Tuesday, November 10, 2020

Were Republican-Leaning Voters Lying to Pollsters?

There is something in polling called the generic ballot.  Voters are asked if they support a Republican or Democrat for Congress.  No name is attached to the poll question.  The answer pollsters get has for decades closely reflected the national partisan vote for Congress.  Here is the final generic poll average v. election result for the last 6 presidential election years:

Year              Final Generic Ballot Margin          House Popular Vote Margin      Error
1996                     D+3                                                Even                                     3
2000                     D+1                                                Even                                     1
2004                     Even                                               R+3                                      3
2008                     D+9                                                D+11                                    2
2012                     Even                                               D+1                                      2
2016                     D+1                                                R+1                                      2

Now let's look at 2020:

2020                    D+7.3                                              D+1                                      6.3          

I'm not sure if there's ever been as big a difference between the generic ballot and the actual vote.

It is often argued that the "shy Trump" voter is why the polls were off in 2016.  The premise is faulty though because the 2016 polls weren't off.  The national polls came within 1 point of predicting the final vote margin.  The results in all the battleground states were well within the margin of error of the polling

It is just that pollsters had the wrong candidate ahead in three states - Pennsylvania, Michigan and Wisconsin.   But while Trump overperformed the polls in those states (though, again, within the margin of error), he underperformed the polls in several other states. Several studies show the "shy Trump" voter was not a real thing.

Nonetheless, pollsters made adjustments following 2016 campaign, primarily to recalibrate the under-sampling of people without college degrees.  (In 2016, education level was shown to have a big influence on support for candidates.)  The polls during the 2018 midterms were spectacularly good.  But during the 2020 election, they were just as spectacularly bad, truly awful.

"Shy Trump" voter?  The trouble with that theory is that while Trump's support was understated in the polls, support for Republican Senate and congressional candidates was even more understated.  As demonstrated with the generic ballot table above, Republicans were undercounted big time.

Thus, the methodology employed was such that Republicans were under sampled in the polls, or Republican-leaning voters contacted by pollsters were simply lying to them about their preferences.  The latter might well not be that these GOP voters are "shy," i.e. ashamed to announce their preference, but because they harbor a deep disdain for pollsters (a dislike promoted by Trump himself) and simply chose to lie to them about who they would support. 

OOP's short takes:
  • The two Georgia Republican Senators, Kelly Loeffler and David Perdue, are trying to get the Georgia Secretary of State Brad Raffensperger, to resign.   Why?  It appears they want Raffensperger ousted because he did not put his finger on the scale in Georgia so Trump would win the state.  Maybe Loeffler and Perdue should worry more about their January 5th run-off against, respectively, Democrats Raphael Warnock and Jon Ossoff?
  • Speaking of that run-off, the thought is the election will be low turnout and that will favor Republicans.  While I don't think any of the four candidates are strong candidates, I'm not convinced that Republicans will swamp Democrats in turnout in the Peach State. Trump won't be on the ballot and I'm not convinced that GOP leaning voters will go to the polls in huge numbers to keep Mitch McConnell in power.  Likewise, I'm not convinced that the need to have a check on Joe Biden by electing a Republican Senate is a concept that more than a few voters consider when casting a ballot.  Not sure who I'm going to predict, but I expect the race(s) will be closer than the easy GOP win(s) pundits currently are predicting.
  • Donald Trump, Jr. and his girlfriend, Kimberly Guilfoyle, are apparently working behind the scenes to take control over the Republican National Committee.  They want to oust current RNC chair Ronna McDaniel, who was seen as ineffective in getting Trump re-elected.  McDaniel sold her soul to Trump and got nothing in return for her loyalty. Who could have seen that happening?  
  • One name I heard to replace McDaniel is Corey Lewandowski.  I kid you not.  Corey is not competent to run a 7/11.  So let's make him the operational head of the GOP!
  • On a recent political podcast, a panel of seasoned political experts - mostly Republican types who had worked in numerous campaigns - were asked what percentage of the GOP electorate would believe Trump's narrative that the election was stolen through fraud.  The consensus was that after a month of peddling the "stolen election" line, 20% to 30% of Republicans would come to believe it was true.  LOL.  I knew it was not going to take a month and I expected the number to be about 70%.  It did not take 24 hours for me to be proven right.
  • I've heard some analysts say that if Trump doesn't bow out gracefully, it could cost him to lose support.  Really?  Do I have to list all the things Trump has done over the last four years for which he didn't lose support?  Pretty sure Trump can overcome being labeled, correctly, as a sore loser.  Sadly, the Trump cult lives on.

No comments: