00:00:00
make sure you shine a light on how
00:00:03
things are going in different groups
00:00:04
because scrutiny increases the social
00:00:07
pressure people feel to make different
00:00:10
kinds of hiring and promotion decisions
00:00:12
and attend to a lack of diversity so
00:00:14
scrutiny matters highlighting social
00:00:16
norms when they're positive social norms
00:00:18
meaning hey you're the only group in the
00:00:19
company that hasn't yet promoted a woman
00:00:21
to this a senior ranks that kind of
00:00:24
scrutiny and those kinds of social norms
00:00:26
matter
00:00:27
welcome to the ripple effect the podcast
00:00:30
that takes you on a journey through the
00:00:32
minds of work and faculty I'm your host
00:00:34
Dan Loney and in each episode we'll be
00:00:36
diving deep into the inspiration behind
00:00:39
the groundbreaking research that Wharton
00:00:41
professors have conducted and exploring
00:00:43
how their findings resonate with the
00:00:45
world today we'll be covering a diverse
00:00:48
range of topics bringing you the latest
00:00:50
insights and knowledge that you can
00:00:52
apply to your life into work so get
00:00:54
ready to dive into new ideas with the
00:00:57
ripple effect
00:00:59
[Music]
00:01:00
so you have done a lot of research
00:01:03
around gender in the workplace and
00:01:06
that's obviously a very important topic
00:01:08
these days a lot of questions have been
00:01:10
answered there are still a lot of
00:01:12
unanswered questions out there what are
00:01:14
some of those that have kind of inspired
00:01:16
your research in this area
00:01:19
yeah I think some of the important the
00:01:22
most important questions are frankly how
00:01:24
do we solve it how do we solve the issue
00:01:26
that women are still underrepresented at
00:01:28
the top of organizations and uh we are
00:01:32
trying so many things is it that we need
00:01:34
to
00:01:36
eliminate bias and change the
00:01:38
individuals and the organizations to be
00:01:40
more open-minded about uh working with
00:01:43
women and that's one possibility is it
00:01:46
that we need to change the structure of
00:01:47
organization so we can make them more
00:01:49
accommodating to the different needs and
00:01:51
preferences and strengths and challenges
00:01:54
that women face is it that we need to
00:01:58
eliminate the biases that women face not
00:02:01
by changing the shape of the
00:02:04
organization but the way decisions are
00:02:06
made inside the organization so there's
00:02:07
all these different possibilities in
00:02:10
terms of what can fix what I think most
00:02:13
people at this point agree is a real
00:02:15
problem because
00:02:17
enormous amount of talent is not being
00:02:19
tapped to generate the best outcomes for
00:02:21
organizations and we the truth is we
00:02:24
still don't know we don't know the right
00:02:26
combination of solutions but my research
00:02:28
and the research of many who I
00:02:30
collaborate with is trying to pick away
00:02:32
at some of the big questions there and
00:02:35
and get some answers
00:02:36
so going back a few years you had done a
00:02:39
study on the effectiveness of diversity
00:02:42
training if you can't go back and take
00:02:44
us through that study what were you
00:02:47
setting out to do what did you learn and
00:02:50
and really particularly about how these
00:02:52
programs affect women thank you for the
00:02:54
show and this is a really exciting
00:02:56
project that I got to work on with a
00:02:58
massive team at Wharton several years
00:03:00
ago and it came about because at the
00:03:03
time I was co-directing the Wharton
00:03:05
people analytics initiative along with
00:03:08
Adam Grant and Angela Duckworth and Kade
00:03:10
Massey and others and the four of us in
00:03:12
particular felt that one of the most
00:03:13
pressing questions in people analytics
00:03:16
that needed to be answered was whether
00:03:18
or not these diversity trainings that so
00:03:20
many organizations were pouring money
00:03:22
into we're adding value and also you
00:03:24
know could we build one that was really
00:03:26
effective that was our real goal as well
00:03:28
let's let's prove whether or not this
00:03:30
works and let's build the best version
00:03:32
of a diversity training we can given
00:03:34
what we know about the science of
00:03:36
discrimination and bias in organizations
00:03:38
so we found first of all an incredibly
00:03:42
Brave organizational partner it was a
00:03:44
Fortune 500 company that was ready to
00:03:46
team up with us and do this project and
00:03:48
the reason I call them Brave is because
00:03:50
it's a really challenging thing to do as
00:03:53
an organization to open yourself up to
00:03:55
to testing to see whether or not there's
00:03:58
potential gender bias and if you can
00:04:00
combat it in your organization it opens
00:04:02
you up to lawsuits and a lot of
00:04:03
organizations are not brave enough to
00:04:05
team up with academics and do that kind
00:04:07
of science so we found an amazing
00:04:08
partner we had a at the time first year
00:04:11
PhD student passionate about this topic
00:04:13
Edward Chang at the Warren school now
00:04:15
he's a professor at Harvard Business
00:04:17
School we're very proud of having
00:04:18
mentored him in the amazing work he did
00:04:20
leading this and we built a roughly one
00:04:24
hour online diversity training program
