PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Share a file Manage my documents Convert Recover PDF Search Help Contact

1 s2.0 S1048984317305179 main .pdf

Original filename: 1-s2.0-S1048984317305179-main.pdf
Title: The queen bee_ A myth? The effect of top-level female leadership on subordinate females
Author: Paulo Roberto Arvate

This PDF 1.7 document has been generated by Elsevier / , and has been sent on pdf-archive.com on 15/04/2018 at 22:14, from IP address 147.91.x.x. The current document download page has been viewed 712 times.
File size: 586 KB (16 pages).
Privacy: public file

Download original PDF file

Document preview

The Leadership Quarterly xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

The Leadership Quarterly
journal homepage: www.elsevier.com/locate/leaqua

The queen bee: A myth? The effect of top-level female leadership on
subordinate females

Paulo Roberto Arvatea, , Gisele Walczak Galileab, Isabela Todescatc

São Paulo School of Business Administration and Center of Studies on Microeconomic Applied, Getulio Vargas Foundation, Rua Itapeva, 474, 01332-010 São Paulo, SP,
São Paulo School of Business Administration, Getulio Vargas Foundation and Institute of Education and Research, INSPER, Brazil
São Paulo School of Business Administration, Getulio Vargas Foundation, Brazil


We investigate the effect of female leadership on gender differences in public and private organizations. Female
leadership was constructed using a quasi-experiment involving mayoral elections, and our research used a
sample of 8.3 million organizations distributed over 5600 Brazilian municipalities. Our main results show that
when municipalities in which a woman was elected leader (treatment group) are compared with municipalities
in which a male was elected leader (control group) there was an increase in the number of top and middle
managers in public organizations. Two aspects contribute to the results: time and command/role model. The
time effect is important because our results are obtained with reelected women – in their second term – and the
command/role model (the queen bee phenomenon is either small, or non-existent) is important because of the
institutional characteristics of public organizations: female leaders (mayor) have much asymmetrical power and
decision-making discretion, i.e., she chooses the top managers. These top managers then choose middle managers influenced by female leadership (a role model). We obtained no results for private organizations. Our work
contributes to the literature on leadership by addressing some specific issues: an empirical investigation with a
causal effect between the variables (regression-discontinuity design – a non-parametric estimation), the importance of role models, and how the observed effects are time-dependent. Insofar as public organizations are
concerned, the evidence from our large-scale study suggests that the queen bee phenomenon may be a myth;
instead, of keeping subordinate women at bay, our results show that women leaders who are afforded much
managerial discretion behave in a benevolent manner toward subordinate women. The term “Regal Leader”
instead of “Queen Bee” is thus a more appropriate characterization of women in top positions of power.

“There is a special place in hell for women who don't help each
other!” These words, which were spoken by Former Secretary of State
Madeleine Albright urging other women to support Hillary's candidacy
in the last USA presidential election, had a great repercussion in the
world's press (including the New York Times, The Guardian, and TIME
magazine). Once the election had been set and following this premonition, would a portion of American women have their place in hell
guaranteed (and another portion take their place in heaven)?
Thankfully, between the heaven and hell of the declarations, there is an
empirical purgatory trying to understand if and under what conditions
women support each other in different areas of society (such as in
politics, business, government).
Our work is an empirical investigation that seeks to shed some light

on what is apparently a well-established effect, the QUEEN BEE phenomenon – QBP (Derks, Ellemers, Van Laar, & De Groot, 2011; Derks,
Laar, Ellemers, & Raghoe, 2015; Derks, Van Laar, Ellemers, & De Groot,
2011; Faniko, Ellemers, & Derks, 2016). Our investigation focuses on
women in leadership; with our empirical strategy, we have strong
control over the environment for estimating the causal effect of a
woman in power on other females. Up to this point, the literature on
leadership has not decisively addressed the issue of endogeneity bias
(Antonakis, Bendahan, Jacquart, & Lalive, 2010). In the presence of this
bias, which bedevils much of the observational and correlational research on which the validity of the QBP phenomenon rests, it is impossible to know what the causal relation is between a woman in a
position of power and gender-oriented outcomes.
To identify the effect of female leadership independent of the endogeneity bias due to reverse causality and omitted variables, we use

Corresponding author.
E-mail address: paulo.arvate@fgv.br (P.R. Arvate).

Received 28 July 2017; Received in revised form 8 March 2018; Accepted 13 March 2018
1048-9843/ © 2018 Elsevier Inc. All rights reserved.

Please cite this article as: Arvate, P.R., The Leadership Quarterly (2018), https://doi.org/10.1016/j.leaqua.2018.03.002

