Popular in Social Sciences

Books

City Time: On Being Sentenced to Rikers Island

Independent Scholar David A. Campbell

Culture and Society

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

New York University Harvey Moltoch

Economics

Gender and Economics

University of California Santa Barbara Shelly Lundberg

American Politics

Keeping the Republic

Boston College Marc Landy

Banks and Policy Organisations

Greening the financial sector

European Investment Bank Matteo Ferrazzi

British Politics

Entrepreneurial thinking: the politics and practice of policy impact

University of Sheffield Matthew Flinders

Culture and Society

Carceral Entrapments: Views from the Prison/Street Interface in India

Indian Institute of Technology Delhi Mahuya Bandyopadhya

Does the Share of Female Managers Affect the Firm’s Geographic Spread?

View more insights in Social Sciences - Economics - Labor Economics

View more figures from University of Birmingham

View the page of Panos Fouskas

Read the Study

Panagiotis Fouskas aims to add a new thread on the debate of gender quota by assessing gender diversity in senior managerial positions.

Image courtesy of interviewee. June 19, 2024

Generating Workspace

Export as PDF

This video discusses the distinctions between the concepts of gender and sex, emphasizing that gender pertains to a broad spectrum of self-identification and preferences, whereas sex refers to biological characteristics. It highlights the importance of gender diversity, particularly in professional contexts, and examines the barriers women face in labor market access and career progression.

The work presented builds on existing literature linking gender diversity in management roles with firm performance, acknowledging mixed findings regarding causality. It aims to investigate the effect of the share of female managers on the geographic spread of firms, proposing two hypotheses: one suggesting a positive relationship due to women’s strengths in forming connections, and another indicating a negative association based on evidence of women's risk aversion and reluctance in competitive environments.

Using data from the Workplace Employment Relations Survey, the study finds that a higher proportion of female managers correlates with a greater likelihood of firms operating in local markets rather than internationally. The analysis reveals significant differences in the proportion of female managers between domestic (approximately 35%) and international firms (approximately 25%). The results show a negative association where an increase in female managers is linked to a smaller market presence.

The study controls for competition and maintains that findings remain robust despite variations in competitive intensity. This video concludes with reflections on the implications of quotas for female representation in management, detailing potential benefits in community engagement versus drawbacks concerning merit and efficiency.

Limitations of the research are acknowledged, suggesting avenues for further studies including better data access and cross-country comparisons to examine cultural impacts on the observed relationships.

Language:

Export as PDF

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

00:00.395 --> 00:03.295
Let me start by, making,

00:03.735 --> 00:07.935
a small di a small distinction between, two key terms

00:08.685 --> 00:11.655
that are, used in, the article

00:11.755 --> 00:14.255
and are also used in a common language.

00:14.875 --> 00:18.775
Um, but there is an important distinction between these two

00:18.775 --> 00:20.735
that some people may not know about,

00:20.795 --> 00:23.695
and these are of course, gender and sex.

00:24.555 --> 00:28.695
Um, so when we talk about gender, obviously we give

00:28.795 --> 00:31.975
to someone, to an individual, the possibility

00:32.075 --> 00:36.815
to self-identify within a broad spectrum

00:36.885 --> 00:39.855
that goes beyond the binary framework.

00:40.115 --> 00:43.935
Uh, instead, when we talk about sex, sex relates

00:43.995 --> 00:46.375
to the biological characteristics and

00:46.375 --> 00:50.295
therefore relates obviously to the binary, framework.

00:50.395 --> 00:54.255
So in a sense, when, talking about gender,

00:54.595 --> 00:58.855
we can say that gender, reveals preferences

00:59.435 --> 01:02.735
of the individual and, preferences.

01:02.735 --> 01:07.175
Obviously in economics, we allow them to vary over time

01:07.475 --> 01:11.015
or we allow them to vary due to certain events.

01:11.795 --> 01:13.295
Um, and as a result,

01:13.295 --> 01:15.175
therefore one can conclude

01:16.325 --> 01:20.615
that gender can be more variable, if you like,

01:21.135 --> 01:22.455
compared to sex.

01:22.995 --> 01:24.375
Uh, generally speaking,

01:24.435 --> 01:28.895
we talk about gender diversity rather than diversity in,

01:29.195 --> 01:30.255
in sex, of course.

01:30.355 --> 01:33.015
So the language that we use is framed around gender,

01:34.455 --> 01:38.135
although many studies including our study,

01:38.525 --> 01:40.655
uses sex rather than gender.