00:04:26
was primarily focused on introducing
00:04:29
people to the idea of gender bias
00:04:31
teaching them about
00:04:33
the science showing that women are often
00:04:37
facing backlash when they do things like
00:04:39
negotiate for higher salaries that uh
00:04:42
there are implicit biases that influence
00:04:45
the way we judge other people even if we
00:04:47
don't explicitly intend to discriminate
00:04:50
that we have different associations with
00:04:51
women we expect them to spend time in
00:04:53
the home uh and it's easier for us to
00:04:56
sort words that we associate with women
00:04:58
with words we associate associate with
00:05:00
doing work in the home
00:05:02
um than it is for us to sort those words
00:05:03
with with men it's easier for us to sort
00:05:05
words related to Career with
00:05:09
um men than it is for us to sort those
00:05:11
words with women and there's speed tests
00:05:12
called implicit Association tests that
00:05:15
you can take that show you that even
00:05:16
when you'd say I a hundred percent
00:05:18
support women at work I do not hold any
00:05:20
belief that women need to be at home as
00:05:22
opposed to in the office uh in spite of
00:05:25
that
00:05:26
you'll still show these implicit
00:05:28
associations because it's like smog it's
00:05:31
been around your whole life and you
00:05:32
absorb these stereotypes that are in the
00:05:34
world so we had participants take one of
00:05:36
these tests and then we armed them with
00:05:39
tools that they could use to actually
00:05:41
try to combat bias at work we uh pointed
00:05:45
out to them for instance that one
00:05:46
strategy that's been proven effective
00:05:48
for reducing gender bias is say you're
00:05:50
evaluating CVS and you're worried you
00:05:52
might
00:05:53
be giving priority to men rather than
00:05:56
women who are otherwise
00:05:58
equally qualified for a job just because
00:06:00
of your implicit biases well one thing
00:06:02
you can do is you can evaluate CVS
00:06:04
without names associated with them it's
00:06:06
called blinding that's a way that you
00:06:07
can eliminate bias from that process so
00:06:10
we gave them a series of tools of that
00:06:12
sort and walked them through some
00:06:13
scenarios so they could think about how
00:06:15
to use them
00:06:16
we're really proud of what we built
00:06:18
so let me go back for a second because
00:06:20
the company that you worked with Fortune
00:06:22
500 company and you call them Brave how
00:06:25
did you convince them in the first place
00:06:27
to do this study because this could be
00:06:29
as you said some fairly sensitive areas
00:06:32
that they would be going into yeah it's
00:06:34
a really great question it was honestly
00:06:36
a company that was really intrinsically
00:06:38
motivated they care deeply about being a
00:06:41
thought leader in this space and they
00:06:43
were hopeful that if we could build
00:06:44
something together and I should say they
00:06:45
contributed a lot of insights because
00:06:48
um ideally in my from my perspective and
00:06:51
I know for my collaborators perspectives
00:06:52
as well when we're trying to build
00:06:54
something that can go in and improve
00:06:55
organizations it's really helpful to
00:06:57
have people inside those organizations
00:06:59
thinking through these issues with you
00:07:01
not just scientists who sit all day and
00:07:03
read academic papers and teach but
00:07:05
people who are on the front lines so
00:07:07
they they wanted to contribute to
00:07:08
building the best possible training that
00:07:10
could exist and then proving out you
00:07:12
know whether it added value and if it
00:07:13
did they would they would be a thought
00:07:15
leader in being able to share those
00:07:18
insights with the world and say that
00:07:20
that they were on the The Cutting Edge
00:07:21
so that was the motivation with the
00:07:23
recognition that there was a big risk if
00:07:25
things didn't work or if things looked
00:07:27
bad in the organization of of some kind
00:07:30
of lawsuit ensuing so it was a it's a
00:07:32
tricky balance to walk and I will say in
00:07:34
the end I'm not allowed to say the name
00:07:35
of the organization that we partnered
00:07:37
with and that was one of the
00:07:39
stipulations that we all agreed to to
00:07:40
protect them when we started the project
00:07:43
so when you're talking about diversity
00:07:45
training programs are there downsides
00:07:48
that that you see to that uh type of uh
00:07:51
concept and are there limitations as
00:07:54
well
00:07:54
yeah well it's a fantastic question
00:07:57
whether or not there may be some
00:07:59
downsides to diversity trading programs
00:08:01
and that's really what we set out to
00:08:02
test by developing this partnership with
00:08:05
the Fortune 500 company doing a
00:08:06
randomized controlled trial to evaluate
00:08:10
whether or not randomly assigning some
00:08:13
people to complete a diversity training
00:08:16
program and others to complete an
00:08:20
unrelated program for an hour would
00:08:22
produce benefits or not for women at the
00:08:25
organization we did this with thousands
00:08:26
of employees and we measured a number of
00:08:28
different outcomes we looked first at
00:08:32
attitudes at the end of both programs
00:08:34
both the diversity training and uh I'll
00:08:36
call it a placebo training program that
