Here’s the thing about that Mitch McConnell – Amy McGrath poll
Anyone who’s been closely following the 2020 U.S. Senate contest in Kentucky knows that Mitch McConnell is at least vaguely in trouble in his race against Democratic candidate Amy McGrath. But wait a minute here, because a new poll this morning says McConnell is suddenly seventeen points ahead of McGrath. This revelation has set off a firestorm of confusion and panic.
Here’s the thing. You know how Palmer Report constantly reminds everyone not to take individual polls out of context? This is why. Imagine flipping a coin once, seeing that it came up heads, and concluding that coins always come up heads. That’s roughly the same thing as looking at one poll in a race, and assuming that poll is giving you the full picture of what’s going on in that race.
You only ever want to look at the polling averages to get a solid picture of what’s going on. When an election is over and you look back at the polls, you’ll nearly always find that the final polling averages were correct within their own stated margin of error (this was true even in 2016). But polling averages paint a boring story and are terrible for ratings, so the media likes to hype individual polls – the more outlandish the better. That’s why today’s seventeen point poll is getting so much attention; it makes for a shocking headline.
Usually I can shoot down this kind of outlier hype by simply pointing to the fact that the polling averages tell a very different story. In this instance it’s trickier because there hasn’t been frequent polling in this particular Senate race. In fact the last time we saw polling in Kentucky was when Amy McGrath and Charles Booker were still in a primary challenge. But the (admittedly scant) polling averages back then showed that they were both within high single digits of McConnell. There is absolutely no reason to believe that McConnell has somehow magically doubled his lead at a time when nothing has occurred in real-world politics that would shift the race in his favor. Nor is there any basis for interpreting this as a sign that the Democrats somehow chose the wrong candidate. Again, McGrath and Booker were polling similarly against McConnell.
That doesn’t mean that today’s result is a fake poll. With few exceptions, the pollsters are always trying to get it right. But their methodologies are based on educated best guesses, and none of them can be guaranteed to individually get it right. Again, this is why you look at the polling averages; it’s a way to take every pollster’s best effort and make sense of the overall picture. But then the media gets involved and takes what is clearly an outlier poll, and plays it up as if it were the only poll, because that’s what gets the most ratings attention.
The polls aren’t the problem; the media is. The media’s ongoing misrepresentation of the polls isn’t a reason to ignore the polls entirely. It’s just that if you want to know what’s going on in a race, you have to ignore everything the media says about the polls, and instead look at the polling averages yourself. As for the McConnell-McGrath race, today’s poll tells us nothing. If a few more polling outlets release results from Kentucky this week, then we can look at the averages, determine outliers, and make sense of the data. For now, this is just a lot of hype over gibberish.
