There is actually some interesting analysis on Warren's last election:
https://fivethirtyeight.com/features/4-things-elizabeth-warrens-last-election-can-tell-us-about-2020/
Compared to Clinton, she did worse with rich voters and better with poor ones. Not much diff on minorities.
I think the main risk with Warren is if she scares off Romney-Clinton type voters too much. I think she has enough appeal to WWC voters for a Democrat. Not that she has a lot of it, but Democrats don't need a lot of it to win a general election anymore.
I generally agree with this article.
The point 1 is true, Warren would undoubtedly do far worse among +150k $ income voters than Clinton, I could see Darien voting for Trump if Warren is the D candidate.
Point 3 is also true, Warren would do very well in university counties where she would fired up progressive voters, Dane county would give her more than 75% in my humble opinion.
Point 4 is also mostly true, she would do well among minorities voters but she would likely underperform Obama among them.
But I really disagree with Point 2, Warren would be a horrible candidate for Obama/Trump voters. The author of the article is explaining that Warren would do well among these voters because she overperformed Clinton in Western Massachusetts, but western Massachusetts has more in common with Vermont than with Northeastern PA, the reason why Clinton did badly in western MA is that this area has a lot of die hard Berniecrats who voted third parties ; but these voters are not really comparable to Obama/Trump voters from Macomb county. It’s not a good comparison.
How can you say this though? Do you have compelling evidence that she'd do poorly among Obama/Trump voters?
