A pair of columns on issue polls in the Trump era
One on the technical challenge, the other on reading the numbers
For the last two weeks, I’ve focused on issue/policy polls for my National Journal column.
The first, published last week, asks whether we are underestimating support for Trump’s policies as we tend to underestimate Trump’s support in elections. In working through that question, I discuss why we might still underestimate the right even though we’re not trying to predict who will vote (nonresponse is nonresponse, whether all voters or likely voters). But the bigger issue in how we assess policies is question wording.
The second column, published yesterday, looks at how public opinion has shifted in the 10 years since Trump descended on the golden escalator and threw our politics into an alternate dimension. I’m convinced we fell into a wormhole that day.
Anyway, Trump accelerated polarization trends by party and education - with the result of a lot of poll results on presidential approval and the more contested policy issues looking quite different than they used to.
Is it enough to declare an issue "winning" or "good" (or "losing" or "bad") for someone based on a plurality or a narrow majority? And what constitutes substantial movement when presidential approval stays between 40-45%?