00:08:39
focused on entirely unrelated material
00:08:41
we asked people questions about their
00:08:43
attitudes towards women in the
00:08:44
organization and in general and and you
00:08:47
know ways of supporting them we had some
00:08:50
little scenario questions where we asked
00:08:52
in this scenario you know how might you
00:08:54
try to problem solve and the scenarios
00:08:56
related to supporting women so those
00:08:59
that's a simple sort of attitudinal or
00:09:02
scenario-based way of evaluating whether
00:09:04
the training had an impact but we also
00:09:06
really cared about Downstream
00:09:08
consequences would women be treated
00:09:10
differently in the organization would
00:09:12
people uh make more of an effort for
00:09:14
instance when they had an opportunity to
00:09:16
nominate women for awards would they be
00:09:18
more willing to Mentor women when
00:09:21
mentoring opportunities came up and so
00:09:23
we had a number of different measures
00:09:25
where we looked at exactly those kinds
00:09:28
of outcomes mentoring sign ups
00:09:31
Award nominations and even what we call
00:09:35
an a little audit experiment where
00:09:37
people an email went out and invited
00:09:39
people to offer support and information
00:09:42
to a new female employee at the company
00:09:45
or male employee and we looked at was
00:09:47
there a difference in the willingness to
00:09:49
support new female versus male employees
00:09:51
as a function of whether you'd been
00:09:53
through the training or not
00:09:54
okay so these are all different measures
00:09:57
and the question is what did we find
00:09:59
does it backfire does it work does the
00:10:01
diversity training produce the outcomes
00:10:02
we had hoped for all right I wish I had
00:10:05
a really simple answer for you if I had
00:10:07
to boil it down to one thing I would say
00:10:09
diversity training wildly underperforms
00:10:12
expectations but it's subtler than that
00:10:14
so
00:10:16
um on average we see that it has bigger
00:10:19
effects on attitudes than on actions
00:10:21
very little evidence of movement in
00:10:24
terms of behavior in terms of the
00:10:27
attitudinal lift though we do see that
00:10:29
it's having an impact particularly on
00:10:32
those who uh
00:10:35
the subpopulation they belong to
00:10:38
suggests they had more room for growth
00:10:40
because the average attitudes in that
00:10:42
subpopulation started out lower at
00:10:44
Baseline
00:10:45
um this means it's helping maybe change
00:10:47
the attitudes a bit more for uh men in
00:10:50
international settings uh in particular
00:10:53
how about Behavior change what I think
00:10:55
is really interesting about the behavior
00:10:57
changes was actually the opposite so the
00:10:59
people who whose Behavior was shifted
00:11:01
most tended to be the people whose
00:11:03
attitudes were most aligned to begin
00:11:05
with with the training
00:11:06
um in fact women and women are the the
00:11:09
group that ended up changing their
00:11:10
behaviors the most but one of the
00:11:13
interesting things is that some of the
00:11:14
behavior change we saw was not the kind
00:11:18
we were expecting so we were expecting
00:11:20
to lead more people to sign up to Mentor
00:11:23
women if they'd gone through the
00:11:24
training for instance and we do see that
00:11:27
but we actually also saw women looking
00:11:30
for mentorship themselves at a higher
00:11:32
rate when they'd gone through the
00:11:33
training which was really interesting so
00:11:36
one of the key results and PS this also
00:11:38
happens with minorities who go through
00:11:39
the training is that we may through the
00:11:41
training have made it more Salient to
00:11:44
women and minorities that they needed
00:11:46
sponsorship that there were threats to
00:11:48
their success in the organization and
00:11:49
they needed to look out for themselves
00:11:51
and find other women and minorities to
00:11:53
support them that's not normally what we
00:11:55
think of as the goal of a diversity
00:11:57
training right the goal of a diversity
00:11:58
training is how do we ensure that our
00:12:00
Workforce
00:12:01
um particularly those in positions of
00:12:02
Power are providing more support to
00:12:05
women and minorities instead what we're
00:12:06
doing is alerting women and minorities
00:12:08
to look out for themselves so it's not
00:12:11
necessarily a bad result that may be
00:12:14
beneficial but it's not the intention of
00:12:16
the program it's it's not the reason
00:12:18
that
00:12:19
billions of dollars are being poured
00:12:21
into these kinds of uh programs at
00:12:24
companies around the world so I guess
00:12:27
there is part of this that a lot of
00:12:31
people will think about the quote
00:12:32
unquote fixing the people but I would
00:12:34
think there's also have to be a focus to
00:12:37
a degree on fixing the system as well
00:12:40
correct
00:12:41
yeah I think that's one of the most
00:12:43
important takeaways from all the work
00:12:45
that I have done and all of the work I
00:12:47
have read in the last several decades of
00:12:50
blossoming amount of research in this
00:12:53
area of how do we increase diversity in
00:12:55
organizations that when we try to fix
00:12:57
people and say you know we're just going
00:12:59
to make the managers better at treating
00:13:01
women and minorities with respect or
