Much ado about nothing?
"Our 2021 data shows that guests perceived to be Black were able to successfully book the stay of their choice 91.4 percent of the time, versus 94.1 percent for guests perceived to be White,"
Unless my government education math is wrong, that is only a difference of 2.7%. The three other races identified are all in the middle of that range which makes the Black vs White the outlying statistics. The average is 93.1%. So Blacks are 1.7% less than average and Whites are 1% more than average.
The associated document indicates that is from 750k randomly selected 2021 reservation requests. So they did select from a large sample, which is good. I couldn't find any reference for a margin of error though. The basis is "Guests perceived" to be a certain race. That is a subjective measurement and not an objective measurement, which makes the data susceptible to a margin of error. What I perceive to be a certain race is different from what you will perceive. To which race is a 50/50 mixed race person 'perceived' to belong?
I don't deny the numbers exist, but I do question how big of a deal this is. These are small single percentage point differences. Not double digit percentages. There MUST be some variance, that is how statistics work. One statistic will be at one end, and a different statistic will be at the other end. And the nature of perception introduces a margin of error.
My point is the data does not support a response with damning condemnation of AirBNB. It seems this data actually validates there is not a massive systematic discrimination issue.