No DNC Bump & Dancing Inside the Margins

Quicker than usual; I have to be ready to Teach on Monday. There are actually two sets of results here. I’ll explain.

Results for 28 August 2024:

As usual, we collected data on Wednesday. We’re in virtually the same place as we were last week. Here Harris wins about 62% of the simulations and Trump takes about 37%. The likelihood of a tie is still at about 1%. Let’s look at the state shifts.

State	Change
Arizona	Barely Democratic to Barely Republican
Maine AL	Leaning Democratic to Strongly Democratic
Maine 2nd	Barely Republican to Leaning Democratic
North Carolina	Barely Democratic to Toss Up
Wisconsin	Toss Up to Leaning Democratic

There is more dancing of the results that live inside the margin of error. The Maine shifts are the most significant here and that’s why the probability barely moved even though Trump pulled ahead with the electoral count. The light blue, light red, and tan states all have probabilities at or near 50/50; those will bounce back and forth in the simulations but the dark blue and the dark red? Those stay on one side for the majority of the trials. If something moves into one of those categories we’re likely to see it.

So, we’re holding steady with the odds of a Harris victory sitting at 3 to 2; there’s no real evidence of a bump from the Democratic convention. Globally the numbers still seem to be trending in Harris’s direction.

Results for 29 August 2024:

Several polls that showed a movement towards Harris in the swing states arrived on 29 August. Here are the results from that next day. I won’t give these results their own write-up because I played with the parameters; there are no apples-to-apples comparisons.

Last Words

These results are making me think about my methodology; I’d like to tamp down the variance and I’m not sure the colors are giving the best indication of the state of the race. Still, I don’t want to switch things up too many times.

So, I’m pondering. I may just let the status quo ride and make changes for 2028. If I make more alterations I’ll use them until Election Day. My plan is to have everything sorted by Wednesday.

This is what the stock market calls a “correction.” Harris Still Favored

I’m watching night 3 of the Democratic National Convention and Oprah just exited the stage. It was electric; the crowd was as energized as it has been all through the convention. Today we’re looking at a set of data that was collected almost entirely before the convention.

The State of the Race

The slowdown in the campaigning is over and the DNC seems to be sucking up all the air in the election. Well, maybe it’s just sucking up all of my attention. (Boy, that was just a hell of a response for Mayor Pete.) The excitement in the Arena is palpable and infectious. There was little to no bounce coming out of the RNC. I don’t think the same will be true of the DNC.

Trump and Vance are on the campaign trail, and Trump seems more active than last week, but I don’t know if it’s enough. More on that after the analysis.

Results:

A minor change in the methodology; with this update, I’ve separated out the congressional districts in Maine and Nebraska. Those are modeled separately and this will make the projections a bit more accurate. Since Electoral-Vote.com doesn’t have that data I’ll be taking it from Projects.FiveThirtyEight.com. Plus the first and third districts in Nebraska needed a smidgen of algebra because I only had separate polling on District 2. Aside from that, as always, an explanation of how the map was created and our scheme for finding our polling averages can be found here: Methodology.

You probably know how this goes by now. We entered the polling data into our model and ran 40,000 simulated elections. If the election were held today, Vice President Harris would have roughly a 62% chance of winning while Donald Trump would win about 37% of the time. The likelihood of an Electoral College tie is still approximately 1%.

This is a “polls only” approximation and so it’s sensitive to changes in polling numbers. That leads to some volatility. For example, let’s look at Virginia which moved from “Leaning Democratic” to “Barely Democratic.” Here’s what that looks like in their data.

That is not a dramatic change, we’re looking mostly at the result normal of statistical variation.

Given this data, Harris still leads and the odds of her winning the election are about 3 to 2.

Trends:

The averages have shifted back in Trump’s direction and it’s put the odds for Harris and Trump back to where they were two weeks ago. Here are the changes from last week.