01:41.235 --> 01:46.215
So we employ here, as a, as an explanatory variable,

01:46.915 --> 01:49.135
the share of female managers.

01:49.835 --> 01:54.095
Um, but of course we acknowledge that policies

01:54.275 --> 01:59.095
or priorities that they help to achieve a better,

01:59.385 --> 02:02.655
ratio, if you like, equality between, men

02:02.655 --> 02:07.415
and women are a good step towards the broader aim

02:07.475 --> 02:09.095
of, gender equality.

02:10.315 --> 02:13.055
Um, so literature

02:13.395 --> 02:16.655
and data in general, they show that,

02:17.025 --> 02:21.775
women are more disadvantaged compared to men when it comes

02:21.875 --> 02:24.935
to access into the labor market,

02:24.935 --> 02:26.375
but also career progression.

02:27.875 --> 02:32.015
Of course, you would expect in a meritocratic environment,

02:32.715 --> 02:37.375
an inclusive environment, you would expect both men

02:37.375 --> 02:39.935
and women to have the same, opportunities.

02:40.195 --> 02:41.455
Of course, in practice,

02:41.515 --> 02:44.655
as we know from the evidence, this is not true.

02:45.115 --> 02:49.535
So you can have, for example, let's say more developing,

02:49.835 --> 02:54.615
economies where there are more clear remnants of,

02:55.565 --> 02:58.855
patriarchal, patriarchal societies.

02:59.385 --> 03:02.885
You would expect the women to engage more

03:02.885 --> 03:06.325
with the household work and men to be the breadwinners,

03:07.025 --> 03:10.525
and instead you can have some other societies where the idea

03:10.585 --> 03:13.605
of a working woman is more diffused

03:13.785 --> 03:16.525
and has matured in that society, and

03:16.525 --> 03:19.845
therefore it would be easier for the women to,

03:19.915 --> 03:21.925
obviously access the labor market,

03:22.145 --> 03:25.725
but also, progress in their, career.

03:26.545 --> 03:30.245
So the argument towards,

03:30.395 --> 03:34.005
greater gender diversity, of course relates to,

03:34.665 --> 03:38.765
the greater decree of fairness, that we would like

03:38.765 --> 03:41.805
to see in a society, but also inclusivity.

03:43.445 --> 03:47.425
Um, and at the same, at the same time, it also relates

03:47.645 --> 03:51.225
and links to any associated benefits

03:51.375 --> 03:54.145
that may emerge for greater diversity.

03:55.365 --> 04:00.065
Um, and that may also be the key to incur change

04:00.195 --> 04:01.545
where, where it's needed.

04:01.645 --> 04:03.585
So, for example, by highlighting

04:03.805 --> 04:07.265
and emphasizing any associated benefits

04:07.455 --> 04:11.825
with greater gender diversity, it might be easier to

04:12.385 --> 04:13.985
convince policymakers

04:14.205 --> 04:19.025
or the key stakeholders, to put forward, priorities

04:19.205 --> 04:22.945
and initiatives that will, allow for greater diversity.

04:24.005 --> 04:28.145
Our work builds obviously upon, previous work that

04:28.665 --> 04:33.425
investigated the relationship between, gender diversity

04:34.325 --> 04:37.265
in the boardroom and firm performance.

04:38.125 --> 04:41.985
Um, that early literature looks to identify

04:42.205 --> 04:45.225
and document the association,

04:45.365 --> 04:49.745
but without necessarily delving into the issue of causality.

04:49.925 --> 04:53.705
So to put it simply, what triggers what?

04:53.805 --> 04:58.625
So is it that gender diversity makes firms better,

04:59.285 --> 05:01.305
or is it simply that firms

05:01.455 --> 05:04.585
that are better are better performing firms

05:04.685 --> 05:05.985
or better managed firms?

05:06.695 --> 05:11.665
They seek to have a more diverse, portal.

05:12.565 --> 05:16.225
Um, obviously what you need is you need better data.

05:16.565 --> 05:20.505
You need better techniques to, to be able to uncover

05:21.085 --> 05:23.225
the causality in this relationship.

05:23.725 --> 05:27.265
Um, in general, of course, we have a good body

05:27.325 --> 05:31.985
of literature, that shows a positive association

05:32.095 --> 05:33.905
between the two variables.

05:34.365 --> 05:37.865
But of course, this may need to be taken with,

05:38.055 --> 05:39.385
with a grain of salt

05:39.415 --> 05:43.425
because some of these models may be unspecified.

05:44.085 --> 05:45.785
Uh, so for example, it could be the case

05:45.815 --> 05:48.065
that key variables like age

05:48.205 --> 05:50.905
or experience may, may have been overlooked.

