29 September 2016

What A Change A Debate Makes (UPDATE even in U.S. Senate races!)

Going into Monday's debate with Donald Trump, the 538 election model "now cast" of who would prevail if the election was held on that day based averages of state surveys, gave Trump at 47.9% chance of winning the Presidency on September 26 (with new data on that day still reflecting surveys taken before the debate on the evening of September 25), if the election were held then.

Post-debate polling have dropped his odds to 27.6% and falling, a 20.3 percentage point plunge in three days!

There has really been no news event in the relevant time period other than the debate that could explain this dramatic change in public opinion.

But, the pre-debate and post-debate media coverage does seem to also reflect greater media resolve to call out Trump's failings. Multiple media outlets calling him a liar on the eve of the debate and the moderator in the debate also called him out on this score. USA Today "antiendorsed" Trump in its first Presidential endorsement in 34 years. An Arizona paper endorsed Clinton with the first Democratic Presidential endorsement in more than a century.  The Harvard Republicans club declined to endorse Trump, as did all of the still living former Republican Presidents and all 100 of the CEOs of the Fortune 100.

Some of this is in response to criticism from the left about the failure of the media to clearly explain how far off the reservation Trump's conduct has been. But, a lot of the shift of intelligent conservatives in politics and the media come from pure disgust with Trump as a person.

As a result, until the debate, lots of swing voters in Republican leaning areas have not really had to confront many of Trump's failings as they consider their Presidential preferences, but they are now hearing the unanimous chorus against him as they start to consider the matter more seriously and are seeing for the first time personally his shortcomings.

Then again, maybe tens of millions of people simply read my post-debate rant at this blog and changed their minds based upon that data point.

Clinton doesn't have to hold the lead much longer, either.  Iowa's in person early voting starts today and voters in Colorado will get their ballots in less than three weeks.

UPDATED to reflect falling Trump odds based upon new survey data put into the 538 election model during the course of the last few hours today (which left Trump 3 percentage points worse off than earlier today).

Even more striking is the shift in the odds of Republicans retaining control of the U.S. Senate, which, theoretically, from a high school civics perspective, should have nothing to do with Monday's Presidential debate.

The Republicans had a 52.0% chance of retaining control of the U.S. Senate in the 538 election model peaked on September 25 (i.e. immediately before the debate).  This has dropped to 27.0%, a decline of 25.0 percentage points in the same four day time period based upon how a candidate in the same election who isn't running for U.S. Senate performed in a debate.

Republican candidates for U.S. Senate have actually been hit more heavily by Trump's poor debate performance on Monday night than Trump himself!

This powerfully proved the recent political science research that I mentioned in a post a couple of weeks ago, showing that the biggest factor in how voters vote in down ticket partisan races is their Presidential race preference.

No comments: