Transcript Orchestrating Impartiality: The Impact of “Blind” Auditions on Female

Orchestrating Impartiality:
The Impact of “Blind”
Auditions on Female
Musicians
By: Claudia Goldin and
Cecilia Rouse
The Typical Orchestra
» About 100 members but may vary
between 90 and 105
» The positions are nearly identical
between orchestras and over time
» What does this mean to the study?
Orchestras Examined
» Top Tier (the “Big Five”):





Boston Symphony Orchestra (BSO)
Chicago Symphony Orchestra
Cleveland Symphony Orchestra
New York Philharmonic (NYPhil)
Philadelphia Orchestra
» Second Tier




Los Angeles Symphony Orchestra (LA)
San Francisco Philharmonic (SF)
Detroit Symphony Orchestra
Pittsburg Symphony Orchestra (PSO)
Noteworthy Data
» Currently, NYPhil has the highest
percentage of females at 35% after being
the lowest (generally at zero) for decades
» Female increases in orchestras is even
more remarkable because of low
turnover rates.
 An average of 4 “new hires” per year for topranked Chicago and NYPhil and about 6 for
the others.
 The number of new hire was small after 1960
and declines over time.
Changes in Gender Share
of New Hires
» Since the early 1980’s the share
female among new hires has risen.
 About 35% for the BSO and Chicago
 About 50% for NYPhil
» Compare this to the rate of less than
10% before 1970.
Main Question:
» Did the screen matter in a direct manner
or did the increase in women hires come
as the result of a host of other factors,
including the appearance of impartiality
or an incresed pool of female contestants
coming out of music schools?
 So, where the most selective music schools
producing substantially more female students
in the early 1970s?
Could Changes in Audition
Procedures Unveil Hiring Bias?
» Quotes concerning women in orchestras:
 “women have smaller techniques than men”
 “are more temperamental and more likely to
demand special attention or treatment”
 “the more women (in an orchestra), the
poorer the sound”
» Zubin Mehta
 Conductor of LA Symphony (1964-1978) and
NYPhil (1978-1990)
 “I just don’t think women should be in an
orchestra.”
The Audition Process: Today
» The orchestra advertises auditions
» Interested applicants submit resume and
often a tape of compulsory music
 In some, may audition live, even if previously
rejected based on tape
» 3 Rounds:
 Preliminary
 Semifinal
 Finals
» Two Main Types:
 Live
 Blind
Live Auditions
» The semifinal round is sometimes live
» The final round is almost always live and
involves the attendance and input of the
music director.
» Committee advances all deemed
qualified, so no limit to number of
musicians in each round
» The final round generally results in a hire,
but sometimes does not
Blind Auditions
» Almost all preliminary rounds are now blind.
» A screen is used to hide the identity of the
player from the committee.
» Examples:
 Large pieces of heavy (but sound-porous) cloth
 Sometimes suspended from the ceiling of the
symphony hall
» Similar to large room dividers
» To ensure even more impartiality, some roll
out a carpet leading to center stage to
muffle footsteps that could betray gender
» Only the personnel manager knows the
mapping from number to name
Is Blind Really Blind?
» Can trained musicians discern
candidates based on different
playing styles?
» Ultimately, it would be vary rare for
this to happen, as it would have to
be a well-known candidate with an
unusually distinctive musical style.
Data and Methods
» Actual audition records of 8 major symphony
orchestras
» To preserve full confidentiality, do not reveal
names of orchestras
» Rely on first name of the musician in
determining sex
 Determined sex of 96% of those auditioning
» Excluded incomplete auditions
» Considered each round separately
» Restrictions exclude 294 rounds and 1,539
individuals
» Final sample: 7,065 individuals and 588 audition
rounds (from 309 separate auditions)
Data and Methods Cont…
» Second data set:
 Orchestra personnel rosters
 Musician considered new to the orchestra if
he or she had not previously been a regular
member of that orchestra
 Excluded temporary and substitute musicians,
as well as harpists and pianists
 Final sample of 1,128 new orchestra members
from 1970 to 1996
Effect of Screen of Likelihood
of Being Advanced
» Raw data:
 Without a variable for orchestral “ability”,
women fare less well in blind auditions than
otherwise
 The relative of success of female candidates
(compared to male) appears worse for blind
than for not-blind auditions and holds for
every round
» Limited Sample:
 For those who auditioned both with and
without a screen, women’s success rate is
almost always higher in blind auditions
Effect of Screen of Likelihood
of Being Advanced
» Examples:
 Preliminary round with no seminfinals:
» Blind:
 28.6% of women advance; 20.2% of men advance
» Not-blind:
 19.3% of women advance; 22.5% of men advance
 Since these are the same women, a woman
enhances her own success rate by 9.3% by entering a
blind preliminary round
» These differences are large relative to the average
rate of success
 Women’s success rate increased by 14.8% for
blind finals and overall success rate is 1.6 times
higher for blind auditions
Controlling for “Ability”
» Used an individual fixed-effects strategy
to control for contestant “ability” that
does not change with time
 Can do so because 42% of individuals in
sample were in more than 1 round in the data
set and 6% competed both with and without
a screen
 24% of the musicians competed in more than
one audition
“Ability” Findings
» The screen had a positive effect on the
likelihood that a woman is advanced
from the preliminary round (when there is
no semifinal) and from the finals
 For preliminaries not prceded by a semifinal,
woman had a 11% better chance of selection
 For those in the finals, chances of winning
increased by about 33%
» The effect in the semifinal round remains
strongly negative.
Potential Biases
Could the females that are improving over
time be those that switch from not-blind to
blind auditions?
»