The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

does not have direct authority over private organizations.
It would be ideal to observe what occurs within organizations in
“each position” (e.g., CEO, middle managers) from the highest to the
lowest levels, but we do not have this type of information. As it is
reasonable to believe that an investigation into different earnings levels
reflects the organizational hierarchies, we use this fact to establish what
the top, middle managers, and lower positions are. Higher salaries
mean a top management position, middle salaries mean a middle
management position, and lower salaries mean lower positions in the
Briefly, our results suggest that the QBP may be a myth. We find that
there is a pro-female causal effect of female leadership in public organizations: in other words, we find a larger number of women than men
in both top and middle management positions. We find no robust evidence to show that this result extends to lower positions. We interpret
the first result (at the top level) as command and the second result as the
RM effect reflecting a female leader's influence. Thus, QBP is either nonexistent, or less than the command and RM effect.
In public organizations, the top manager can be chosen directly by
the mayor and indirectly by the same mayor by way of political
agreement with different levels of government (state and federal government) if the public organization is in a municipality but is not owned
by the municipal government (e.g., patronage). Middle managers depend on the internal dynamic of organizations: top managers choose
middle managers and are “influenced” by the mayor as leader (via RM)
in their choice for these positions. Therefore, we expand our understanding of the process of change by investigating different pathways to
gender-related outcomes (Fischer, Dietz, & Antonakis, 2017).
The most robust effect favorable to women occurs when the same
woman is reelected, that is, she serves two consecutive terms in office.
The time effect as to how long it takes for leaders to assert their choices
is also an issue that is not well investigated in leadership literature (see
Antonakis, Day, & Schyns, 2012; Fischer et al., 2017). Delayed effects
exist because the choice of leaders, the implementation of new preference proposals and the change in women's preferences in organizations caused by “the RM effect” (mainly, the results at the intermediate
level in public organizations) all take time. In fact, as indicated by the
eponymous title of the article by Shamir (2011) “Leadership takes
Contrary to what we find in public organizations, our results show
that there is no observed improvement for women in private organizations. The choice of top managers in private organizations is different
from those in public organizations. Our non-result for private organizations can be related to the work by Bertrand, Black, Jensen, and
Lleras-Muney (2014). In Bertrand's work, the change proposed by
Norwegian legislation (2003) relating to newly-appointed female board
members is the only change observed. We cannot correlate the private
result with QBP because we do not observe the emergence of female
leaders in organizations. Thus, there is no positive effect in private
firms, which makes sense, at least in the short to medium term, given
that majors do not have much command and there is no RM effect on
gender-related outcomes in private organizations.
Our work is organized as follows. We present a review of the related
literature on both QBP, RM, and our main hypothesis in Section 0. In
Section 0, we present the institutional background, dataset and empirical strategy of our study. In Section 0, we report our results. Finally,
in Section 0, we summarize our findings and discuss their implications.

the procedure established by Lee (2001); Lee and Card (2008); and Lee,
Moretti, and Butler (2004). Basically, we study the effect of a female
mayor chosen in a gender race–where a man is in first place and a
woman in second place, or vice-versa–by a very small margin of votes.
If this margin is close to zero, this type of election mimics an experiment because the final result under these conditions is almost random.
Mayors are visible and uncontestable leaders with much asymmetrical
power (Rucker, Dubois, & Galinsky, 2010; Sturm & Antonakis, 2015)
and decision-making discretion (Finkelstein & Hambrick, 1990;
Finkelstein & Peteraf, 2007; Hambrick & Finkelstein, 1987); that is,
they have the means to asymmetrically enforce their will (preferences)
over others, and mayors can significantly shape an organization. In
adopting this causal identification procedure, we can compare the difference in outcomes in municipalities that have a female leader (i.e.,
treatment group) vis-à-vis municipalities with a male leader (i.e., control group).
A female leader, such as a mayor, permits us to observe gender
differences in heterogeneous environments on municipalities because
she may both impose her choice by command and influence on preferences lower down the ranks in public organizations; and her influence on other women in private organizations.
However, as the existing literature on top-level female leadership
suggests, women heading up organizations may provoke the so-called
QBP. The QBP is a situation in which women who succeed in maledominated settings play a negative role in the advancement of their
female subordinates (Derks et al., 2011).
In contrast to what the QBP may suggest, we add to the leadership
literature the importance of influence through “the role model (RM)
effect”. Hoyt (2005) and Hoyt and Blascovich (2007) report that
women may react to the tendency of thinking that only men are suitable
for management roles by demonstrating greater confidence and performing better. There must be a factor, however, that provokes this
reaction. In line with social cognitive theory (Bandura, 1977, 1986,
1992, 1999, 2005), we believe that the existence of female leaders may
increase women's self-esteem and encourage them to enter historically
male-dominated environments. Thus, an elected female leader may
influence and work as a RM who triggers a positive dynamic within
public and private organizations, which reduces gender-related differences. There is a tradition in political science and economics literature
showing that the RM effect has an influence on other women (Atkeson,
2003; Beaman, Chattopadhyay, Duflo, Pande, & Topalova, 2008;
Carroll, 1994; Hansen, 1997; Schlozman, Burns, & Verba, 1994). The
results of our investigation allow us to infer what occurs in organizations when female leaders are quasi-randomly appointed.
We conduct our research in Brazil because, to our knowledge, there
are no other empirical studies with a database as large as the one we
use; also, despite the effort involved in conducting our study, it is entirely replicable. Brazil has approximately 5600 municipalities averaging 20,000 inhabitants each, which ensures an investigation having
sufficient data points and statistical power to detect any effects. Mayors
hold an important political position in Brazil (Miguel, 2003). To verify
the changes that occur in organizations following the appointment of a
female leader, we consult a database containing the individual information of workers in approximately 8.3 million registered firms
(private and public organizations) at the municipal level. The proportion of women in the labor market is higher (59%) in Brazil than in
other developed countries, such as France (52%) and the United
Kingdom (57%).
The purpose of our study is to observe changes in gender results at
different positions in public and private organizations: top managers,
middle managers, and lower positions. Public organizations have
command: in other words, a mayor may choose the top-managers.
Furthermore, as leader she can influence women in lower positions in
her own organization for leadership, such as, for instance, for the position of middle manager. We also study the effect of female leadership
in terms of influence in private organizations, because the female leader

