Outside the Beltway

Kennedy’s 5%

Where will it go?

Steven L. Taylor · Thursday, August 8, 2024 · 4 comments

I have long been operating under the assumption that there is no way that Robert Kennedy, Jr. is going to end up winning anywhere near 5% of the national popular vote, although he has been consistently polling in that range. Here’s the latest for FiveThirtyEight.

RCP has a similar result in their 5-way poll:

I was reminded of all of this given the latest RFK, Jr. story (via NBC): Robert F. Kennedy Jr. posts video saying he put a young dead bear in Central Park.

Independent presidential candidate Robert F. Kennedy Jr.acknowledged Sunday that he abandoned a young dead bear in Central Park after he initially planned to skin the cub for meat. 

Kennedy said in a three-minute video on X that The New Yorker magazine found out about the incident, the date of which is unknown, and asked him for confirmation. Kennedy described driving north of New York City to go falconing with a group when he saw a woman in a van hit and kill a young bear.

“So I pulled over and I picked up the bear and put him in the back of my van because I was going to skin the bear, and it was very good condition,” Kennedy said in the video, talking to Roseanne Barr. “And I was going to put the meat in my refrigerator.”

An infinite number of monkeys with typewriters may well want to have a word about the script for Hamlet that they just worked out, but there is simply no way they are going to generate a tale of RFK Jr.’s roadkill bear-for-meat story which includes a reference to falconing as told to Roseanne Barr.

If a man who has acknowledged that a worm ate part of his brain and who had admitted to leaving a dead bear in Central Park (a bear that, mind you, he was planning to eat–in the clip he notes that taking such roadkill in New York state is legal) can get 5% of the popular vote, even with that last name, I may have to stop writing about elections.

That he knows the roadkill rules may be linked to how he got that worm, but I digress.

I would note that the average two-party national popular vote share in the United States for the presidency since 1988 has been 94.46%, and that includes 1992 when it was 80.46%, and 1996 when it was 89.95%. Both of those years Ross Perot was on the ballot (i.e., outliers). Since 2000 the two-party vote share has averaged 96.48%. It was 98.16% in 2020.*

All of that is to say that it is unlikely that the total third-party vote share will be over 5%, let alone for one candidate. I will note that the only non-Ross Perot year in this set of data where the third-party vote share exceeded 5% was 2016 when it was 5.73%. In that year we not only had the contest of two highly unpopular candidates (Trump and Clinton) but also had Gary Johnson (3.28%), Jill Stein (1.07%), and Evan McMullen (.54%) running more successfully, as a whole, for third-party runs in a situation where a lot of people both did not like Trump and also thought he was going to lose (and, indeed, he did lose the popular vote). Johnson, in particular, did especially well in a historical sense.

Getting back to 2024, it is not unusual to see third-party candidates poll higher than their likely vote totals at this point in the process. It is typically a place for people who really haven’t made up their minds to park their opinions (or people who aren’t ultimately going to vote showing up in the polling). I will be interested to see if the stability in these numbers continues once we are post-DNC and into the general election campaign proper (which, has not yet started for those of you scoring at home).** At that point in time I would also expect likely voter screens to start shifting those numbers.

Based solely on historical trends, I am expecting RFK, Jr. to win 1.5% of the vote or less (and that is me giving him juice because of his name and because I am probably being too conservative in my prediction). Indeed, his name is a major part of why he is polling as well as he currently is.*** I think, too, it has been a function of the doubler-hater phenomenon (i.e., the unpopularity of both Trump and Biden). I always assumed that most of those voters would either be washed out via likely voter screens or accrete to Biden and Trump. I expected more to go to Biden because I thought (and still think) that Trump has a lower ceiling than Biden.

Of course, Harris mixes all of this up. It seems likely that her odds of attracting some of the double-haters than does Trump.

At a minimum, it seems that 2024 should be closer to 2020 in terms of the third-party vote if anything because 2016 is a more anomalous case and because the undecided/swing voters know what a Trump presidency looks like when in 2016 not only had they not experienced it, most people didn’t believe it would happen.

I recognize that each cycle is different, and ongoing resentments over inflation are a negative for Harris. Still, I think that the excitement that we are seeing (which can be measured both in polls moving Harrisward, but also in terms of her fundraising successes to date) suggests that there is some reason to assume that the marginal anti-Trump voter is coalescing in her direction, which would suggest an outcome more like 2020 than 2016 as far at the third-party vote goes.

Of course, state-level effects are their own issue, and I do not think there is enough data to address that question at the moment.

Still, if you had “roadkill bear” on your bingo card, you win!

*I had those numbers handy for another project I was recently working on, and was focusing on 2004-2022, but I had the numbers back to 1988 in the spreadsheet, so share them here.

**I would note that the UK and France both announced and ran full campaign and voting cycles in just the last couple of months. American exceptionalism can be exhausting at times.

**That 5.3% is considered polling “well” is a comment on third-party support in the US.