Florida Man vs. RFK: The Most Overrated Third-Party Candidate Ever
No, RFK Jr. doesn’t matter … at least not yet.
In theory, I should love Robert F. Kennedy, Jr. On the surface, he’s the perfect Florida Man even though he has never lived in the Sunshine State. He’s a 70-year-old Baby Boomer who hates the COVID-19 vaccine. He has “thoughts” on the JFK assassination. (He’s just asking questions.) He probably takes TRT based on his appearance. He is a fan of the Joe Rogan Experience and Tucker Carlson, and he is coasting off his father’s dynastic name and family fortune. Robert sounds like a Florida Man to me!
However, in 2024, Robert F. Kennedy does not matter — and I am tired of political analysts pretending like he does.
The Water Cooler Barrier and Your Aunt Barb:
I get it. He’s a flashy news story. He has a dynastic political name. Not only is his last name Kennedy, but he also shares the exact name as the second most famous Kennedy: Robert F. Kennedy, the candidate’s father. Robert Sr. could have perhaps won the presidency if he was not tragically assassinated on the presidential campaign trail on June 6, 1968.
By this point, RFK Jr. has broken what I call the “Water Cooler Barrier”. The relevance of RFK Jr. has now penetrated into normie discourse. He is no longer a curiosity for political junkie nerds online. Your vaguely apolitical co-workers now might start talking about him at the water cooler. They follow politics in a cursory manner, but they are just “fed up” with “both sides”. These people don’t know who Jill Stein or Cornel West is, but they do know RFK Jr.
During the Fourth of July party at your uncle’s house this year, your younger female cousin (on summer vacation) will bring her boyfriend from FSU. With a slight lisp because of the Zyn pursed in his upper lip, the boyfriend will ask you: “What do you think about RFK? He makes some good points.” You see. Your cousin’s boyfriend Trevor doesn’t like either side.
Your uncle’s Boomer wife (second marriage) will corroborate what Trevor is saying, but she is doing so from a different context. Her name is Barbara but goes by “Barb”. She is a 63-year-old real estate agent from Sarasota. Barb has interest in RFK Jr. not because she saw him on Joe Rogan’s and Theo Von’s podcasts (as Trevor did) but because she has a vague nostalgia for the Kennedy family from her childhood.
Barb is originally from Massachusetts, but she came to Florida in 2005 to take advantage of the bustling housing market and low taxes. Barb might have voted for Ross Perot in 1992, but she can’t remember. She voted for Barack Obama twice but, then, voted for Donald Trump twice. Barb laments that other countries are “taking our manufacturing jobs”. At the same time, she has a used 2009 Toyota Land Cruiser in the driveway.
RFK’s Ballot Access Problem:
Here’s the problem. Luckily, for Trevor and Barb, they are registered voters in the state of Florida. They will see “Robert F. Kennedy, Jr.” on the ballot underneath the names of Donald Trump and Joe Biden. Will they end up actually voting for RFK Jr.? History says no. Leaners towards third parties usually end up “falling in line” with one of the two major party nominees. I will bet that Barb will vote for Donald Trump, and Trevor might forget to make it to the polls.
Barb and Trevor have the freedom to toy with the idea of voting for RFK Jr. as an alternative to Biden and Trump, but there are Barbs and Trevors in every other battleground state. Many of these states do not yet have RFK Jr. on the ballot for November, and time is running out for him.
The map above shows Kennedy’s ballot access as of now in each state. Look at the green-colored states. Michigan is probably the only true battleground. Florida and New Hampshire are close, but they firmly lean in one direction. As of now, the 2024 model published by FiveThirtyEight gives Trump a 75% chance in Florida and Biden a 75% chance in New Hampshire.
For all of the other states, many will do as much as they can to blockade third-party candidates. Kennedy’s campaign might claim that they are getting on the ballots in the non-green states. They very well might, but — because of the structural disadvantage of third parties — it is not a rubber stamp. States will do whatever to gum up the byzantine process mired with red tape. Look at Nevada. The Silver State did not accept the petition, and now, it is in dispute. What is to stop that from happening in all the states in which they claim they are just awaiting certification? In other words, I cannot take Kennedy seriously in a state until he officially gets on the ballot.