05:51.125 --> 05:53.705
Now, as is always the case with empirical work,

05:53.965 --> 05:55.545
the sample at hand matters.

05:56.405 --> 06:00.245
Uh, so if you change your your sample, it could be the case

06:00.355 --> 06:03.205
that the results are not the same anymore.

06:03.945 --> 06:07.685
Uh, and indeed here we have some contradicting results.

06:07.745 --> 06:11.885
So for example, using some Danish data,

06:12.705 --> 06:16.605
I remember Rose 2007 finds,

06:17.105 --> 06:20.885
no association between, gender diversity

06:21.425 --> 06:24.085
and firm performance,

06:24.305 --> 06:26.965
and also using the US data.

06:27.245 --> 06:28.765
I remember the work of Adams

06:28.765 --> 06:32.045
and Ferrera, who find a negative,

06:32.505 --> 06:33.765
association.

06:34.825 --> 06:39.005
Now in general, this, this literature has been used

06:39.065 --> 06:43.405
as a springboard to investigate other associated effects

06:43.555 --> 06:45.805
with, with gender diversity.

06:46.785 --> 06:50.515
Um, and we have, for example, evidence

06:50.515 --> 06:54.915
of better corporate governance, responsible business,

06:55.295 --> 06:58.795
positive motivational effects,

06:59.495 --> 07:03.795
organizational innovation, just to mention a few.

07:04.415 --> 07:08.905
Um, so we are looking to extend this,

07:09.345 --> 07:11.025
literature by this literature

07:11.145 --> 07:15.025
and associated benefits by looking at the relationship

07:15.025 --> 07:17.665
between, the share of female managers

07:18.325 --> 07:20.945
and the geographic spread of the firm.

07:21.125 --> 07:25.225
So our dependent variable is the geographic spread

07:25.295 --> 07:27.745
that captures the geographic segment.

07:28.125 --> 07:29.905
So that is where the product

07:30.725 --> 07:34.785
or the service of the firm is essentially, offered.

07:35.645 --> 07:38.985
And what we do is we put forward two competing hypothesis.

07:39.285 --> 07:43.745
So a positive association, that may be due

07:43.745 --> 07:47.825
to the fact that, women has been documented

07:47.935 --> 07:50.145
that are better in forming ties.

07:50.445 --> 07:54.705
So therefore that makes them, more suitable as managers,

07:55.245 --> 07:58.925
for firms that they look to expand,

07:59.185 --> 08:03.485
or also to the fact that having greater diversity,

08:03.835 --> 08:07.405
essentially increases the capacity of the firm to tackle

08:08.195 --> 08:10.965
some business complexities that may emerge

08:11.595 --> 08:15.005
when we are looking to expand, our business.

08:15.825 --> 08:17.005
Of course, on the other hand,

08:17.265 --> 08:19.245
you can have a negative association,

08:19.865 --> 08:21.765
and that is, again, underpinned

08:21.785 --> 08:23.605
by some evidence in the literature.

08:23.625 --> 08:27.405
So we have some evidence from the behavioral literature

08:27.915 --> 08:32.605
that shows that women are in general more risk averse,

08:33.345 --> 08:36.205
but also that they are more reluctant

08:36.385 --> 08:38.565
to engage in competitive environments.

08:38.565 --> 08:41.645
And these are environments where, are

08:42.165 --> 08:45.085
possibly you have firms that are selling abroad,

08:45.105 --> 08:47.605
and the competition is, higher.

08:48.545 --> 08:50.125
We are using here,

08:50.825 --> 08:53.725
the workplace employment relations

08:54.555 --> 08:58.045
Surveys, the worst 2000, 11,

08:58.775 --> 09:03.365
which at the time was, the most up to date,

09:03.795 --> 09:06.645
version in this, survey series,

09:06.745 --> 09:11.485
and I think is still the most up to date, dataset.

09:11.865 --> 09:13.405
Uh, it is a rich dataset.

09:13.785 --> 09:15.725
Uh, it brings together firms

09:16.145 --> 09:18.885
and employs something that is not common.

09:19.065 --> 09:22.045
So obviously our dataset refers to the uk.

09:22.985 --> 09:27.285
Um, so most of the time researchers, they will have to

09:27.935 --> 09:32.845
merge different datasets to answer similar, questions,

09:32.905 --> 09:34.245
but you don't have to do that

09:34.355 --> 09:36.885
with the workplace employment relations survey.

