The 2024 US election ordeal is over. Now people want to know what happened and why. Every major and minor pundit in the country is weighing in. You can find reports from news outlets like Newsweek and NPR, think tanks like the Brookings Institution, and more.

These early analyses are mostly based on exit polls, which everyone agrees are crap. We need to wait several months for proper statistical data to be gathered and analyzed. But of course, the pundits aren’t waiting.

So why should I? After all, I’ve got an opinion and a keyboard. And do you really need any other qualifications to be a pundit these days, especially a minor one? Without further ado, then, here’s my unsolicited post-election punditry.

What changed between 2020 and 2024?

I’m interested in the differences between the 2020 and 2024 presidential elections. Here’s what I found from looking at published vote counts.

A decline in total votes cast: Overall about 3.25 million fewer people voted for President in 2024 than in 2020. About 158.4 million Americans voted for President in 2020, and 154.2 million voted in 2024. I’m using the words “votes cast” or “total votes” rather than “turnout” because I don’t have figures for turnout. Please see the Methodology section below for more details about why. At the state level, total votes declined in 33 states and increased in only 18. (51 states? Yeah, the District of Columbia is counted here because it has three votes in the Electoral College.)

A popular vote win, but not a landslide: Donald Trump did win the popular vote this time, the first Republican presidential candidate to do so since 2004. Harris got about 6.2 million fewer votes in 2024 than Biden did in 2020, and Trump got just over 3 million more votes in 2024 than in 2020. This represents a swing of about 6% in the popular vote towards Trump. However, nationwide, his margin of victory was a little less than 1.5%. In the Electoral College, this gets magnified into a much larger victory by the winner-take-all allocation of electoral votes in every state except Maine and Nebraska. Still, this was no landslide victory for Trump.    

Trump’s percent share of the popular vote improved in all 50 states from 2020 to 2024, even in states he didn’t win. For example, Trump’s share increased by 4.0% in California and 5.6% in New York though he still lost both states.

Harris increased her percent share in only one state from 2020 to 2024: Utah. It didn’t help. Trump still won Utah by 21.6%.

A swing state sweep: Trump won all the marbles here, flipping six swing states – Arizona, Georgia, Michigan, Nevada, Pennsylvania and Wisconsin – and taking North Carolina again. Despite lower vote counts nationwide, total votes cast for President went up in all seven swing states from 2020 to 2024. Clearly both campaigns invested a lot of money and energy in these contests.

It paid off for Trump. His vote count and share increased in every swing state. He got his vote out. Harris’s vote count increased in Georgia, Nevada, North Carolina and Wisconsin, but by less than Trump’s, so her share declined in these states.

3^rd party spoilers: Did votes for 3^rd party candidates make a difference to the outcome? The short answer is no. I looked at whether re-apportioning votes for 3^rd party candidates to either Trump or Harris would have changed the results. In only two states, Michigan and Wisconsin, did 3^rd party candidates get more votes in total than the winning candidate’s margin of victory. In Michigan, 3^rd party candidates got a total of 111,017 votes while Trump’s margin of victory was 80,103 votes. And in Wisconsin, 3^rd party candidates received 57,083 votes while Trump carried the state by just 29,397. So if you made the improbable assumption that, absent 3^rd party candidates, the vast majority of those votes would have gone to Harris, she would theoretically have won those two states and got 25 more votes in the Electoral College. But that still wouldn’t have been enough to beat Trump.  

Stay-at-home spoilers: Did the decline in votes cast from 2020 to 2024 change the outcome in any states? Here too the short answer is no. New Jersey was the only state where the decline in total votes cast for President was greater than the winning candidate’s margin of victory. Harris won New Jersey by 252,498 votes, but there were 278,732 fewer votes cast there in 2024 compared to 2020. So again, if you made the improbable assumption that most of those missing voters would have cast their ballots for Trump, he could theoretically have won the state. But that would only have increased his showing in the Electoral College and not changed the outcome overall.

Polarized and staying that way: The table below counts the states according to the margin of victory by the winning presidential candidate. A couple of things to note here. First, there was little change between 2020 and 2024. They say there’s a “great sorting” going on in America where people are moving to states that align with their political views. In other words, red states are becoming redder and blue states are becoming bluer. The table doesn’t show this, but you might need to look back farther than two elections to see a real trend. However, the table does show how polarized we are. 38 states had a margin of victory greater than 10% in 2024, and 37 did in 2020. The chances of flipping a state with a margin over 10% is basically zero. For comparison, all seven swing states had margins of victory of less than 3% in both 2020 and 2024, except Arizona in 2024 where the margin was still just 5.5%. I don’t see this deep polarization changing anytime soon.

Margin of victory	2020	2024
Less than 5%	8	8
5% – 10%	6	5
10% – 20%	17	18
20% – 30%	11	10
30% – 40%	7	7
Greater than 40%	2	3

Demographic splits: I didn’t look at breakdowns of the vote by race, ethnicity, gender, income, urban/rural location etc. As I said earlier, exit polls are notoriously unreliable so we’ll have to wait a few months for proper demographic analysis to be completed. However, the shift to Trump in every state indicates a broad nationwide trend. I would guess that trend is reflected to some extent across most demographic groups.