00:13:04
we're going to make the organization
00:13:05
friendlier by by changing
00:13:09
um the way that we talk about diversity
00:13:12
these
00:13:14
Solutions I'm put I should put Solutions
00:13:16
in air quotes there these Solutions
00:13:19
generally if not
00:13:21
lived up to expectations
00:13:23
for instance our diversity training well
00:13:25
I should say a key limitation of the
00:13:28
research we did is that we're looking at
00:13:29
a one-hour online training you might see
00:13:32
something really different if you did a
00:13:34
week of training with a trained
00:13:35
facilitator and and you know really
00:13:37
hammered these points differently so
00:13:40
that's a limitation of the work but
00:13:42
still every study I have seen points in
00:13:44
the same direction of if we just try to
00:13:46
you know fix attitudes change beliefs um
00:13:50
try to change people the result it's
00:13:53
just much harder to do that this is true
00:13:55
not only when it comes to
00:13:58
gender diversity issues in organizations
00:14:00
which is what we're talking about now
00:14:02
but any kind of human biases and I tend
00:14:05
to study decision making more broadly I
00:14:07
look at gender diversity but I also look
00:14:09
at
00:14:09
other biases and judgment besides biases
00:14:13
against certain groups of people that
00:14:14
can lead us to make mistakes they're
00:14:16
incredibly hard to train away what we
00:14:18
find in both situations is that works
00:14:20
what works much better than training is
00:14:23
changing systems so Systems Support
00:14:25
better decisions
00:14:27
and and that's really what we're finding
00:14:29
time and again don't fix the person fix
00:14:30
the system they're embedded in so the
00:14:32
system is more is better structured to
00:14:34
support the outcomes we want to see and
00:14:37
I can talk more about what that means
00:14:39
exactly if you'd like well and can you
00:14:42
also uh talk about this concept of
00:14:44
isolated Choice effect and and how that
00:14:47
does work as well
00:14:49
so this is an example of changing
00:14:51
systems as opposed to changing people
00:14:53
and it's a project that actually I just
00:14:55
want to highlight we just talked about
00:14:57
this wonderful doctoral student from
00:14:58
Wharton who came in as a first year LED
00:15:00
this amazing team
00:15:02
um and in testing the value of diversity
00:15:05
training his name is Edward Chang and
00:15:07
Edward also is one of the co-lead
00:15:10
authors on this work on isolated Choice
00:15:12
effect along with another former Wharton
00:15:15
PhD student Erica kyrgios who's now a
00:15:18
professor at Chicago Booth so the folks
00:15:20
who were training and doing this work
00:15:21
are now thought leaders at all of our
00:15:23
peer institutions which just makes us
00:15:24
incredibly proud so um Edward and Erica
00:15:27
LED this project along with Anish Rai
00:15:30
and myself where what we were trying to
00:15:32
do is figure out whether or not we could
00:15:34
restructure the way that selection
00:15:37
decisions are made to improve outcomes
00:15:40
for women
00:15:41
what do I mean by selection decision so
00:15:43
you can think of a selection decision as
00:15:44
who am I going to hire for this job or
00:15:46
who am I going to put on this
00:15:47
prestigious uh committee or panel or you
00:15:51
know put up in front of my organization
00:15:52
and highlight as a as a star so who is
00:15:55
going to get selected for opportunities
00:15:57
that are important and and one way we
00:15:59
could make those kinds of selections is
00:16:03
um normally when we think about
00:16:04
promotions hires we can make those
00:16:06
decisions one at a time right we have
00:16:09
at the Wharton School we hire a bunch of
00:16:11
new faculty every year and you know we
00:16:14
often will have a separate search for
00:16:16
each and every faculty hire you know now
00:16:18
we're looking for someone in marketing
00:16:19
now we're looking for somebody in
00:16:21
operations and we're gonna look for one
00:16:23
person this spring in operations another
00:16:24
person in the fall so we isolate those
00:16:26
choices one Higher at a time
00:16:29
but another way you can make choices in
00:16:31
hiring decisions is actually in sets you
00:16:33
could say we're gonna have a cluster
00:16:35
higher we're gonna hire five new faculty
00:16:37
in this department right in the spring
00:16:39
we're gonna hire um or we're gonna put
00:16:41
five people in this award category as
00:16:44
opposed to one at a time so we have
00:16:46
options because any organization that is
00:16:49
growing is you know
00:16:50
it has opportunities and they can be
00:16:52
clustered or not and what we
00:16:55
hypothesized is that when people hire or
00:16:58
promote or select people for
00:17:00
opportunities one at a time
00:17:04
they focus on
00:17:08
just the attributes of the person in
00:17:11
front of them and don't think globally
00:17:13
about how that person is contributing to
00:17:16
the diversity of the organization
00:17:17
they're just looking at the one person
00:17:19
that's their focus when they hire in
00:17:22
sets our hypothesis was they're going to
00:17:25
because of the nature of the set attend
00:17:28
to what the group looks like right so if
00:17:31
I hire five people in a row at the
00:17:33
Wharton School one at a time I probably
00:17:35
