In the days leading up to November 3rd, I have been pouring through the polling from the same period in 2016. Yes, I know that many of us are sick of talking about the 2016 polling – how it was wrong, the media bias, and everything else. But when comparing the Trump/Clinton voting to the polling today, I noticed many vital differences that all of us should note.
So, let’s rewind. On Election Day in 2016, former Secretary of State Hillary Clinton was ahead in the polls. It was the prevailing theme of all the media coverage heading into that Tuesday.
Just take a look at the chart I made using data compiled by RealClearPolitics. In the swing states, Clinton led on Election Day by 0.56%. She wound up losing the swing states by 0.95% – a shift of 1.5 percentage points.
That was huge, but it wasn’t what I noticed.
What I noticed were the polling averages missed on four swing states. The polling average had Clinton winning Michigan by 3.4 points, but she lost by 0.3, which was a difference of 3.7. She was up in Pennsylvania by 1.9 and lost by 0.7 – a 2.6-point change. The worst was Wisconsin: Clinton led by 6.5% and lost by 0.7% for a margin of 7.2%.
However, the averages didn’t only miss on Clinton. In Nevada, Trump led by 0.8% only to lose by 2.4% – a net of 3.2%. The impact on the Electoral College swung the election, transferring 46 votes from Clinton to Trump while doing vice versa with 6 for a net difference of 40 votes.
But that wasn’t what I noticed either.
What I did notice should keep President Donald Trump up at night in fear. I shared three charts above for a reason. I want you to look at the third chart (pictured right). It shows the difference between where the candidate was in the polling average versus the actual percentage of the vote received in the election:
In all but two of the swing states, Clinton received a higher percentage of the vote than her polling average by 1.95%. She was polling at 48.0% in Minnesota but won only 46.4% of the vote – underperforming by a margin of 1.6 points. The same occurred in Wisconsin, where she polled at 46.8% but only received 46.5% for a decrease of 0.3 points. It also cuts both ways. In Nevada and New Mexico, Trump had 0.3% less of the vote than the polling averages held. In only four of the 34 swing state opportunities, a candidate got a lower percentage of the actual vote than the polling average indicated.
So, the question is: Where are the averages today?
If an estimated three percent of the vote goes third party in each swing state, a candidate polling above 49% is almost a shoo-in to win the state. This metric means the three key Rust Belt states – Pennsylvania, Michigan, and Wisconsin – are out of reach for Trump. Former Vice-President Joe Biden is receiving a polling average above that benchmark: 50.5% in Michigan, 49.6% in Pennsylvania, and 50.3% in Wisconsin.
This shifts the election but keeps states that Clinton won like Minnesota and Nevada in play. However, this also leaves other Republican-held states the Trump won in danger, with Trump polling at 48% in Texas, 47.2% in Georgia, and 48.0% in Florida. If the trend holds, Trump has a better chance of winning Nevada and Minnesota than he does the Rust Belt stated mentioned earlier that’ handed him the Presidency.
With all of that in mind, it won’t be Pennsylvania or Florida deciding the election, but North Carolina. Right now, the RCP average has Trump in the lead by a slim 0.6 of a point. If the Tar Heel State goes to Biden, it looks like the election may be over on November 3rd. Unlike Pennsylvania and other states that may take days to tabulate their mail-in votes, we should know the results of North Carolina reasonably early. If Trump loses North Carolina, Pennsylvania will not matter.
This state should be keeping the President up at night.