LILY JAMPOL: So I want to ask you guys a question.

Do you-- yes or no-- consider yourselves, generally speaking,

to be nice?

A nice person?

AUDIENCE: Yes.

LILY JAMPOL: Yes.

Of course you are.

[LAUGHS] Now I'm going to upset you

by telling you about how sometimes being nice

can have unintended and ironic consequences for the workplace.

So in the 2009 film "Up in the Air,"

George Clooney plays a character who is hired by other companies

to go and fire their employees so

that they don't have to do it.

Besides being a pretty good film,

what the plot captures perfectly is

that giving feedback to others is pretty terrible for everyone

involved, especially if it's negative feedback.

In fact, research has found that supervisors

will delay giving feedback, not give it at all,

or even distort the contents of that feedback

so that it's not quite as harsh.

What we know as white lies.

And this tension between truthfulness but also

not wanting to hurt people's feelings

can be exacerbated when the recipient of that feedback

is particularly empathy-eliciting.

And historically speaking, women have stereotypically

been seen to be less competent than men

but also more emotionally unstable and vulnerable.

And these beliefs, despite being often subconscious or implicit,

can lead to protective or sometimes even patronizing

attitudes and behaviors towards women.

So my question was, if it's hard to give feedback

and women are seen with more empathy,

does that mean that women are going

to be told more white lies about their performance than men?

And in order to test this question,

we put people in a situation that

would mimic this tension between truthfulness and empathy.

And we asked them to give feedback directly

to some other participants.

So we told our participants that they

were going to be working remotely

with two other students, who we identified as AM and SB,

and who were, unbeknownst to them,

actually computer-generated programs.

So they weren't real participants.

And we asked them to evaluate essays

that AM and SB had written and evaluate them on a 0

to 100 scale and tell us what they thought

about these particular essays.

So this is phase one.

Once they had submitted those ratings to us, in phase two,

we told them, now you're gonna have a chance

to give feedback directly to these other students.

And this time, we had the other fake students

introduce themselves using their real names, Andrew and Sarah,

which we used as a subtle manipulation of gender.

So now they were both assigned genders.

And then we asked the participants to rate

the essays again using the exact same scale

as they had before, only this time they

would submit those ratings directly to the participants.

And then we measured the degree of white lie

by taking the difference between phase two and phase one.

And what we found is despite there being

no difference in the phase one ratings-- that is,

before they knew the gender, people were rating equally bad

the essays, about 35% on a 100% scale.

In phase two, people were telling Sarah

that she was doing more than 15% better than they had told us

they thought she was doing, indicating

that at least to some degree, Sarah was getting white lies.

And we've replicated this in other studies since then.

And we also asked people, did you

know that you lied more to Sarah than to Andrew?

And people at least reported that they

had no recollection of having given biased feedback to one

participant over the other.

All right.

So this isn't necessarily a problem

if women actually do prefer hearing white lies, right?

[LAUGHTER]

Well, you could argue that that is a problem.

But in a follow-up study, we decided to ask them,

do you prefer hearing white lies?

And we asked both women and men.

It turns out that no, everybody actually expressed a preference

for hearing the truth.

So at least on a nominal level, people

desire to hear truthful feedback, even if it was harsh.

Now it seems that women are being

told more white lies than men.

But there are other broader implications

of this besides just this incongruency problem.

First and foremost, a lack of information

means that women-- and possibly other disadvantaged groups--

are not getting the information that they need in order

to improve.

And if they do detect that supervisors are lying,

this could also lead to resentment and demotivation.

And all this has implications not just

for individual performance but also for the performance

of the workplace and the organizations.

And I have to thank Nancy for teaching me

how to draw hair on stick figures using PowerPoint.

It's pretty awesome.

All right.

So what can we do?

What are some practical interventions?

First and foremost, and this has been

brought up a bit with the conversation

on subconscious bias, but increasing awareness

is really important.

Particularly important is having everyone, both employees

and employers, understand that good people

with good or chivalrous intentions

can have ironic consequences for equality in the workplace.

Secondly, some practical interventions we could do

would be to have employees ask for truthful feedback.

So in feedback situations, this might alleviate the pressure

there is on supervisors to be nice instead of be truthful.

And likewise, supervisors could prompt employees

to express their preferences during those feedback

situations.

Another action that could be taken

is to have supervisors concentrate on the broader

goals of the organization and the concrete improvements

that employees could make, given truthful feedback.

And not just on avoiding negative interactions

with people when they're face to face with them.

Also we've found that people who've had experience

giving feedback tend to show this bias a little bit less.

So perhaps running simulations or training programs,

putting supervisors in a series of being

able to deal with different kinds of feedback situations

would more adequately prepare them

to be able to give unbiased feedback.

Ultimately, there's been a lot of progress

in terms of equality.

However, the glass ceiling still exists

for women in the workplace.

Making broad structural changes, such as maternity care,

are really important and necessary steps.

However, understanding how biases-- subtle ones,

like the one that I've just shown you-- play out

through common and normal workplace routines

is also really, really essential.

It's not just for breaking down those barriers

and identifying these biases, but also

for developing interventions that

will make the workplace a more equal and hospitable place

for everyone.

Thanks.

[APPLAUSE]

[MUSIC PLAYING]