don't even notice
00:17:37
how those five people look at it as a
00:17:39
group because I'm zoomed in looking at
00:17:40
the candidate set and choosing oh this
00:17:42
is the person I think is best for this
00:17:44
role this role and so on but if I hire
00:17:45
five at a time instead I'm thinking wow
00:17:49
oh gosh how did I end up hiring five
00:17:51
people who were all from the exact same
00:17:53
University and the exact same have the
00:17:56
exact same you know dissertation advisor
00:17:58
uh they all look the same they all walk
00:18:01
the same and talk the same I might
00:18:03
notice if there's a lack of diversity
00:18:04
because
00:18:05
hiring in a set forces me to attend to
00:18:08
that and that is what we found in study
00:18:10
after study actually when we randomly
00:18:11
assign people to make selection
00:18:12
decisions one at a time versus in sets
00:18:16
people select more diverse pools of
00:18:19
candidates when they're choosing from
00:18:20
exactly the same applicant pool for each
00:18:23
hire they make more diverse Selections
00:18:25
in sets than in Singletons the set
00:18:28
forces a focus on the aggregate on
00:18:31
whether you're creating a pool that has
00:18:33
diversity that that supports your values
00:18:35
that um that reflects the diversity you
00:18:38
want to have in uh an effective
00:18:41
organization but when you hire an
00:18:43
isolation you don't have that so that's
00:18:44
a structural change it's not I'm not
00:18:47
changing the people who are making
00:18:48
higher decisions hiring decisions I'm
00:18:50
not training them I'm not suggesting to
00:18:51
them to focus on diversity but what
00:18:54
naturally happens when we look at sets
00:18:57
is we think about diversity and when we
00:18:59
look at Singletons we don't and PS this
00:19:01
is something that's been known for a
00:19:03
while when people make product choices
00:19:04
right you're you're ordering snap you
00:19:06
know you get to order five days of lunch
00:19:08
for the week ahead all at once or you
00:19:10
order Monday Tuesday Wednesday Thursday
00:19:11
Friday people order more diverse menu
00:19:14
sets if they're choosing all once
00:19:16
because they're like oh I don't have
00:19:17
salmon every day but if they're if
00:19:19
they're ordering every day they pick the
00:19:21
same dish every day in and day out
00:19:23
because they're they're focusing just on
00:19:24
that one choice so it's interesting that
00:19:26
the same phenomenon that would lead to
00:19:28
different behaviors when we're thinking
00:19:30
about products also leads to these
00:19:32
different patterns of hiring behavior
00:19:33
that can be so important if you want to
00:19:35
create a diverse organization
00:19:38
you've also looked at how social norms
00:19:40
uh affect a group composition and tend
00:19:44
to contribute to underrepresentation of
00:19:46
women what did you find happening there
00:19:49
yeah yeah so you will laugh because yet
00:19:53
who was leading this work my amazing
00:19:55
former student Edward Chang who's now
00:19:57
professor of the Harvard Business School
00:19:58
so I I do a lot of work with doctoral
00:20:01
students on this topic and Edward was an
00:20:03
incredibly productive
00:20:05
um student who really is passionate
00:20:07
about understanding gender diversity or
00:20:10
in organizations and we had a great
00:20:11
series of collaborations
00:20:14
um this is actually though I have to say
00:20:15
this is a project the idea came from my
00:20:17
husband so I just have to give him a
00:20:19
shout out he's a physics Professor here
00:20:20
at Penn and he said to me
00:20:23
Katie I have noticed that it seems like
00:20:26
when a physics department decides they
00:20:29
lack gender diversity they panic
00:20:32
they put a whole lot of effort into a
00:20:36
higher finding stealing a woman from
00:20:38
another top institution they get one and
00:20:41
then they breathe a sigh of relief say
00:20:42
our problem is solved and they never
00:20:44
think about it or talk about it again
00:20:45
and he said does that happen is it just
00:20:47
like are people just trying to grab
00:20:49
tokens and then they they put a check
00:20:51
box and they quit and I said you know
00:20:53
that's a really interesting question I
00:20:55
have this amazing new doctoral student
00:20:56
his name is Edward Chang let me see if
00:20:58
we can come up with a way to test your
00:21:00
hypothesis but in a setting that we
00:21:02
think might be more consequential even
00:21:04
than academic physics departments
00:21:06
um we decided to take it to the
00:21:07
boardroom
00:21:08
and what we wanted to do and this is
00:21:10
Joint work I should say with um midupak
00:21:12
andola of Colombia and Dolly chug of NYU
00:21:16
is we grabbed the universe of data on
00:21:19
who sits on corporate boards at Fortune
00:21:22
500 companies we actually also looked at
00:21:24
for the fortune 1500 in the study but we
00:21:26
zoomed in particularly on the Fortune
00:21:28
500 and some of the analyzes and
00:21:31
um we looked for at the distribution of
00:21:34
the number of women on boards and we
00:21:36
realized that if there were cliffs in
00:21:39
that distribution meaning if there was a
00:21:42
point at which you could clearly see a
00:21:45
giant discontinuity in the
00:21:47
representation of women that might
00:21:49
suggest something about organizations
