As the final days of the 2024 presidential campaign draw to a close and final polls are published, the race looks to be too close to call. That may or may not be true.
Advertisement
The “too close to call” CNN pollster Harry Enten best exemplifies school. He shows that Harry Truman’s search for a one-armed economist also applies to pollsters. Here we have the classic on the one hand, we have this:
But on the other hand, we have this:
One of Reagan’s pollsters says the race isn’t close at all; see Why the race isn’t as close as you think: With one week to go, analyst CRAIG KESHISHIAN predicts the polls are missing a hidden voter surge | Daily Mail Online.
There are some new polls out today that further muddy the water. A new Quinnipiac Poll, a poll that is usually Democrat-friendly though not as slavishly so as Morning Consult, Has Trump moving ahead in Pennsylvania and nationally.
Advertisement
Another by CNN shows Trump tied nationally but Harris with solid leads in two of the three states that will probably decide the election.
It boggles the mind to try to comprehend how Trump is tied nationally, tied in Pennsylvania, and down by five in Wisconsin and Michigan.
Instead of being whipsawed by data points, I think we can learn a lot more by looking at macro trends. Macro trends are easier to identify (for instance, we’re all going to die) but the micro trends in the larger picture are much more difficult (will I die in an incident involving hookers and lines of blow?). While CNN’s Enten talks about ” No party has ever been underestimated three presidential cycles in a row,” he ignores the unique laboratory that we have at our disposal: for the first time in the history of polling, we have the same candidate running for the same office in three presidential elections. I think Enten also goes off the rails in conflating off-year and presidential elections. The two events are not similar in national intensity.
National Polls
First off, let’s take a look at the big picture.
As of this writing, Harris leads by 1.3 points in the 538 average, and Trump leads by 0.4 points in the RCP average. That looks like a tie, but is it?
In 2016, at this time, Clinton led national polls by 3.2 points, and the final count had her up by 2.1 points, but she lost the Electoral College vote. Trump overperformed by 1.1 points.
Advertisement
The 2020 campaign had Biden leading Trump in the final polls by 8.4 points in the 538 average and 7.2 points in the RCP average. The final count gave Biden a 4.5-point victory, but Trump still overperformed the polls by about 3 points.
Trump’s RCP lead in the polls is the first time a GOP candidate has led in national polls at this point in a presidential race since 2004. Given Trump’s history of underperforming in the polls, his lead is probably about 1.5 points greater than the averages.
Swing State Polls
National polls are fine, but as Hillary Clinton found out the hard way, running up a lead in New York and California is pretty meaningless if you can’t hit 270 in the Electoral College. Barring the total collapse of one side or the other, this race will be decided by a handful of states.
Georgia
Both 538 and RCP have Trump leading in Georgia by 1.8 and 2.3, respectively. In 2016, Trump led the final RCP average by 4.8 points, carrying the state by 5.1. In 2020, Trump was down by 1.2 points in the 538 average and led the RCP average by one point. Biden “won” the state by 0.3 points. Currently, Trump leads the 538 average for Georgia by 1.8 points and the RCP average by 2.3. Given the history, this is probably pretty close to how things will end up, even if I think Trump will go much better than a 2-point win; see Latest Georgia Poll Spells Doom! for the Harris Campaign.
Michigan
2016 marked the first time a Republican had won Michigan since GHW Bush’s drubbing of Michael Dukakis. The final RCP average showed Clinton leading by 3.4, but Trump eked out a 0.3-point win. In 2020, the 538 average forecasted Biden by 7.8, and RCP predicted a 4.2-point Biden win. Trump lost by 2.8 points. The current 538 average has Harris up by 0.8, and RCP shows a 0.3 lead. Past performance indicates Trump will do about 3 points better in Michigan than the final averages.
Advertisement
Nevada
Nevada probably isn’t a “must win,” but it is definitely a “nice to have.” A strong Trump win in 2024 could herald a realignment of a Democrat fiefdom into a reliably Republican state. In 2016, Trump lost the state by 2.4 points despite the RCP average having him up by 0.8. In 2020, he had the same margin of loss, but this time, 538 predicted a 5.3-point Biden win, and RCP called it exactly right. The present average has Trump leading by 0.3 at RCP and Kamala ahead by 0.1 at 538. All the on-the-ground reports and voting data indicate that Trump will win the state. I think this one is too close to call.
