Popular in Social Sciences

Political Science

V-Dem’s Conceptions of Democracy and Their Consequences

University of Notre Dame Michael Coppedge

Central Monetary Authorities

Macroeconomic modelling of CBDC: a critical review

European Central Bank Ulrich Bindseil

Social Identities and Divisions

A new explanation of antisemitism: Jew hatred as a civilisational phenomenon

Boston University Liah Greenfeld

Central Monetary Authorities

Financing Private Credit

Federal Reserve Bank of New York Nina Boyarchenko

Economics

Failing Banks

Massachusetts Institute of Technology Emil Verner

Culture and Society

The City as a Growth Machine: Toward a Political Economy of Place

New York University Harvey Moltoch

American Politics

Gender Diversity and Diversity of Ideas

View more insights in Social Sciences - Economics - Labor Economics

View more figures from Cornell University

View the page of Michèle Belot

Read the Study

Next in Economics

Do renewable energy investments create local jobs?

Natalia Fabra | University Carlos III Madrid

An early indicator for anomalous stock market performance

Thomas Gries | Universität Paderborn

Diversity in employee representation is often advocated for its potential to promote the diversity of ideas, and thereby innovation. Michèle Belot discusses the phenomenon of ‘idea homophily’, which is a tendency to be more interested in ideas closer to one’s own.

Image courtesy of interviewee. February 14, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video video discusses the role of bias in the selection of ideas and candidates within the economy and academia, focusing on how these biases may influence employment opportunities and the diversity of ideas contributed by various demographics. Economists are interested in innovation and the processes through which new ideas emerge, especially in relation to job creation and societal value.

The speaker highlights that individuals often favor ideas similar to their own, which can affect hiring decisions and academic publications. They emphasize the issue of underrepresentation in various fields, noting particularly the dominance of white males in many sectors. Despite efforts to include diverse groups, biases may still lead to a preference for ideas and candidates that resemble existing norms.

They present research findings indicating that women specializing in traditionally male-dominated fields tend to perform well in job markets. However, they stress the complexity of interpreting these results due to confounding factors related to individual capabilities and employer preferences.

An experimental study is outlined, evaluating how subtle incentives can encourage engagement with underrepresented ideas. The results reveal significant gender differences in preferences for topics, with men leaning towards technology and business, while women gravitate towards health and environment-related discussions.

The speaker concludes that while diversity initiatives are essential, they should not overlook merit and quality. Policies aimed at improving diversity should focus on broadening the spectrum of fields and ideas rather than merely increasing representation. This video video includes raising awareness of biases and encouraging an openness to ideas from different backgrounds to leverage the full benefits of diversity.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.105 --> 00:02.085
So we are interested, we are economists,

00:02.465 --> 00:05.685
so we are interested in ideas

00:05.865 --> 00:08.965
and innovation in general in many sectors of the economy.

00:09.115 --> 00:10.565
It's important to think about

00:11.065 --> 00:12.885
how do you come up with new ideas.

00:13.075 --> 00:16.805
It's very important to create employment, create new jobs,

00:16.985 --> 00:20.685
and also just to add value in, in society

00:20.825 --> 00:22.005
and, and in the economy.

00:22.625 --> 00:24.405
And so we are interested in

00:24.665 --> 00:26.765
how do we actually come about come,

00:27.545 --> 00:29.005
how do new ideas come about?

00:29.545 --> 00:32.245
And often you think about, well, people have ideas

00:32.385 --> 00:33.565
and implement these ideas,

00:33.945 --> 00:36.085
and then they make their way into society.

00:37.025 --> 00:40.565
Now, this process of actually selecting ideas

00:40.785 --> 00:44.845
and implementing ideas is maybe not as innocuous

00:44.905 --> 00:48.405
as we think, because there's often actually employers

00:48.465 --> 00:51.005
who select, you know, who they are going to hire.

00:51.385 --> 00:54.325
And then when you think about academia, we have a lot of,

00:54.585 --> 00:56.765
you know, institutions in place

00:56.765 --> 01:00.485
that basically select which ideas are gonna be published,

01:00.665 --> 01:03.445
which ideas are patented and so on.

01:04.145 --> 01:07.485
So we're interested in the fact that, the people

01:07.665 --> 01:10.565
who might be doing this selection obviously have their own,

01:10.625 --> 01:12.725
you know, preferences and interests

01:12.725 --> 01:15.685
and beliefs in what might be good ideas.

01:16.505 --> 01:19.565
Um, and our conjecture was that maybe,

01:19.905 --> 01:23.765
and this would apply to all of us, we may have a little bit

01:23.765 --> 01:28.245
of a bias towards ideas that are close to our own ideas.

01:28.625 --> 01:30.165
So it's almost like a totology,

01:30.345 --> 01:32.605
but of course, I'm gonna be more interested in ideas

01:32.715 --> 01:34.685
that are close to my own ideas.

01:35.265 --> 01:38.405
And so, if you think about an employer who is interviewing,

01:38.785 --> 01:41.085
you know, a number of, of candidates,

01:41.345 --> 01:44.365
and all these candidates have different ideas,

01:44.625 --> 01:47.365
you may think that maybe they're gonna be more enthusiastic,

01:47.365 --> 01:49.845
more excited by candidates who have ideas

01:49.875 --> 01:52.925
that really resonate with, with the person

01:53.025 --> 01:54.205
who is making the decision.

01:54.825 --> 01:57.685
Um, which is this idea of idea mo fully.

01:58.185 --> 02:02.165
Now, of course, because in many sectors of the economy,

02:02.995 --> 02:05.125
well, we have a history of, of course,

02:05.145 --> 02:08.445
the labor force being more white male, dominated.

02:09.025 --> 02:12.445
Uh, then the, the, the question is whether this actually

02:12.655 --> 02:13.965
influences,

02:14.215 --> 02:17.885
which ideas gets elected even when there are efforts

02:17.985 --> 02:21.925
to actually try to bring in, underrepresented groups.

02:22.145 --> 02:23.245
And so this is something

02:23.245 --> 02:26.125
that we know is very much in the air in many sectors

02:26.125 --> 02:29.325
of the economy, in academia, there's a lot of attention

02:29.625 --> 02:33.485
to try to bring in underrepresented groups in academia.

02:33.665 --> 02:36.605
And then you sort of think, well, this could be a good idea

02:36.605 --> 02:40.285
because also just for the, the, the purpose

02:40.425 --> 02:44.365
of actually generating new ideas, novel angles and so on.

02:44.985 --> 02:48.885
Um, but if, if there is at the same time this tendency

02:48.945 --> 02:53.365
to try to go for things that resemble all the ideas,

02:53.515 --> 02:55.765
then you don't get the benefit of the,

02:56.025 --> 02:57.565
the full benefit of diversity.

02:58.305 --> 03:00.545
So this is a, in a nutshell, the sort

03:00.545 --> 03:01.985
of the context of the study.

03:02.485 --> 03:06.865
And we were interested in studying this, with real,

03:07.005 --> 03:08.905
you know, data from the real world.

03:09.085 --> 03:10.985
But, we focus on academia

03:10.985 --> 03:14.025
because this is a, sector we know well

03:14.025 --> 03:15.985
and where ideas are very important.

03:16.805 --> 03:20.625
And then we also have additional evidence from an experiment

03:20.625 --> 03:22.865
where we are just really interested in

03:22.865 --> 03:24.825
what ideas people engage with

03:25.605 --> 03:28.705
and whether the ideas they're happy to engage

03:28.735 --> 03:32.145
with are in a way malleable if we give them a little bit

03:32.145 --> 03:34.985
of a nudge, a little bit of an incentive to actually engage

03:35.015 --> 03:36.025
with other ideas.

03:36.395 --> 03:37.905
There is a lot of literature,

03:37.965 --> 03:40.345
and actually quite recently, perhaps even more,

03:40.345 --> 03:42.705
this is really a booming, you know, area.

03:43.445 --> 03:47.745
Um, of course we know that they are, you know,

03:47.745 --> 03:50.105
there are disparities between men

03:50.105 --> 03:52.865
and women, like in other sectors of the economy.

03:53.405 --> 03:57.385
So it's harder for, women, for example, to progress

03:57.385 --> 04:00.465
through, the academic profession

04:00.605 --> 04:02.145
to reach, you know, higher levels.

04:02.275 --> 04:06.905
There are fewer, female professors than they are,

04:06.925 --> 04:08.985
female, junior professors.

04:09.085 --> 04:11.945
And so as if you look at every level, you see that

04:12.445 --> 04:15.465
as you go higher up, then there are less and less women.

04:15.885 --> 04:19.225
It is also true for other, underrepresented,

04:19.525 --> 04:20.665
minorities.

04:21.365 --> 04:23.625
Uh, and there are lots of, you know, discussions about

04:23.625 --> 04:24.785
what that may be the case.

04:24.845 --> 04:27.385
And this is what the literature is really been tackling.

04:27.525 --> 04:28.945
You know, what are the, the barriers?

04:29.445 --> 04:34.185
Is it actually on the side of, the, the women themselves

04:34.285 --> 04:36.505
who are actually maybe have different preferences

04:36.765 --> 04:41.185
or might actually, be, finding some characteristics

04:41.245 --> 04:44.505
of the academic world, not so attractive, whether it's,

04:44.765 --> 04:46.305
the fact that it's competitive

04:46.605 --> 04:50.425
or the fact that, you actually have to go

04:50.425 --> 04:54.105
through a tenure track which overlaps with, years

04:54.155 --> 04:58.065
where many women would like to have children and, and so on.

04:58.085 --> 05:00.665
So this is a, a, a quite a broad literature.

05:01.165 --> 05:06.105
Now, our work is really, to try to think about barriers

05:06.255 --> 05:08.065
that a barrier that has maybe

05:08.645 --> 05:11.145
not been receiving that much attention.

05:11.645 --> 05:16.265
And, which is this idea that, well, the, the,

05:16.285 --> 05:18.745
the sort of, we are moving from an environment

05:19.095 --> 05:21.825
that has been built by, you know, men

05:22.445 --> 05:25.545
and then we are trying to move into a, an environment

05:25.545 --> 05:29.585
that, welcomes, you know, other, other groups.

05:29.765 --> 05:31.745
And I think that there is a genuine, you know,

05:32.505 --> 05:34.665
interest in getting, you know, these other groups.

05:35.205 --> 05:37.705
But at the same time, we think there might be,

05:37.705 --> 05:42.065
we don't always realize maybe, the sort of, biases

05:42.095 --> 05:45.625
that you, we might have towards, for example, what are the,

05:45.845 --> 05:47.625
the ideas, the fields that we should,

05:47.625 --> 05:49.625
should be investing in and working on?

05:49.845 --> 05:53.305
So our first, study is a study, from the field.

05:53.405 --> 05:56.265
So that means we have real world, data,

05:56.985 --> 05:59.205
And we look at placement.

05:59.265 --> 06:00.725
So we look at the first job

06:01.275 --> 06:05.565
that a PhD candidates get when they graduate,

06:05.665 --> 06:09.685
in economics from the top institutions in the us.

06:09.745 --> 06:12.685
So we have looked at the top 33 institutions in the US

06:13.345 --> 06:15.805
and we look at where they find a job

06:15.945 --> 06:18.605
and whether in particular they're likely to find a job

06:18.625 --> 06:21.965
as assistant professor, which is probably the most,

06:22.025 --> 06:25.085
you know, the best job you can hope to get when you,

06:25.565 --> 06:28.885
graduate as a PhD, student in economics.

06:29.345 --> 06:32.645
Um, so this is a desirable, type of position.

06:33.225 --> 06:35.925
And we look at, whether men

06:35.925 --> 06:38.925
and women are equally likely to find such jobs,

06:39.425 --> 06:42.565
but also more, perhaps more relevant to the topic

06:43.265 --> 06:47.085
is whether, the area they specialize in matters.

06:47.945 --> 06:50.005
Uh, and so the hypothesis was

06:50.005 --> 06:53.525
that if there is something at work, like ideal moly,

06:54.145 --> 06:56.845
you might expect that maybe the people

06:56.945 --> 07:00.805
who would do best in this market are people who have id

07:00.985 --> 07:04.565
or who specialize in areas that are similar to

07:04.565 --> 07:05.645
what already exists.

07:06.505 --> 07:10.445
Um, and so in particular, male fields in particular,

07:11.305 --> 07:14.765
and so that's the, the, the hypothesis we tested.

07:14.985 --> 07:18.285
And what we found is indeed that this is the case.

07:18.425 --> 07:21.165
But what was also interesting is that the group

07:21.165 --> 07:24.285
that is really advantaged at the moment, who seems

07:24.285 --> 07:26.085
to be doing much better than everyone else,

07:26.665 --> 07:30.605
are actually women who specialize in male fields.

07:31.385 --> 07:33.325
Uh, so there is actually an advantage.

07:33.415 --> 07:37.445
Women are doing better than men on the, on the market,

07:37.745 --> 07:40.525
but especially if they are actually,

07:40.525 --> 07:41.805
specializing in male fields.

07:42.505 --> 07:45.245
Now, the problem with this data, the field data, is

07:45.245 --> 07:47.405
that it's confounding all sorts of effects.

07:47.505 --> 07:50.685
No. So we still don't know whether it's actually the choice

07:50.705 --> 07:52.005
of the women themselves

07:52.105 --> 07:54.565
or whether it's something on the side of the employer.

07:55.345 --> 07:57.285
Um, we don't know whether

07:57.465 --> 08:00.565
for some reason the women specializing in in male fields

08:00.575 --> 08:02.205
might be much better, you know,

08:02.205 --> 08:05.205
or much might have characteristics that make them much more,

08:05.785 --> 08:07.965
attractive than than others.

08:08.305 --> 08:10.245
We try to control for that.

08:10.425 --> 08:14.045
But it's still, it is very hard to have a very nice clean,

08:14.145 --> 08:16.685
you know, setting where you can say, well,

08:16.685 --> 08:18.365
these are really the same people.

08:18.955 --> 08:21.085
They just differ by gender

08:21.385 --> 08:23.165
and by the field they specialize in.

08:23.305 --> 08:24.965
So it's very hard to, to do that.

08:25.545 --> 08:28.125
And that's what, the experiment allows us to do.

08:28.585 --> 08:32.285
So we designed an online experiment where instead

08:32.285 --> 08:35.525
of choosing between, you know, human candidates who apply

08:35.545 --> 08:39.445
to a job, you choose between TED Talks basically,

08:39.785 --> 08:42.485
and you choose, you know, what talk you're gonna listen to.

08:43.065 --> 08:45.685
And these are, again, these are gonna be participants

08:45.705 --> 08:49.165
who are college educated, who are recruited from a platform

08:49.165 --> 08:50.605
that is a research platform.

08:51.065 --> 08:55.205
And so they are, then asked, well, you are supposed

08:55.205 --> 08:56.605
to choose between these talks.

08:56.625 --> 08:59.605
We give you an incentive to choose a talk that is

08:59.605 --> 09:01.045
of high quality.

09:01.545 --> 09:05.565
And this is something that we, measured separately

09:05.595 --> 09:07.685
with a separate sample of experts.

09:08.885 --> 09:10.265
And so we give them an incentive.

09:10.325 --> 09:12.665
We tell them, you will actually receive money,

09:12.965 --> 09:15.505
you receive more money if you choose a talk

09:15.505 --> 09:16.625
that is of high quality.

09:17.725 --> 09:20.545
And we also want you to engage with the idea.

09:20.685 --> 09:22.225
So after the talk,

09:22.315 --> 09:25.945
after watching the the talk, you will also have questions.

09:26.925 --> 09:30.545
Um, and then, so that's the fourth first, the baseline.

09:31.045 --> 09:35.625
And what we see is very, very strong differences, which

09:35.685 --> 09:37.305
to some extent depressed me a little

09:37.325 --> 09:38.345
bit when I look at this.

09:38.345 --> 09:41.825
But it's true. You see very big differences between men

09:41.825 --> 09:44.665
and women in what types of talks they watch, right?

09:44.765 --> 09:47.865
So men are much more likely to go for talks that are related

09:47.885 --> 09:50.345
to technology and business,

09:51.245 --> 09:54.985
and women are much more likely to watch talks that relate

09:55.005 --> 09:57.145
to environment and and health.

09:57.725 --> 10:01.185
So we have, you know, this is a little bit the sort of,

10:01.645 --> 10:04.705
test of the hypothesis that first of all,

10:04.705 --> 10:06.665
they are big differences in what men

10:06.665 --> 10:08.985
and women are interested in to start with.

10:09.885 --> 10:13.425
But then maybe the sort of nice, sort of,

10:13.575 --> 10:15.465
next step is to say, well,

10:15.465 --> 10:19.625
what if I'm give you a little nudge to actually look at

10:19.775 --> 10:23.625
what you talk that is, either by a woman, so

10:23.625 --> 10:24.785
that is one treatment,

10:25.325 --> 10:28.625
or I'm giving you, a little incentive

10:28.885 --> 10:33.225
to watch a talk that is actually in a field that women tend

10:33.225 --> 10:34.225
to specialize in.

10:34.365 --> 10:36.865
So health and the environment, for example.

10:38.495 --> 10:40.555
And so we have different groups of people

10:40.695 --> 10:42.795
who face these different, incentives,

10:43.335 --> 10:45.195
and then we look at how it affects

10:45.305 --> 10:46.835
what they choose to watch.

10:47.655 --> 10:49.835
And the incentive is relatively small.

10:49.935 --> 10:51.755
So this is something we want to stress.

10:52.295 --> 10:55.835
It sort of echoes, you know, the idea of this,

10:55.835 --> 10:59.955
diversity initiatives that we have in, academia

11:00.335 --> 11:04.555
or in many sectors is that we understand it is good to try

11:04.575 --> 11:07.195
to attract underrepresented groups,

11:07.735 --> 11:11.515
but we also really want to preserve, you know, the goal

11:11.515 --> 11:14.275
of going for high quality candidates and so on.

11:14.335 --> 11:18.515
So on the, these incentives for, bringing in,

11:18.515 --> 11:19.995
diverse groups tend

11:19.995 --> 11:22.675
to be quite small also in the real world.

11:23.755 --> 11:24.935
But what we see is

11:24.935 --> 11:28.615
that this actually really shifts massively what people are

11:29.295 --> 11:30.935
engaging, engaging with.

11:31.835 --> 11:33.775
And in fact, but,

11:33.795 --> 11:35.415
but again, we replicate some

11:35.415 --> 11:38.895
of the findings we have in the field, data in the sense

11:38.895 --> 11:41.775
that when I'm asking, you know, a a sample

11:41.835 --> 11:46.495
of male respondents to actually listen to a talk by a woman,

11:47.605 --> 11:49.815
they, they will choose a talk by a woman,

11:49.955 --> 11:52.775
but still, you know, will prefer the male field.

11:53.545 --> 11:57.045
So, so the ones that are in over demand are the women

11:57.185 --> 11:58.405
who are in the male field.

11:58.465 --> 12:02.005
So this is sort of, if you want nailing the point that

12:02.695 --> 12:05.445
maybe in the real world there might be some of

12:05.445 --> 12:06.565
that going on as well.