00:21:52
satisficing trying to sort of reach a
00:21:54
specific point and then quitting on
00:21:56
their efforts to achieve diversity in
00:21:58
the boardroom along the lines of my
00:22:01
husband's hypothesis in physics
00:22:02
departments so
00:22:05
um we actually did something that I'm
00:22:07
very proud of we created this little
00:22:08
simulator it's like taking all the board
00:22:11
directors and and all the fortune 1500
00:22:13
companies and playing musical chairs
00:22:15
with them and we'll say you know imagine
00:22:17
there's no these are the people who are
00:22:19
on boards and well even you know people
00:22:21
are on multiple boards will keep them on
00:22:22
multiple boards but we're going to
00:22:23
Resort them and reshuffle and we'll do
00:22:25
it a thousand times and we'll figure out
00:22:27
what would the distribution of women
00:22:29
look like if this is how boards were
00:22:30
created with no attention to the uh
00:22:34
diversity just taking the people who are
00:22:37
available and and shuffling them
00:22:39
randomly and then we're going to compare
00:22:41
that random shuffle to what we actually
00:22:44
see and see if it looks like there's any
00:22:46
contortions any Cliffs suggesting a
00:22:49
magic number above which boards stop
00:22:50
making an effort to achieve women and
00:22:52
when we compare the two distributions
00:22:54
the musical chairs that's gender neutral
00:22:56
and the reality we see a huge gap the
00:23:00
huge gap is actually not at the magic
00:23:02
number one it's at the magic number two
00:23:05
so boards you can see this giant
00:23:07
discontinuity boards are racing to get
00:23:09
exactly two women and then they stop
00:23:10
trying there's a a giant drop off
00:23:13
relative to what you'd expect if they
00:23:15
were just you know seeding women uh at
00:23:18
the same rate as as any other member of
00:23:20
the population that's out there and
00:23:21
available for board seats
00:23:24
um and interestingly when we do a
00:23:26
historical analysis we can go back in
00:23:28
time with our time machine because
00:23:30
boards the seed who's Seated on boards
00:23:32
has been tracked for ages we can
00:23:34
actually see that there was a transition
00:23:35
point from tokenism it used to be the
00:23:37
ones where the magic number when boards
00:23:39
quit to tucanism
00:23:41
um about a little over a decade ago uh
00:23:45
so
00:23:46
um what's going on you might be asking
00:23:49
it turns out the social Norm at this
00:23:51
point is most boards have two of them
00:23:53
that's the average number and if you
00:23:55
have less uh you are deviating from all
00:23:58
the others and you might be called out
00:23:59
in the media and uh you might be you
00:24:03
know labeled the kind of company that's
00:24:05
not supporting women and putting them on
00:24:07
your boards and so we actually see that
00:24:10
um companies that tend to be more
00:24:11
scrutinized seem to show this Cliff to a
00:24:14
greater degree companies of the Fortune
00:24:16
500 which are particularly under the
00:24:18
microscope more than the 1500 show a
00:24:20
greater degree of this bias and we ran
00:24:22
little experiments where we showed
00:24:25
um when you're choosing who to add to a
00:24:27
group outside of the boardroom right
00:24:30
just any sort of selection decision
00:24:32
social Norm information is very Salient
00:24:35
and leads to these kinds of clustering
00:24:36
effects nobody wants to be the outlier
00:24:38
has fewer women than usual because they
00:24:41
worry that that will yield you know
00:24:44
negative repercussions so I think this
00:24:46
is another really interesting finding
00:24:48
and what's important about it to me is
00:24:50
what it shows is the power of scrutiny
00:24:52
that one of the reasons that
00:24:54
organizations especially at the highest
00:24:55
levels are attending to diversity and
00:24:58
and making an effort is they don't want
00:25:00
to be called out and that gives us power
00:25:02
because if we as a as a community we is
00:25:05
a society want to see greater diversity
00:25:09
then we need to continue to point out
00:25:13
when organizations aren't achieving it
00:25:15
and that is highly motivating and leads
00:25:17
them to better Behavior now does that
00:25:19
mean that we're going to see some
00:25:20
bizarre
00:25:22
actions as a result like clustering and
00:25:24
sort of everybody else has two so we're
00:25:26
okay if we have two yeah it does
00:25:30
um but net net it means that the
00:25:31
pressure and the scrutiny are working
00:25:33
and we're seeing this upward pressure
00:25:35
continue and the number of women on
00:25:36
boards the average keeps creeping up and
00:25:38
just as we saw a Tipping Point where
00:25:40
eventually um tokenism gave way to
00:25:42
tokenism hopefully we'll get to the
00:25:44
point where it's humiliating to have a
00:25:46
board with uh less than three women and
00:25:48
we need to put the same pressure I
00:25:50
should say in terms of board diversity
00:25:52
in my opinion on um on on minority
00:25:55
representation uh so I think we're doing
00:25:57
better with gender representation and
00:26:00
creating that scrutiny right now than
00:26:01
than we have with
00:26:03
um creating scrutiny around minority
00:26:05
representation so what then should
00:26:07
organizations do uh in terms of
00:26:12
confronting gender bias what are the
