The Case for Believing National Polls

by John English

I’ve noticed whenever I post or tweet about polls on the 2020 presidential election, someone will chime in with something “The polls were wrong in 2016!” They actually weren’t. The national polls were correct. It was some of the state polls that threw everybody off.

In RealClearPolitics’ final average, they had Clinton at 46.8% and Trump at 43.6%, a difference of 3.2% with polls that averaged a margin of error of +/- 2.8%. This also meant that 9.6% of the electorate was voting third-party or hadn’t decided in the few days before the election. The final results had Clinton at 48.2% and Trump at 46.1%, a difference of 2.1%. So the national polls were off by 1.1% in an MOE of 2.8%. Which means they were correct. Clinton gained 1.4% from the undecideds while Trump gained 2.5%, while 5.7% opted for third-party/write-ins, etc. (Compare that to the 1.7% from 2012.)

So how did Trump still win the presidency? Let’s look at some battleground state polling.

Wisconsin had Clinton at +6.5%. Trump won by 0.7%.
Michigan had Clinton at +3.4%. Trump won by 0.3%.

Within the margin of error but breaking Trump’s way:

Pennsylvania had Trump down at -1.9%; he won by 0.7%.
North Carolina had Trump up at +1%; he won by 3.7%.

The close ones didn’t all swing Trump’s way. He was up 0.8% in Nevada, lost to Clinton by 2.4%, for example.

But on its final night, RCP had Clinton with 202 electoral votes, Trump with 164, and 171 were toss-ups. They had Wisconsin’s 10 in Clinton’s camp, but those went to Trump. Everyone knows the rest. You need 270 electoral votes to win, and Trump won 306-232. (Final count 304-227, with 7 faithless electors.)

So moving forward, if we’re comparing Biden vs. Trump to Clinton vs. Trump, things are looking very encouraging for Biden.

Clinton / Trump
Apr 30 2016 – 47.4 / 40.1 — 7.3%
May 30 2016 – 43.8 / 42.8 — 1%
Jun 30 2016 – 44.6 / 39.8 — 4.8%

Biden / Trump
Apr 30 2020 – 47.4 / 42.1 — 5.3%
May 30 2020 – 48.4 / 42.5 — 5.9%
Jun 30 2020 – 49.5 / 40.3 — 9.2%

2016 was a year in flux. No incumbent. The two least popular candidates in US presidential election history. Biden doesn’t have many of the negatives Clinton had. Trump’s base is titanium solid, but he hasn’t done anything to expand it. Does that make a Biden win guaranteed? Never. And in 2016, at one point in July, Trump led Clinton by 1.1%, and the week before the election, Clinton was only up by 1.2%. I’m just not seeing the clamoring for a third party or “Anyone else!” this year like we did in 2016. Trump isn’t going to lose many of the votes he won in 2016, but I’m seeing that Biden is going to get many more votes than Clinton did. And Clinton won 2.9 million more votes than Trump, just not enough of them were in the right states. This is a year where the referendum is on Trump, and his job approval numbers are well below where his predecessors were at this point in their respective presidencies. His 538 approval average is at 40.5% after 1258 days on the job. Obama was at 47.8%, George W. Bush at 45.3%, Bill Clinton at 53.8%. Now George HW Bush was at 35.8%, and he only won 37.4% of the popular vote in his 1992 re-election bid.

Liked it? Take a second to support Utah.Politico.Hub on Patreon!

Related posts