State	Change
Georgia	Toss Up to Leaning Democratic
Michigan	Barely Democratic to Toss Up
Nevada	Barely Democratic to Barely Republican
North Carolina	Toss Up to Barely Democratic
Pennsylvania	Leaning Democratic to Toss Up
Virginia	Leaning Democratic to Barely Democratic

As I mentioned above most of these look like what happens with results within the margin of error. The possible exception there is Georgia where the new poll looks a lot different from the previous one. My best guess is that this is merely a more accurate description of the state of the race than a rightward shift but it could be either. More polling will determine that. It turns out that one consequence of my choice of using a ten-day window for poll numbers is some volatility as we’re frequently replacing a single poll with a different single poll. I hope we get more state polls as election day draws closer; that will tamp down the variation.

The odds of a Harris winning sitting at 3 to 2 seems more reasonable than the 3 to 1 we found last week.

Last Words

I’m now writing on Thursday Morning. The end of Night 3 at the DNC was nearly flawless. Walz gave a powerful “pep talk” that received what might have been the best crowd response of the day. It was also steeped in Americana. A high school football coach who loves his family, helps his neighbors, and helped win a state championship surely has appeal to voters all across the political spectrum. And watching his family was heart-warming. Trying to reconcile that with a “soulless communist who will destroy America” just does not compute.

Neil Young’s “Keep on Rocking in the Free World” is a perfect song to end the evening.

As I mentioned, I haven’t seen much coverage of the Republican campaign. I’ve seen a tiny rally featuring Vance, part of a small, weird, appearance by Trump in Howell, Michigan, and an AI-generated image implying that Taylor Swift is endorsing Trump. In addition, a Trump staffer released a now-deleted video of Trump backed by Beyoncé’s song “Freedom,” Trump had claimed that he is better looking than Kamala Harris and that there’s no way Harris’s acceptance speech will get better ratings than his speech to the RNC. Trump’s interest in crowd sizes remains an ongoing theme. Maybe I’ve seen more than I thought.

Let’s take those one at a time.

• The Vance rally just seems like more of the same.
• In his appearance in Howell, Michigan Trump seemed lethargic. If any other candidate had made that speech Trump would have been all over them for being “low energy.” Also, “Kamala crime wave?” Seriously?
• I’m pretty sure the Taylor Swift thing was a misstep.
• I’m certain using Beyoncé’s “Freedom” was a misstep. There’s already a cease and desist order. A lawsuit will probably follow.
• “I’m better looking than Kamala Harris” is just sad.
• And bragging about the ratings of his acceptance speech is just dumb. There will be objective evidence tomorrow and my money here is on the new kid. If Trump’s right this is a non-story. If he’s wrong it’s an embarrassment. There’s no upside.
• Trump’s obsession with crowd sizes is getting funny. The best example of that is probably Walz and Harris appearing remotely to the DNC after the roll call vote from the very same venue that hosted the RNC. Both venues were packed and there was no footage of anyone leaving early. If nothing else this was some world-class trolling.

Trump had a pretty effective line of attack against President Biden, but it’s now been a month since the president dropped out. The Trump Campaign still does not have a coherent strategy to run against Harris. Falsely calling her a communist or blaming her for a nonexistent crime wave might play to his base but it won’t play with independents and it won’t bring any new voters to his cause. We’ve seen it all before.

Harris’s Momentum Continues

Let’s GO!

The State of the Race

The campaign seems to have slowed down somewhat while the Democrats gear up for their convention. Vice President Harris and President Biden will appear together today (8/15) in Maryland while Governor Walz is on his own tour of the Northeast. JD Vance is in Pennsylvania today while yet another controversial quote has surfaced.

Donald Trump has been a bit more active than he had been. This week we’ve seen his interview with Elon Musk on X and a last-minute rally in North Carolina which was supposed to be focused on the economy. The economic content was sparse while Trump frequently veered off-topic and attacked his opponents.

