,
5 tweets,
2 min read
Read on Twitter
A huge amount of political reporting, and maneuvering has relied on the assumption that Trump did very well in the Rust Belt in 2016. But if you look at the absolute votes shares over time, you can see it's JUST NOT TRUE. In most states, Bush in 2004 did better than Trump did. 
The problem, as the graphs should make perfectly clear, is that the Democratic vote share absolutely collapsed in 2016 - probably a combination of anti-incumbent voting and Clinton's personal unpopularity.
So why the persistent myth that Trump is popular in the Rust Belt? Two reasons: first, it's where his RELATIVE strength is - in most states, he badly underperformed Romney and Bush, rather than performing on par with them. But that's different than saying he's popular there.
Second, the geographic composition of the vote shifted dramatically, with cities and suburbs moving Dem and white exurbs and small white towns moving GOP. On net, the shift benefited Democrats - but the Rust Belt has disproportionately many of the latter type of community.
I think it's also fair to say that shifts in small white towns resonate more in the political media than shifts that happen in the (far more common) diverse middle- and working-class suburbs. Bleaksville, Ohio is the "heartland" in a way that Maple Grove, Minnesota isn't.
