That defensive hand-wringing aside, it’s still the case that something happened to Warren. In early October, she was surging in the polls, rising in RealClearPolitics’ average of polls to match (and even briefly surpass) former vice president Joe Biden. At the time, we noted a recent history of people surging into ties — and then fading away. (Ben Carson in the 2016 Republican contest, for example.) That’s exactly what happened to Warren: a peak and then a fade.
Polling from Quinnipiac University released Tuesday allows us to look a little more closely at how Warren’s fortunes have fared. In that poll, Warren slipped from a clear first at the end of October to a tie for third with Sen. Bernie Sanders (I-Vt.). Biden regained his lead, but South Bend, Ind., Mayor Pete Buttigieg climbed into second. (Sen. Kamala D. Harris of California, who surged after the first Democratic debate, continues to poll in the single digits.)
We learn a bit more when we look at the shifts by demographic. Here is age, ideology, income and race. Take a look.
Two things stand out pretty clearly in many of those charts: Warren’s drop and Buttigieg’s increase. Among “very liberal” respondents, both Biden and Buttigieg appear to have gained at Warren’s expense. Her support among moderates, once respectable, collapsed entirely. It’s worth noting that the shift against Warren came after significant focus on her approach to Medicare-for-all — which Quinnipiac suggests has relatively low support among Democratic moderates.
Buttigieg didn’t gain any ground among black Democrats (something his campaign is no doubt concerned about), but he did see a marked increase among whites, primarily at Warren’s expense. He also gained from Warren’s plunge among voters over 65, while Warren’s drop among younger voters was to Biden’s benefit.
Over and over, the pattern is the same: Warren goes down and Buttigieg goes up. That’s reflected more subtly in the national polling average, in which Buttigieg’s gains are more subtle relative to Warren. Some of the volatility in the Quinnipiac polling is, of course, a function of margins of error. The margin of sampling error is nearly five points among the Democrats, meaning that shifts of less than that aren’t statistically significant.
The most important caveat here is the one we began with: The field is awfully fluid. There’s plenty of time for Warren to have a second surge — as Buttigieg did. Plenty of time for her team to print out this article and use it as an example of the shortsighted punditry of the chattering class.