Mostly Peaceful Election Scenarios
Bringing back the Smooting of nationwide violence for a prediction
In June of 2022, HWFO posited a new measure for civil unrest, the “MPS20,” or one-percent-of-the-mostly-peaceful-summer-of-2020. Read about it here:
At the time, we compared the Floyd protests, which totaled 5000 demonstrations, 20 million protesters, 14,000 arrests, two billion dollars worth of damage, an average of $100 worth of insured damages per protester, and twenty deaths, against other natural and manmade disasters, such as hurricanes and terrorism. We determined that the Floyd protests were approximately equivalent to a Category 2 hurricane.
I’ve been told from multiple sources, including SHOT Show vendors who supply law enforcement, that police in the United States are already preparing for a similar event after the 2024 election no matter who wins. In this article we will examine the likelihoods of different election outcomes, prognosticate what sort of protests each might provoke, and set an expected value for death and destruction in December 2024 based on that prognostication. Forward this to any insurance adjusters or police captains you may know.
People I’ve spoken with mostly hold a blanket presumption that the election will be between Biden and Trump, but in aggregate there are many things that could prevent that matchup, so folks should first consider the chance that it doesn’t transpire at all. We can do this with mathematics by pulling some highly nefarious and not-at-all reliable numbers out of a hat. These are the possible ways I see that a Biden-Trump matchup might not transpire:
Trump may be banned from running, legally, using the 14th Amendment as a vehicle. After some deep examination, I don’t think this is likely but it is definitely within the realm of possibility even without a conviction because of how the 14th Amendment is written. Let’s call this 10% for now.
Trump or Biden may die of natural causes. They’re both very old, and actuarial tables show that a 77 year old has a 3.39% chance of kicking off naturally, while an 81 year old has a 5.25% chance of keeling over.
Either could be shot. In our history, we’ve had four successful presidential assassinations out of 46 total presidents, for a net chance of 8.7% that either gets plugged.
Biden could be replaced by DNC back room deals, probably by Newsom. The chances of this shrink with every passing day, but Biden’s mental health shrinks as well, so let’s throw 25% at this chance for the time being.
Since each of these things must not happen in order for Trump or Biden to make it to nomination, we multiply (1-X) by (1-Y) by (1-Z) etcetera, following the rules of probability, to determine the chance that none of them happens. Then we can multiply that by the chance that each wins their respective nominations provided none of those things occur (let’s say 100% for Biden and 95% for Trump) and get the chance that either is on the ballot. It’s lower than you think.
Chance Biden is on the ballot: 64.88%
Chance Trump is on the ballot: 75.42%
Trump edges Biden out in this calculation mostly because 81 year olds die more often than 77 year olds.
Since these two probabilities are independent, the chance that both are on the ballot is simply them multiplied together, or 48.93% You read that right. If the presumed odds are correct above, you have a coin flip’s chance of seeing a ballot without the expected Trump v Biden matchup, because of death, dismemberment, legal shenanigans, and political dealings. In order to hike the number over 50% you’d have to downgrade the chance either gets shot, and why would you do that when comparing modern political divisions to historical ones?
The one legitimate scenario I’m omitting is if somehow Trump isn’t prohibited from running, but he loses the nomination, and then runs as a Ross Perot style independent. In that case the Blues 100% win and the Reds 100% lose because of first past the post game theory, but I’m neglecting it purely because it makes the math tremendously harder and there’s not much chance of it anyway. In fact, I didn’t want to write this article until Trump had things largely mopped up because the math is too hard.
Trump v Biden Odds
Current (2/1/2024) election odds from Oddschecker.com have a Trump win at +120 and Biden win at +200, for an implied chance of 45% for Trump and 33% for Biden, but those odds also include other unlikely candidates. If we strip those out and distribute those chances into a two man race, that adjusts to 58% for Trump and 42% for Biden, in a head to head. Let’s pretend the bookies know what they’re doing.
If true, the chance we both (A) have a Trump v Biden ballot and also (B) Trump wins is 28.23%. The chance that the ballot is Trump v Biden and of Biden winning it is 20.7%.
Based on our odds above, the chance Trump faces someone else is 26%, the chance Biden faces someone else is 16%, and the chance neither Trump nor Biden are on the ballot is 9%, which is what most Americans would actually prefer.
It’s much more difficult to determine what would happen in a Trump v NotBiden or a Biden v NotTrump election. Haley seems to poll much higher vs Biden than Trump does, probably because so many people hate Trump, but there are probably people (such as pro wrestling fans) who would show up to vote for Trump but would stay home if he wasn’t on the ballot, and those probably aren’t accounted for properly in polling. There’s no good way to tell, so either we have to abandon the experiment here or throw the same Trump v Biden odds into the other scenarios and see what shakes out. That’s not a great assumption but it’s not a horrible one given how deep the partisanship runs in the USA today. That gives us:
Someone other than Biden beats Trump: 11.21%
Trump beats someone other than Biden: 15.28%
Biden beats someone other than Trump: 6.75%
Someone other than Trump beats Biden: 9.20%
Trump’s replacement beats Biden’s replacement: 4.98%
Biden’s replacement beats Trump’s replacement: 3.65%
These are the eight scenarios, and they add to 100%, so we know at a minimum we described the math problem properly.
Expected Riot Value
We will now jump back to our New Unit of Measure for nationwide riot damage, imprisonment, and death, the Mostly-Peaceful-Summer-of-2020. (MPS20)
Recall, the MPS20 scale sets the Summer of 2020 at a “100,” and references other events against that. The Rodney King riots were a 91, but transpired almost entirely in Los Angeles. Hurricane Floyd (category 2 on US landfall) was a 150, Hurricane Sally (also Category 2) was an 89, and the 9-11 bombings were a 3431. The Floyd Riots were about one third of a 9-11 event.
Trump Beats Biden will probably have a level of rioting much closer to a Summer of 2020 total, but the Summer of 2020 was inflated by people going stir crazy from lockdown driven anxiety. Let’s give it a 30, presuming Covid lockdowns, unemployment, and social media freakoutery exacerbated people’s behavior by a factor of a little over x3.
Biden Beats (someone else) will probably have less than if Biden beats Trump, call it a 2.
(Someone else) beats Biden will likely have zero rioting, since a lot of Democrats don’t like Biden, and most of their rioting has purely to do with how much they hate Trump.
Trump beats (someone else) will likely have a similar amount of rioting than if Trump beat Biden, since the blue rioters in particular aren’t pro-Biden, they’re anti-Trump. Give that a 30.
(someone else) beats Trump will have at most the same value as if Biden beats Trump, so call that a 3 as well.
If neither Biden nor Trump are running, I don’t foresee any rioting.
We calculate expected value by doing a weighted average, like this:
What do these numbers mean? If you are an insurance adjuster or police captain attempting to prepare for election aftermath, you probably shouldn’t be worried about a 2020 level event. Across all outcomes, the average damage, death, and anarchy should only be about 14.15% of what we saw in 2020, with most of the heavier outcomes concentrated in two possible eventualities. If we as a country can keep it under 700 demonstrations, 2.8 million protesters, 2,000 arrests, $283 million dollars in damages, and 3 deaths, we’ll have beaten the odds.
If you disbelieve the results of this analysis, then you can go through the exercise yourself with different presumed probabilities, but you’re likely to land on something curiously similar. And if you become a paid subscriber, I’ll build a Google Doc to do it for you. Send me a message.