09:37.545 --> 09:40.245
It allows you, for example, to to link

09:40.345 --> 09:43.285
to look into situations where you would like

09:43.305 --> 09:45.085
to see workplace practices

09:45.505 --> 09:47.685
and how these affect fair performance.

09:48.545 --> 09:52.165
Um, and similar to the question that we are doing,

09:52.265 --> 09:55.205
for example, where again, we are looking,

09:55.985 --> 09:57.965
so firm level characteristics

09:58.505 --> 10:03.485
and how this, relate to, managerial,

10:03.585 --> 10:05.165
characteristics as well.

10:05.785 --> 10:09.405
Um, so there are two versions of,

10:10.065 --> 10:11.165
of this dataset.

10:11.185 --> 10:12.645
So there's the restricted version,

10:12.985 --> 10:15.165
and there is a publicly available, version.

10:16.025 --> 10:18.045
Um, the restricted version

10:18.905 --> 10:22.525
in includes also the financial performance questionnaire,

10:23.515 --> 10:28.265
which we, don't have access to since we are using only

10:28.885 --> 10:30.465
the publicly available version.

10:31.365 --> 10:33.105
The limitation of that is

10:33.105 --> 10:37.265
that you cannot use objective measures of firm performance.

10:38.005 --> 10:39.985
Uh, instead in our, research,

10:40.045 --> 10:42.625
we are only using the subjective measures

10:42.645 --> 10:44.145
of firm performance.

10:44.845 --> 10:46.225
Uh, as previously said,

10:46.485 --> 10:50.065
the dependent variable is the geographic spread of the firm.

10:50.205 --> 10:54.385
So that is the, the geographic segment at which the,

10:54.535 --> 10:57.305
product or the service is, offered.

10:57.845 --> 10:59.745
And we have four tiers here.

10:59.805 --> 11:02.585
So we have, the local level, tier one,

11:02.925 --> 11:06.785
and then from there you move on up to regional tier two,

11:07.505 --> 11:11.305
national and international, which is tier, four.

11:11.365 --> 11:15.305
So in a sense, our, dependent variable

11:15.925 --> 11:18.705
has a natural, order there,

11:18.925 --> 11:22.785
and that allow us to use what we call,

11:22.885 --> 11:25.025
an anor logic model.

11:26.045 --> 11:29.945
Um, and we do that using both weighted

11:30.205 --> 11:32.345
and unweighted, samples.

11:32.495 --> 11:36.105
Obviously, the advantage of using weighted sample is that,

11:36.525 --> 11:38.145
excuse me, that you make the results

11:38.475 --> 11:39.745
nationally representative.

11:40.725 --> 11:43.345
We start as far as results are concerned.

11:43.345 --> 11:46.265
So we start by splitting first of all the sample

11:47.865 --> 11:50.355
into firms that they cover, the domestic market,

11:50.695 --> 11:51.795
and, firms,

11:52.695 --> 11:55.805
Uh, that serve mainly the international market.

11:55.905 --> 11:58.085
So therefore, we have two subsamples there.

11:58.705 --> 12:00.845
And what we are doing is we are trying

12:00.865 --> 12:02.245
to see whether the share

12:02.245 --> 12:06.365
of female managers is different in these two subsamples.

12:06.985 --> 12:10.765
So what we find is we find that for the firms

12:10.875 --> 12:15.405
that operate domestically, so that would include firms

12:15.555 --> 12:18.245
that sell locally, regionally

12:18.505 --> 12:21.765
or nationally, we find that the share is

12:21.785 --> 12:23.725
around 35%.

12:24.595 --> 12:28.615
Uh, instead, the share of, female manager for firms

12:28.615 --> 12:32.735
that operate internationally is lower at 25%.

12:33.395 --> 12:35.895
So took key observations here.

12:35.995 --> 12:40.655
So obviously moving from a domestic to international domain,

12:41.315 --> 12:43.295
the share drops,

12:43.635 --> 12:46.655
and that drop, which is 10 percentage points roughly,

12:47.845 --> 12:49.505
is statistically significant.

12:49.965 --> 12:53.945
Uh, and the other observation is that in both sub samples,

12:54.725 --> 12:57.505
we are obviously far from achieving a quality

12:57.615 --> 12:59.425
that would be a ratio 50%.

13:01.125 --> 13:04.225
Um, now our baseline regression

13:04.415 --> 13:07.425
that looks at the association between the share

13:07.465 --> 13:08.505
of female managers

13:09.085 --> 13:13.585
and geographic spread, reveals a negative association.

