Poll Vaulting

The business of creating news

2024 is an Olympic year. 

This summer in Paris, world-class athletes will hurtle down a track carrying a long pole that they’ll use to propel themselves over a bar suspended several feet above the ground. Do it right, and you might soar to glory. A misstep could be very damaging to your health. 

2024 is also a presidential election year. 

And that means a very different type of “poll” vaulting. In this one, precision is impossible. The exercise has a built-in margin of error. But that doesn’t stop news organizations from leaping from the results into speculative coverage. 

If you have been anywhere near the news lately, you are undoubtedly aware of The New York Times poll which showed Donald Trump leading Joe Biden by 5 points (48 to 43). When it came out this past Saturday, it was like a careening pinball, lighting up social media, cable talk shows, chat groups, and all the other virtual and in-person news nodes wired into political coverage.

Many Democrats desperately seeking to avoid despair promptly raised issues with how the poll was conducted — the statistical significance of the sample size, questions about who answers calls from unknown numbers, and doubts about the composition of the respondents, which seemed to skew more Republican than the electorate who show up at the actual polls to vote. 

Skeptics also pointed to other data that suggest caution, like how Democrats keep overperforming in special elections. There was also the 2022 midterms, which polls suggested would be a “Red Wave.” When the results came in, Democrats more than held their own, and the only red wave should have been embarrassed faces from all those who got it wrong. 

That said, Democrats do themselves no favors by trying to put a rosy lens on a fraught political environment. But do you know any Democrat who needs a poll to tell them the election will be far too close for comfort? And far closer than one would hope, considering the morally bankrupt, autocrat-loving, democracy-destroying con man who is being propelled to the Republican nomination. However, it should be noted that Trump’s vote totals in the primaries are underperforming much of the polling. 

Whether The Times poll and others are flawed I will leave to mathematical and political experts to parse. As a journalist, however, I can speak firsthand about what polls do to news organizations and the resulting coverage of our political process. 

The short answer is: It’s not good at all. 

In the wake of its poll, The Times has spent three days with dramatic headlines extrapolating the results into breathless news coverage. Because the poll showed Biden not doing well, the coverage had to match. For example, the reports include testimony from voters that support the poll’s findings. But in a country of more than 330 million, a reporter can always find an anecdote to fit a preconception. 

Let’s be clear: A poll is a manufactured news event. And I would argue that means it shouldn’t be considered a news event at all. Journalistic organizations should not be in the business of creating news, especially in ways that they have the power to control. 

A poll isn’t an objective observation of nature. It is infused with the biases of the polling organization which determines a series of variables — Who do you poll? How do you poll them? What questions do you ask? What is the makeup of the sample size? These decisions will shape the results. Maybe they are based on reasonable guesses as to what would represent the likely electorate, but they are still, at best, educated assumptions. 

And let’s consider the questions that are asked and how they shape what follows. 

One issue that got a lot of coverage because it was featured in the poll is concern about Biden’s age. Like many other press outlets, The Times seems to consider it one of the most important issues of the election. As I wrote in this newsletter a month ago, it is a legitimate concern but also one that is being magnified by the press with little nuance in the coverage. 

So it is no surprise that, armed with data they generated from questions they decided to ask, The Times has featured Biden’s age as a primary takeaway about his poor standing. To which Trump has basically said, “Thank you very much.”

This coverage is a choice. Please make no mistake about it. Consider this very perceptive note I saw on social media: 

The NYT commissioned this poll; they refuse to commission a poll asking Trump voters whether the rapes he’s committed make him an ineffective leader or could cost him their vote. Which obviously means the NYT thinks age is more disqualifying than committing multiple sexual assaults.

A jury has, after all, found that Trump sexually abused E. Jean Carroll. That’s a fact. By contrast, no medical professional has found Biden has cognitive decline. 

Now, let’s consider some other questions The Times could have asked. 

• Does Trump’s embrace of the murderous autocrat Vladimir Putin make you concerned about his judgment on national security?

• Do you think a president who lies about election results and summons an insurrectionist mob is suitable for office?

• Now that Trump’s business was found to have committed criminal tax fraud and falsified business records, do you think that Trump is fit to lead the nation? 

The New York Times proudly heralds its slogan “All the News That’s Fit to Print.”

But what happens when that news is generated by a manufactured event? And what happens when you exclude “All the Questions That Should be Asked?”