2024 Election Life and Death Game Theory: Introduction


Assumption #1: Voting is an affirmative action. You have to be willing to do it. When I say, “if it’s raining, I’m not going to waste my time voting for Joe Biden and Tammy Murphy,” it will not change the outcome of the Presidential or, probably, even the New Jersey Senate election. But the turnout will be lower than a sunny day would be. Similarly, “if it’s only drizzling, I’ll probably vote for Andy Kim, at which point I might as well vote for Biden,” sets another level. The number of contests, the preference for the candidates, the weather all affect voter turnout because voters have to want to vote for or against someone or something. (See, for instance, the bid to get a measure on the Florida ballot to make women’s reproductive health care something that the state constitution guarantees.)

This is one of the key reasons that Dr. Black keeps noting that “orange man bad” is a lousy GOTV strategy. DJT’s supporters especially want to vote for him. Dave Chappelle called him “an honest liar” and then explained:

Dave Chappelle Explains It All to You

Don’t get me wrong. That same logorrhea motivated people against Trump in 2016, 2018, 2020, and even 2022, and it will motivate again in 2024. Even as the White voters (especially not-so-poor ones) showed up for him–and promise to do so again–Trump motivated people to vote in both directions.

Last Time Was Different

Donald Trump’s total votes in 2020 would have won any previous election by a wide margin. (Data for the next few paragraphs can be found here and at the corresponding Wikipedia page, among other places.) The previous record was 69,498,516, for Barack H. Obama in 2008; Trump’s total would give him over 51.5% of the vote, all else equal. Obama’s re-election in 2012 and Hillary Clinton’s 2016 total are the second- and third-highest totals before 2020, both under 66,000,000. (Willard M. Romney was the first losing candidate to break 60 million votes; John McCain in 2012 was a little over fifty thousand short of that, while John Kerry in 2004 was less than one million away.)

What made 2020 different? In a word, COVID, which the U.S. was still treating seriously at the time. Early voting, voting by mail, drop-off ballots, and other measures to reduce the difficulty of voting. The result was that about 27,500,000 more people voted in 2020 than in 2016. Basically an 11% increase, with voter participation going from just over 60% to 66.6%, the highest total since well before American women were legally allowed to vote. (The turnout for 2008 was, until 2020, the highest of the post-Watergate era. The people did not let a crisis go to waste.)

Setting Voter Expectations

Assumption #2: nationwide voter participation will be lower than it was in 2020. Not just because people who were motivated to vote against Trump in 2020 may be disenchanted or because there are fewer Trump supporters than they were four years ago. In several states, it will be harder to vote. Absent a crisis, people at the margin will be less motivated.

This is also where the Electoral College becomes more important. The Democrats need 270 EC votes to win. The Republicans only need 269; tie goes to the House, where voting is by state. Start from 2020, flip Pennsylvania and Michigan, keep everything else the same, and a Republican becomes President. Even if there is a greater drop in enthusiasm for Trump, the question is where as much as how many.

So from the start, we are not expecting, even with a population increase, so many voters as in 2020. I strongly suspect DJT lost more voters than Biden, even after the past five months. Certainly Nimrata Haley’s primary campaign suggests that possibility. Does Biden lose more support as the genocide in Gaza continues, especially given the planned Republican ads? To be generous, let us set Assumption #3: Enthusiasm for both candidates declines proportionately. That is, all else remains equal.

The Dying Game

What happens if all else isn’t equal? Let’s play “ask an actuary.” Given that the peak average life expectancy for a 65-year-old U.S. male has never been above 2019’s 18.2 years, there is a distinct if moderate chance that one or both candidates will die before November. At the least, the chance of one of the candidates dying before November is within two standard deviations

Biden has the best health care in the world. He certainly appears reasonably fit. However, he is three years closer to that average than DJT. A white male in his condition and socioeconomic class should live more than an average lifespan. Still, the presidency is a stressful job if you’re trying to do it right.

DJT is further under that average but in much worse condition physically; he’s certainly more likely to have a life expectancy below average. It’s not the way to bet–even odds would be that both of them live to the inauguration next year–but the chance is present and very much non-zero.

If I had to bet, I would say Trump is more likely to die before 20 January 2029, even if he were elected and Ronny Jackson rejoined the White House.

The Game Theory of Death has several possible modes. Scenarios over my next several posts which I plan to finish before that election.