Yesterday I was talking to a friend of mine who’s a political moderate and who wondered whether I was starting to feel sympathy for Trump.
“No, not at all!” I protested. “I’ve never felt any sympathy for him, and he doesn’t ask for it. If anybody can fight his own battles, it’s Trump.”
And yet, as day after day goes by and I see the constant barrage of negative spin on everything the man says and does, I’ve begun to think that in addition to my weariness with the attacks, I’ll come to feel a certain amount of sympathy. And I can’t possibly be alone in this.
Take this rather subtle version, which I just read. It quotes Charles Krauthammer:
Conservative commentator Charles Krauthammer says President-elect Donald Trump has already lost any goodwill following his White House win.
“The shortest honeymoon on record is officially over,” he wrote in Thursday’s edition of his column.
“Normally, newly elected presidents enjoy a wave of goodwill that allows them to fly high at least through their first 100 days. Donald Trump has not been sworn in, and the honeymoon has already come and gone.”
It’s that word “lost” that’s so clever of the writers. Such a little word, isn’t it? But so meaningful. How can something be “lost” that never existed? How can something be “lost” that never had a chance of existing?
The Hill article then goes on to quote Krauthammer talking about Trump’s flaws, as well as the fact that, as Krauthammer writes, Democrats have been “pushing one line after another to delegitimize the election.” But the impression is that Trump initially had a chance to have a honeymoon, both with the people and the press, and that he has squandered it.
In service of that argument, Krauthammer cites polls:
The result is quantifiable. A Quinnipiac poll from November 17 through 20 ”” the quiet, hope-and-change phase ”” showed a decided bump in Trump’s popularity and in general national optimism. It didn’t last long. In the latest Quinnipiac poll, the numbers have essentially returned to Trump’s (historically dismal) pre-election levels.
The Hill echoes with some more poll evidence:
A Gallup poll released Friday found Trump has the lowest recorded approval rating of any president in the final days before first entering office.
Fifty-one percent of Americans disapprove of Trump before his Jan. 20 inauguration, while only 44 percent approve of him.
That last bit caught my eye, and made me think only 51%? I thought his disapproval was much higher than that.
And indeed it was, for most of the campaign season.
Fortunately, RealClearPolitics still has (as of this writing, anyway), a list of old polls that can be found here. If you scroll down there, you can see that in a poll from late last February (2/26-3/3)taken by Gallup, Trump’s disapproval was 63%. In May (5/1-5/22) Gallup had him at 60% disappoval, and in 6/13-6/19 at 64%. In July it was 59%, in three separate August polls it ranged from 63 to 61, and in September there were three polls with the same exact range. In October, 65, 66, and 62. And then in a poll taken in the days right after the November election, Trump’s disapproval scooted way down (that’s sarcasm, by the way) to 55. His approval ratings during all that time had usually been around 33 or so, climbing to 42 only in that same immediate post-election poll.
The most recent Gallup poll isn’t in that RCP link (at least, I can’t find it). But it’s cited in article after article in the press. I won’t bother to link to them all, but there are plenty of them (here’s a typical one), and they all describe it as the lowest ever for a president at the beginning of his term.
Interestingly, most don’t give a link to the poll, but here it is. You can see that the disapproval/approval is 51/44, and a month ago it was 48/48. Way way down in the results (which are only a summary), the margin of error is stated to be plus or minus 4. So whatever change has occurred in the last month is within the margin of error, although no one points that out that I’ve seen.
More importantly, to my way of thinking—and outside the margin of error—is the large improvement in Trump’s pre- vs. post-election ratings, involving both the lowering of his disapproval by more than 10 points and the raising of his approval by a similar margin. My guess is that this represents mostly an approval increase among conservatives and moderates.
As for those Quinnipiac polls cited by Kauthammer, they’re pretty interesting, too. He wrote that the latest Quinnipiac poll shows Trump’s disapproval “numbers have essentially returned to Trump’s (historically dismal) pre-election levels,” after a bump in the November 17-20 poll. Again, however, there’s no link to those polls. But if you go to that RCP page, you’ll see that the most recent Quninnipiac poll was taken on 1/5-1/9 and showed a disapproval/approval spread of 51/37. That’s somewhat of an outlier in that its approval rating is quite a bit lower than in the other polls taken at the same time, but its disapproval rating is similar to that found in other polls, including the Gallup poll already discussed.
So, what of that November 17-20 bump? Again, if you look at the chart and compare that November Quinnipiac poll to other polls taken at pretty much the same time, you will see that the Nov. 17-20 poll (which showed 46 disapproval and 44 approval) was the most Trump-favorable poll of all, with the others varying quite a bit from each other, but all of them finding a disapproval rating of 52 or more, which is similar to Trump’s disapproval rating of 51 in the most recent Quinnipiac poll.
So as far as I can see, this squandered-bump theory appears to be wrong.
And what about Trump’s disapproval ratings returning to his “(historically dismal) pre-election levels”? Nope. Not one poll I could find had a pre-election disapproval level of 51% for Trump—not Quinnipiac and not anyone else. Again, that RCP chart is informative, and if you bother to do what I did (I went back to December of 2015, checking all the disapproval ratings from then to the election, which took place almost a year later) you can see that the pre-election polls agreed that Trump’s disapproval was in the high 50s and the 60s, with nothing at 51 prior to the election. And that includes Quinnipiac.
So one can only conclude that Trump’s disapproval ratings have fallen since the election, and his approval ratings have gotten higher. That’s pretty much across the board, in all the polls. The small fluctuations and variations appear to represent mere noise and not to be consistent or meaningful.
You may think I’ve spent way too much time dealing with these numbers. You may be right. Who cares about polls?
Well, lots of people. And—to return to the original point of this post—pundits use them to create an impression of something and to describe and document a supposed phenomenon. They rely on people not going back themselves to check the work they’ve done. And most people won’t—because, let me tell you, it’s tedious work to do.
It may even be tedious to read. For that, I apologize. But this sort of distortion gets my goat, and is extremely frequent. On both sides, let me add.
And it might even get me to feel sympathy for Trump.