13:13.765 --> 13:17.145
So what that means, it means that as the share

13:17.165 --> 13:21.585
of female managers increases, it is more likely

13:21.895 --> 13:25.145
that the firm will, sell their product

13:25.445 --> 13:27.785
or service in a smaller segment.

13:27.965 --> 13:29.625
So it's more likely, for example,

13:30.095 --> 13:33.985
that they will sell it locally rather than regionally.

13:34.245 --> 13:36.665
At the same time, it's more likely that they will sell it

13:37.275 --> 13:39.585
nationally rather than internationally.

13:40.405 --> 13:43.225
Um, so when we are looking at, the,

13:43.285 --> 13:48.105
the predictive margins of, of the model, so that is when

13:49.505 --> 13:54.165
all the covariate of the model are held at the mean value,

13:54.905 --> 13:58.845
we get the probability of, roughly

13:59.505 --> 14:04.485
90% that the firm will sell the product,

14:04.865 --> 14:08.055
domestically,

14:08.435 --> 14:11.655
and only 10% that the firm will,

14:11.945 --> 14:14.175
offer the product internationally.

14:14.915 --> 14:18.975
Um, if you look at especially the domestic market

14:18.975 --> 14:22.815
that include includes the three tiers there,

14:22.955 --> 14:24.815
we reveal a u-shape pattern.

14:25.635 --> 14:30.295
So that is obviously the probability is higher

14:30.915 --> 14:33.655
for selling at the local market as the share

14:33.655 --> 14:35.935
of female female managers increases

14:36.355 --> 14:41.095
and is roughly, 40%, if I remember correctly,

14:41.395 --> 14:46.295
and it goes higher, for regional,

14:46.555 --> 14:49.015
it is drops down to 20%,

14:49.075 --> 14:52.525
and then it goes up again for national to 30%.

14:52.545 --> 14:57.085
So as aship, as I said, there is a a u-shape pattern there.

14:57.785 --> 15:00.805
Um, I mentioned previously that one

15:00.805 --> 15:05.565
of the underlying mechanisms for this negative association,

15:06.385 --> 15:10.405
can be the competitive nature of, women.

15:11.185 --> 15:14.405
Um, and we are trying to test that by, again,

15:14.505 --> 15:19.165
by splitting our sample into a sample where,

15:19.425 --> 15:23.525
the managers feel that the firm is facing high competition

15:24.225 --> 15:28.805
and also, a sub sample where,

15:29.305 --> 15:34.085
the manager thinks that the firm is facing low competition.

15:34.085 --> 15:36.165
So we have these two subsamples

15:36.165 --> 15:38.725
and we retest our model there

15:39.345 --> 15:43.085
to see if this predictive probabilities change at all,

15:43.145 --> 15:44.965
and we find that they don't.

15:44.965 --> 15:47.965
So the finding is essentially robust

15:49.445 --> 15:52.695
when it comes to incorporating, competition.

15:53.395 --> 15:58.175
Uh, now obviously there can be some, some explanation as

15:58.175 --> 16:02.695
to why there is this higher concentration of,

16:03.065 --> 16:06.695
women in firms that operate at a smaller scale.

16:07.555 --> 16:10.535
Um, and obviously we said also previously

16:10.535 --> 16:14.375
that the literature has uncovered that it is more difficult

16:14.395 --> 16:19.015
for women to progress to higher managerial positions,

16:19.115 --> 16:22.615
and that might be part of, of the explanation.

16:23.315 --> 16:27.615
Um, at the same time, it can be also a personal choice.

16:27.955 --> 16:31.295
So in the sense that it could be that the, the women

16:31.635 --> 16:35.135
for some reason, they prefer the job flexibility

16:35.395 --> 16:38.175
or the better work life balance that comes

16:38.765 --> 16:43.535
with the managerial positions at the smaller, level,

16:44.535 --> 16:47.505
because possibly this may allow them,

16:47.685 --> 16:49.865
to give greater support to the family.

16:50.005 --> 16:54.465
So it might be something that applies to, women in

16:54.985 --> 16:56.825
families with young kids, for example.

16:57.565 --> 17:00.905
Um, also, we know from the literature

17:00.905 --> 17:04.585
that women in general are more prone than men

17:05.205 --> 17:08.745
to engage in communities, in community activities.

17:09.365 --> 17:13.785
Um, and therefore it could be more likely for them to work

17:14.485 --> 17:15.865
in, companies

17:15.925 --> 17:20.665
or, firms, where they, they have activities

17:20.665 --> 17:21.785
that relate to

17:21.785 --> 17:24.385
what is happening on the place where they live.