The Spoiler Effect:
Why is Kennedy’s access of interest from state to state? Kennedy has 0% chance of winning the presidency and, also, a very low chance of winning any states. This is true of pretty much all third-party candidates in the modern era. Texan billionaire Ross Perot was the last largely successful third-party presidential candidate. In the 1992 election against President George H.W. Bush and Arkansas Governor Bill Clinton, Ross Perot won 18.9% of the popular vote yet still did not win any state. Alabama Governor George Wallace was the last third-party presidential candidate to win any state1. He was a Democrat but ran under the “American Independent” label in the 1968 election against Vice President Hubert Humphrey (Democrat) and former Vice President Richard Nixon (Republican), but that was a different era in a regionalized South opposing the Civil Rights Act of 1964.
Instead of worrying about Kennedy’s chance of winning any states, we have interest in his ability to act as a spoiler candidate. In other words, his presence on the ballot takes away votes from one of the two major party candidates — either Biden or Trump. President George H.W. Bush blamed Ross Perot for eating into his vote and, ultimately, causing him to lose the presidency against Bill Clinton. Bill’s wife Hillary blamed Green Party candidate Jill Stein for losing the 2016 election against Donald Trump. After Al Gore’s loss to George W. Bush in the 2000 election, supporters of the then-incumbent vice president blamed third-party candidates Pat Buchanan and Ralph Nader (for two different reasons).
On the margins, a big name like Robert Kennedy can make a big impact. I am not refuting that reality. Hillary Clinton might have made a decent point. If she had won the razor-thin “Blue Wall” states of Michigan, Pennsylvania, and Wisconsin — she would have beaten Trump. In each of those states, Jill Steins’s popular vote share exceeded the margin2 by which Donald Trump beat Hillary Clinton.
Furthermore, Robert Kennedy goes way beyond Jill Stein. Because he shares a name with his father, Kennedy has infinitely more name I.D. in 2024 than Stein had in 2016, when Clinton claimed she “spoiled” the vote — but despite the astronomical difference in name I.D., Stein had one major advantage over Kennedy. She had full, equivalent ballot access in 44 states and had additional write-in access in three additional states.
.svg){kind=link}
Compare this map with Kennedy’s map for the 2024 election as of now, which I placed earlier in the article. He has nowhere near the footprint that Stein’s Green Party had. The darker green color represents states in which Jill Stein’s name appeared on the ballot underneath the names of Donald Trump and Hillary Clinton. The lighter green color represents those in which she just had write-in status but no name printed on the ballot.
Looming ballot deadlines for RFK:
The clock is ticking for Kennedy to qualify in some of the most important states if he wants to make an impact. I made the following chart, which shows Kennedy’s status in some of the most important states (the ones that Larry Sabato defines as not solid for either side). For each state, I show the:
number of days before the deadline to appear on the ballot
number of electoral votes
win probability for either Biden or Trump from FiveThirtyEight as of today
tipping point probability (the probability that a state is the state whose electoral votes decide the winner of the election)
Kennedy still has a lot on the table in terms of states in which he could affect the result of the election (or “spoil”), and he has missed two very important states: North Carolina and Texas. Even though Trump is performing strongly against Biden in both states, North Carolina and Texas each have so many electoral votes that they have a high chance of determining the outcome of the election.
The “closeness” of a race is not the exclusive factor of determining if a state is important. Compare Texas to Nevada. The Silver State is much more competitive than the Lone Star State. More specifically, Trump has a 78% chance of winning Texas while he only has a 56% chance in Nevada, but Texas has a higher chance of deciding the election because it has 6.7 times more electoral votes than Nevada does.
Meanwhile, let’s look at how the other third parties are doing with regards to qualifying.
As seen in the chart above from Wikipedia, both the Green Party ticket and the Libertarian Party ticket have access to more electoral votes than Kennedy currently does. Not only do I think that Kennedy is underrated because he, on his own, has not achieved ballot access in many states — but I think so also because two other parties have more access than he does. The Libertarian Party has significantly more.
Beyond just ballot access, the Green Party and Libertarian Party have already made impacts on presidential elections, particularly, in 2016 as I have already articulated. As for the Green Party, specifically, it is the same woman running who ran against Clinton and Trump in 2016. Although Stein still does not have the name I.D. that Kennedy has, she still has way more name I.D. than she did in 2016, when Clinton claimed that she spoiled the election.
Ironically, Clinton probably bolstered Stein ahead of 2024. Stein arguably accrued most of her current name I.D. because of Clinton and other media outlets talking about Stein after the election. Look at Stein’s stretch of relevance on Google Trends from late November onward (after the election on Tuesday, November 8).
