Tuesday, November 14, 2017

What did Donald Trump do today?

He suggested that perhaps "some people" thought his actual approval rating was 20 points higher than it is.

Trump tweeted this early Tuesday morning at the close of his Asia trip: 
Rasmussen is, in fact, not especially highly regarded for its tracking poll. Fivethirtyeight.com gives it a C-plus, behind outfits like Gallup (B-minus, which has Trump at 38%), Ipsos/Reuters (A-minus, 37%), and Quinnipiac (A-minus, 35%). The fivethirtyeight.com average of tracking polls has Trump at 38.1% approval, unprecedentedly low for a president this early in his first term.

By "the last time" Trump presumably means the election polls, which are meant to predict how people will vote rather than whether they approve of an incumbent president. Rasmussen's actual election day forecast was fairly accurate: it had Hillary Clinton winning the popular vote by 2%, whereas the average prediction among all national pollsters had Clinton winning by 3%. She won by 2.1%. Of course, Trump also claims to believe that he won the popular vote if three to five million "illegal votes" he somehow knows were cast exclusively for Clinton are excluded.

Trump didn't specify who the "some people" were who thought his approval rating was above water, but it's possible to guess.

Why is this a problem?

  • Politicians fudge numbers sometimes, but a president who cherry-picks a poll number that still has more than half of Americans disapproving of his performance is in a bad place.
  • Numbers are not #fakenews just because they make a president look bad.
  • People who voted for Trump and then lost confidence in him might be surprised to learn they don't exist.