Theoretical overview and hypothesis
The Queen Bee Phenomenon in Business
Much of the research on female leadership is based on assumptions
of sisterhood and solidarity between women (see Mavin, 2006; Mavin,
2008). Women consider other women to be their natural allies. However, the expectation that women will align themselves with other

The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

presence of women in top management positions would extend “beyond
the walls of any single corporation": women leaders are role models and
mentors to other women and can break down stereotypes, encourage
young women to pursue careers in business, and break down the wage
gap between men and women. Still, such observational studies, and the
many others that claim to have documented a QBP have not been undertaken using appropriate designs to determine the causal effect of
top-level female appointments on female-related work outcomes.

women may not materialize and in these cases the “ally” label is replaced by the “queen bee” label. Some even suggest that women are
more like “evil stepmothers” than “fairy godmothers” and therefore will
be eternally punished for not supporting other women. For example,
Margaret Thatcher, the UK's first female prime minister, received from
the world's press (see BBC News's Reality Check team, 2018; Moreton,
2015) the “queen bee label” for not promoting the careers of other
women in her cabinet. However, many specificities may have explained
Thatcher's choices at that time; moreover, that was just one case, which
obviously may not be generalized.
The queen bee label is given to women who distance themselves
from other women in organizations where the majority of leadership
positions are held by men. Such women apparently seek individual
success by adjusting to the predominantly masculine culture in the
organization (see Kanter, 1977 and 1987; Staines, Tavris, & Jayaratne,
1974). Negative relations between women in organizations–that is, not
building alliances–have been highlighted in the literature for several
decades. Women in male-dominated organizations, aid and abet the
status quo by turning against other women, ignoring derogatory remarks about them and contributing to the derogation of these other
women by being disloyal to them (Nieva & Gutek, 1981).
In a general sense, female leaders are expected to be more understanding, more nurturing, more giving and more forgiving than men
(O'Leary & Ryan, 1994). However, when displaying such characteristics, they fail to meet the perceived requirements of the managerial
role, which mainly calls for masculine characteristics (Mavin, 2006).
This stereotyped view of women–that is, the perception of women as
being “less adequate” for a position–might lead to detachment from
their identity reference group, that is, other women. It also leads to
double bind due to descriptive, but also prescriptive, stereotyping because women cannot violate social role expectations (Eagly & Karau,
2002). To reach a senior position woman need to prove that they are
different from the other women in that environment; thus, in addition
to alienation they also describe themselves as possessing characteristics
considered to be masculine (e.g., assertive, competitive, risk-taking),
and this, in addition to having a stereotyped view of other women
(Ellemers, Heuvel, Gilder, Maass, & Bonvini, 2004).
An important argument in the literature about the QBP is that it is
not men or women who suppress better job outcomes for women; rather
it is the presence in the organization of gender stereotypes that hinders
the success of women in their careers and compels top-level female
leaders to behave in a hostile way toward subordinate females.
According to the QBP, women in the work environment are responsible
for “undermining” the careers of other women, but this phenomenon is
a response to gender inequality. Apparently, the QBP is present among
women who have low gender identification with their reference group
(i.e., other women). This low gender identification is manifested in
environments where women have experienced gender inequalities in
their careers. After being encouraged to remember the gender inequalities and bias they experienced in their careers, women then take
on stereotypically masculine characteristics, which involves emphasizing being different from other women and minimizing the presence
of gender inequality (Derks et al., 2011; Derks et al., 2011; Faniko et al.,
2016). The QBP is, therefore, thought to be a response to gender inequality, a consequence of gender discrimination experienced in the
workplace, and not a cause of it (Derks et al., 2015; Derks, Van Laar, &
Ellemers, 2016). Of course, once in place, the QBP would add to a simultaneity effect by increasing anti-female bias.
The QBP is, however, a questionable phenomenon because it is
difficult to establish a causal relationship between female behavior and
the low participation of women in top management positions. Deloitte's
study (see Women in the boardroom: A global perspective), for example,
makes us wonder if QBP “may have lost its sting”. There is evidence in
this study that when organizations have women in top leadership positions the number of board seats held by women is almost double that
of organizations with men at the top. Moreover, the effect of the