12:07.065 --> 12:10.845
Uh, is that, when we sort of have all these initiatives

12:10.875 --> 12:15.085
that aim at, you know, bringing in diverse groups,

12:15.185 --> 12:18.885
we might still want to find, you know, the women, the women

12:19.105 --> 12:23.045
or the underrepresented, ethnic groups in,

12:23.625 --> 12:26.085
in areas that are typically, you know,

12:26.085 --> 12:27.245
the ones of white men.

12:28.145 --> 12:31.245
And this is what we find interesting is, is,

12:31.755 --> 12:32.965
hopefully, you know,

12:33.275 --> 12:36.685
basically raising the awareness on, on this issue.

12:37.265 --> 12:40.365
Uh, for us, in many departments meetings,

12:40.385 --> 12:42.165
we actually often discuss this.

12:42.265 --> 12:46.965
And, and it's very hard to be, you know, going a bit beyond,

12:47.105 --> 12:49.005
you know, your general interests.

12:49.025 --> 12:50.965
Of course you have your own preferences.

12:51.585 --> 12:54.365
Uh, and, and we all, you know, we all have that.

12:54.625 --> 12:56.005
And so it's not,

12:56.005 --> 12:58.285
it's not just about blaming the white men,

12:59.195 --> 13:00.685
it's just we all have these biases,

13:01.025 --> 13:02.765
but, it just happens to be the case

13:02.765 --> 13:05.805
that now there's one group that is just more represented.

13:06.665 --> 13:10.365
But I want to be somewhat pessi, optimistic about

13:10.905 --> 13:14.085
how malleable, these preferences might be.

13:14.625 --> 13:17.245
And I think that this is maybe the hope that

13:17.265 --> 13:19.125
by raising a little bit the awareness

13:19.825 --> 13:22.845
and maybe giving a little bit of notch to actually perhaps

13:23.395 --> 13:25.365
instead of saying, well, let's go

13:25.645 --> 13:27.805
and try to hire, you know,

13:27.805 --> 13:29.645
people from underrepresented groups.

13:30.285 --> 13:33.565
I think it would be maybe a better idea to think about

13:34.095 --> 13:38.405
let's try and hire, in groups that are, in,

13:38.425 --> 13:41.485
in fields that are typically the ones

13:41.595 --> 13:44.165
that underrepresented groups specialize in.

13:44.765 --> 13:47.325
I do think that we can think about policies.

13:47.425 --> 13:51.525
We are particularly, concerned about policies that push

13:51.785 --> 13:55.765
for the hiring of underrepresented groups, per se,

13:56.555 --> 13:59.565
without thinking about maybe what we want.

13:59.745 --> 14:01.165
And particularly if we want

14:01.165 --> 14:05.005
to really benefit from diversity, we might want

14:05.005 --> 14:07.525
to actually think about, let's broaden our horizon.

14:08.055 --> 14:11.045
Let's try to actually, bring in,

14:12.415 --> 14:15.525
bring in ideas from, from, groups

14:15.555 --> 14:16.965
that are underrepresented

14:16.985 --> 14:18.245
and broaden our horizon,

14:18.675 --> 14:20.765
broaden the fields we are looking at.