00:26:13
recommendations that you believe they
00:26:15
should consider
00:26:17
few recommendations so one thing I want
00:26:19
to start with is I don't think you
00:26:21
should use diversity training as your
00:26:22
solution that is not going to fix your
00:26:25
problem
00:26:27
um to a thing that you can do is make
00:26:30
sure you shine a light on how things are
00:26:33
going in different groups because
00:26:35
scrutiny increases the social pressure
00:26:37
people feel to make different kinds of
00:26:40
hiring and promotion decisions and
00:26:42
attend to a lack of diversity so
00:26:44
scrutiny matters highlighting social
00:26:46
norms when they're positive social norms
00:26:48
meaning hey you're the only group in the
00:26:49
company that hasn't yet promoted a woman
00:26:51
to this a senior ranks that kind of
00:26:54
scrutiny and those kinds of social norms
00:26:56
matter
00:26:57
um third when you can hire in sets
00:26:59
rather than in Singletons that is going
00:27:01
to lead to Greater diversity in your
00:27:03
hiring pool because only when we hire in
00:27:05
sets do we seem to really attend to
00:27:07
these issues of diversity and then a
00:27:09
final Point that's unrelated to my
00:27:11
research but I think an incredibly
00:27:12
important finding that should be used
00:27:14
more here is that there's been some
00:27:16
really wonderful research done led by
00:27:19
um Joyce he of UCLA and Sonia Kang of
00:27:22
the University of Toronto showing that
00:27:23
uh women
00:27:25
because of probably stereotypes and the
00:27:30
backlash they sometimes face when they
00:27:33
um self-nominate for things like
00:27:34
promotions are very much less likely to
00:27:39
put their hand up when it's time to be
00:27:41
promoted even when they've performed at
00:27:43
the same level as others and this this
00:27:46
is research also by Muriel neederly of
00:27:47
Stanford showing women under compete
00:27:49
relative to men when they have the same
00:27:50
credentials so what Sonia and Joyce did
00:27:55
is they they experimented with changing
00:27:58
organizational defaults meaning in most
00:28:00
organizations if you want a promotion
00:28:03
you have to say I would like to apply
00:28:04
for this promotion but what if you
00:28:06
flipped the structure
00:28:08
and you said everyone who has met a
00:28:10
certain qualification threshold is going
00:28:12
to be
00:28:13
considered for a promotion you you're
00:28:16
welcome to opt out if you really don't
00:28:17
want to be leveled up in this
00:28:18
organization there may be personal
00:28:19
reasons you really don't but but you
00:28:22
will be considered unless you request
00:28:24
that you not be considered and guess
00:28:26
what women
00:28:28
ended up being in the consideration set
00:28:30
and therefore promoted dramatically more
00:28:32
in the second system where the default
00:28:34
the the situation is that you'll be
00:28:38
considered whether you volunteer or not
00:28:39
so thinking about the fact that women
00:28:41
may be more hesitant to put their name
00:28:43
forward they may be less willing to
00:28:45
compete and creating structures where
00:28:47
it's easier because the friction is in
00:28:49
the other direction where everyone's
00:28:51
going to compete and you have to
00:28:53
actually exert some energy to avoid it
00:28:55
that can create a More Level Playing
00:28:57
Field
00:28:58
as for women in the workplace what you
00:29:01
think they can do to try and level that
00:29:03
playing field
00:29:06
well
00:29:07
one thing that I think is really
00:29:09
important is having a strong group of
00:29:12
mentors we know that's important and
00:29:13
that women tend to have weaker networks
00:29:16
than men
00:29:18
um because uh frankly we we uh we know
00:29:21
about homophily which is a tendency to
00:29:23
affiliate with others who are like
00:29:24
ourselves and if you're at the top of an
00:29:26
organization and you're a woman you'll
00:29:27
find fewer people who are like you you
00:29:29
can affiliate with and chat with at the
00:29:31
water cooler so that's a challenge
00:29:33
um and and recognizing that and actually
00:29:35
making extra efforts to to look out for
00:29:38
opportunities to connect with others to
00:29:40
make sure you have strong networks and
00:29:43
strong bonds and that's really important
00:29:47
um I actually have a group in my own
00:29:49
life that is a no club and this is based
00:29:53
on Research that was done by Linda
00:29:55
Babcock at Carnegie Mellon University
00:29:57
and collaborators showing that women are
00:30:00
too willing to say yes to non-promotable
00:30:02
tasks at work we're too quick to do sort
00:30:06
of the office housework taking those
00:30:07
votes in a meeting organizing the
00:30:08
holiday party the kinds of things that
00:30:10
aren't ultimately rewarded but can be
00:30:12
very time consuming and
00:30:15
um Linda and her collaborators their
00:30:17
solution to this and they've written a
00:30:18
wonderful book by the way called the no
00:30:20
Club was to create a group of women who
00:30:22
were at a similar career stage who
00:30:24
helped support each other in saying no
00:30:26
when it was necessary because
00:30:27
interestingly even though we're bad at
00:30:29
saying no for ourselves women are are
00:30:31
just as good at men as saying no for
00:30:33
others recognizing when others should