Female Leaders and Role Models
Organizational and government leaders are responsible for making
far-reaching decisions that can influence many aspects of society.
Individuals in key leadership positions may seize the opportunity to
change things for the better because leaders have the opportunity to be
proactive (Parker, Williams, & Turner, 2006). Despite the importance of
leadership for institutions and for ensuring the most capable are appointed to power, the existence of female leaders in a labor market is
very unequal when we compare it with the male situation. Women are
severely under-represented among business and political leaders, and
whereas they are generally increasing their share in the labor market in
different countries, female leaders are not particularly well represented
in the upper echelons; only 19% of firms have female top managers and
only 23% of the seats in national parliaments are held by women
(Miller, 2017).
Moreover, one can assume that male and female leaders behave
differently and favor those of the same sex in the group; given that
males dominate in many consequential settings suggests that it is
women who will pay the price in the long run as the male-oriented
hierarchical structures are reinforced. Another important argument,
therefore, is that if top leaders can reshape the allocation of resources
and influence the outcomes of large numbers of other people, current
gender imbalances in leadership may create additional distortions in
the overall distribution of wealth, power and wellbeing (Miller, 2017).
An explanation for this dearth of female leaders is the belief that
only men are suitable for managerial roles. According to Hoyt (2005)
and Hoyt and Blascovich (2007), top management positions require
achievement-oriented aggressiveness and emotional toughness. An incongruity exists between women's qualities and the leadership role,
which makes it more difficult for women to attain top leadership positions, and more difficult for these women to be viewed as effective in
these roles. Women are also negatively stereotyped if they violate
prescriptive expectations (Eagly & Karau, 2002).
Although the incongruity exists, many authors suggest that its prevalence depends on women's responses to this situation: The reaction of
women to the socially accepted ideas that men are more suitable for
managerial roles and that the qualities associated with women are incompatible with the qualities necessary for leadership roles (i.e., such
roles require more agent-like qualities). Redressing the situation can
occur by influence from one institution (e.g., observing more female
leaders in public institutions and politics) to other institutions (e.g.,
private), or encourage more women to enter public service and thereby
reduce descriptive stereotyping and what is considered normal. This
means that the more female individuals are seen in positions of power,
the less prevalent is the male-oriented stereotype of leadership in those
institutions and the more the cognitive structures chip away at the
“think manager think male” stereotype (the “think manager-think
male” view is a global phenomenon, especially among males according
to Schein, Mueller, Lituchy, & Liu, 1996).
The mechanism by which these effects may occur include (a) raising
self-efficacy, that is, belief about one's competence for addressing specific tasks; and (b) counteracting negative stereotypes of women from
observing female gender role models (Hoyt, 2005; Hoyt & Blascovich,
2007; Hoyt & Simon, 2011). Within this framework, research needs to
advance toward understanding how these positive pro-female influences (role model) can be triggered.

The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

considered unique or unusual, such as when gender issues are central to
the campaign agenda (Campbell & Wolbrecht, 2006; Wolbrecht &
Campbell, 2007). This type of visibility suggests that the candidacy of a
woman is sufficiently important to potentially result in victory. This
effect may shape the agendas of future election campaigns and have a
large impact on the political socialization of young women.
In short, we believe that women in power are likely to provide a
nurturing ground for other women to succeed and that these effects
stem from the direct effects that women in power exert in their spheres,
or under the influence of the RM effect. In considering our development
and these studies, the main hypothesis we examine is the following:

With regard to the role model argument in social cognitive theory
(Bandura, 1977, 1986), there is an element that can activate women's
responses to female leaders as role models. Although there are several
ways of developing leadership (Antonakis et al., 2012), action by
learning vicariously from observing a female leader, this is also a possibility. Other women observe and remember the sequence of events
that legitimized this leader and use this information to guide them in
their subsequent behaviors. It is possible, therefore, to establish a link
between a person's perceived self-efficacy stemming from learning vicariously, and future actions taken to emulate the role model.
Individual action can be replicated in the aggregate behavior of
individuals and, as a consequence, of an organization. Exposure to female leaders challenges the incongruity that exists between female
leaders and the male stereotype of a leader and can help lessen the
gender difference in organizations. Theoretically, this increases the
sense of female self-efficacy and thereby improves their aspirations and
may even encourage them to enter historically male-dominated environments. Evidence of this development pattern is found in political
science and economics literature (Beaman et al., 2008; Carroll, 1994). A
great number of studies from different settings confirm that women in
important political positions become positive role models for other
women and young girls due to an increase in the intention to change.
We will show some work on this below.
Carroll (1994) showed that the presence of women in political
leadership positions transformed beliefs about the appropriateness of
politics for women, thereby increasing interest in political issues. Additionally, the research carried out by Koch (1997), Sapiro and
Johnston Conover (1997), Hansen (1997), and Atkeson (2003) suggested that the presence of female candidates under certain conditions–legislative elections–has a positive effect on women's political
engagement, resulting in broader political discussion. Likewise, Fox
(1997) showed that political campaigns that have female candidates
tend to have a greater impact among families, generating further discussion and greater interest from young women in political matters.
Moreover, Campbell and Wolbrecht (2006) showed that the presence of
women in government can raise the level of involvement of young
women in politics, which then positively affects their likelihood of
political participation.
In a similar vein, Koch (1997), like Schlozman et al. (1994), stated
that women report greater knowledge of campaign details (e.g., candidate names and campaign schedules) when female candidates run for
leadership positions. However, Koch (1997), Sapiro and Johnston
Conover (1997) and Hansen (1997) believe the “role model effect” is
limited to the degree of visibility of the candidate. Atkeson (2003)
analyzed senatorial and gubernatorial campaigns and found that the
presence of female candidates affected the engagement of other women
only in campaigns that were highly competitive.
The role model effect can stem from other channels too. For instance, Beaman et al. (2008) showed that increased public exposure to
female leaders tended to improve the perception of the effectiveness of
female leaders, thereby weakening stereotypes about gender roles in
both public and private spheres. They also observed that there was a
greater tendency for voters to vote for a woman if those same voters
had been previously exposed to a government with a female leader.
This work showed that the effect of female leaders on other women led
to an effect that was more widespread than merely affecting women.
Research has also explored the position of women as an unusual
phenomenon (Campbell & Wolbrecht, 2006). They showed that it is
“unusual” for young women to become candidates for office and that
female candidates evoke a greater impact when the number of women
in politics is low. As more women participate in politics, however, their
involvement becomes less unusual and, therefore, there is a perceived
reduction in public interest. Thus, the increasing presence of women in
politics may have less of an impact on the aspirations of younger
women over time.
Gender is a relevant factor when the candidacy of a woman is