Admittedly, Stein’s relevance peaked on Election Day, but Stein was still more relevant on Google in late November and early December (after the election) than she was in the weeks leading up to the election.
Moreover, Kennedy could easily access the ballot in more of these important states as to make an impact on the ultimate outcome of the election, but he is not there yet. Therefore, we should not take him seriously until he gets there. It is not a fait accompli that he achieves ballot access in those outstanding influential states.
RFK’s Detriment to Accurate Election Forecasting:
Shakeup at FiveThirtyEight
For months, I have had these thoughts about Kennedy at a cursory level with regards to his impact on the Electoral College, but what really piqued my interest in delving deeper was the launch of the 2024 presidential forecast on FiveThirtyEight. This electoral analysis website under the ABC News moniker has always had major traffic since The New York Times acquired them ahead of the 2012 presidential election. In 2013, ESPN — a subsidiary of Disney — acquired FiveThirtyEight from The New York Times. Although FiveThirtyEight has since fallen under the Disney umbrella, it shifted from ESPN to ABC News in 2018.
Nevertheless, this year’s forecast has a huge difference from every other election forecast ever published by FiveThirtyEight. Founder Nate Silver is no longer leading the construction of the model because of layoffs by Disney in April 2023. In the following month, on May 19, ABC News hired G. Elliot Morris as “director of data analytics”, essentially, to replace the function that Nate Silver had at the outlet. Silver still owns his model, so the 2024 election model from FiveThirtyEight does not use Silver’s algorithm, built over the course of 16 years (ever since the 2008 election). Morris originally worked at The Economist. He built the election model that they had for the 2020 election. Later this month, Silver will eventually launch a competing model to the subscribers to his Substack newsletter Silver Bulletin.
Feud between Nate Silver and Elliott Morris:
ABC’s decision to replace Silver with Morris is, especially, interesting because of open online quarrels between the two. Silver elaborated on this feud pretty well in one of his early Substack articles “Polling averages shouldn't be political litmus tests”, published on July 1, 2023. In the article, Silver says about Morris:
Morris, meanwhile, is one of only something like four people whom I have blocked on Twitter. This is an extremely high bar for me — it’s easier to just mute or unfollow someone. I only undertake after many rounds of going back and forth with someone when I feel like arguing with them has become an asymmetrical quagmire. So take that as a signal that I don’t intend this a back-and-forth.
Still, I felt like I needed to say something, and explain myself a little more thoroughly here than I originally did on Twitter. The reason is that Morris’s approach has the potential to create negative downstream consequences for me because of possible brand confusion. There are still people who think I write for The New York Times, which I last worked for 10 years ago. So if “FiveThirtyEight” is in the news because Morris or ABC News is making decisions I strongly disagree with, that creates a potential issue.
Nate Silver also elaborates on the issue on an interview on a very good episode of the Substack-based podcast House of Strauss, hosted by Ethan Strauss — Substacker and former NBA beat writer for the Golden State Warriors. It is a very good interview of the post-Disney Nate Silver, now unfettered from any corporate censorship from ABC News superiors. Strauss’s podcast and Substack are just great overall.
Moreover, Nate Silver has accused Elliott Morris of being a leftist partisan, whose bias will consequently distort the output of the new FiveThirtyEight model. I will also say Morris has personally blocked me on Twitter due to critiques that I made of him on September 24, 2023.
My doubts with the 2024 FiveThirtyEight model (Elliott’s version)
Going back to the effect that Kennedy might have on the election, I want to look at parts of Morris’s new model. I am curious to see how Silver responds to this issue in his own private model. Personally, I probably will be looking at Silver’s model on Substack more once he releases it although Morris’s model will probably be good enough despite my issues. At the end of the day, Morris definitely makes a model much, much better than I ever could. I just suspect that Silver can make one even better than Morris can.
Perhaps I have a bias because Morris individually blocked me on Twitter due to minimally critical tweets that I made of him. I wish that I could still see those tweets of mine on Twitter, but they no longer exist because I made them in reply to Morris, who deleted them.
Regardless of anything personal about Morris, I want to make a fair and earnest critique of his inclusion of Kennedy in the model. Morris has clearly not researched Kennedy’s ballot status in any state, which could have been easily found on Wikipedia. As I wrote earlier, Kennedy has missed the ballot deadlines in two very consequential states: North Carolina and Texas.
