Trump’s bad polls? Before you panic, dig into the details
After many declarations that President Trump was vanquished on the basis of an NBC/Wall Street Journal poll showing him 14 points behind, it turns out that the poll was totally bogus.
Four years ago, in October 2016, an AP-GfK poll also showed that Trump had fallen 14 points behind Hillary Clinton right after their final debate — 12 days before the election. We all know how that turned out.
The first clue about the new poll’s problems should have been on the first line: the poll was conducted by “Hart (!) Research Associates/Public Opinion Strategies.” Peter Hart is a Democratic operative, and, if the poll was supposed to be “bipartisan,” it certainly doesn’t show.
It polled only 800 voters, a relatively small national sample. Its sample was “registered voters” rather than “likely voters,” which is thought to give Biden an artificial 2-point advantage. Only 2% of the sample failed to complete high school.
But the real kicker is way down on page 12, where it is revealed that self-identified Democrats in the sample outnumber Republicans by 9% (45-36%) and “Strong Democrats” outnumber “Strong Republicans” by 6% (28-22%).
In other words, what the poll showed is that Biden voters favor Biden. If you subtract the 9% party differential and the 2% “registered voter” bump, Trump trails by only 3%, which probably means he wins the Electoral College (and a second term).
But even beyond this, Trump brings advantages to the table that pollsters generally fail to measure:
- THE ENTHUSIASM GAP: In two USA Today surveys, highly motivated Trump voters outnumbered highly motivated Biden voters, 50-29% and 43-32%.
- THE “GROUND GAME” GAP: The Trump campaign has prepared for a door-to-door effort for four years. One month before the election, the Biden strategists have suddenly discovered that their wholly computer- and telephone-based effort is coming up short.
- THE IMPACT OF CLOSED CAMPUSES: These are huge sources of Democratic support in virtually every battleground state. In Pennsylvania, for instance, registration by 18- to 24-year-olds is 3% down.
WHAT TRUMP MUST DO TO WIN
1. PREVAIL IN 22 “RED” NON-BATTLEGROUND STATES, PLUS FLORIDA, NORTH CAROLINA, OHIO, AND IOWA
Most observers expect Trump to win these states. If he doesn’t, it is hard to see how he can prevail in the Electoral College.
FLORIDA: First of all, Trump is 17 points ahead of where he was vis-à-vis Clinton in vote-rich Miami-Dade. Nationally, Trump’s support among Hispanics is thought to be as high as 37%. In Florida, with Hispanics, he is 23 points ahead of where he was with Clinton. Even ABC had Trump four points up in September in the Sunshine State, and left-leaning USA Today had a tie on October 6.
NORTH CAROLINA: This has been a statistical tie in the liberal polls for months. But the Democrat turnout model — busing black churchgoers to the polls — is collapsing. And the Democratic Senate candidate — Cal Cunningham — is in the middle of a sexting scandal.
OHIO and IOWA: Although isolated polls have shown these states in contention, most assume they will repeat their 2016 margins for Trump. Within the past few days, Trafalgar gave Trump a 4-point lead in Ohio.
2. PICK UP 23 ADDITIONAL VOTES ELSEWHERE
PENNSYLVANIA (20): Registration among 18- to 24-year-olds (who lean Democratic) is down 3%. Fracking is a major issue. Bucks County, in the Philadelphia suburbs, may go “red.”
MICHIGAN (16): The Macomb County pollster who predicted that Michigan would go for Trump in 2016 is predicting the same for 2020.
ARIZONA (11): Republican senator Martha McSally has climbed from a 17-point deficit against anti-gun activist Mark Kelly to within the margin of error, according to some polls. Kelly has a massive amount of anti-gun money but is entangled in a scandal involving Communist Chinese infiltration of his space firm.
- WISCONSIN (10)
- MINNESOTA (10)
- NEVADA (6)
- NEW HAMPSHIRE (4)
- MAINE 2 (1)
- NEBRASKA 2 (1)
THE BOTTOM LINE
No one can predict what the next month will bring. With two presidential debates, the Barrett nomination, and the uncertain situation involving COVID, there are probably more “unknowns” than there are “knowns.”
But those suggesting that the election is over might remember how they were embarrassed in 2016 and consider the real possibility that this may happen again.