H1. Female leaders reduce gender differences in organizations.
Institutional Background, Dataset and Empirical Strategy
Institutional Background
Brazilian Political System
Brazil is a federalist country with three levels of government,
namely, federal, with 27 states and 5565 municipalities. There are three
electoral district sizes in the country. For local elections, the district is
the municipality, where elections for mayor and councilors are held.
For national and state elections, the electoral district is the state, where
federal, district, and state deputies, governors, and senators are elected.
The electoral district of the country as a whole exists only for presidential elections. Elections are held every two years in Brazil, with
local elections occurring every four years and elections for president,
governors, senators and federal, district and state deputies occurring
mid-term to local elections. Except for senators, who are elected for
eight-year terms, all office-holders in the Executive and Legislative
branches have four-year fixed terms. The constitutional amendment of
1997 established that Executive officials, including mayors, governors
and the president, can only be reelected once, which is the term limit
rule. Federal, District, and State Deputies and Councilors are elected by
way of an open-list proportional representation system; voters can
order the list of either candidates or parties. Although the parties initially order candidates, like countries such as Belgium, Austria, the
Netherlands, Switzerland, and Luxembourg do, voters can still alter the
order of candidates by what is known as preference voting. There is no
term limit for legislative members.
Since the 1988 Constitution, mayors have been chosen in one-round
elections using a majoritarian system in municipalities with fewer than
200,000 registered voters. Mayors are chosen in run-off elections in
municipalities with over 200,000 registered voters if no candidate
achieves a majority of valid votes in the first round (50% plus one of the
valid votes). We only use data for municipalities below the 200,000voter threshold. In doing so, we exclude approximately 120 of the
largest municipalities from the initial country sample. The reason for
this decision is to avoid strategic voting behavior, when voters do not
necessarily reveal their preferences in the first round (Fujiwara, 2011).
In the first-round election, voters choose the party or candidate of their
preference. In the second-round election, some voters do not have a
party of their preference. Given that the vote is compulsory, these voters can vote strategically. They can choose a party or candidate that is
not as bad for them in line with their initial preferences in the firstround. Thus, voters reveal their preference for parties and candidates
only in the first-round.
A. Electoral Gender Quota
Law 9100 of 1995 established an electoral gender quota in Brazilian
elections. In accordance with the law, all parties and coalitions had to
“to reserve” 20% of their candidacies for women. Law 9504 of 1997
changed the percentage of “reserve” candidacies for women from 20%
to 30%. Law 12,034 of 2009 subsequently changed the word “reserve”

The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

to “filled by” 30% female candidates, with a maximum of 70%. This
change was important because parties and coalitions did not understand the reserve requirement as a “commitment”. In the judgment of
the Superior Electoral Court (Resolution No. 23,373, 2011), the court
established that if female candidates do not represent 30% of the candidates of the parties or coalitions, all the candidates of these parties or
coalitions will be contested. Therefore, institutionally the electoral
gender quota was valid only after 2011. Before, there was an established idea that the quota may or may not be complied with by the
party or coalitions. The electoral quota was technically non-mandatory
for coalitions or parties in the period of our investigation.

State ownership of large-scale enterprises in Brazil occurs at the
federal, state and municipal levels. Federal state-owned enterprises may
be found in a great number of Brazilian municipalities. The Empresa
Brasileira de Correios e Telégrafos, for example, is set to expand the postal
service through its business units to all municipalities with a population
of > 500 residents by 2018.1 State-owned enterprises at the state level,
such as the Companhia de Processamento de Dados do Estado de São Paulo
(PRODESP)2 [data processing] and the São Paulo Desenvolvimento Rodoviário S.A. (DERSA) [highways], both owned by São Paulo, which is
richest state in Brazil, are located in several municipalities in that state.
Whereas mayors cannot occupy an office that involves a formal
position of command in a state enterprise (e.g., CEO), they delegate
power to trusted collaborators3 by making these collaborators the directors of state enterprises with power to appoint scores of other subordinates. According to Schneider (1991), appointments create a different informal control hierarchy: A trusted collaborator occupying a
top position gives the appointer a type of control, to the extent that they
are appointing someone who will exhibit predictable behavior. The
great appeal comes with the appointer's ability to gain control over
uncertainties and elicit the desired performance. The prior personal
bond between appointer and appointee establishes the subsequent
working relationship between government and enterprise and gives the
appointer control over management of the firm.
Geddes's (1994) approach is probably the most comprehensive reference to the topic of Brazilian appointment strategies. According to
Geddes, politicians who appoint able administrators committed to
achieving political goals have a greater probability of carrying out
successful policies. She argued that appointment strategies can be
predicted on the assumption that politicians want to continue to exercise political power in the future.
To conduct her tests, Geddes (1994) developed a quantitative
measure, the Appointment Strategy Index, of the extent to which politicians used competence and personal loyalty as the basis for selecting
administrative personnel in Latin American countries in the 1945–1993
period; these countries included Argentina, Chile, Brazil, Colombia,
Venezuela and Peru. The Appointment Strategy Index is based on answers to eight questions by the forty-four constitutional governments
(e.g., Was the criterion for choosing administrative personnel primarily
partisan rather than competence-driven? Was there a concentration of
appointments among party or coalition members? Were the publicsector jobs of party members protected?). Scores on this index depend
on the number of negative answers to the questions for each administration.
The Brazilian Appointment Strategy Index indicates that the political selection of top administrative personnel by politicians emphasizes
competence and loyalty in some periods and remaining within the
politician's party or coalition in other periods; note, loyalty and holding
the coalition together take precedence over competence.