Exploring Innovation Bias in Academia and Economics

00:00:00 - So we are interested, we are economists, so we are interested in ideas and innovation in general in many sectors of the economy. It's important to think about how do you come up with new ideas. It's very important to create employment, create new jobs, and also just to add value in, in society and, and in the economy. And so we are interested in how do we actually come about come, how do new ideas come about? And often you think about, well, people have ideas
00:00:33 - and then they make their way into society. Now, this process of actually selecting ideas and implementing ideas is maybe not as innocuous as we think, because there's often actually employers who select, you know, who they are going to hire. And then when you think about academia, we have a lot of, you know, institutions in place that basically select which ideas are gonna be published, which ideas are patented and so on.
00:01:07 - who might be doing this selection obviously have their own, you know, preferences and interests and beliefs in what might be good ideas. and our conjecture was that maybe, and this would apply to all of us, we may have a little bit of a bias towards ideas that are close to our own ideas. So it's almost like a totology, but of course, I'm gonna be more interested in ideas that are close to my own ideas.
00:01:38 - you know, a number of, of candidates, and all these candidates have different ideas, you may think that maybe they're gonna be more enthusiastic, more excited by candidates who have ideas that really resonate with, with the person who is making the decision. which is this idea of idea mo fully. Now, of course, because in many sectors of the economy, well, we have a history of, of course, the labor force being more white male, dominated.
00:02:12 - influences, which ideas gets elected even when there are efforts to actually try to bring in, underrepresented groups. And so this is something that we know is very much in the air in many sectors of the economy, in academia, there's a lot of attention to try to bring in underrepresented groups in academia. And then you sort of think, well, this could be a good idea because also just for the, the, the purpose
00:02:44 - but if, if there is at the same time this tendency to try to go for things that resemble all the ideas, then you don't get the benefit of the, the full benefit of diversity. So this is a, in a nutshell, the sort of the context of the study. And we were interested in studying this, with real, you know, data from the real world. But, we focus on academia

Examining Innovation and Gender Disparities in Academia

00:03:14 - and where ideas are very important. And then we also have additional evidence from an experiment where we are just really interested in what ideas people engage with and whether the ideas they're happy to engage with are in a way malleable if we give them a little bit of a nudge, a little bit of an incentive to actually engage with other ideas. There is a lot of literature, and actually quite recently, perhaps even more, this is really a booming, you know, area.
00:03:47 - there are disparities between men and women, like in other sectors of the economy. So it's harder for, women, for example, to progress through, the academic profession to reach, you know, higher levels. There are fewer, female professors than they are, female, junior professors. And so as if you look at every level, you see that as you go higher up, then there are less and less women.
00:04:19 - minorities. and there are lots of, you know, discussions about what that may be the case. And this is what the literature is really been tackling. You know, what are the, the barriers? Is it actually on the side of, the, the women themselves who are actually maybe have different preferences or might actually, be, finding some characteristics of the academic world, not so attractive, whether it's, the fact that it's competitive
00:04:50 - through a tenure track which overlaps with, years where many women would like to have children and, and so on. So this is a, a, a quite a broad literature. Now, our work is really, to try to think about barriers that a barrier that has maybe not been receiving that much attention. And, which is this idea that, well, the, the, the sort of, we are moving from an environment
00:05:22 - and then we are trying to move into a, an environment that, welcomes, you know, other, other groups. And I think that there is a genuine, you know, interest in getting, you know, these other groups. But at the same time, we think there might be, we don't always realize maybe, the sort of, biases that you, we might have towards, for example, what are the, the ideas, the fields that we should, should be investing in and working on?
00:05:53 - So that means we have real world, data, And we look at placement. So we look at the first job that a PhD candidates get when they graduate, in economics from the top institutions in the us. So we have looked at the top 33 institutions in the US and we look at where they find a job and whether in particular they're likely to find a job as assistant professor, which is probably the most,

