People who want to trash this study with the same old "Have they thought of $obvious_counterargument?" should first browse through their GitHub repo: https://github.com/the-markup/investigation-wheres-my-email

I did and I have gone through the mbox files and I cannot find anything that would explain the curious difference. That is not to say that there isn't a rational explanation, the probably is, but please give the study authors some credit. They aren't dumb.

I think Alex Stamos' take on this is pretty good: https://twitter.com/alexstamos/status/1232687398250639361?s=...

I’d imagine this is almost entirely due to scale - Yang probably sends each email to <100k users while Warren/Bloomberg have 25MM+ list sizes.

Typically an email sent to 25 million people is less important than one sent to 100k, and so should be sent to promotions.

I doubt that. 100k is already very, very far into the range of likely promotion.

I would prefer to send them all to a Politics inbox, to be honest.

It's just an algorithm. Not a big conspiracy. You can definitely influence in which folder you end up. Seems like Buttigieg's team has done a better job at it.

Oh, yes? Pete B is clearly the cabal candidate, so why not assume it's part of the conspiracy?

Google has clearly lied (based on the blacklists that whistleblowers have released) about their political bias.

Sure, however this is the kind of cases where "the system" (democracy) would benefit from an equal treatment of all, regardless of their capacity to game gmail.