Written by Scott Rasmussen
One of the most controversial things you can say in the heat of this election season is that—at this point in time— nobody knows who is going to win.
The reason it’s so controversial is that fans of both candidates are extremely confident their team will win. Ninety-four percent (94%) of Harris voters think she is likely to win and 89% of Trump voters say the same about their own candidate. Solid majorities on both sides see victory as very likely.
Adding to the problem is the fact that hardly anybody on either side can imagine how a reasonable person could even consider voting for the other candidate. Both assume that as election day draws closer that reality will help drive their team to victory.
The reality, however, is that both professional and armchair analysts are placing too much certainty on the precision of polls while acknowledging too little unpredictability on the impact of future events.
With that in mind, here are five good reasons why it’s impossible to know who will win at this time.
The National Polling shows a race that is too close to call. As I write this, the Real Clear Politics average shows Harris up by two points. In 2004, 2008, and 2016 that average was very close to the actual results. However, in 2020, that average overstated Biden’s margin of victory by 2.7 points. Eight years earlier, in 2012, it understated Obama’s margin of victory by 3.2 points. In other words, based upon recent history we could anticipate the average being off by up to three points in either direction (a range from Trump +1 to Harris +5). That sounds like a reasonable margin of error to apply.
The Swing State Polling confirms that the race is too close to call. The RCP average shows the race to be within two points in all seven swing states. In other words, if they are all off two points in the same direction, either Trump or Harris could win all seven states and a significant Electoral College victory. Those who mistakenly assume that polling error can only go in one direction are making the same mistake that Election Forecasters made in 2016. They assumed any polling error would benefit Clinton.
The Electoral College Impact is impossible to quantify. People on both sides of the great political divide talk about the GOP’s rumored Electoral College advantage. That’s primarily because Trump lost the popular vote by two points in 2016, but won the Electoral College. The reality is that the decisive states in both 2016 and 2020 were decided by extremely small margins. For example, Trump won Michigan in 2016 by 11,000 votes out of 5 million, Biden won Georgia in 2020 by 12,000 votes out of 5 million. No poll or other analysis can predict such narrow margins.
Big things in the real world could shake up the race. It is possible that truly unexpected events—and the candidates’ reactions to them—would shift the race in one direction or another. What might such an event be? None of us can know because they’d be unexpected!
Small things in the real world could decide the race. With close races, anything can happen. A heavy snowstorm in Michigan or Wisconsin—or the impact of Hurricane Helene in North Carolina—could impact turnout. A court ruling on whether undated mail-in ballots should be included might be the difference in Pennsylvania. The list could go on and on. If the race is really as close as it appears, anything could make the difference.
As we look ahead, it’s possible that things will shift and one candidate will emerge as more likely to win. But it’s also possible we won’t know who is going to win when we wake up on Election Day.
If the election does remain too close to call, there are only a couple of safe predictions. First, the winning side will be frustrated by the fact that their narrow victory doesn’t give them limitless power to reshape the nation. Second, the losing side will think they were robbed and refuse to accept the legitimacy of the new president.
Scott Rasmussen is the President of RMG Research, Inc. and founder of the Napolitan Institute.