Polling and the Weather Forecast: A Lesson in Uncertainty

Polling is like the weather forecast: it offers a range of outcomes. It may seem unlikely to rain where you are, but it’s still possible, just as somewhere nearby could stay completely dry. Polling models work the same way, providing a range of possibilities, with the most likely outcome somewhere in the middle. However, journalists often don’t report or explain this range, and some polling companies don’t fully address it either.

When the herd (the electorate) moves, it usually moves together—yet occasionally, it splinters, adding layers of uncertainty. This unpredictability is what allows people to sometimes defy the odds and come out on top. The margin of error means the actual result could land on either side of the median outcome or even on the outer edges of what’s possible. Think of it as science fiction: in The Hitchhiker’s Guide to the Galaxy, when travelling across space, you might theoretically turn into an apple. Highly unlikely, but technically possible!

What Do the Current US Election Models Tell Us?

So, how does this translate to the current U.S. election landscape?

The current model from FiveThirtyEight has Harris at 268 electoral votes and Trump at 270. Just a few days ago, it was the other way around. It’s a remarkably close race, one we haven’t seen for decades. In practice, this means the herd is taking different paths, with some states following separate trajectories from others. Why? This could be due to local issues or unique demographics at play, which are often difficult to model accurately.

Certain demographic and local factors influence swing states differently, adding complexity to polling models. For example, Arizona’s growing Latino voter base leans conservative in some areas, potentially benefiting Trump. Meanwhile, Pennsylvania’s situation is unique, with its declining manufacturing jobs and large Arab population possibly making it more competitive. In other states, recent events or changing demographics play a big role; a hurricane’s impact on voters or a high Black population could sway support towards Harris, especially given her profile as a Black female candidate. Each of these factors highlights why swing states can defy broader polling expectations and swing differently depending on local issues.

The Swing State Landscape and Possible Outcomes

Based on current numbers, swing states could align—or they might diverge entirely. Swing voters typically follow similar trends, but today’s models suggest they’re all over the place. The final result could be a landslide, a narrow victory, or a razor-thin margin. The most unlikely outcome is a tie, which would be decided by Congress. The only time this happened was in 1800, between Thomas Jefferson and Aaron Burr.

In practical terms, we could see anything between 320 and 270 electoral votes for either side, making predictions a toss-up due to how close it is. And despite the possible range of outcomes, most commentary doesn’t explore these diverse possibilities in-depth, though they should.

Swing states, which often switch between the two main parties, are pivotal in deciding elections. These states can shift as demographics evolve—much like music tastes changing over time. Currently, there are seven key swing states: Arizona, Georgia, Michigan, Pennsylvania, Wisconsin, North Carolina, and Nevada. Pennsylvania holds the biggest prize in terms of electoral college votes, which is why it’s often rated the most crucial.

At the moment, Harris has more paths to the White House than Trump. In fact, nine out of twenty-one possible routes for Harris don’t require Pennsylvania, whereas only five out of twenty-one for Trump don’t need Pennsylvania. Care to guess the tipping point state?

Current polls suggest a 51-56% chance of Trump winning and a 49-45% chance for Harris. This modelling assumes only seven states are in play. I’d say Harris has the advantage based on the seven swing states, yet betting on either side winning is difficult due to the close margins. The only thing we know for certain is that results may not be known for days if trends follow recent averages. The final outcome will either take one candidate to the White House or the jailhouse—but we’re likely heading to the courtroom first.

In a race this close, every voice truly counts. So, whatever the forecast, cast your vote—because, in the end, even a small shift in the herd can tip the balance.