There remains a major divide in the way national news outlets call presidential elections

“When you come to a fork in the road, take it.”

– Attributed to Yogi Berra and, contrary to some Yogi-isms, believed to be authentic because his home in Montclair, New Jersey could be reached by bearing left or right at a fork

After the 2016 presidential election, it was clear that the established way of calling the outcome state by state needed to be corrected. More and more voters voted by mail or early in person. This change accelerated in 2020 during the pandemic. Election day polls no longer made sense as a key to a final forecast by the “decision offices” of data providers and their customers.

At that point, members of the 15-year-old media collective National Election Pool split into two competing groups.

The Associated Press developed a new method called AP VoteCast and sold it to major news organizations, including Fox News and The Wall Street Journal. With complex screening checks to identify those who will vote, VoteCast relies heavily on self-reporting in a survey.

The other major national broadcasters in the pool (CNN, ABC, CBS and NBC), as well as the Washington Post and BBC, chose instead to stick with an established provider, Edison Research, whose decision was to keep exit polls the focus of the to keep calls but to modernize it.

I recently spoke with Joe Lenski, Edison’s executive vice president, who leads the company’s election research. He offered both an explanation and a defense of the approach.

Thanks to Edison’s groundwork, Lenski said, it is expected that 50% of American voters will cast their ballots the old-fashioned way, in person, on Election Day this year. Another 15 to 20% will vote early in person and the remainder will vote by mail.

“So we have two-thirds of the voters that we know are voters — we talk to them as they leave the polls,” Lenski said. In his opinion, this is a big advantage in terms of reliability.

As for the mail-in voting portion, there’s no escaping self-reporting, but Lenski said a series of calls and screening questions yield a representative and reliable sample – tested in the 2022 election and presidential primaries, where it performed well.

I wondered if Edison’s estimates of the percentages voting for these three different species might be wrong. If so, wouldn’t that affect the prediction of the outcome?

Maybe, Lenski said, but the counts of how many vote early and how many mail-in ballots are ordered are available well before Election Day, so the models can be adjusted.

Whichever provider a network or newspaper chooses, they combine this input with their own election analysis. The result is what you see on the air, with analysts like NBC News’ Steve Kornacki or CNN’s John King. Political leaders also have influence behind the scenes.

With the sun shining, both Edison and The Associated Press want an extra measure of certainty before calling a specific state or the overall winner of the election in what could be a very close election.

AP Editor-in-Chief Julie Pace told me that the wire service’s standard remains that a race can be called if analysts believe the final candidate has no path to victory. But this point can of course be adjusted from election cycle to cycle.

Similarly, Lenski told me, “If we aim for 99.5% security in 2020, this time it needs to be even closer to 100%.”

Pace and Lenski also contrasted the contested present with the way it was practiced in the 20th century, with crucial results.

“It used to be that if AP called a race, that was it,” Pace said.

Lenski recalled starting in 1988 as a 23-year-old low-level employee for CBS’ election desk. “We accidentally called a few states for Dukakis – but it wasn’t a big deal.” The mistake didn’t matter, as Democratic presidential candidate Michael Dukakis lost in a landslide to Republican George HW Bush.

The bottom line is that there is a possibility that states will be called differently, or at least at different times, as the customer lists remain the same and the different methods continue to exist.

This was dramatically the case in 2020. Fox News has its own extensive data and analytics operation led by Arnon Mishkin, a superstar in the field. The conservative broadcaster called Arizona for Joe Biden on election night, as did the Wall Street Journal and NPR. That sparked howls of protest from candidate Donald Trump and other Republicans.

It took nine days for the National Election Pool group to follow suit.

While the competition between the units is mutually respectful and reserved, both sides still maintain that they have Arizona in the right.

Fox and the AP said Biden won as announced, albeit by a narrow margin.

Lenski told me that Edison’s models showed that mail-in votes counted later the following week would be a major blow to Trump. They did, he said, “and in the end we had the right” not to call.

I don’t know enough about the math involved to say how rare the kind of discrepancy that occurred in Arizona is. I expect there is a good chance of a repeat if the election is on the line.

Like Yogi’s House, the fork in the road between the two players in the decision table data in 2016 could just as easily lead the competing entities to the same spot next month.