»
»
»
Could the growth rate of their “ability” be faster
than that of men?
What if those individuals who get hired at
their first audition are more able musicians
than those who audition multiple times
Could the orchestras that use screens be less
discriminatory against women than those
that do not?
What if they did a bad job picking the sex of
participants based on their first names?
Effect of the Screen on the
Hiring of Women
» Compare auditions that are blind versus those
that do not use the screen at all or for early
rounds only.
 Women are about 5% more likely to be hired than
are men (not significant)
» Effect is nil when there is a semifinal round.
» Impact for all rounds is about 1% (large standard errors and
not significant)
 But, given that the chance of winning an audition is
3%, the authors need more data to be able to
estimate significantly and even a 1% increase is
large
 In addition, these estimates demonstrate a
difference stimulated by a completely blind process
Example for Clarification on
Hiring Results
» Two Regimes:
Not Blind: no screen; Blind: with screen
Not Blind: 20% female candidates
Blind: 30% female candidates
In 1970s, 10% of new hires were woman
Assume 30 candidates enter each audition
Success rate for not-blind will be .0166 for female
and .0375 for typical male
 In blind regime, % of new hires that are female is
35% (approx. figure for last 10 yrs)
 The success rate must now be .0389 for females
and males decreased to .0310
 Therefore, success rate for woman had to increase
by almost 2.2% when going from not-blind to blind






Conclusions from Example
» The point estimate of 1% for the result of the
effect of screening describes half of the increase
in the success rate
» The increase of being hired out of audition
accounts for 66% of the total increase of females
among new hires
» Half of the 66% comes from the switch to blind
auditions
» Other half could be from a greater acceptance
of female musicians by music directors
» The remaining 34% is a result of a higher
percentage of females as a part of the applicant
pool.
Conclusions
» The audition process began changing in the
1970s and physical screens were increasingly
used to conceal candidate’s identity and
ensure impartiality
» The weight of the evidence persuaded the
author’s and is what they decided to
emphasize
» They find that:
 the screen increases - by 50% - the probability that
a woman will be advanced from certain
preliminary rounds
 Increases by severalfold the likelihood a woman will
be selected in the final round