Well, in this study, they explicitly had equality - all four groups were as likely to succeed. So any difference in hiring was actual bias (or statistical noise).