Quantitative Editor, @FiveThirtyEight | PhD, @LSEGovernment | 
in 
, via 
| she/her
You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Late deciders swung Super Tuesday for Biden. What do we know about them? @geoffreyvs, @baseballot and I dug into the exit polls:
1. White voters swung the most towards him -- he got a majority of the late white vote.
https://fivethirtyeight.com/features/what-we-know-about-the-voters-who-swung-super-tuesday-for-biden/ …pic.twitter.com/adQ2dXvONe
2. Late deciders were more likely to be moderates, maybe because Buttigieg/Klobuchar supporters *had* to change their vote late (MN had 55% late deciders!). But Biden picked up late support across the ideological spectrum:pic.twitter.com/XAw1r7QmAT
3. The Northeast had lot of late deciders (49%), which helped Biden win states where he wasn't the favorite, like ME and MA. (He even seems to have won the late vote in Vermont!) Fewer late deciders in the South & CA, though Biden won the ones that did decide late.pic.twitter.com/szTNgY0P4G
Plus a bonus table showing the percentage of different groups that decided in the last few days 
pic.twitter.com/jQeHc8wyXN
Do we have any data parsing levels of late deciders (e.g., late deciders who were almost certain of their vote versus late deciders who really had no idea for whom they would vote)?
In the exits, not really. I'd love to know the Klobuchar/Buttigieg/Bloomberg switchers vs truly undecideds coming in.
This year, I've learned to model very late deciders as a separate group, rather than dividing them between candidates if I want to understand the potential variability of a specific day. In reality, one gets a huge share: IA Pete, NH Amy, SC/ST Joe.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