Results:

An explanation of how the map was created and our scheme for finding our polling averages and can be found here: Methodology.

As usual, we entered the polling data into our model and ran 40,000 simulated elections. If the election were held today, it appears Vice President Harris would have about a 74% chance of winning while Donald Trump would win about 25% of the time. The likelyhood of an Electoral College tie is roughly 1%.

Remember that this is an estimate and it is only based on the current polling numbers. As we’ve seen these can change dramatically over three weeks; and we’re nearly three months from the election.

Given this data, Harris leads and the odds of her winning could be as high as 3 to 1.

Trends:

The race continues to move in Harris’s direction. This is evident if we look at the movements of the state polls between last week and this. There’s one exception to the trend;

Michigan changed from Strongly Democratic to Barely Democratic. This was expected; the polling average last week was based on a single poll that was a clear outlier. That poll, which had Harris up by ten fell out of the average was replaced by two more typical results.

The other shifts tell a consistent story.

• Texas: Strongly Republican to Leans Republican
• North Carolina: Barely Republican to Toss Up
• Arizona: Barely Republican to Barely Democratic
• Pennsylvania: Toss Up to Leans Democratic

These all show movement toward Harris. In addition, within the last week, Harris is now leading:

• in national popular vote polls and polling averages,
• in states with a total of more than 270 electoral votes,
• on forecasting sites like 538 and The Cook Political Report, and
• on betting sites like Polymarket.

The comparative attendance and enthusiasm at Democratic vs. Republican Rallies also speak in Harris’s favor.

The momentum remains on Harris’s side.

Last Words

The Harris campaign has consolidated its lead over the last week. It’s no surprise; just over three weeks ago most voters believed that the presumptive nominees of both major political parties were too old to effectively serve as president. The voters wanted a new candidate in the race. With President Biden’s decision to withdraw his candidacy, they got one. The political landscape shifted and the Democrats are capitalizing on that by running an effective, aggressive, and positive campaign. Likely, they will further strengthen their position throughout their convention. But conventional wisdom would then suggest that things will shift back toward being competitive as the campaign begins in earnest and Harris and Walz have to add some specifics to their policy proposals.

But maybe not.

Donald Trump spent months attacking President Biden and implying he should withdraw from the race. He should have been prepared when it happened. But he wasn’t.

At the RNC, with the possibility of Biden’s withdrawal growing more likely, Trump could have picked a running mate that could bring more voters into their base. But he didn’t.

In the aftermath of an assassination attempt, Trump could have changed tactics, eschewed political violence, and become a candidate aiming to unite the nation. But he couldn’t.

Trump could have gone on the campaign trail to court voters and make the case for the Republican ticket every day since 21 July. But he wouldn’t.

Instead, he picked a running mate who appealed to the extreme edge of his base and who is alienating voters, childless cat ladies, and post-menopausal females.

And he didn’t turn away from political violence. When he mentioned Paul Pelosi at the RNC he pantomimed a hammer. “Bam, bam, bam!”

And there’s nothing new in his rhetoric. He hasn’t even bothered to change his talking points. the “Worst President in American History,” became the “Worst Vice President” and “Lyin’ Joe Biden” has turned into “Lyin’ Kamala Harris.” Simplistic.

And the lies are all the same no matter what Trump wants to tell us. The 2020 Election wasn’t stolen. Trump didn’t have the biggest inaugural crowd ever. He wasn’t the best President for African Americans since Abraham Lincoln. We are actually better off than we were four years ago. The list goes on.

This election could still go either way, but it’s getting close to the point where it will take something as dramatic as a major party candidate withdrawing from the election for that to happen. But it’s getting late to try that; ballots are closing.

If Donald Trump doesn’t get back on the campaign trail and give the country something more than the same old grievances, the same old falsehoods, and the same old show this election could be a rout. We’re tired of it.

Harris Takes Lead

Hello again!