00:30:34
make a certain decision
00:30:36
um we're very good at arguing for other
00:30:38
people
00:30:39
so uh and we're good at arguing for
00:30:42
ourselves too when we're comfortable
00:30:43
doing it it's just there's a lot of
00:30:46
pressure against it and often we conform
00:30:49
or or fold in the face of that pressure
00:30:51
so I actually have a no Club in my life
00:30:53
it includes a couple of other female
00:30:56
academics at a similar career phase and
00:30:58
when we Face challenges about things
00:31:00
that we're not sure is this a yes or
00:31:02
isn't a no and frankly advice in general
00:31:04
we reach out to each other and it has
00:31:05
been the most amazing resource for a few
00:31:08
reasons
00:31:09
um one is the expected I expected it
00:31:11
would be a great resource because they'd
00:31:13
give me wisdom and sort of Consulting
00:31:15
from brilliant people it also creates
00:31:18
social connection and makes me enjoy my
00:31:21
work more to have these awesome
00:31:22
supporters in in my camp and to be
00:31:25
supporting these other terrific women
00:31:27
um but the third reason which I didn't
00:31:29
anticipate but his really been there too
00:31:32
is that every time they ask me for
00:31:35
advice I'm learning so they will face a
00:31:38
challenge they will need to figure out
00:31:41
you know should I do this or not and
00:31:42
from an arm's length View and when it's
00:31:43
not me I Can See Clearly what the answer
00:31:45
is and then once I've said oh no you
00:31:48
should definitely say no to that
00:31:49
opportunity I'm confident that this is
00:31:51
the wrong thing for you when I face a
00:31:53
similar challenge it's built my
00:31:54
confidence and competence to respond and
00:31:57
handle that situation myself and what's
00:31:59
really interesting actually is I've
00:32:00
gotten to be involved in some research
00:32:02
led by Lauren s Chris Winkler now at
00:32:04
Kellogg former Wharton postdoc showing
00:32:07
that when we give advice to others it
00:32:09
actually improves our own outcomes not
00:32:11
just in the domain of you know this no
00:32:14
club that I'm describing where a lot of
00:32:16
the decisions have to do with gender and
00:32:18
work but you know high school students
00:32:20
who we randomly assign to give advice to
00:32:22
their younger peers get better grades
00:32:24
because they just coach someone else on
00:32:26
how to get good grades and when you
00:32:27
coach you learn so I think a really
00:32:29
interesting thing that women can do is
00:32:30
actually build these kinds of advice
00:32:32
clubs so that they have stronger
00:32:34
networks stronger social connections
00:32:36
they're benefiting from others who can
00:32:38
see clearly in places where they can't
00:32:39
how to not conform to stereotypes but
00:32:42
make those decisions that are better for
00:32:43
their career and in advising others
00:32:46
they're going to gain benefits to
00:32:48
competence confidence and social support
00:32:52
so I would advise all women to have
00:32:54
advice clubs and PS men should have them
00:32:56
as well but this may be a particularly
00:32:58
useful tool for women who can't say no
00:33:00
where do you think that the gender
00:33:03
equality
00:33:04
question ends up taking us let's say a
00:33:07
decade from now
00:33:10
honestly I'm such an optimist especially
00:33:13
after doing for instance the work we
00:33:15
discussed on
00:33:16
toucanism and corporate boards and
00:33:19
seeing how things have evolved now
00:33:21
they've evolved slowly I wish they'd
00:33:23
evolved a lot faster but we're seeing
00:33:25
progress in the right direction and I
00:33:28
think the pressure and scrutiny have
00:33:30
accelerated
00:33:32
um so I'm hopeful I'm feeling very
00:33:34
hopeful that things are going to
00:33:36
continue to get better and there's
00:33:38
certainly a lot of scientific attention
00:33:40
on this question when I was a graduate
00:33:43
student first starting to do some
00:33:44
research related to race and gender bias
00:33:46
and and how can we solve it it was not a
00:33:49
very popular area and I think as it's
00:33:52
become easier and easier to collect
00:33:54
massive data sets as a b testing has
00:33:57
become more straightforward to do in
00:33:58
organizations and more accepted maybe in
00:34:00
part because of the you know Tech being
00:34:02
such a a big part of
00:34:05
um the corporate world and and so much a
00:34:07
b testing there it's becoming more part
00:34:08
of the culture we're accelerating
00:34:11
insights about what works and as those
00:34:14
insights become more widely adopted I
00:34:16
think a along with the increased
00:34:18
attention to these issues and desire for
00:34:21
equality I'm optimistic that we can use
00:34:24
science to get where we want to be
00:34:26
[Music]
00:34:27
Katie thanks very much for your time
00:34:29
today all the best
00:34:31
thank you so much for the great
00:34:32
questions and for having me on the show
00:34:34
thank you Katie Milkman who's a
00:34:36
professor of operations information and
00:34:38
decisions here at the Wharton School
00:34:41
thank you for listening to the ripple
00:34:43
effect we hope you found this episode
00:34:44
informative and engaging don't forget to
00:34:47
subscribe and leave us a review so that
00:34:49
we can continue to bring you the best
00:34:51
Insight from the warden School
