Paul Rosenberg has an interesting and somewhat reverential interview on Salon with Rachel Bitecofer, a Christoper Newport University political scientist. Bitecofer, as some of you may recall, was out early with a very accurate and detailed model-based House prediction for 2018.
To which I'd say: good for her but it doesn't mean that everything she now says and predicts is correct. Unfortunately, she now seems a bit more confident in her judgments and predictions that she probably should be.
The negative partisanship stuff she alludes to in the interview is real and she is right to take is seriously. But her methodology doesn't really show some of the things she thinks it shows. This is particularly true of her pronouncements on the overwhelming role of turnout in the 2018 election. The best, most careful and sophisticated analysis of 2016-2018 changes, using individual level vote data, is not consistent with her pronouncements. This is the Catalist analysis which shockingly she does not even mention. That analysis showed that 90 percent of the Democrats' margin advantage in 2018 was due to people who voted for Trump in 2016 but Democratic in 2018. Whatever that is, it's not turnout.
That's the trouble with believing your own press clippings--you tend to overstate your case. I would take her view on the centrality of base mobilization and the alleged efficacy of turning up the "progressive" dial to maximum in 2020 with an entire cellarful of salt.