In the Washington Monthly’s continuing quest to immunize America against “Mad Poll Disease,” Contributing Writer Robert J. Shapiro reminds us that the history of June polls is not all predictive of November outcomes.
Aside from that history, that the 2024 polls to date—and in turn, the poll averages—are shockingly inaccurate in one surprising way.
But first, here’s what’s leading the Washington Monthly today:
***
After the Trump Verdict: This Week’s Presidential Contest Polls Mean Little to Nothing: Shapiro’s look back at the track record of June polls. Click here for the full story.
Biden’s Immigration Policy Is Not Trump’s: My assessment of the president’s new executive order and how it is nothing like Trump’s anti-immigrant nativist vision. Click here for the full story.
***
One challenge for pollsters is determining how third-party and independent candidates will affect the presidential race.
As Jacob Indursky reported back in December, “third-party candidates almost always shed their early support” as Election Day grows closer and worries of wasted votes increase.
Still, the einclusion of minor candidates in polls can give us some sense of how well the major party candidates are retaining support from their respective coalitions, and what slices of the electorate they need to convince.
However, to fulfill that objective, pollsters need to include all the minor candidates with substantial ballot access.
Yet almost no pollster has included a Libertarian candidate in national presidential surveys.
This is despite the fact that the Libertarian Party has the most ballot lines of any third-party operation. According to a Ballot Access News edition dated June 3, the Libertarians are on the ballot in 37 states so far. (The Libertarians hit 50 states in the last two elections, but there is some question whether its current leadership can replicate the feat.)
Yes, it’s true that the Libertarian Party only picked its nominee, Chase Oliver, about two weeks ago. But the Green Party won’t pick its nominee until next month, yet Jill Stein is typically included in multi-candidate polls.
Most pollsters haven’t bothered to include any stand-in Libertarian. In the national surveys collected by FiveThirtyEight since May 1, only one included Oliver and one included a different Libertarian candidate, while 13 polling outfits have conducted trial heats with Stein, Robert F. Kennedy Jr., and Cornel West.
I have long maintained Kennedy should not be in national polls until he has shown he is likely to be on most state ballots, and he has not yet done so. The New York Times’s ballot tracker says Kennedy is certified in six states so far. (The Boston Globe reported last month on the Kennedy campaign’s penchant to claim ballot access prematurely.)
But let’s set aside that argument for now and accept that Kennedy—thanks to his name, his fundraising, and his poll numbers—will continue to be included in polls.
Stein’s Green Party has a pretty good ballot access track record, clearing 40 states in 2012 and 2016 (but not 2020). She may have difficulty this year, as she’s complained that Kennedy has bought up many of the professional signature gatherers on whom the Greens typically rely. But it’s not completely unreasonable for pollsters to give her the benefit of the doubt for now.
But there is zero justification for including West.
Not only is he certified in very few states (the Times says four, Ballot Access News says six, the West campaign says eight), but his campaign barely has any money—just $18,693 cash on hand as of last month, with $28,850 in debts.
And West’s inclusion appears to have some impact in shaping perception of the race.
The Real Clear Politics two-way average as of June 6 has Trump ahead by half of a percentage point, but in a five-way race—without any Libertarian candidate—Trump is up by 2.1 points.
West’s average share is 1.5 points, almost equal to the difference of the two-way and five-way margins.
In fact, the combination of Stein and West—totaling 3.3 points—appears to be the major reason Biden performs worse in the five-way average than the two-way average.
In other words, Kennedy appears to draw more from Trump, but that effect is more than offset by the inclusion of Stein and West.
What about the FiveThirtyEight average? Strangely, that outfit has no two-way or five-way average, only a three-way average with Trump, Biden, and Kennedy.
And that is not strictly based on polls that include Kennedy, but also polls that don’t. The explanation of the average’s methodology reads, “polls that don’t include him as an option are likely overestimating support for Biden and Trump. So we apply a third-party adjustment to polls that don’t ask about third parties.”
A three-way race is an artificial construct, ignoring the parties with a history of successful ballot access and including a candidate with a lot left to prove.
Inclusion of Oliver probably would not immediately make much of a difference; he got zero percent in the one poll that included him. But a lack of attention by pollsters leads to a lack of attention by reporters, and vice-versa.
If campaign reporters begin to recognize that they are giving far more attention to minor candidates with far less ballot access than Oliver, perhaps that will change, and we will see if Oliver might siphon off votes from Trump.
Until that happens, we don’t have any data that sheds light on how all of the potentially significant minor candidates could impact the outcome.
FIND THE MONTHLY ON SOCIAL
We’re on Twitter @monthly
We’re on Threads @WAMonthly
We’re on Instagram @WAMonthly
We’re on Facebook @WashingtonMonthly
Best,
Bill Scher, Washington Monthly politics editor
