Friday, July 3, 2020

Is Stan Greenberg Right to Say We Should Believe the Polls This Time?

I largely agree with Greenberg's analysis in this Atlantic piece. I thought it would be interesting to go through the piece and annotate his argument a bit where I have something to add.
"[T]his moment is very different. To start, during the summer and fall of 2016, Clinton never had the kind of national poll lead that Biden now has. She led by an average of four points four months before the election and the same four points just before Election Day. This year, after Biden effectively clinched the nomination, he moved into an average six-point lead over Trump, which has grown to nearly 10 points after the death of George Floyd and the weeks of protests that have followed. The lingering apprehension among Democrats fails to recognize just how much the political landscape has changed since 2016. We are looking at different polls, a different America, and different campaigns with different leaders."
This is all correct. It is now incorrect to say Clinton followed the same trajectory and still lost.
"The Clinton campaign’s worst blunder came in September 2016, when the candidate described “half of Trump’s supporters” as “deplorables” and walked right into the white working-class revolt against elites. Her primary campaign against Bernie Sanders had exposed a lack of enthusiasm for her in white working-class suburbs that Barack Obama had won. Her campaign hoped to make up for the lost votes with landslide wins among women, voters of color, and voters in big cities. White working-class voters noticed the lack of respect, and Trump ran up startling margins with them: He won these men by 48 points and women by 27, according to exit polls."
I agree with this; anyone who doesn't think Clinton committed political malpractice with how she handled white working class voters in 2016 has rocks in their head. However, note that the exit poll figures quoted by Greenberg are almost certainly too high. Better data however still show an immense swing away from the Democrats among this demographic.
"And the white working-class shift toward Trump is the biggest reason the national polls overestimated Clinton’s margin by two points and the state polls by much more. Mostly using exit polls from prior elections as their guide, pollsters—including me—had overestimated the number of four-year college graduates in the electorate. Getting that wrong mattered a lot in an election where the white working class was in revolt. Crucially, many pollsters, including me, have adjusted our assumptions about the makeup of the November 2020 electorate."
Correct. Though of course I had been pointing this out for 20 years or so prior to 2016 and no one seemed to think t was a problem worth attending to.
"So one reason to trust my polls more now than in 2016 is this change: Four years ago, those without a four-year degree made up 48 percent of my survey respondents; today they account for 60 percent. Whites without a college degree were 33 percent of my surveys; today they are 43 percent. That is a huge change—an elixir against being deceived again. The pain of Trump’s victory and disastrous presidency has concentrated the minds of campaign staff and the polling profession in ways that give me confidence that Biden’s lead in the polls is real."
Better late than never. Let's hope other pollsters have followed the same trajectory.
"Much more devastating to Trump’s prospects is waning support from women who form a majority of the white working class. Without strong support from these voters, Trump cannot win. Right now, Biden is losing them by only seven points in my same battleground poll."
This is correct and is a point I have stressed many times. It really is true that Trump cannot win with his current support levels among this demographic.
"Recently, Trump’s average approval rating has slipped a bit to about 41 percent, while his disapproval rating has jumped to about 56 percent. That looks a lot like the 14-point margin for Biden over Trump in the most recent New York Times poll."
This is consistent with the 538 running poll average on approval. The importance of such low approval cannot be overemphasized--approval numbers like this are a death sentence.
"In the next four months, many things could put Biden’s current lead at risk. On occasion, between now and November, Biden will garble his words in an interview or make some public statement that many people will struggle to understand. He will surely sound out of touch or offend one group or another. Younger voters and Sanders primary voters do not appear to be rapturously excited about Biden. Calls for defunding the police reveal genuine fractures in the Democratic Party."
I am less worried about the youth vote at this point. The defunding the police movement could be more of a genuine problem but Biden seems to be fairly deft in how he's handling it.
"Even before the pandemic, the American political landscape had changed dramatically since Trump’s election, and not in ways that favor the incumbent. Biden’s big poll lead should not make Democrats complacent, but neither should members of my party shake their heads and think, Here it comes again. Rather, the current polls should persuade Democrats to work for the greatest possible rejection of a widely distrusted U.S. president and the political party that enables him."
What he said.
THEATLANTIC.COM
These aren’t Hillary Clinton’s numbers. Biden has a wide lead because the landscape has changed.

No comments:

Post a Comment