In a State: Assessing WH2020

In a State: Assessing WH2020

Why Trump approval ratings are a better pointer than state polls

I do have some fairly strong views on the election. In particular, I would advise you to cast aside state polling on Biden vs Trump and look instead at Trump’s favourability on a state-by-state basis.

Why? Because US pollsters use very strict turnout filters. Didn’t vote in Election 2016… well, they largely assume that you won’t vote in 2020. This matters, especially as President Trump didn’t win 2016, Hillary lost it.

Let me explain. Mitt Romney. Nice chap. He once sat on the table next to me at a restaurant on the slopes in Deer Valley. Generally considered to be a weak candidate.

But here’s the thing. Mr Romney got a greater share of the US electorate to vote for him than President Trump. In 2012 there were 235,248,000 eligible voters in the US. Mr Romney got 25.9% of them. In 2016, there were 250,056,000 eligible voters and President Trump got 25.1% of them,

In some states, despite populations that had grown since 2012, President Trump got fewer votes than Mitt Romney.

Here’s one – Wisconsin.

Traditionally Democratic. Nice place. Big on cheese, I believe.

In 2012, Mitt got 1,407,966 votes there and lost by quite a large margin. Four years later, President Trump got 1,405,284 votes and won the state by a whisker. The issue for Ms Clinton is that she was a voter repellant. Democrats simply didn’t turn out for her.

In 2018, two years after President Trump scooped Wisconsin, there was a Senate race in the state. The Democrats, fools that they are, put up a lesbian candidate. As everyone on here knows, neither women nor homosexuals go down well in the rust belt, and she was roundly defeated.

Oh wait,

No she wasn’t. She won.

And she won with more votes than President Trump got in 2016. I want you to think about that for a second. A Senatorial candidate, in a midterms year, got more votes than the winning Presidential candidate got two years earlier. If anyone can find another example of that, I will send along a 20 pound Amazon voucher. I expect I’ll keep my money.

The point of this meandering is a simple one. In a whole bunch of states, it was not President Trump who brought massive numbers of Republicans to the ballot box, it was Ms Clinton who scared Democratic ones away.

Which brings us back to why you shouldn’t trust state opinion polls too much. A voter who chose Obama in 2008 and 2012, and who stayed home in 2016, doesn’t get counted. And there are a lot of voters like that in states like Wisconsin and Michigan.

To see who’s going to win, look instead at President Trump’s approval ratings. These are surveys of all voters, whether they trekked to the ballot box in 2016 or not. In Wisconsin, the President is on -10. That’s actually worse than the -9 he was registering on the eve of the Republicans getting smacked there in 2018.

The best place to see month-by-month, state-by-state approval ratings is Morning Consult: https://morningconsult.com/tracking-trump-2/. They show Trump in trouble in Michigan and Wisconsin (both -10), but holding on in Arizona (+1). Iowa looks dicey for the President (-5), while Minnesota and Virginia (both -5) are both sufficiently close that they might go from Democrat to Republican.

If anyone wants a wager, I would reckon that the approval numbers will prove significantly more predictive of individual state results than opinion polls – does anyone fancy being on the other side?

Robert Smithson

Comments are closed.