It’s helpful to remember that news reporting on anything involving numbers is usually relying heavily on hyperbolic horse shit.

Here is the “source data” (LOL) for this: https://pos.org/wp-content/uploads/2024/12/FI14730D-CNBC-AAES-Q4.pdf

There are some key points that really matter here:

1. What method was used to select the “1002 adults nationwide”? This is not described. Was it people with landlines that skew heavily older and conservative? Was it an even mix of blue and red states? IOW, was it even really random? I could not find this information anywhere.
2. This polling was conducted by an outfit named “Public Opinion Strategies” (appropriately, POS). Since we apparently can’t know what sampling method was used (please! correct me if I’m wrong), we can maybe derive something from the either the organization doing the polling or the data itself. 2a: The data shows a slight shift to identifying as Republican (42/39). Anecdotally we can maybe surmise that people who say they are “strictly independent” may lean right-wing. Other demographic info suggests a slight bias toward older and less educated. 2b: In 2024 POS has been heavily used by Republicans. You can find this data on Open Secrets. Basically every single entity that gave them money that I looked up is Republican.

So, tl;dr until any poll can tell me their methodology and confirm that their sampling is truly random I’m just going to laugh at the “conclusions” made by the media.