New national polls show voters are more upbeat about President Donald Trump’s performance and more pessimistic about the Democrats’ chances of taking back the House. Or not.
An April 8-11 Washington Post-ABC News poll showed Trump’s job approval rating at 40 percent, while 56 percent disapproved of his performance.
An April 8-11 NBC News/Wall Street Journal survey showed almost identical numbers, with the president’s job approval at 39 percent and his disapproval at 57 percent.
But when you look below the topline numbers or start comparing survey results to earlier polls, things start to get a little more, well, complicated.
The Post-ABC survey found Trump’s job approval had improved from its January poll (when the president’s job approval stood at 36 percent), while the NBC/Journal survey showed Trump’s approval had worsened since its previous poll in March, when it stood at 43 percent.
This no doubt caused some head-scratching by news anchors and talking heads who have relatively little experience with or understanding of survey research — and who insist on focusing on every bump in the road rather than on the road itself.
Hold the horses
In fact, it’s best not to try to explain every data point.
“Margin of error” exists for a reason, and treating every survey number as if it is a perfect reflection of reality is an invitation to an emotional breakdown.
Most results from reputable national surveys show the president’s job approval in the 38 percent to 42 percent range. Try to live with that until there is a clear breakout in one direction or the other.
It’s best to approach survey results with a dose of common sense. That may make those of us seeking complete objectivity a bit uncomfortable, but it is still necessary given the increasing difficulty in getting accurate results at a time when many voters refuse to respond to public opinion surveys.
I find it difficult to believe the president’s job approval has been moving around much — or that it has improved significantly over the past few months.
Few Trump enthusiasts are defecting from him, and his opponents are as “locked in” as can be.
Yes, good news for the White House can move the president’s numbers up a couple of points for a few days, and bad news can cost him a few points for a matter of days or weeks. But public opinion always seems to return to the same 38 percent to 42 percent job approval range we have seen for many months.
A horse race?
Of course, if you want to be confused, there is always the “generic ballot.”
The NBC/Journal poll from April showed Democrats with a 7-point advantage, 47 percent to 40 percent, while the Post-ABC survey put the Democrats’ advantage at 10 points among voting-age adults but at only 4 points among registered voters.
But when you adopt a longer time frame, things suddenly get more complicated.
The Post-ABC surveys of adults from April, January and November showed Democrats with a low double-digit lead in the generic ballot. But the Democratic advantage in the generic ballot among registered voters plunged from 12 points in January to just 4 points in the latest survey.
While I’m certain you can come up with some elaborate explanation for the shift among registered voters, I wouldn’t waste the time.
It seems very unlikely that there has been a fundamental shift in sentiment among registered voters. It’s much likelier that the Post-ABC result is a statistical hiccup that doesn’t mean much.
That’s why I was surprised how The Washington Post played the results, suggesting the findings were a “signal to party leaders and strategists that they could be premature in anticipating a huge wave of victories in November.”
Maybe that 4 percent number is a signal. And maybe it isn’t.
It could simply be that the January survey was misleading or the April survey understated the Democratic advantage. Or a little of both.
If I were you, I’d wait for the next round of generic ballot tests from the major pollsters before getting too excited about the most recent Post-ABC generic ballot result.
More horsing around
Of course, if you really want to drive yourself crazy, you can consider Quinnipiac University’s generic ballot polling.
An April 6-9 Quinnipiac poll showed Democrats with a 3-point advantage, 46 percent to 43 percent. Another Quinnipiac poll conducted three weeks earlier gave them a 6-point edge.
Two weeks before that, Democrats had a 10-point advantage, and two weeks before that, Quinnipiac found them with a 15-point generic ballot lead.
That means that between mid-February and early April, a period of about seven weeks, the Democrats’ advantage in the congressional generic ballot shrunk from 15 points to 3 points.
Feel free to believe that if you want, but it strikes me as unreasonable given that the overall news flow hasn’t been favorable for Trump, and recent special elections suggest growing Democratic enthusiasm and strengths.
I have spent years watching polls, and I value them immensely. My repeated warnings about reading too much into the responses to any single question doesn’t change that.
I’d merely like to see everyone show a bit more skepticism and caution, rather than simply regurgitating the numbers and accepting them at face value.
Note: This column first appeared in Roll Call on April 19, 2018.