Trump’s approval rating as a case study in public opinion

Donald Trump’s popularity is really quite stable. For a while, it looked as if he was losing a point or so per month, but that trend has reversed. From the perspective of January 2018, the flatness of the line is striking.

Then again, presidential popularity is usually correlated with economic performance. A strong correlate of approval for an incumbent president is satisfaction with the direction of the country. Since the economy is humming along and satisfaction with the direction of the country is modestly risingone would expect Trump’s popularity to be 7-17 percentage points higher than it is.

To be clear, I don’t think that the economy is producing fair results. But historically, measures like satisfaction and consumer optimism usually correlate with presidential approval. Trump has broken that pattern.

Also, a prevailing model in political science holds that our demographic identities come first. They lead us to affiliate with political parties that seem to represent or encompass those identities. Our attitudes toward politicians are then strongly colored by our partisan affiliations.  But party identification sometimes changes faster than the demographic composition of the country. I’ve created the following graph of party affiliation using Gallup data (moving averages over 7 months). It shows that there’s not been that much change over time–the y-axis goes from 20%-50%–but Republican identification (the red line) has fallen since Trump was elected. First Independents (gray) seemed to gain at the expense of Republicans and Democrats, but lately it’s been Democrats (blue) who have increased their share.

I’d conclude that underlying factors–demographics, economics, and partisanship–do explain most of a president’s support. But they don’t fully explain it, and Donald Trump is demonstrating that you can alienate a lot of people from yourself and your party if you really act like a jerk. This is kind of a perverse finding (doing a very bad job can cause damage), but it’s still evidence that rhetoric and intentional action matter, regardless of what else is happening in the world. It lends support to a theory I have long suspected: agency is often hard to detect because most people who lead major organizations and movements are pretty competent, and their efforts tend to cancel out. Trump is an exception that shows that intentional behavior and competence mattered all along.

If the economy continues to prosper and Trump doesn’t behave even worse, I suspect we will see some improvement in his popularity. The underlying circumstances will count more and more. On the other hand, if the economy hits some bumps, he’s vulnerable. (But that is truly not to be wished for, because too many people will suffer.)