Formal Labor Market
The Brazilian labor market has both formal and informal elements.
The formal market is regulated and comprises natural and legal persons
both registered according to the Brazilian legal system, whereas the
informal market has neither registered individuals nor employment ties
for professionals, nor legal obligations for companies.
As in other countries, Brazilian firms are either public or private.
The rules and conditions for employment in the public sector differ
from those in the private sector. Formal public employment includes
special, statutory, and contractual employment and employment by
appointment. Employment by appointment refers to employees appointed by government officials. Special employment refers to workers
hired under exceptional circumstances (e.g., cases in which there are
urgent needs or that require extraordinary skills) and who have a predetermined term of employment specified in their contracts. Statutory
employment is governed by a set of special rules, including lifetime
contracts from which employees cannot be discharged and pension
schemes that are far better than those available to formal employees in
the private sector. Finally, contractual employees are subject to the
same rules as employees in the private sector, this type of employment
being governed by the Consolidation of the Labor Laws (Consolidação
das Leis do Trabalho), which was enacted in 1943 to consolidate all
Brazilian labor legislation.
The 1988 Constitution established a minimum monthly salary for
formal workers in the public and private sectors. The minimum salary is
fixed for full-time work of 8 h per day, or 44 h a week. Employees who
work part-time–a maximum of 25 h per week–can receive less than one
minimum salary. The minimum salary is the lowest salary that an
employer can legally pay their employees for a full-time job, and the
lowest price for which a person can legally sell their labor. The Nominal
Minimum Wage was BRL 937.00 (BRL 31.23 per day) on January 1st
2016. This amount is approximately US$292.82 per month (exchange
rate: US$ 1/BRL 3.20).
Command in Public Organizations
The influence of mayors on the management of public-sector organizations can be observed in the appointments made. Appointments in
these Brazilian state-owned enterprises are direct in municipal-owned
enterprises and indirect at the federal and state levels. Indirect appointments are fundamentally associated with the distribution of public
resources and the partisan-electoral arena, in other words following the
practice of political “patronage” (Barbosa & Ferreira, 2017). Essentially,
this concept is an exchange of favors between federal/state politicians
and of public resources between municipal politicians–jobs in stateowned enterprises, allowances or tariff protection for particular industries, construction projects in particular districts–by way of votes or
other forms of political support (see Gordin, 2002Graziano, 1976). The
literature justifies this exchange of public resources by way of votes or
other forms of political support: Voters are heterogeneous in their affinity to parties. They also care about private benefits, and this interest
tempers their basic party loyalties. Consequently, voter willingness to
compromise their party affinities in response to offers of private benefits gives rise to this exchange (see Cox & McCubbins, 1986; Dixit &
Londregan, 1996; Lindbeck & Weibull, 1993).

We collect microdata from three different sources for our key hypothesis: the Supreme Electoral Court (Tribunal Supremo Eleitoral - TSE),
the Annual Report of Social Information (Relação Anual de Informações
Sociais - RAIS), and the Brazilian Institute of Geography and Statistics
(Instituto Brasileiro de Geografia e Estatística - IBGE).
The TSE began publishing electoral data electronically only after
1996, following introduction of the electronic ballot by Law 9100/95.
The 1996 municipal elections were the first to have electronic ballots in

According to Portaria, MC N° 6206 of 13/11/2015 there are 4261 municipalities.
For example, developing the “Poupatempo Program”, a program for reducing the
state's service bureaucracy.
As stated by Schneider (1991), trusted collaborators are people who have school or
professional ties and other personal relationships with the politicians and thus are appointed to “positions of trust”.


The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

earning < 0.5 minimum salaries. One last factor to note is that public
organizations are found in fewer municipalities than are private organizations.

