I would love to see this same experiment performed with other factors like different color skin and sex/gender. How better to find actual evidence of hiring bias than with a giant automated freelancing system?
White sounding names get 50% more callbacks then black sounding names when sending out identical resumes.