Just looking at the polling average on 538 today, which currently has Biden up 10.4 in the national average, which is not only the widest margin Biden has ever held, but also wider than any margin Clinton held.
However, the probability for Clinton was actually higher at this point. One of the reasons is that their new metric for the unpredictability of the news cycle (measure by full-column headlines in the NYT) finds this a more tumultuous time than previous election cycles.
However there's something that is, to me, problematic with this approach: the polling average has been incredibly stable despite the number of major news stories. In 2016, every major news story seemed to result in big swings, sometimes for Trump, sometimes for Clinton. This year, the polling average seems to be largely impervious to the news cycles, until last week when several massive negative stories for Trump resulted in what is so far a modest gain for Biden (but may continue to grow). And that's still not the sort of swings we were seeing in 2016. It's hard to imagine any news story that results in a huge pro-Trump swing at this point.
Another thing that's keeping Biden's odds down a bit are that the state-level races suggest a closer race than the national polls: Silver was saying on their podcast that the state level races suggest that the national race is actually somewhere closer to 8 rather than 10. Can't argue with that, since state-level is more critical to the actual election odds.
Iowa is continuing to look like a really solid shot at a senate flip for Democrats. Quinnipiac and Yougov both found Greenfield at +5 and +4, and she's led every recent poll (except for a tie from Hart Research, which was still a +5 move toward Greenfield since their previous). I expect Iowa is one of those states where down-ballot Republicans are freaking out a bit at the effect of the president's drag on their own popularity.
|