One of the best-known studies of bias in hiring is the “blind audition” study. This study, conducted in 1997, explored hiring practices at American symphony orchestras – specifically, whether “blind auditions”, when musicians play for the hiring committee while hidden behind a screen, made a difference in how many female musicians were hired. The “blind audition” study demonstrated how bias could affect hiring decisions, even when the hiring process was designed to be as neutral and objective as possible. 

However, the study only addressed gender bias in hiring. Now there are suggestions that the findings from that study could be built on to address racial and ethnic bias. Although, like the “blind audition” study, these suggestions are based in the world of symphony orchestras, they have relevance to any kind of workplace.

Claudia Goldin and Cecilia Rouse, the authors of the “blind audition” study, were curious as to why the number of women in US symphony orchestras had dramatically increased from the 1950s to the 1980s, even though the percentages of women graduating from classical music schools did not significantly change during that time.

They suggested that orchestras are excellent places to study the process of hiring decisions, because orchestras have a stable structure. The number of players in a symphony orchestra is fairly standard, around 100 musicians, and the number of “jobs” within the orchestra – the number of musicians playing each type of instrument – rarely changes. So variations in the number of male and female musicians in an orchestra usually can’t be attributed to changes in the size of the orchestra, or to changes in the work its members do.

The data for Goldin and Rouse’s study came from 11 US orchestras. The data included audition records (the list of musicians who applied for jobs in the orchestra) from the mid-1950s to the 1980s; the lists of members of the orchestra (in other words, the musicians that were hired); and how each orchestra hires new members. Most orchestras have a multi-step hiring process, starting with initial applications for the position, and then a series of live auditions, with applicants eliminated after each round of auditions.

Goldin and Rouse looked at whether each orchestra used “blind” auditions for some or all of the steps in its hiring process, and, if it did, when it started using them.  Some orchestras have a completely “blind” audition process, with the musicians concealed behind screens every time they play for the hiring committee. (Before they play, the musicians are assigned identifying numbers, so that the committee can’t tell a musician’s gender from their name.) Other orchestras use “blind” auditions just for the later rounds of auditions, and some use them only once, when the finalists for the job are playing for the committee one last time. Some orchestras also take extra steps to make the audition as truly “blind” as possible – for example, by putting a carpet at the doorway of the audition room, so that female and male musicians’ footsteps sound the same when they enter and leave the room. 

Goldin and Rouse compared the number of female job applicants at each orchestra across time with the number of female musicians in that same orchestra across time, and then compared those numbers to whether and how the orchestra used “blind” auditions during the hiring process. They found that female musicians were 50% more likely to get past the preliminary stages of the hiring process when “blind” auditions were used early in the process. They also found that female musicians were more likely to be hired when the last audition before a hiring decision was a “blind” audition. These results suggested that gender bias was affecting hiring decisions, even if the members of hiring committees were not consciously intending to be biased.

A few weeks ago, in light of the Black Lives Matter movement and the conversations that has generated around the problem of pervasive racial discrimination, the New York Times ran a very interesting story on racial equity and orchestra auditions. The story suggested that while “blind” auditions might improve gender diversity in orchestras, they didn’t seem to improve ethnic or racial diversity. A 2016 study showed that while the representation of non-white musicians was increasing in US orchestras, less than 15% of orchestral musicians in the US are non-white. The Times asked classical musicians, conductors, and orchestra administrators what could be done to make orchestras both gender-diverse and racially diverse.

Here are some of their suggestions:

• Do away with final auditions entirely. Have the finalists actually play with the orchestra for a few weeks, and see if they can do the job.
• Even if auditions are blind, information collected earlier in the hiring process (e.g. names) can indicate the applicant’s racial or ethnic background. Establishing demographic quotas for hiring, especially when all the candidates have comparable experience and skills, may be the only way to overcome unconscious or conscious bias resulting from that information.
• Be consistent in how candidates are recruited and evaluated. For example, even when a vacant position in an orchestra is advertised as being open to all qualified applicants, hiring managers often recruit additional candidates they know or who are recommended to them. The hiring managers may also give those candidates special treatment during the hiring process, such as allowing them to audition separately from other candidates. Recruitment and evaluation processes need to be the same for for all candidates.
• Measure and track the gender and racial demographics of applicants at each stage of the hiring process, including those who are eliminated after each audition. This could help to identify unseen barriers in the process that discriminate against non-white candidates.
• Establish minimum levels of demographic representation among the applicants chosen to audition (e.g. 25% Latinx musicians), and also establish a policy that the hiring process cannot proceed further until those levels are met.
• Consider factors other than the candidate’s playing ability when deciding who to hire. For example, can the candidate teach younger musicians? Can they represent the orchestra in the community? Are they entrepreneurs who can help the orchestra pursue new opportunities or directions? These types of skills can help the orchestra in many ways beyond the individual’s musical contributions. Expanding hiring criteria in directions such as these also takes the focus of hiring away from the single-minded focus on who is “best”.
• Diversify the repertoire of orchestral music to include more works by non-white composers, and encourage and support young musicians of colour.

And maybe most importantly,

• Ask members of the underrepresented demographic groups what they think needs to change.

Ultimately, it may not be possible to make any hiring process completely “blind” to the impact of gender or racial bias. But being aware of the potential for bias, and of the conditions that may create bias, are both excellent strategies to reduce any impact that bias might have in the hiring process.  And even though these suggestions focus on how orchestras hire their members, any organization looking to increase its gender or racial diversity  can learn something from these ideas.