17:24.695 --> 17:27.145
Just to wrap up, essentially in this work,

17:27.335 --> 17:30.545
what we are doing is we are looking the relationship,

17:30.855 --> 17:32.825
between the share of female managers

17:32.965 --> 17:34.305
and the geographic spread.

17:35.085 --> 17:38.585
Uh, we have uncovered, a negative association

17:38.615 --> 17:43.305
between these two, variables of interest, which means

17:43.305 --> 17:46.945
that as the share of female managers increases,

17:47.405 --> 17:49.525
it is more likely that the product

17:49.905 --> 17:53.645
or the service will be sold, will be sold,

17:53.745 --> 17:55.045
at the local market,

17:55.185 --> 17:58.845
or at least in a smaller geographic, segment.

17:59.025 --> 18:00.885
We also control for competition,

18:01.305 --> 18:05.685
and we said that the results are robust, to that.

18:06.425 --> 18:09.965
Um, now obviously in terms of a

18:10.795 --> 18:15.565
legislative action towards, quota, possibly,

18:15.745 --> 18:20.525
in, senior managerial, managerial roles,

18:21.015 --> 18:22.485
there can be mixed effects.

18:22.745 --> 18:26.445
Uh, we think so on one side.

18:27.185 --> 18:29.965
Um, so as we mentioned earlier,

18:30.105 --> 18:33.685
as women are more prone towards community activities,

18:34.065 --> 18:37.485
having more women in larger cooperation means

18:37.555 --> 18:40.565
that there can be some positive welfare effects

18:40.565 --> 18:42.725
because women are likely to engage

18:43.695 --> 18:45.685
these firms in philanthropy

18:45.785 --> 18:48.085
or other related community activities.

18:48.085 --> 18:50.885
So there can be some positive effect coming from there.

18:51.825 --> 18:54.965
Um, but there can also be, on the other hand,

18:55.445 --> 18:59.725
negative effects because, if there are certain quotas

18:59.725 --> 19:02.845
that need to be met, then the appointments,

19:02.845 --> 19:05.525
they will not be made based on merit.

19:06.265 --> 19:09.405
And as a result, it could be that they, the firms

19:10.075 --> 19:13.605
they don't operate on their productive from dear.

19:13.705 --> 19:17.205
So inducing this way, some inefficiency.

19:18.345 --> 19:21.805
So obviously now this work is not without limitations.

19:22.265 --> 19:26.535
It can be extend, further, hopefully

19:26.605 --> 19:29.775
with a better dataset that has,

19:30.695 --> 19:32.215
a time dimension perhaps.

19:32.435 --> 19:36.055
And this would allow the researchers to make some

19:36.055 --> 19:38.735
of the covariates predetermined

19:38.735 --> 19:40.615
and tackle the issue of causality.

19:41.515 --> 19:44.735
Um, at the same time, as I mentioned previously,

19:45.195 --> 19:48.735
we don't have access to objective measures of,

19:48.805 --> 19:49.935
firm performance.

19:50.035 --> 19:54.375
So therefore another data set that gives that access can,

19:54.515 --> 19:56.895
test the robustness of our results.

19:57.755 --> 20:02.615
Um, and perhaps more interestingly, one would like to

20:02.915 --> 20:05.295
to do a cross country analysis

20:05.995 --> 20:10.415
and see whether, cultural characteristics moderate,

20:11.275 --> 20:13.335
the relationship, the association

20:13.335 --> 20:15.495
that we found in our article.

00:00 - Let me start by, making, a small di a small distinction between, two key terms that are, used in, the article and are also used in a common language. but there is an important distinction between these two that some people may not know about,

00:20 - and these are of course, gender and sex. so when we talk about gender, obviously we give to someone, to an individual, the possibility to self-identify within a broad spectrum that goes beyond the binary framework. instead, when we talk about sex, sex relates to the biological characteristics and therefore relates obviously to the binary, framework. So in a sense, when, talking about gender, we can say that gender, reveals preferences

00:59 - of the individual and, preferences. Obviously in economics, we allow them to vary over time or we allow them to vary due to certain events. and as a result, therefore one can conclude that gender can be more variable, if you like, compared to sex. generally speaking, we talk about gender diversity rather than diversity in, in sex, of course.

01:30 - So the language that we use is framed around gender, although many studies including our study, uses sex rather than gender. So we employ here, as a, as an explanatory variable, the share of female managers. but of course we acknowledge that policies or priorities that they help to achieve a better, ratio, if you like, equality between, men and women are a good step towards the broader aim of, gender equality.