A couple of thoughts on this. First, we should all stop thinking of demographic groups as solid voting blocks. There is no such thing as “the Hispanic vote” or “the Black vote” or “the women’s vote” anymore, if there ever was. It’s insulting to assume that a person’s gender, skin color, education or other characteristic determines how they vote. Equally important, it’s a deeply flawed foundation on which to build campaign strategy.

Second, any hope among Democrats that an increasingly diverse electorate will automatically translate into solid majorities for their party looks pretty shaky now. Democrats may still get the lion’s share of votes from non-white communities, but they can no longer take them for granted. They’re going to have to earn those votes. I think that’s healthy.

Final thoughts

Around two billion people voted in over sixty elections worldwide in 2024. Incumbents fared badly in most of them. The US was no exception. As the Pew Research Center reports, voters around the world expressed their frustration about many of the same issues as they did here. If anything, Democrats did better than incumbents in other countries.

Still Democrats have serious work to do. While Trump’s margin in the popular vote is by no means insurmountable – he hasn’t put together a durable coalition in my view – it’s also clear that Harris didn’t successfully address key issues that were important to voters like inflation, immigration and a general rise in disorder (petty crime, homelessness in cities, etc.). Fundamentally, Harris represented the status quo in an election where voters wanted change.

Democrats probably squeezed every last vote out of people who dislike Trump’s bigotry, corruption, criminal convictions and threats to democracy. Yet it wasn’t enough. I’m dismayed by this fact, but I have to admit I’m in the minority.

So now there must be a reset. In my view, Democrats have been putting this off since 2016. I believe Hilary Clinton lost to Trump for largely the same reasons that Harris did: she failed to present a compelling message that addressed important issues for a majority of voters. Yes, misogyny may have played a role too. Biden won in 2020 largely due to Trump’s mishandling of the COVID pandemic, but Biden’s victory may have misled Democrats into thinking Trump was beatable on character alone. And better-than-expected performance in the 2022 midterms may have convinced some that a deep rethink wasn’t necessary. 2024 has surely disproved those ideas.

Either Democrats had the wrong policies, or they didn’t talk about them in a way that connected with voters, or both. I personally think many of the policies implemented by the Biden Administration, especially the Inflation Reduction Act, were on the right track and will bring long term benefits to the country. Trump will get to take credit for them if he doesn’t screw them up. But long-term benefits don’t address today’s issues. And you can’t tell people the economy is doing great when their lived experience tells them otherwise. (You also can’t gaslight voters into believing that an octogenarian leader has the strength for another four years when they can clearly see he doesn’t.) More importantly, Trump won’t be on the ballot in 2028, unless the Supreme Court goes completely off the rails and invalidates the 22^nd Amendment.

Democrats have their work cut out for them.

Methodology

Any pundit worth their salt must disclose their methodology, so here’s mine.

I’ve compared the results of the 2020 and 2024 elections for President and Vice President only. I’ve ignored all down-ballot votes. I used the following sources for election results data:

2020 election: Federal Elections Commission. “Federal Election Results 2020: Election Results for the U.S. President, the U.S. Senate and the U.S. House of Representatives.”, Oct. 2022, https://www.fec.gov/resources/cms-content/documents/federalelections2020.pdf, p. 7.

2024 election: Wikipedia contributors. “’2024 United States presidential election.” Wikipedia, The Free Encyclopedia, 10 Dec 2024, https://en.wikipedia.org/w/index.php?title=2024_United_States_presidential_election,  accessed 17 Dec 2024.

2024 data will not be final until all states have certified their results.

For Maine and Nebraska, which allocate electoral votes by district, I show only the total vote counts for each state.

I did not attempt to measure voter turnout. Turnout turns out to be kind of complicated. Should you use registered voters or voting-eligible population (VEP)? If you use VEP, how do you estimate the population of eligible voters in each state? How do you count convicted felons who can vote in some states but not others? The University of Florida’s Election Lab has good data on this. But I’m only a minor pundit so I just consider the votes actually cast for President and Vice President.

I have not considered ballots cast by US citizens overseas unless they have already been included in their home state tallies.

Want to check my work? Click this button to download the Excel spreadsheet I created.

Thanks for reading.

---

If you enjoyed this post, please subscribe to Unsolicited Feedback.

Related Links

Bruni, Frank, et al. “‘Democrats Have Gone Beyond Soul Searching to Soul Spelunking’: 4 Writers Dive Deep Into the Party’s Distress.” New York Times, 23 Dec 2024, https://www.nytimes.com/2024/12/23/opinion/democrats-election-future.html.
Frank Bruni interviews three Democratic insiders on what went wrong in 2024 and what the party should do to recover. One of the people interviewed claims that, based on AP VoteCast data, 19 million Biden voters sat out the 2024 election. I haven’t been able to find out where this number comes from.

Klein, Ezra. “Democrats Need to Face Why Trump Won.” New York Times, 18 Mar. 2025, https://www.nytimes.com/2025/03/18/opinion/ezra-klein-podcast-david-shor.html.
Interview with Democratic pollster David Shor. This one is also worth watching on YouTube because they review some charts and graphs from Shor’s analysis of the 2024 election. https://www.youtube.com/watch?v=Sx0J7dIlL7c