North Carolina
Even though Hurricane Helene devastated the GOP heartland of Western North Carolina, that may have served as an in-kind campaign contribution to Donald Trump. Right now, Republicans lead Democrats by over 40,000 votes in early voting, and Kamala’s campaign seems to be shifting money out of the state.
So, what does the polling look like?
Donald Trump led by 1 in the final 2016 RCP average and won the state by 3.7 points. In 2020, he trailed in the 538 average by 1.8 points and was 0.2 points up in the RCP average. He won by 1.3 points. Currently, he’s ahead by 1.1 at 538 and 0.9 at RCP. Trump is probably leading by 3 in real life. The Harris campaign knows that, and that’s why they are taking money out of North Carolina.
Advertisement
RELATED:
Code Red at Kamala HQ With Worries of North Carolina ‘Slipping’ Away (and Here’s Why)
Update: More Compelling Evidence That Kamala Harris Feels Like North Carolina Is Slipping Away
Pennsylvania
They don’t call Pennsylvania the “Keystone State” for nothing. If, as expected, Georgia and North Carolina go Red next Tuesday and there are no unexpected losses, a win in Pennsylvania would put Trump back in the White House. Pennsylvania is also perceived as the easiest of the three Rust Belt swing states—Michigan, Pennsylvania, and Wisconsin—for Trump to win.
In 2016, RCP had Trump down by 1.9 points in its final poll average, and Trump won by 0.7 points. In 2020, both major poll averages showed Trump behind; 538 by 4.7 points and RCP by 1.2 points, and Trump lost the state by 1.2 points. The current average at 538 has Trump down by 0.3 and RCP has him up by 0.4. RCP was very accurate in 2024, and I think Trump has the advantage.
Wisconsin
Wisconsin is the “white whale” for the GOP. It’s always “close but no cigar.” Trump’s victory there in 2016 was the first time a Republican had carried the state in the last twelve elections since Reagan racked up a 9-point win in 1984, Trump’s 2016 victory was by 0.7 points though RCP had forecasted a 6.5-point Democrat win. In 2020, Biden squeaked by with a 0.7-point win despite leading in the poll averages by 8.4 points at 538 and 6.7 points at RCP. RCP has Trump leading by 0.5, and 538 has him up by 0.4. If history is a guide, Trump will win Wisconsin by around 5 points.
Advertisement
Enten may be right that no party has ever under-polled in three consecutive presidential elections, but no other party has run the same candidate three times in a row. The bottom line is that Trump is at the best place a Republican candidate has been in national polls since George Bush’s 2004 reelection campaign. He looks to have taken Georgia and North Carolina off the “swing state” map and is running much stronger in the Rust Belt than he did in either 2016 or 2022.
Earlier today, the Washington Post ran an interesting infographic of polls versus reality. It compares the Washington Post’s poll average to the outcome in 2016 and 2020. READ: Harris vs. Trump 2024 presidential polls: Who is ahead? – Washington Post.
If the 2024 election looks like 2016, this is how Trump will do. He will lose Nevada and sweep the Rust Belt.
Should 2024 resemble the 2020 election, this is how it plays out: Trump runs the table.
No matter how you cut it, polling, which has historically been Democrat-friendly, says Trump is the odds-on favorite to win. That said, never forget the financial disclaimer statement, “Please remember that past performance may not be indicative of future results.”
Final Thoughts
Polls are just models that frequently work on the principle of GIGO, that is, Garbage In, Gospel Out. I’ve frequently commented on X about the ridiculous polls where the crosstabs of Men and Women contradict the topline findings of that poll. Polling averages are better than individual polls, and when compared to quantifiable outcomes, we can draw inferences.
Advertisement
Never forget that polls don’t drive real election results; results are created by people who get out to vote rather than sitting at home reading polls. In the words of Saint Augustine, “Pray as though everything depended on God. Work as though everything depended on you.” Work like hell to elect Donald Trump, make time to go to church before Election Day and pray like the nation’s existence depends on it, and vote.