02:10 - so literature and data in general, they show that, women are more disadvantaged compared to men when it comes to access into the labor market, but also career progression. Of course, you would expect in a meritocratic environment, an inclusive environment, you would expect both men and women to have the same, opportunities. Of course, in practice, as we know from the evidence, this is not true.

02:45 - So you can have, for example, let's say more developing, economies where there are more clear remnants of, patriarchal, patriarchal societies. You would expect the women to engage more with the household work and men to be the breadwinners, and instead you can have some other societies where the idea of a working woman is more diffused and has matured in that society, and therefore it would be easier for the women to, obviously access the labor market,

03:22 - but also, progress in their, career. So the argument towards, greater gender diversity, of course relates to, the greater decree of fairness, that we would like to see in a society, but also inclusivity. and at the same, at the same time, it also relates and links to any associated benefits that may emerge for greater diversity. and that may also be the key to incur change where, where it's needed.

04:01 - So, for example, by highlighting and emphasizing any associated benefits with greater gender diversity, it might be easier to convince policymakers or the key stakeholders, to put forward, priorities and initiatives that will, allow for greater diversity. Our work builds obviously upon, previous work that investigated the relationship between, gender diversity in the boardroom and firm performance. that early literature looks to identify

04:42 - and document the association, but without necessarily delving into the issue of causality. So to put it simply, what triggers what? So is it that gender diversity makes firms better, or is it simply that firms that are better are better performing firms or better managed firms? They seek to have a more diverse, portal. obviously what you need is you need better data. You need better techniques to, to be able to uncover

05:21 - the causality in this relationship. in general, of course, we have a good body of literature, that shows a positive association between the two variables. But of course, this may need to be taken with, with a grain of salt because some of these models may be unspecified. so for example, it could be the case that key variables like age or experience may, may have been overlooked.

05:51 - Now, as is always the case with empirical work, the sample at hand matters. so if you change your your sample, it could be the case that the results are not the same anymore. and indeed here we have some contradicting results. So for example, using some Danish data, I remember Rose 2007 finds, no association between, gender diversity and firm performance, and also using the US data.

06:27 - I remember the work of Adams and Ferrera, who find a negative, association. Now in general, this, this literature has been used as a springboard to investigate other associated effects with, with gender diversity. and we have, for example, evidence of better corporate governance, responsible business, positive motivational effects, organizational innovation, just to mention a few.

07:04 - so we are looking to extend this, literature by this literature and associated benefits by looking at the relationship between, the share of female managers and the geographic spread of the firm. So our dependent variable is the geographic spread that captures the geographic segment. So that is where the product or the service of the firm is essentially, offered. And what we do is we put forward two competing hypothesis.

07:39 - So a positive association, that may be due to the fact that, women has been documented that are better in forming ties. So therefore that makes them, more suitable as managers, for firms that they look to expand, or also to the fact that having greater diversity, essentially increases the capacity of the firm to tackle some business complexities that may emerge when we are looking to expand, our business. Of course, on the other hand,

08:17 - you can have a negative association, and that is, again, underpinned by some evidence in the literature. So we have some evidence from the behavioral literature that shows that women are in general more risk averse, but also that they are more reluctant to engage in competitive environments. And these are environments where, are possibly you have firms that are selling abroad, and the competition is, higher.

08:48 - We are using here, the workplace employment relations Surveys, the worst 2000, 11, which at the time was, the most up to date, version in this, survey series, and I think is still the most up to date, dataset. it is a rich dataset. it brings together firms and employs something that is not common. So obviously our dataset refers to the uk.

09:22 - so most of the time researchers, they will have to merge different datasets to answer similar, questions, but you don't have to do that with the workplace employment relations survey. It allows you, for example, to to link to look into situations where you would like to see workplace practices and how these affect fair performance. and similar to the question that we are doing, for example, where again, we are looking,

09:55 - so firm level characteristics and how this, relate to, managerial, characteristics as well. so there are two versions of, of this dataset. So there's the restricted version, and there is a publicly available, version. the restricted version in includes also the financial performance questionnaire, which we, don't have access to since we are using only

10:28 - the publicly available version. The limitation of that is that you cannot use objective measures of firm performance. instead in our, research, we are only using the subjective measures of firm performance. as previously said, the dependent variable is the geographic spread of the firm. So that is the, the geographic segment at which the, product or the service is, offered.

10:57 - And we have four tiers here. So we have, the local level, tier one, and then from there you move on up to regional tier two, national and international, which is tier, four. So in a sense, our, dependent variable has a natural, order there, and that allow us to use what we call, an anor logic model. and we do that using both weighted and unweighted, samples.