the vast majority of the contests. Candidate information (mainly their
gender) is incomplete for the 1996 election. Our quasi-experiment focuses on the first municipal election in 2000; the immediate term after
the election started in 2001 (2001–2004); the next election that occurred in 2004; and the subsequent term, which began in 2005
(2005–2008). By choosing data for municipalities below the 200,000voter threshold, we exclude approximately 120 of the largest municipalities from the initial country sample.
The RAIS database contains information about workers employed in
formally-constituted firms in Brazil. Formal workers are employees who
have a formal labor contract, which in Brazil is defined as having a
booklet (i.e., the carteira de trabalho) that registers a worker's entire
employment history in the formal sector. Formal firms are those registered with the tax authorities, which means they possess the tax
identification number required for Brazilian firms (i.e., the Cadastro
Nacional de Pessoa Juridica – CNPJ). Our database contains no informal
firms; 40% of the economically active workforce in Brazil is involved in
informal employment (Census, 2010).
We know the municipality, wage information, and the gender of
employees. To carry out our investigation, we assume that the salary
information reflects the hierarchical position of workers in firms. For
example, individuals receive low salaries at the bottom of the hierarchy
(e.g., cleaning workers), intermediate salaries are middle managers,
and individuals with high salaries at the top managers (e.g., CEO). We
build the position of individuals and show results considering three
levels, observing the ratio between the groups (women and men): low
position (earning up to 1.5 minimum salaries), middle (earning between 10 and 20 minimum salaries), and top (earning over 20
minimum salaries). As over 20 minimum salaries is the highest-level
piece of information we obtained, “over 20” is regarded as the salaries
of top managers. Lower than this, it is salary of intermediate position:
middle managers. However, what is the empirical salary difference
between low and intermediate positions given that the separation point
between ranges is not so obvious? We would have found it difficult to
separate and classify these positions if we had had a large number of
significant intermediate results. This would have been particularly
difficult if we had had significant results close to 10: between 9.51 and
10, for example. This did not occur. The non-significant result of the
intermediate ranges, between 1.51 and close to 10, helped us establish
the difference in positions between low and middle level careers.
We aggregate the salary information by gender at the municipal
level annually. Then we build the average results for each municipality
in the three terms: 1997–2000, 2001–2004, and 2005–2008. The
1997–2000 term is important for the internal validity of the quasi-experiment (Eggers, Fowler, Hainmueller, Hall, & Snyder, 2015; Imbens &
Lemieux, 2008). The 2001–2004 term is the result immediately after
the 2000 election, our quasi-experiment election. The 2005–2008 term
serves to evaluate the gender difference of workers in firms in the following term when the mayor in the electoral quasi-experiment is reelected. By establishing different moments in time, we expect to contribute to the “time” dimension that is lacking in literature, as indicated
by Fischer et al. (2017). Finally, we extract municipal data from the
census that was held in 2000, which was produced by IBGE.
Table 1 shows the variables used in our empirical investigation,
their creation, and sources.
Tables 2A and 2B show the descriptive statistics of the variables of
three different groups of margins of victory (30%, 10%, and 5%) and
the significant difference between them when the margin of victory is
5%. The statistical difference between the variables is small.
There is variation in the number of observations between the variables that reflect gender differences (e.g., the ratio of female to male
workers with different earnings). This variation reflects the fact that
there are no organizations paying certain salary levels in some municipalities. For example, if a municipality has no large organizations
then it is unlikely to have organizations paying their employees > 20
minimum salaries. In contrast, there are municipalities with no workers

Empirical Strategy
Our empirical strategy is designed to investigate whether a female
leader, compared to a male one, improves the position of female
workers in organizations over which she has “command/influence”
(i.e., as an elected mayor), or for which she only has influence, a role
model (i.e., in private organizations). This test is no easy task, given
that the non-observable characteristics that define a leader can interfere
in this relationship. In line with previous literature, for instance, personality traits (Fleeson, 2001; Fleeson & Gallagher, 2009) and biological differences (Ashton, 2007) may explain leadership, and voters
may be influenced by these differences in their choices. Moreover, it is
possible to assume reverse causality between the variables. The direction of our investigation is x (modelled independent variable) → y
(dependent variable) and not y → x. This reverse causality may occur
because improvements in the situation of female workers compared to
male workers in a municipality may explain the emergence of a female
leader, that is, because of a pro-women environment. An investigation
that does not consider this possibility, therefore, is likely to contain
To solve this problem, we adopt a sharp regression-discontinuity
approach, as proposed by Lee et al. (2004) and Lee and Card (2008).
Municipalities where a woman wins by a large margin are likely to be
different from those where a woman wins by a small margin. However,
when we narrow our focus to those municipalities with close-run
elections, it becomes more plausible to believe that election outcomes
are determined by idiosyncratic factors. The female leader is elected
without the influence of observable and non-observable variables, and
we can observe the influence of the leader on our variables of interest
(x → y).
Following the above arguments, we focus our investigation on
municipalities with close-run mayoral races (i.e., a victory margin close
to zero) in which the two candidates receiving the most votes are a
woman and a man. In other words, we are modeling a gender race to
observe what occurs when the winner is “randomly” appointed.
Our treatment variable Dit is a dummy variable that equals one
when a woman defeats a male opponent in municipality i in year t. The
control group (Dit = 0) is formed by the municipalities that elect a man.
The running variable is the margin of victory (Marginit), which is defined as the percentage of votes between female and male candidates
for mayor. Thus, the relationship between Dit and Marginit can be
written as follows:

1 if Marginit > 0
Dit = ⎧

⎩ 0 otherwise


The cut-off point at which the margin of victory equals zero corresponds to a single criterion for determining the candidate's gender. The
impact of a local female leader on Yit+τ, the dependent variable that
represents the gender ratio for each municipality–the main variable in
our investigation–is defined by parameter β, which is an average
treatment effect near the cut-off point. This effect can be written as


lim E (Yit + τ | Marginit ) −

Margin ↓ 0

lim E (Yit + τ | Marginit )

Margin ↑ 0


Our causal identification strategy, therefore, is to detect the “jump”
in the dependent variable at the discontinuity point that can be attributed to the effect of crossing the discontinuity point. The treatment
effect is causally interpreted because for a woman winning in municipality i there should be no systematic differences between the observable characteristics of municipalities, the electorate, and so forth,
and the non-observable characteristics of the leader observed by voters.