Gender Dynamics in Academic Job Market

00:06:25 - graduate as a PhD, student in economics. so this is a desirable, type of position. And we look at, whether men and women are equally likely to find such jobs, but also more, perhaps more relevant to the topic is whether, the area they specialize in matters. and so the hypothesis was that if there is something at work, like ideal moly,
00:06:56 - who would do best in this market are people who have id or who specialize in areas that are similar to what already exists. and so in particular, male fields in particular, and so that's the, the, the hypothesis we tested. And what we found is indeed that this is the case. But what was also interesting is that the group that is really advantaged at the moment, who seems
00:07:26 - are actually women who specialize in male fields. so there is actually an advantage. Women are doing better than men on the, on the market, but especially if they are actually, specializing in male fields. Now, the problem with this data, the field data, is that it's confounding all sorts of effects. No. So we still don't know whether it's actually the choice of the women themselves or whether it's something on the side of the employer.
00:07:57 - for some reason the women specializing in in male fields might be much better, you know, or much might have characteristics that make them much more, attractive than than others. We try to control for that. But it's still, it is very hard to have a very nice clean, you know, setting where you can say, well, these are really the same people. They just differ by gender and by the field they specialize in. So it's very hard to, to do that.
00:08:28 - So we designed an online experiment where instead of choosing between, you know, human candidates who apply to a job, you choose between TED Talks basically, and you choose, you know, what talk you're gonna listen to. And these are, again, these are gonna be participants who are college educated, who are recruited from a platform that is a research platform. And so they are, then asked, well, you are supposed to choose between these talks.
00:08:59 - of high quality. And this is something that we, measured separately with a separate sample of experts. And so we give them an incentive. We tell them, you will actually receive money, you receive more money if you choose a talk that is of high quality. And we also want you to engage with the idea. So after the talk, after watching the the talk, you will also have questions.

Gender Differences in Talk Preferences and Engagement

00:09:31 - And what we see is very, very strong differences, which to some extent depressed me a little bit when I look at this. But it's true. You see very big differences between men and women in what types of talks they watch, right? So men are much more likely to go for talks that are related to technology and business, and women are much more likely to watch talks that relate to environment and and health.
00:10:01 - test of the hypothesis that first of all, they are big differences in what men and women are interested in to start with. But then maybe the sort of nice, sort of, next step is to say, well, what if I'm give you a little nudge to actually look at what you talk that is, either by a woman, so that is one treatment, or I'm giving you, a little incentive
00:10:33 - to specialize in. So health and the environment, for example. And so we have different groups of people who face these different, incentives, and then we look at how it affects what they choose to watch. And the incentive is relatively small. So this is something we want to stress. It sort of echoes, you know, the idea of this, diversity initiatives that we have in, academia
00:11:04 - to attract underrepresented groups, but we also really want to preserve, you know, the goal of going for high quality candidates and so on. So on the, these incentives for, bringing in, diverse groups tend to be quite small also in the real world. But what we see is that this actually really shifts massively what people are engaging, engaging with.
00:11:33 - but again, we replicate some of the findings we have in the field, data in the sense that when I'm asking, you know, a a sample of male respondents to actually listen to a talk by a woman, they, they will choose a talk by a woman, but still, you know, will prefer the male field. So, so the ones that are in over demand are the women who are in the male field. So this is sort of, if you want nailing the point that
00:12:05 - that going on as well. is that, when we sort of have all these initiatives that aim at, you know, bringing in diverse groups, we might still want to find, you know, the women, the women or the underrepresented, ethnic groups in, in areas that are typically, you know, the ones of white men. And this is what we find interesting is, is, hopefully, you know,

Addressing Biases in Hiring Practices

00:12:37 - for us, in many departments meetings, we actually often discuss this. And, and it's very hard to be, you know, going a bit beyond, you know, your general interests. Of course you have your own preferences. and, and we all, you know, we all have that. And so it's not, it's not just about blaming the white men, it's just we all have these biases, but, it just happens to be the case that now there's one group that is just more represented.
00:13:10 - how malleable, these preferences might be. And I think that this is maybe the hope that by raising a little bit the awareness and maybe giving a little bit of notch to actually perhaps instead of saying, well, let's go and try to hire, you know, people from underrepresented groups. I think it would be maybe a better idea to think about let's try and hire, in groups that are, in,
00:13:41 - that underrepresented groups specialize in. I do think that we can think about policies. We are particularly, concerned about policies that push for the hiring of underrepresented groups, per se, without thinking about maybe what we want. And particularly if we want to really benefit from diversity, we might want to actually think about, let's broaden our horizon. Let's try to actually, bring in,
00:14:18 - that are underrepresented and broaden our horizon, broaden the fields we are looking at.

Faculti AI can make mistakes

Research

Policy

Analysis