Looking at the other two averages, I’m pretty uncomfortable with both of them, and all of a sudden I’m glad to be doing my own average, just to have an independent assessment of where the race is. Because both of them seem to be kinda lopsided in the polls they rely on right now, just in opposite directions.
I’ll start with The Economist’s average, which has Warren up by 2. If you click through, you’ll see that 5 of the most recent 10 polls in their average are YouGov polls. And I’m hardly the only observer who’s noticed that YouGov has a pretty visible pro-Warren house effect, compared to other pollsters. (The Economist is the sponsor of YouGov’s weekly poll, btw.)
They also “include only surveys from pollsters who conduct their interviews over the phone with a live interviewer—rather than with automated machine recording—or that use rigorous and well-documented online methods” which is a reasonable standard, which apparently neither Morning Consult nor HarrisX live up to, since they’re not included at all in The Economist’s average. (I’m rather suspicious of both outfits for various reasons, but I don’t feel I’d be on solid ground to exclude them due to those suspicions.) Since they both have heavy pro-Biden leans (and MC has a big pro-Sanders lean too), that makes a difference.
Which brings me to RCP, which suddenly has Biden up by 6. They weight all their polls equally, and two of their five polls right now are HarrisX and Morning Consult, with their heavy pro-Biden leans (MC hasn’t had Biden below 31% since April, and HarrisX currently has Biden at 35% and Warren at 17%, which is not unusual for them) and low 538 ratings.
So these two iffy polls count 40% of their average right now; due to downweighting for being B- and C-rated polls per 538, they count 18% of my average this week, and they’re only that high because of how few recent polls there have been.
As far as my own average is concerned, I think Biden’s 28.7% is a fluke of the scarce polls, and I’d bet his average will drop a point or two once the post-debate polling is published. Warren’s probably about right, midway between the other two.
But I think I have it right, and both of them are wrong, on Bernie’s drop, and how far Harris is down now. My sense from the polls is that Bernie’s heart attack is costing him some of his less intense fans. And I don’t know what the deal is with Harris, but she’s sinking. I’m surprised (not so much now specifically, but rather on an ongoing basis), because I expected her to be more formidable a candidate than she’s been.