Harris in the Lead
Fueled by Tight Polls in the Southeast
August 31, 2024
Uh oh. Harris takes the lead.
For the first time this election cycle, a Democrat leads in the odds to win the White House, albeit 52% to 47% – a statistical tie.
But Trump still leads in the Electoral College?
It does not exactly take an eagle-eyed observer to notice that Trump still maintains a lead in the electoral college…, so how can he be trailing in the electoral college odds?
With Harris’ having two-to-one odds in Michigan and pretty good odds in Pennsylvania1 and Trump having two-to-one odds in Nevada and stellar odds in Arizona, the race as it stands turns on the other three swing states. Each of which are approximately 50-50. With Harris winning Michigan and Pennsylvania in as many simulations as she is currently, she only needs to win one of these three remaining states in order to win the electoral college, but they all three lean slightly in Donald Trump’s favor, which means they are added to his electoral college tally even though they are far from a sure bet.
If we were to nudge the Georgia odds by 0.2% in Harris’ direction. It would flip the electoral college count, but have an unnoticeable effect on the odds to win the electoral college. That is because such a small change – even in a very contested state – does not greatly reshape the odds overall.
Two-to-one odds is not a guarantee in any of these states, but hopefully that explanation makes it more clear as to why she has an ever so slight edge in the odds while he still leads in the electoral college count.
Southeastern Shake-Up
It wasn’t Harris’ lead in the national popular vote that did that. Rather, it was polls narrowing in Georgia and North Carolina that cratered Trump’s lead.
Trump’s already narrow lead in North Carolina waned in recent days with the release of the Fox and Emerson polls. Both polls showed a 1-point race, with a different candidate leading in both surveys. However, the adjusted margin – the adjustment made to correct for overrepresentation of political parties in a poll – brought both of the margins to 0.1% and 0%, respectively. These two polls, along with the three others released earlier this month, come together to give Trump a 0.7% lead in the Tar Heel State. This state is critical for Trump, as he relies on it being in his column in many of his narrow victory scenarios. He needs it for the NC+GA+PA path. He needs it for the NC+GA+AZ+WI path. He also needs it for the NC+PA+WI+NV path. Losing here would put Trump’s odds on election night in dire straits.
Thankfully for Trump, pollsters are horrible at polling North Carolina. If you scroll down to RealClearPolitics’ polling history for North Carolina in 2016 and 2020, you’ll notice that polling does not start to even consistently approximate what would end up being the final outcome in North Carolina until the last couple weeks before election day. Even a week out, we were still getting Clinton +6 and Biden +6 polls. And both times, the aggregate managed to underestimate the former president.
Georgia went from having Trump lead around 3-4% to it being separated by 0.1%. This follows two strong Georgia polls released for Harris this week, each of which show her with a small lead. That lead actually grew when corrected for overrepresentation, so that actually her adjusted margin (the one that gets used by the simulations), is nearly double.
The curious thing about these two polls is that they both show the same alarming (and probably false) thing: Harris leads Trump with independents by double digits (15% and 16%).
The three other polls released in Georgia this month show a competitive race for independents. The New York Times had them tied among them, and Emerson and Fabrizio showed Harris with a 5-point lead – something easily covered by the sub-group margin of error.
This is a different animal, however. And one that I suspect is not reflective of Georgian voters. In 2020, a particularly down year for Republicans in Georgia one might say, Trump lost independents to Joe Biden by 9% according to CNN exit poll estimates. Call it what you want, but this was in the middle of a pandemic and BLM riots. The country, stirred up by Democrats and the media2 for political gain, was chaotic and that inculcates negative attitudes towards incumbents.
The state’s not what it used to be, but independents in Georgia are not voting to the left of independents in Nevada, Arizona, Wisconsin, Pennsylvania, Michigan, and Minnesota. Either all of those polls are wrong or these two or wrong, and I know which way I am leaning.
Even if the two candidates split independents or if Harris has a narrow edge, Trump can still win the state by a couple percentage points if Republicans turn out like they did in 2020.
The concerning thing for both candidates actually is that polling in the aggregate has closely approximated the final outcome in the state. The final eight polls in 2016 showed Trump with a lead of 4.1% while the final seven polls in 2020 showed Trump with an even 1% lead. Both times, the polls actually underestimated the winner, but the difference in the polling aggregate compared to the final outcome is pretty respectable considering just how off they’ve been in other states like Wisconsin and Michigan.
All that is to say, that massive lead with independents will not pan out. If it was going to, we’d see that kind of lead everywhere – it wouldn’t be localized. Independents in one place are not that different from independents in another. But, if come mid-October and we’re still seeing a narrow one-to-two percent race in Georgia, that’s probably real, and you should count on a nail biter in the Peach State. This isn’t Wisconsin, folks – we don’t get 8-point Democratic leads that manifest on election day in 1% wins or losses.
Arizona and Kennedy’s Endorsement
Lastly, I want to mention Arizona. The impact of RFK Jr. removing his name from the ballot here and endorsing Trump cannot be understated. Like many of its western neighbors, Arizona has an affinity for third party candidates. It had one of the highest rates of support for third parties in 2016 – 7.3%. This is also a state where independents boast a plurality in voter registration. Third party candidates, led by RFK Jr., were regularly clearing 10% of the vote, cresting even north of 20%. Third party support does fade as an election cycle goes on, but when Harris entered the race, we saw a marked downward shift in third party support. Much of that movement, as we have seen in the national polling, did go to Harris.
However, we now have our first poll that was fully conducted after the Trump-Kennedy alliance formed. That poll from Emerson showed support for independents plummeting once again – now, down to less than 2%. And in that poll’s adjusted margin we now see support for the Trump ticket that rivals his margin in the state from 2016. That nearly 5% lead is monumental and is consistent with the shift in neighboring Nevada, another state that’s rather fond of third parties.
The southwest is shaping up well for the former president at the moment while the southeast shows signs of difficulty, but considering the polling shortcomings discussed there, the trouble may be fleeting.
Looking Ahead
It is worth noting that 6 of the 7 core swing states are at or below a 2% separation. We have seen bigger polling misses in recent years to cover that kind of spread. None of those states – or even Arizona with its Trump +5 lead – is truly out of reach for either candidate.
We still have a long way to go until election day. Two – maybe three – debates. Many more Kamala Harris interviews, I’m sure.3 A (potential) sentencing hearing. And who knows what other fun electoral surprises this year has in store for us.
Two full months remain in this race as we close out August. To put into perspective all the crazy things that could happen in the next two, just look at the last two: the former president is grazed by a bullet in an assassination attempt, the former president drops out of the race, a major party selects a new nominee without a single vote cast, and a Kennedy endorses a Republican.
Strap in!