Ok, so maybe I should have inserted the word "systematic" before the word "bias".
Anyhow, that brings me back to my thought experiment. If quality is the determining factor in what gets published and conceptions of quality are independent of the various categories we are slotted into, then it seems like it wouldn't matter if all editors were white, male, able-bodied, heterosexual and upper-class. (Or if all editors were black, female, disabled, lesbian, and working class). We'd end up with pretty much the same poems being published and no systematic distortions.
I guess that's just not something that seems likely to me. But maybe that's a bias on my part.
And if it is, I'll put it down to my age, education, nationality, social class, (dis)abilities and life experience, of course
-Matt