Harry Enten is a senior political data reporter for CNN and the host of the ‘Margins of Error’ podcast, where he specializes in data-driven journalism.
The 2024 U.S. presidential election continues to be the closest of the century. In fact, it is the closest race for the White House in the past 60 years.
Polling since the Sept. 10 debate between Donald Trump and Kamala Harris shows that while the U.S. vice-president seems to have opened up a slight national edge over her Republican rival, their race remains well within the margin of error and too close to call. This is especially the case when looking at the Electoral College.
Consider the polling that came out Sunday from CBS News and NBC News. Their surveys were some of Harris’ best to date, and yet she is only ahead by four and five points, respectively. The Democratic nominee’s largest leads in the CBS News/YouGov and NBC News polls in 2016 and 2020 were at least double where Harris is now.
To put the new Sunday polls into further context, consider all the national surveys conducted since the debate. This includes the aforementioned polls and surveys from ABC News/Ipsos, Fox News and The New York Times/Siena College. On average, according to the latest CNN Poll of Polls, Harris is ahead by three points.
This matches what we’ve seen all year: Neither candidate has been able to open an advantage of five points or more in the national polling. This includes the period when U.S. President Joe Biden was the likely and then the presumptive Democratic nominee.
The fact that no one has led by at least five points this cycle is noteworthy because it’s incredibly rare. Even in races that end up being very close, one candidate at some point almost always builds a significant advantage. This year, most voters seem locked in.
Even Harris’ dominating debate performance over Trump – according to the voters – has only appeared to move the dial by a few points.
You’d have to go back to the 1960 campaign to find a race in which the major-party nominees were always within five points of each other in an average of the national polling. Every presidential year since then has had at least three weeks when one candidate was up by 5 points or more.
A 3-point edge in the national polls is far from secure for Harris. Since 1948, the average difference between the polls on the eve of the election and the Election Day result has been 3 points. Some years, like 2020, the error rate is even higher.
(This far out from the election, the average difference between the polls and the eventual result would be, not surprisingly, greater.)
Electoral College math
But perhaps the more important reason this election is too close to call is that this isn’t a national election. Instead, it is a race to 270 electoral votes through the Electoral College.
Trump is likely to be in a better position in the Electoral College than the popular vote because of his coalition (e.g. white voters without a college degree are overrepresented in key battleground states). One estimate from my old colleague Nate Silver suggests that Harris would need to win the popular vote by greater than 3 points to be considered a clear favorite in the Electoral College.
She isn’t there yet.
Indeed, neither Harris nor Trump has a big leg up when you look at the state-level data. Per CNN’s current race ratings, Harris starts at 225 electoral votes to Trump’s 219. Seven states and the one electoral vote in Nebraska’s 2nd Congressional District remain up for grabs.
Harris seems to be doing slightly better than Trump in three of the seven states: Michigan, Pennsylvania and Wisconsin. This northern battleground path is similar to the one that Biden’s campaign had hoped to achieve back in the spring.
But when I say Harris does “slightly better” than Trump, the emphasis is on the word slightly. Harris is polling about two points above Trump in all of them.
We’re talking about races well within the margin of error and with no clear leader.
Meanwhile, Trump does slightly better than Harris in two of these states: Arizona and Georgia. But like with Michigan, Pennsylvania and Wisconsin for Harris, Trump is doing a point or two better, on average, than Harris in the polling from these two states.
If you assigned the electoral votes to the candidate with a greater than one-point advantage at this point in the polls, Harris would be at 269 to Trump’s 246.
Nevada and North Carolina are within a point and way too close to call, just like the other five states. But for the sake of this exercise, we’ll give Trump North Carolina, a state he’s won twice before and where the polling average has the former president polling above Harris by mere decimal points. Under this scenario, he would get to 262 electoral votes.
The limited data we have about Nebraska’s 2nd District indicates that Harris is favored there (the Cornhusker State is one of two states, along with Maine, that splits some of its electoral votes by congressional district). Biden would have won the current incarnation of the 2nd District by 6 points in 2020 – a significantly wider margin than what we’ve seen in the seven battleground states this year. Most modeling, as well as the betting markets, has Harris ahead in this Omaha-area seat.
A win in Nebraska’s 2nd District would likely give Harris exactly 270 electoral votes when added to Michigan, Pennsylvania and Wisconsin. Talk about barely getting by!
There is a catch here, however. Some Nebraska Republicans, urged on by Trump, want to change the state’s method of allocating electoral votes to a winner-take-all format.
No Democrat has won Nebraska on the presidential level since 1964.
If such a last-minute rule change were to happen, that would bring Trump up to 263 electoral votes to Harris’ 269, and the election would then come down to Nevada and its 6 electoral votes. The last published poll that meets CNN’s standards for publication was our own poll conducted by SSRS last month that found Harris at 48 per cent and Trump at 47 per cent, well within the margin of error.
In other words, a Trump win in Nevada is quite plausible and would leave us looking at a 269-269 tie.
That would throw the presidential race to the U.S. House, where each state delegation will get one vote. Trump would likely be favoured under that scenario because Republicans are more likely to continue holding more state delegations in the chamber than Democrats come January.
Regardless of who wins Nevada, we could be waiting a while on the Silver State to count its ballots. And given how long that has taken in past close races, we could be waiting days, with the presidency in the balance.
The bottom line is that this year’s presidential race is as close as it could be. One tiny shift in either direction could make all the difference in the world.