The political pundit class was jolted into a bit of a frenzy this morning after the release of the New York Times/Sienna’s latest presidential poll, which showed Donald Trump leading Joe Biden nationally by 5 percentage points among registered voters and 4 percentage points among likely voters. People like Bill Kristol have already used the results of the poll to suggest that Biden must step aside immediately. As I have stated multiple times, that would throw the election to Trump. It’s a really bad idea.
I’m extremely angry about the results, too, but only because of the fact that NYT/Sienna engaged in true statistical malpractice - and it wasn’t hard to find within a single data point in the poll’s crosstabs: Density. One of the demographic questions asked in the poll to its respondents was the kind of place you live: The city, the suburbs, or a rural area. As expected, the results skewed more Republican as population density decreaesd:
![](https://substackcdn.com/image/fetch/w_720,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F89fc482a-9871-492a-8313-d5e330bcb796_1068x682.png)
![](https://substackcdn.com/image/fetch/w_720,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Facbbed8c-18fb-45b8-9d7d-0e8fc100232f_888x701.png)
But hiding in plain sight just below the Biden-Trump numbers was the proportion of urban/suburban/rural voters in the poll. And that’s when I noticed something really, really wrong. Rural voters made up 35-36% of respondents, while urban voters only accounted for 22-24%. Immediately after observing this, I checked the Times’ result-adjusted 2020 exit polls, and found exactly what I thought I would see:
It turns out, the NYT/Sienna poll oversampled rural voters by a whopping 84% in excess of their true proportion of the electorate. Let’s be clear about something: THAT’S INSANE. No wonder Joe Biden is getting clobbered in the polls; the people answering these polls are a bunch of Trump lovers in rural areas. And this would also account for Biden’s supposed “slippage” among black, latino and young voters; they’re over-polling minorities and young people who live in rural areas. This is probably why Donald Trump has under-performed his average-of-polls in all four early states; Nikki Haley does best among urban Republicans who are being drastically under-sampled.
Here’s what makes my head explode: If you take the results of the urban/suburban/rural percentages between Biden and Trump in the New York Times/Sienna poll, and then readjust them to the proportions of each type of voter in the 2020 exit polling, Biden actually wins among both registered and likely voters.
The evidence is clear: There is a huge polling problem. It is simply not acceptable for a major public opinion pollster to oversample the Trumpiest and most culturally conservative swaths of the electorate by 84%. Think about what that actually means: The influence of rural voters in America’s supposedly #1 pollster has been nearly doubled. And it makes me disturbed that the public discourse surrounding Joe Biden’s re-election campaign is now tied to the results of polls that are at best problematic, and at worst a total distortion. Unfortunately, because these polls are becoming so detached from reality, we are truly flying blind into this election season. I do not know why Donald Trump is the beneficiary of these issues, but until the polls can sort it out (which we won’t know until after the election), I don’t think we can trust the polls. And that goes for polls that show Biden leading Trump, too.
Sadly, that is not the only issue with this poll and maybe not the worst one. The NYT also violated standard pollling procedures by not asking the horse race question first which likely biased the respondents against Biden. The reason I say it is worse is a bad polling sample will happen 5 times out of 100 even if a pollster is doing everything correctly. Messaging up the question is deliberate and, given that they changed the question ordering for this poll compared to their previous poll in December, really looks like the NYT was trying to produce the result, not find the result
You share the specific data points but can you show a link to the actual poll as well? Others may want to do some investigation themselves.