11:32 - Obviously, the advantage of using weighted sample is that, excuse me, that you make the results nationally representative. We start as far as results are concerned. So we start by splitting first of all the sample into firms that they cover, the domestic market, and, firms, that serve mainly the international market. So therefore, we have two subsamples there. And what we are doing is we are trying

12:00 - to see whether the share of female managers is different in these two subsamples. So what we find is we find that for the firms that operate domestically, so that would include firms that sell locally, regionally or nationally, we find that the share is around 35%. instead, the share of, female manager for firms that operate internationally is lower at 25%. So took key observations here.

12:35 - So obviously moving from a domestic to international domain, the share drops, and that drop, which is 10 percentage points roughly, is statistically significant. and the other observation is that in both sub samples, we are obviously far from achieving a quality that would be a ratio 50%. now our baseline regression that looks at the association between the share of female managers

13:09 - and geographic spread, reveals a negative association. So what that means, it means that as the share of female managers increases, it is more likely that the firm will, sell their product or service in a smaller segment. So it's more likely, for example, that they will sell it locally rather than regionally. At the same time, it's more likely that they will sell it nationally rather than internationally. so when we are looking at, the,

13:43 - the predictive margins of, of the model, so that is when all the covariate of the model are held at the mean value, we get the probability of, roughly 90% that the firm will sell the product, domestically, and only 10% that the firm will, offer the product internationally. if you look at especially the domestic market that include includes the three tiers there, we reveal a u-shape pattern.

14:25 - So that is obviously the probability is higher for selling at the local market as the share of female female managers increases and is roughly, 40%, if I remember correctly, and it goes higher, for regional, it is drops down to 20%, and then it goes up again for national to 30%. So as aship, as I said, there is a a u-shape pattern there. I mentioned previously that one of the underlying mechanisms for this negative association,

15:06 - can be the competitive nature of, women. and we are trying to test that by, again, by splitting our sample into a sample where, the managers feel that the firm is facing high competition and also, a sub sample where, the manager thinks that the firm is facing low competition. So we have these two subsamples and we retest our model there to see if this predictive probabilities change at all, and we find that they don't.

15:44 - So the finding is essentially robust when it comes to incorporating, competition. now obviously there can be some, some explanation as to why there is this higher concentration of, women in firms that operate at a smaller scale. and obviously we said also previously that the literature has uncovered that it is more difficult for women to progress to higher managerial positions, and that might be part of, of the explanation. at the same time, it can be also a personal choice.

16:27 - So in the sense that it could be that the, the women for some reason, they prefer the job flexibility or the better work life balance that comes with the managerial positions at the smaller, level, because possibly this may allow them, to give greater support to the family. So it might be something that applies to, women in families with young kids, for example. also, we know from the literature that women in general are more prone than men

17:05 - to engage in communities, in community activities. and therefore it could be more likely for them to work in, companies or, firms, where they, they have activities that relate to what is happening on the place where they live. Just to wrap up, essentially in this work, what we are doing is we are looking the relationship, between the share of female managers and the geographic spread.

17:35 - we have uncovered, a negative association between these two, variables of interest, which means that as the share of female managers increases, it is more likely that the product or the service will be sold, will be sold, at the local market, or at least in a smaller geographic, segment. We also control for competition, and we said that the results are robust, to that. now obviously in terms of a

18:10 - legislative action towards, quota, possibly, in, senior managerial, managerial roles, there can be mixed effects. we think so on one side. so as we mentioned earlier, as women are more prone towards community activities, having more women in larger cooperation means that there can be some positive welfare effects because women are likely to engage these firms in philanthropy

18:45 - or other related community activities. So there can be some positive effect coming from there. but there can also be, on the other hand, negative effects because, if there are certain quotas that need to be met, then the appointments, they will not be made based on merit. And as a result, it could be that they, the firms they don't operate on their productive from dear. So inducing this way, some inefficiency. So obviously now this work is not without limitations.

19:22 - It can be extend, further, hopefully with a better dataset that has, a time dimension perhaps. And this would allow the researchers to make some of the covariates predetermined and tackle the issue of causality. at the same time, as I mentioned previously, we don't have access to objective measures of, firm performance. So therefore another data set that gives that access can,

19:54 - test the robustness of our results. and perhaps more interestingly, one would like to to do a cross country analysis and see whether, cultural characteristics moderate, the relationship, the association that we found in our article.

Faculti AI can make mistakes