Wanna Bet? Trump v. Biden Edition

Wanna bet on who will win the election this year?

Indeed, many do. The presidential election betting market is actually quite large. What’s more, you can use the betting market to gauge the probability – at least, according to those putting their actual money on the line – of who will win the election. Put simply, you can derive this probability by how much it costs to bet on a given candidate.

For example – with Biden, you have to bet more money than you’ll actually net as winnings because he is viewed as the more likely candidate to win. As in, you bet say $185 on Biden to make only $100 if he wins. For Trump, it’s just the opposite. Wager a mere $55 and you’ll win that same $100 if Trump claims victory.

But this is where it gets interesting. According to betting markets, Biden has about a 65% chance of winning (and Trump, thus, 35%). However, Biden leads Trump in national polls by a whopping 9.2 points! This is 3 points higher than Hillary Clinton polled nationally at this same time leading up to the 2016 election. And yet, betting markets were giving Hillary an 88% chance to win at the time in 2016.

So what gives? It certainly seems that betting markets are keen to the fact that polls didn’t do all that well in predicting the election outcome in 2016.

Does Trump really have a chance – and hence, the betting markets are arguably more accurate than the national polls? Or is this clearly going to be a landslide for Biden as the national polls suggest? Just how close might this race be? Let’s dig into some of the key metrics that will help us examine this.

The Lay of the Land

Perhaps the best way to go about understanding the way this year’s election might shake out is to actually drill down to the state level and take a look at which candidates will win which states.

Let’s start with the lay of the land and carve out states that almost undoubtedly will lean Democratic or Republican. According to RealClearPolitics, the US electoral voting map essentially looks like this:

216 votes going to Biden, 125 votes going to Trump, with 197 votes still up in the air as “toss-ups,” otherwise known as battleground states.

Battleground States

You might be wondering – are there really this many toss-ups? Probably not. Let’s next carve out states that aren’t highly likely to be battlegrounds. We’ll attribute states to each political party where there was enough margin of victory for the candidate in 2016 to reasonably assume we can give their party the win again this year.

Take Texas, for example. Is Texas really a toss-up? In 2016, Trump led the polls in Texas by 5.4 points. He ended up winning the state by 9 points. Now, in 2020, he leads again by almost 5 points. Fair enough to give Republicans the win in that state? For the sake of our hypothetical analysis and general reasonableness, let’s do that. It isn’t a perfect technique, but it will make things a bit simpler and more understandable here. This method ends up giving more states to Trump, but in fairness, it’s simply being attributed according to the most recent historical data.

After giving Republicans and Democrats presumed wins in states that might not truly be toss-ups, we have a map that looks like this, with only 6 states remaining as true battlegrounds:

At this point, we’ve got Biden with 234 electoral votes, and Trump with 203. Out of the 538 available electoral votes from the entire US, candidates need 270 to win. With 101 up in the air as true toss-ups, at the very least it seems like it could be either candidate’s game. Ok, so maybe the betting markets are on to something here. Maybe it’s not an obvious Biden landslide like so many expected a Hillary landslide, or as the national polls now seem to suggest.

Next let’s move to some more interesting data. Let’s examine the polling and outcomes in these 6 key toss-up states as it relates to the 2016 election. Here’s a look at that data from RealClearPolitics (you might want to jump down to the text below for the explainer!):

There’s a lot of information here, so let’s break it down. For a given battleground state, this chart first shows the current polling for Biden versus Trump. The “RCP AVERAGE” is the spread by which one candidate leads over the other in the current polls. For example, in Pennsylvania, Biden leads by 6.5 points in the polls (Biden at 50.2, Trump at 43.7). Notice that in each state, Biden leads at the moment, just as Hillary did in each of these states’ polls in 2016.

Next to this is the polling from this same time leading up to the 2016 election (“FOUR YEARS AGO”), and then the outcomes of each state in the 2016 election. (Note that the number in parenthesis next to each state’s name is the number of electoral votes that state has to offer).

What’s first striking is that, compared to the polls at this same time in 2016, Biden actually has a smaller lead on average amongst all of these battleground states. Here’s a look at that statistic in graph form over time:

You’ll notice that Biden leads on average in the battleground states by 4.9 points at present. At the same time in 2016, leading up to the election, Hillary led in these same states by 5.4 points on average. Again, this seems to indicate that this race might be tighter than national polls suggest.

Even more to the point – Trump won each and every one of these battleground states in 2016. It certainly seems like a Trump victory is not out of the cards at this point.

What Does This Battleground Data Mean?

Let’s dig in to the battleground state data a bit further. What is the likely outcome here? If this looks anything like 2016, Trump will carry each of these states again, and ride off to victory with 304 electoral votes.  

Alternatively, we could chalk a win up for Biden in each of the battleground states where he currently has a larger lead over Trump in the polls than Hillary did in 2016. By this method, Biden would take Pennsylvania, North Carolina, and Arizona, ending up with 280 electoral votes and cementing a Democrat victory.

But again – 280 electoral votes is a razor thin margin of victory – only 10 votes over the threshold. And Trump won Pennsylvania in 2016 by more of a margin than Biden has gained over Hillary’s prior lead in that state – meaning Trump could easily still take that state again as well. In that case, Trump would again be victorious, having captured the battlegrounds of Pennsylvania, Florida, Wisconsin, and Michigan, and ending up with 278 electoral votes.

The point here, of course, is that this election seems to be much closer than many believe, particularly if you’re looking at the national polling data. The betting markets seem to be in the know here, as they so often are.

A Few Caveats

It’s worth noting a few caveats to the analysis here. First off, there is certainly no guarantee that the state victories I attributed pursuant to prior victories in 2016 will play out that way. But that’s obvious.

Perhaps more interesting, and less obvious, is that there seem to be fewer undecided voters going into this election. This would seem to suggest there is less room for large surprise results that deviate significantly from the polls. Take the battleground of Michigan, for example. Hillary appeared to lose an over 11 point lead there in 2016. This year, while Biden’s lead seems smaller, fewer undecided voters could mean this lead is actually a better reflection of reality. There simply isn’t room for something like an 11 point surprise swing in Trump’s favor. This may well be due to the fact that Biden is largely a more agreeable candidate than Hillary was. Beyond being the first female president, perhaps itself a hurdle in the eyes of some, Hillary had a uniquely entitled approach to her bid for the presidency. As if she’d already won before the vote (although – can you blame her based upon her polls?). As the well-mannered and likeable old grandpa type, Biden seems to be less polarizing than Hillary was, and could well ride this likeability all the way to victory.

So, have you placed your bet?