The State of the Race

Since our last update, the Democratic Ticket has been campaigning across the country working to extend their momentum from the past two weeks. The big news there is the unveiling of Tim Walz as Harris’s running mate. They’ve been touring the swing states since the announcement and drawing large enthusiastic crowds.

JD Vance has been following them across the country trying to draw comparisons between the two campaigns. It’s been lackluster. The comparison is not helping the Republicans.

Meanwhile, Donald Trump has been mostly absent from the campaign trail; with only one sparsely attended rally in Montana and a press conference. I haven’t watched the press conference yet so I’ll reserve comments for now. The Trump Campaign holding a rally in Montana is mystifying. That’s a move you make when you’re on a clear path to victory and you can afford to spend resources on a Senate seat. His campaign has bigger issues. Was it about trying to get a rally with good attendance for a visual to counter those coming out of the Harris campaign? A positive event to get Trump back on the trail?

The narrative around the debates is mystifying as well. Trump began making up a possible debate on Fox News even before he debated President Biden. But the waffling, then canceling the debate, then insisting on the made-up Fox News debate, and finally canceling again all made Trump look frightened. Making up another debate and then insisting on three debates is a weird thing to do and is too little, too late.

The Harris/Walz tour of the swing states wraps up Saturday in Nevada.

Results:

Once again we entered the polling data into our model and ran 40,000 simulated elections. If the election were held today, it looks like Donald Trump would have about a 35% chance of winning with Vice President Harris winning about 64% of the time. The remaining 1% were Electoral College ties.

Again, I would emphasize that this is an estimate. There weren’t many state polls in the last week and using our scheme had an interesting effect: it turned out that every state ‘average’ was determined by a single poll.

This had its greatest effect in Michigan as the one poll used there is probably an outlier (Harris led by 11), which would make the probability of a Harris victory an overestimate.

Given this data, Harris is in the lead and the odds of her winning the election could be as good as 3 to 2.

Our scheme for finding our polling averages and creating the map can be found here: Methodology.

Trends:

Trump was favored to win in our 21 July probability, and Harris pulled roughly even with him in our 31 July estimate. Now, for the moment, it’s safe to say that Harris has an edge. Strictly looking at the probabilities generated from the models, Harris’s probability of winning the election is roughly where Trump’s was on 21 July, between 63 and 64%. This is a complete turnaround in just under three weeks.

But that’s not the only evidence. Other predictive sites are showing similar results. Nate Silver now has Harris favored to win both the popular vote and the Electoral College. The New York Times/Siena College Poll released this morning has Harris leading Trump by 4 points in Wisconsin, Pennsylvania, and Michigan. The Cook Political Report has reclassified the race in Arizona, Georgia, and Nevada from “Lean Republican” to “Toss Up.”

The momentum remains on Harris’s side.

Last Words

It’s beginning to look like the campaign is the Democrats’ to lose. They have been able to dominate the news cycle and generate enthusiasm since the Republican Convention with President Biden’s withdrawal, followed by Harris introducing herself, the party unifying behind her, and then the announcement of Walz as Harris’s running mate. This could easily continue as the Democratic Convention will lead to hearings about Trump’s January 6th Election Interference Case which will bring us into early voting.

The Trump Campaign still seems unable to find its footing. They continue to throw whatever they can against the Democrats but nothing yet has really stuck outside of their MAGA base. With Vance drawing tiny crowds and Trump largely off the campaign trail it’s hard to see how they can turn this around. Right now it’s easy to find headlines like “Trump Campaign Bleeding Support…” or “Is Trump ‘Losing His Marbles.’“

Still, this election could go either way. Even if the probability of Trump winning the election is really at 35% (and many sites have it higher) that’s still about 10% better than his chances of winning the 2016 election. In that election, he exceeded expectations at every turn. While the Republican Party’s overconfidence in July is partly responsible for the current state of the election other overconfidence could still shake things up dramatically.