Thematic Markets

Thematic Markets

Share this post

Thematic Markets
Thematic Markets
There's a new sheriff in town
Copy link
Facebook
Email
Notes
More

There's a new sheriff in town

Thematic Markets' predicted polling errors presaged a "red wave"

Marvin Barth's avatar
Marvin Barth
Nov 06, 2024
∙ Paid
6

Share this post

Thematic Markets
Thematic Markets
There's a new sheriff in town
Copy link
Facebook
Email
Notes
More
2
Share

Having called the last three elections that befuddled professional pollsters, politicians and data scientists, permit me a gratuitous victory lap. In August, I wrote:

“[M]y analysis [of poll bias] suggests that former President Trump remains in the lead and likely will win the presidential election if he survives to November;...[potentially large] within-demographic bias raises a significant tail risk that...Mr. Trump will not only win a landslide in the Electoral College, but the national popular vote, too, giving him a mandate for his policies.” – And now the fun begins, 21 August 2024

When I revisited my analysis last week, I wrote:

“[Poll bias that I identified] in August appears to have increased since, yet Harris’s lead in the polls has slipped, both nationally and in swing states... Mr. Trump stands a good chance of winning the national popular vote and perhaps all of the swing states...[with] a ‘red-wave’ Republican sweep of the White House and both houses of Congress.” – Observations: Revealed preference meets real limits, Part II, 31 October 2024

My median estimate for polling errors was 2 percentage points, but I noted that the distribution of potential errors was likely skewed towards a much larger win for Mr. Trump, with potential polling errors in the range of 1.8% to 7.0%. Reported results suggest that the national polling error is on the order of 3.4 percentage points and that Mr. Trump winning the national popular vote (Figure 1). The average state polling error was 2.8 percentage points and in no state did it favor Vice President Harris. Further, as my analysis suggested would occur, Ms. Harris badly underperformed President Biden’s 2020 election results by an average of 8 percentage points nationally and an average of 4.8 percentage points across states. Interestingly, the states with both the largest polling errors and underperformance of Ms. Harris versus Mr. Biden were not swing states but deep “blue” states like California (7.8ppts and 12ppts, respectively) and marginally “red” states like Florida (6.5ppts and 10ppts, respectively).

Market opportunities…

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2025 Thematic Markets, Ltd
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share

Copy link
Facebook
Email
Notes
More