North Carolina:
Let’s look at North Carolina because that is more competitive than Texas. Morris is “forecasting” that Kennedy will earn 6.8% of the popular vote in the Tar Heel State. Although Morris projects that Trump will probably win North Carolina, he projects that Trump will not earn a majority of the popular vote in the state. Instead, Trump will win a fairly low plurality of 47.7% of the popular vote in North Carolina.
However, this is a categorically incorrect and terrible forecast for one of the most important states. Morris is not just overestimating Kennedy (and, as a result, underestimating both Trump and Biden). Morris should be giving Kennedy literally zero percent of the vote. Why? Because Kennedy missed the deadline for North Carolina. Kennedy is not even a write-in candidate. Keep in mind that the same thing is happening with Kennedy in Texas.
Okay. Kennedy claims that his petition is “awaiting certification” in North Carolina, but the State Board of Elections has not approved any of the signatures yet. The submission means nothing. Kennedy “submitted” signatures to the state of Nevada, but he is now mired in lawsuits in the state.
As of now, we should act is if Kennedy is not on the ballot in North Carolina, Nevada, Texas, or any other state in which he has not qualified — yet Morris is still predicting that Kennedy will earn 8.0% of the popular vote in Nevada.
The margins are so tiny in these states that Kennedy’s 8.0% in Nevada makes a decent impact on the model. In Nevada, Kennedy’s projected vote share (according to Morris) is 6.7 times that of the projected margin between Trump and Biden. Morris’s model is not looking at if Kennedy is actually on the ballot. He is using polls in each state as if Kennedy is, but that throws a monkey wrench in the forecast because Kennedy’s performance in the polls is very inflated.
Pennsylvania:
Take Pennsylvania as well. According to Morris’s own analysis, Pennsylvania has the highest chance of actually deciding the election at a probability of 13.3 percent.
At the moment, Kennedy has only qualified as a write-in. As a result, technically, Pennsylvanians could vote for Kennedy, but actually writing in a name is likely not to happen when the choices of Trump and Biden are already printed. Despite this fact, Morris projects that Kennedy will earn 5.9% of the popular vote in Pennsylvania.
Final critique of Morris:
Morris always talks about “fundamentals” overcoming the polls. This is why Biden is ahead of Trump in the forecast right now (albeit by 4 percentage points). If Morris fully looked at the polls, Trump would have a 58% probability of winning, yet Morris is not looking at the very fundamental that is election law and missing ballot access. He just indiscriminately puts Kennedy’s poll performance into the model as if Kennedy was just as valid as Trump or Biden, whose parties have actually qualified in all 50 states — not just a measly nine.
Lastly, Morris completely ignores Stein or the Libertarian nominee. Jill Stein has actually run for president before and, arguably, contributed to Hillary Clinton’s unexpected loss to Donald Trump in 2016. She is a legitimate candidate who does have many more states for which she has qualified, yet Morris only looks at Kennedy.
This makes me wonder what would happen if Morris excluded Kennedy in states for which he has not qualified? What would happen if Morris included Stein? Would it hurt Biden in Morris’s forecast? I do not know, but it must make an impact. People assume that Kennedy pulls from Trump voters. We do not know if that is true, but if the conventional wisdom is correct, excluding Kennedy from the forecast could very well help Trump in important states like Pennsylvania or North Carolina. Unfortunately, based on the fact that Morris blocked me, he would never have the opportunity to see such critiques.
“Experts” should not be making such oversights. I am not an expert, but I was able to figure this out from Wikipedia.
Final Thoughts:
Let’s forget about FiveThirtyEight and zoom back out to the macroscopic view. I am not saying that Kennedy cannot have an impact. He could have an impact in the states for which he qualified, but until he qualifies in crucial states beyond Michigan, I am not taking his impact seriously. Things could change, but we need to base analysis on what is certain — not predicting what the State Elections Board in North Carolina may or may not do.
Governor Wallace won the following states: Alabama, Arkansas, Georgia, Louisiana, and Mississippi.
Trump’s margins of victory over Clinton in Michigan, Pennsylvania, and Wisconsin were — respectively — 0.23%, 0.72%, and 0.77%. Stein’s popular votes shares in each of those states were — respectively — 1.07%, 0.81%, and 1.04%.