It’s been a long road with lots of drama and emotion, now it’s time to count the votes and determine the result. Most people I know are too emotional about the outcome to think about things quantitatively, which could lead them to misinterpret the news as it comes out. The key quantitative fact to keep in mind is that while this may be the hardest Presidential election in living memory to predict, the result will probably not be close.
Most hard-to-predict elections are close. If the result will come down to on-the-fence voters in a few swing states then polls, expert opinions and betting markets can’t tell us much.
But 2024 is hard to predict for a different reason, that we’ve never had a similar election and the previous election in 2020 was too strongly affected by COVID to be much of a guide. Voting rules in 2024 are quite different from 2016 and earlier elections, and there have been large demographic and opinion changes in the last eight years. Donald Trump is a unique candidate, and while Kamala Harris is more conventional, her path to the Democratic nomination is unprecedented. We don’t know who is going to vote, or how many, or how they will vote.
It's much easier to predict volatility—or in this case the number of Electoral votes the winner will get—than the identity of that winner. This is a general quant point. Since the Great Depression, the real annual return on the S&P500 has ranged from -41% to +84%, while annual volatility has stayed between 3% and 32%.
In 2016 and 2020, the Presidential winner got 304 and 306 Electoral votes. Looking at polls, expert forecasts, models and betting market prices, my estimate is there’s only about a 30% chance that the election will be this close. I think there’s about a 40% chance that Donald Trump will win with more than 306 Electoral votes and a 30% chance that Kamala Harris will win with more than 306 Electoral votes.
This should change the way you watch the results as they come in. If you expect the election to be close, you’ll assume that each candidate will win their safe states. By most tallies that’s 226 safe Electoral votes for Harris, 219 for Trump, with 93 up for grabs in seven swing states (Michigan, Wisconsin, Pennsylvania, Nevada, North Carolina, Georgia and Arizona). The election will be determined by the swing states, all of which will be close, meaning we may not know the winner until very late—even before the inevitable challenges, arguments and perhaps recounts over days or weeks.
While a close result determined in the swing states is certainly possible, about 30% in my guess, I think it’s more likely that the swing states will swing one way and some of the shakier safe states will be less than safe. If Harris looks like she even has a chance at taking Florida, Alaska, Ohio, Iowa or Texas, she could be looking at a landslide win. If Trump is competitive in Virginia, Minnesota or New Mexico, he may get a popular mandate to make America great again.
It might be easier to understand this in terms of the popular vote, even though it doesn’t matter for the election result. Polls predict Harris will win the popular vote by about 2% which, due to Republican advantages in the Electoral College, translates to a near dead-heat in Electoral votes. But there’s wide uncertainty about that number. If Harris’s lead comes in at 5% or higher, she will likely win more than 306 Electoral votes. If Trump can win the popular vote by 1% or more, he will likely win more than 306 Electoral votes. I estimate that the chances of one of those two events at 70%. If the popular vote comes in between +1% for Trump and +5% for Harris, then we’ll likely have an election as close or closer than 2016 and 2020, one that will be decided in swing states and not known until Wednesday morning or later.
So quants will be watching two things—whether candidates are outperforming polls in the popular vote in early reporting states even if those states are solid for one or the other candidate; and the early-reporting semi-swing states like Florida, Ohio and Virginia. Exit polls and network calls will be useful as well. If there are early calls in swing states Georgia and North Carolina, those could be very significant signals for the winner. If there are delays in calling Florida, Ohio and Virginia, the first two could be very bad news for Trump, the last very bad news for Harris.
I have no idea who will win the 2024 Presidential election, but I would like to see a convincing win by one or the other candidate. Americans have a stark choice—although most people I know would prefer a none-of-the-above option—and it would be nice if we can make it.
