).
-
Sure. But the marginal change in the compositional effect hasn't accelerated much, so you can attribute broad widening since early 90s to that but not the blowout since 2014. All you're saying post then is that the splits have gapped across many groups (age, education, gender)
-
Sure. That’s completely fair. Am not claiming there is no change in these groups voting pattern when properly identified - just seems like that graph is designed to wildly oversell those trends. (where is the chart sheriff?!?)
- 2 more replies
New conversation -
-
-
Likely voters? Registered? What's the N of the two groups? Any cohort age or geographic effects? I'm assuming female group growing ever younger and opposite for male group.
-
Not sure on the N but they're annual averages of monthly polls so it'll be decent I think. Also there'll be cohort effects over the whole series for sure, though I struggle to imagine they're driving the clear gulf over the last 3yrs
End of conversation
New conversation -
-
-
Graph should be rotated 90 degrees
-
But then it wouldn't run left-right! I quite like this way
-
Ok. Time is running down i guess
-
Like in an hourglass
-
These are the days of our lives
End of conversation
New conversation -
-
-
All this shows is universities have become more left wing @david__belle
-
I don't have access to the raw data but I think the speed of the change (both directions) makes it hard to imagine that accounts for all or even most of the change.
-
Why are people so keen to believe that being educated means you've been indoctrinated somehow? It's very conspiracist.
-
what's this got to do with people not being able to think for themselves? so what if your chemistry prof is left leaning?
-
University currently is a leftwing breeding ground. Not sure the point you’re making.
-
what do you mean by 'breeding ground'? I don't remember any left wing breeding going on when I was studying coastal erosion...
-
When were you studying that?
- 2 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.