In the last couple weeks I’ve written a number of not unhappy posts on President Trump’s abysmal approval rating and its downward trend. Here’s something slightly different.
As you can see, we have a somewhat different story here.
In absolute terms, the numbers are still quite bad – net -9 approval. But the trend is different. This comes along with two recent articles by two pretty sharp guys – Nate Silver and Ed Kilgore – arguing that while Trump is unpopular, he’s not quite as unpopular as many Democrats think.
There’s a good point to be made here, even apart from a few good days for Trump in the Gallup tracking poll. Let me try to briefly summarize them.
There’s not just the Gallup poll or Pew or Quinnipiac, each of which show Trump in really bad shape. There are a bunch of other polls and they actually differ significantly. This seems to be for three basic reasons: quality, methodology and population.
Let’s start with quality: the higher reputation polls mostly show Trump doing the worst (good news for anti-Trumpers).
Next, methodology. Live phone polls show Trump worse off than robopolls and online polls. There are arguments that Trump voters resist admitting they support Trump. By that argument, online pollsters may be more accurate. But there’s little actual evidence for this, either in the primaries or in the general election. Live phone polls, meanwhile, are the most tried and tested methodology we have (mostly good news for anti-Trumpers).
Finally, population. Traditionally pollsters poll “adults” to test presidential approval. So if we’re looking at historical standards of comparison, we should look at adults since that’s the way it’s mainly always been done. But voters determine the election, not adults. So if we’re thinking about the 2018 election cycle and other electoral consequences, we should at least be giving a close look at the pollsters already looking at registered voters. And those do tend to be friendlier toward Trump (some good news for Trumpers.)
Note that Pew has Adults and RVs. Trump picks up a few points with RVs. But still generally terrible.
Silver has this chart which illustrates the patterns pretty nicely.
The huge outlier here, of course, is Rasmussen. I’m comfortable largely ignoring Raz since it’s such a historically unreliable poll. The other issue is that they’re measuring “likely voters.” That’s the gold standard in an election season, especially late in a cycle. But I’m not sure I’ve ever seen a pollster use it this early for the simple reason that you have very little ability to judge who and who isn’t likely this early. It seems more like gaming to me. So toss out Raz. But even without it, you still have a pretty big spread.
So what does all of this mean?
I think it means that to be realistic and be looking at all the data, it doesn’t make sense to assume that Trump is approaching net 20 point negative favorability as he has in a few polls. But it seems clear that he’s likely in high single digit unpopularity. Which, again, is awful.
Here’s a key issue to consider. Most Presidents come in high and fall off significantly over their first six months or year in office. If Trump sees a fall off anything like that, he’s in catastrophic danger because he’s starting from deep unpopularity. But maybe Trump has just hit the baseline much earlier than other Presidents. President Obama, after all, spent long stretches of his presidency with a few points net negative approval. I suspect it’s a mix of the two. But these are each important points to keep in mind.
But here’s what I take from all these recent numbers. If we go back to Gallup, Trump’s recent nadir was on February 16th, six days ago. As others have noted, judging by his amazingly low standard, Trump has kept to a somewhat more disciplined routine since about then – no top foreign policy advisors fired amidst a counter-intelligence probe, fewer Twitter tirades, no on-air press conference/mental collapse, etc.
This illustrates an important point for Democrats to internalize: at least so far Trump has needed to sustain a pretty continuous level of just batshit crazy stuff to keep his numbers in the 40% approval range. Even a few days of not delivering the crazy every single day and he stabilizes, at least a bit. That’s fairly amazing. It is a disquieting resilience.
We saw this during the election too. Trump had what might be considered two of the worst weeks in electoral campaign history in the first half of October as he was pretty convincingly revealed as a serial sexual predator, liar and awful person. Three weeks later he was elected President, with a major assist from Vladimir Putin and James Comey.
I think Trump’s nonstop crazy will weigh him down over time. But as much as he tries, he won’t totally self-destruct on his own. Okay, he might. But there’s no betting on it. The real things that will tell the difference in the few 2017 races and especially in 2018 won’t so much be Twitter tirades or even corruption (probably). It will be things like millions of people losing their health care. It will be threats to things like Medicare and Social Security (yes, that’s not done at all). It will be basic pocketbook stuff where Trump’s (and the GOPs) policies are revealed to be deeply damaging to average Americans. Those are things that are about voters, immediate life issues that shape opinions and don’t fade with a few days or the President keeping his head down.
Corruption is important. So is Trump’s demonstrable unfitness to be President. It’s all of the above. But what will really sink him and the GOP is that his policies are just not popular. Anybody who wants to declaw Trump in 2018 should not forget that.
