CNN
—

Nearly two weeks before Election Day, I’m more uncertain about this year’s outcome than any election I’ve covered professionally. That’s partly because of the polls – they’re really close – but it’s also because for every good signal for Donald Trump, there seems to be a good signal for Kamala Harris.

Many Americans believe there is a lot at stake in this election. And yet, to me, it’s still a race with multiple possible outcomes – from a clear Harris victory, to a contest that can’t be predicted until late on election night (or week), to a decisive Trump victory.

Let’s start with a simple suggestion: Harris’ easiest path to victory is by winning the blue wall states of Michigan, Pennsylvania and Wisconsin. If she loses in the Sun Belt battlegrounds (Arizona, Georgia, Nevada and North Carolina) and carries the three Great Lakes states along with Nebraska’s 2nd District and all the other states Joe Biden won in 2020, Harris would exactly Reach 270 electoral votes.

Polling averages in Michigan, Pennsylvania and Wisconsin show that the gap between Harris and Trump is currently less than one percentage point. If that remains the case through Election Day, it would be the first time in at least 50 years that any of these three states had final average margins within one point.

Additionally, margins in these states have been consistently tight since Harris entered the race in July. Neither Trump nor Harris has ever led by 5 points or more in any of these states, reflecting national polls. It is the first time in over 60 years that no candidate is leading by 5 points or more statewide at any point in the race.

Many Republicans hope the close polls point to a landslide victory for Trump next month. The former president performed significantly better in the 2016 and 2020 polls. Should Trump do this again, he would likely claim victory and receive over 300 electoral votes.

However, I would be cautious about assuming that Trump would not benefit from the election. Since 1972, there have never been three presidential elections in a row in which the same party benefited from a failure in the national election. In fact, polls in key battlegrounds in 2022 have underestimated Democrats. If we had an electoral error like we did two years ago, Harris would likely win over 300 electoral votes.

The polling averages for the battleground states are far from perfect. The average miss since 1972 has been 3.4 points, and 5% of the time they miss by more than 9.4 points. Even an average failure in key battlegrounds could turn the election into a bust.

You might be tempted to use macro trends to guess the direction of a possible poll error. Biden’s approval ratings are terrible. No incumbent’s party has ever won another term when the president’s approval rating is this low. No incumbent party has ever won another term, despite so few people believing the country is heading in the right direction.

But Trump may be exactly the wrong candidate to take advantage of these structural advantages. If he wins, he would be the second-least popular candidate since pollsters began measuring candidate popularity in the mid-20th century. The only presidential winner who was less popular was Trump himself in 2016. Consider also that Republicans underperformed in the 2022 midterm elections, despite many macro indicators pointing in their direction.

Complete the party registration. The trends are not as clear as they might seem at first glance. Over the past four years, Republicans have won over Democrats in every key battleground state. This, combined with national party identification trends, would normally lead to Republicans winning this year.

However, it is not clear how many registered Republicans will vote for Trump. It’s possible, as the latest New York Times/Siena College Pennsylvania poll shows, that Harris will win by a higher percentage among Democrats than Trump will win among Republicans. Since there are more registered Democrats than registered Republicans in the Keystone State, such a result would mean Harris most likely wins Pennsylvania.

What’s even more interesting is the fact that we have a close race, even though many Americans have changed their voting behavior compared to four years ago. Trump appears destined to deliver one of the best performances among black voters in many years as a Republican presidential candidate. This is especially true for young black men.

Still, Harris appears to be doing better with white women than any Democratic presidential candidate this century. While her gains among black voters are not as large as Trump’s, white women make up a much larger share of the electorate. Therefore, these shifts can largely cancel each other out.

This means the election likely hinges on the few voters who are still undecided.

More than two-thirds of likely voters believe this is the most important election of their lives, including 72% of Trump supporters and 70% of Harris supporters. The 5% of voters who are currently undecided will decide which of them comes out of this election happy.

Ironically, only 24% of these undecided voters agree that this election is the most important of their lives.

How painful it must be for those who truly care about the election to have it decided by many who don’t care.