Here we go again
Heading into Donald Trump’s Senate impeachment trial, Palmer Report warned you that once it became clear the Republicans were going to acquit him, we’d see the media start trying to convince us that we’re all doomed going forward. Specifically, we predicted that the media would latch onto whatever outlier poll showed Trump’s approval rating skyrocketing, while ignoring any polls that show the opposite, and ignoring the overall polling averages.
Sure enough, that ended up happening today. Gallup has released its latest presidential approval rating poll, and it shows Donald Trump’s approval suddenly spiking all the way to 49%, an all time high for him. Naturally, any number of political pundits – particularly on television – seized on this number as validation of their past week’s worth of claims that Trump’s acquittal is now going to magically get him reelected. The trouble is, these people are either completely clueless about how polls work, or flat out lying.
The new Reuters poll shows Trump’s approval rating at 41%, one point down from a week ago, and the same as it was two weeks ago. This poll shows Trump getting no impeachment trial bump at all, and being as unpopular as ever. But guess which of the two polls the media ran with? The shocking spike Trump supposedly received in the Gallup poll, of course.
The only responsible, accurate, and honest way to report the polls is to place them within the context of the polling averages. Each polling outlet uses a different methodology that produces different, and occasionally gibberish, results. By averaging all of the current polls together, you smooth out any anomalies. When pundits ignore the polling averages in favor of reporting just one random new poll number, they’re reporting the polls inaccurately. And when the pundits ignore the polling averages in favor of reporting the most suspicious outlier poll they can find, they’re being dishonest on purpose.
This can’t be stressed enough: the polls themselves are not the problem. Statistical sampling isn’t perfect, and outlier polls happen. But the reality is that the final polling averages generally prove to be correct within the margin of error (even in 2016, believe it or not). The problem is that the media generally reports on the polls in a highly inaccurate fashion, either due to ignorance of how statistics work, or an intentional attempt at drumming up ratings by only reporting the most shocking statistical outlier available. You can usually trust the polls. Sadly, you just can’t trust the people on TV to tell you what the polls are really said.
