Blinded symphony auditions reduce gender bias

Just a quick post today following from an interesting comment by lindata on my post about gender in science and the Scientiae blog carnival.

lindata points out that blinding evaluators from the identity of musicians during auditions has resulted in increased representation of women in symphonies - I had not been aware that this work has been going on for more than 30 years. Here is a quote from the 17 March 1997 issue of Business Week:

Starting in the 1970s, report Claudia Goldin of Harvard University and Cecilia Rouse of Princeton University, symphonies began to implement major revisions in their audition procedures for job applicants--including the use of ''screens'' to prevent audition committees from actually seeing the musician whose playing they were evaluating. At the same time, the share of female musicians in the top five American orchestras has grown from less than 5% in 1970 to 25% today.

Here is the 2000 press release from a report published by Goldin and Rouse in American Economic Review:

In their study, Rouse and Goldin examined lists of personnel from 11 major orchestras, including the Big Five, and actual accounts of the hiring process maintained by personnel managers in eight major orchestras.

Among musicians who auditioned in both blind and non-blind auditions, about 28.6 percent of female musicians, and 20.2 percent of male musicians, advanced from the preliminary to the final round in blind auditions. When preliminary auditions were not blind, only 19.3 percent of the women advanced, along with 22.5 percent of the men.

Using data from the audition records, the researchers found that blind auditions increased the probability that a woman would advance from preliminary rounds by 50 percent. The likelihood of a woman's ultimate selection is increased several fold, although the competition is extremely difficult and the chance of success still low.

Blinding in hiring for the sciences is obviously impractical. However, some research funding agencies do blind the identity of grant applicants from reviewers - this in an attempt to minimize bias both of gender and age. This is a far more difficult proposition, though, since an applicant's capability to complete a proposed study often requires an examination of their past record of accomplishment as evidenced by publications and preliminary data that cannot always be blinded properly.

However, the message from the studies of symphony candidate selection has certainly revealed how gender-bias can indeed occur.