No Labels is considering a dangerous game

And they’re basing it on a wildly incorrect reading of polling data

You might have missed it amid all the hubbub about former President Donald Trump’s indictment, but news broke this weekend that the organization No Labels is considering running a third party “moderate” candidate for president in 2024. Some sort-of splashy names were bandied around, including former Sen. Joe Lieberman (who was the Democratic vice presidential nominee in 2000), former Maryland Gov. Larry Hogan, and West Virginia Sen. Joe Manchin. 

What’s stunning to me about the No Labels bid is that they are basing a massive risk – a third party could help put Trump back in the White House if he is the Republican nominee by drawing votes from President Joe Biden – on a terrible misreading of their own polling data. 

To be clear, I do not have their polls, I only have the published details they provided to The Washington Post. But I have worked on this problem before in various capacities, and I don’t see anything new in the No Labels pitch. Their argument has four points as articulated in The Post’s article. None survive close scrutiny. 

Point one: No Labels argues that the combination of economic concern, growing independent voter identification, and opinions that the country is on the wrong track give a third party a decent chance. 

Independent affiliation has been up and down over time, and the vast majority will say they “lean” toward a party - the proportion of true independents is quite small. Part of the recent increase is driven by young people, who have also turned out mostly for Democrats in recent cycles. They also don’t vote in the numbers of older generations – which makes them kind of a risky group to count on for a new electoral strategy. Maybe a third party pulls disaffected independents out to vote, but you need stronger evidence for that than just saying they exist.

As for economic concerns and the wrong track question, neither of those support the idea of a successful third party candidacy, partly because both shift with whichever party is in power. Democrats will think the country is going to hell in a handbasket when Republicans are in power. Republicans will think the same when Democrats are in power. It’s true that both measures are particularly negative right now, even among the in-partisans, but again, you’re going to need stronger evidence than just noting the trend in order to say a third party candidacy could be successful.     

Point two: They say the fact that majorities do not want Biden or Trump as nominees indicates third-party success, along with the fact that they have about 60 percent of voters open to a “moderate, independent” candidate if the race is Trump vs. Biden.

The obvious first issue here is that we don’t know that Trump will be the nominee. He is the frontrunner, but given his legal troubles (both state and federal) it might be wise to wait a tick before crowning him just yet. Building an electoral strategy takes time, but basing electoral strategy off preliminary information that could easily change seems dicey.

The second issue is the polling. I’ve analyzed polls that show similar numbers. The thing is that people are really bad at predicting their future behavior when the stakes are low. You can’t get much lower than essentially asking “would you maybe be open to this thing that sounds nice?” Being “open to” a third option is far, far away from voting for them. To earn votes, No Labels would have to overcome the electability perception problem. There have been third party candidates on the ballot in most presidential elections for the last 40 years with no success. Perhaps No Labels thinks they can do that where all others have failed – but there is no evidence so far indicating likely success.

Point three: No Labels’ pollster says their models show about 37 percent of the popular vote would win the electoral college in such a race.  

It takes some pretty creative modeling for 37 percent of the popular vote to win the Electoral College. It’s possible with a winner-take-all system of allocating electors, but it’s clear from the No Labels hypothetical map in the Post article that they are getting this by assuming a uniform national swing toward the middle with few regional or state-level effects. I say that because the map basically just moves states with narrower margins into the moderate column, with only a few exceptions like Utah and Illinois. 

Compare their map to the vote margins in 2020. Maybe it doesn’t take pretty creative modeling after all. This is not a model that should convince anyone to make a high-stakes decision. Remember that Ross Perot won 19 percent of the vote in 1992, and that equated to zero Electoral votes.    

Point four: The polling No Labels shared with The Post showed that a three-way race would land roughly at 33 percent Trump, 28 percent Biden, and 20 percent third party. A two-way race would land roughly 43-42 Trump-Biden. 

That 20 percent is nowhere near the 37 percent they say they would need. What’s more, it advantages Trump–an essentially tied race becomes a five-point Trump advantage. Yet the No Labels affiliates say they wouldn’t do this if it was going to benefit Trump. It’s possible they have more data, but making the claim to not benefit Trump and then only providing this matchup data is bizarre. 

---

No Labels’ logic doesn’t match the data when considered in proper context on any of the four points. If you’re going to come at the third party problem in the U.S., you will need good data and good logic, as well as excellent communication skills. So far, this effort is a bust on all three.