The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

Table 1
Definition of variables, how they are constructed, and their sources.

Margin of victory


PT (Worker´s Party),
(Brazilian Social
Democratic Party), and
PFL (Liberal Front

Difference in the percentage of
votes between female and male
candidates for mayor, considering
the candidates in first and second
places in the first round of
Mayors who completed primary
education but not high school;
mayors who completed high school
but not higher education; and
mayors who completed higher

Construction of variable
We extract the number of votes for
each candidate. We then build the
margin of victory.

Dummy variables with values equal
to 1 when the definition is met and
zero otherwise.

Dummy variables with values
equal to 1 when the definition is
met and zero otherwise. We
include two parties with left-wing
ideology (PSDB and PT) and one
party with right-wing ideology.

Percentage of houses with access to
mains water supply in the
Population, percentage of
municipal population under 1 year
old; percentage of municipal
population 1 to 4 years old;
percentage of municipal population
49 to 59 years old; percentage of
municipal population 60 to 69
years old; and percentage of
municipal population 70 to 79
years old.

Population in thousands

Percentage of women as
a proportion of total

Gender ratio

Superior Electoral Court
(TSE) for the election of
2000 (municipal election
for mayors and

PT (Workers Party), PSDB
(Brazilian Social Democratic Party),
and PFL (Liberal Front Party)

Percentage of population
with access to mains
water supply



IBGE - Brazilian
Institute of Geography
and Statistics – 2000

Percentage of women in total
municipal population
We extract the information for all
firms existing in each municipality
over several years (1996/2008).
Then, we build the different
variables annually. The lagged
variable contains the average
between the years 1996 and 2000 in
the municipality. The moment
immediately after the election
contains the average between the
years 2001 and 2004. The next term
and the mayor being reelected
contains the average between the
years 2005 and 2008.

Ratio of female to male workers in
public and private organizations in
specific variables: Hours worked;
age of workers; average number of
minimum salaries; number of
workers earning up to 0.5
minimum salaries, between 0.51
and 1 minimum salary, between
1.01 and 1,5 minimum salaries,
between 10.01 and 15 minimum
salaries, between 15.01 and 20
minimum salaries, and over 20
minimum salaries.

RAIS – Annual Social
Information on workers
in firms in the formal
sector - produced by the
Ministry of Employment
and Labor. The data are
from 1996 and 2008.
The income values are
deflated by IGP-DI
(2000). The information
is provided annually by
firms (information
centralized in

Note: Ideology of Latin American Parties are classified as established by Coppedge (1997).

taking time into account in empirical investigations is something that is
sorely lacking.

Under certain conditions, municipalities where women barely lose can
serve as a reasonable counterfactual for municipalities where they
barely win; in this way, we have a quasi-experiment, wherein we solve
the endogeneity problem and can make relatively strong causal claims
(Antonakis et al., 2010).
There is the possibility of estimate β with different methods (Calonico,
Cattaneo, & Titiunik, 2014; Imbens & Lemieux, 2008). Without a great
technical difference between them, we opt to use the estimator suggested
by Calonico et al. (2014). This approach shows that the non-parametric
estimation of (2) by local linear regression typically leads to bandwidth
choices that are too “large”, which means that there will be a large
asymptotic bias term; a triangular kernel function, in which the weight of
each observation decays with the distance from the cut-off.
In addition to addressing the issue of endogeneity, we can investigate the influence of the leader on the variables over time (t + n).
Insights about how to study the impact of time in leadership literature
have been discussed by several authors (e.g., Antonakis et al., 2012;
Fischer et al., 2017; Shamir, 2011); however, as we mentioned before,

Validity of the research design
We validate our quasi-experiment using internal and external tests
following the procedures established by Imbens and Lemieux (2008)
and Eggers et al. (2015). First, we determine whether the electoral
process was manipulated. Electoral manipulation for any candidate
(i.e., a woman or man) invalidates the quasi-experiment. We use
McCrary's (2008) test to identify electoral manipulation.
A visual inspection of the histogram of the density of the margin of
victory (Fig. 1 –left side) may show electoral manipulation of the cut-off
point (the percentage margin of votes equals zero) with different class
intervals or “bins” (2 percentage points-pp, 1 pp, and 0.5 pp – left side
of figure). Electoral manipulation occurs when more candidates win the

The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

Table 2A
Descriptive statistics.

Margin of victory
-30% and 30%

Std. Dev.

-10% and 10%


Std. Dev.

-5% and 5%


Std. Dev.


Mayors who completed primary education but not high school
Mayors who completed high school but not higher education
Mayors who completed higher education
Worker´s Party (PT)
Brazilian Social Democratic Party (PSDB)
Front Liberal Party (PFL)
Percentage of total population with access to mains water supply
Percentage of women as proportion of total population
Gender ratio of hours worked – Formal Workers
Gender ratio of age – Formal Workers
Percentage of municipal population less than 1 year old
Percentage of municipal population 1 to 4 years old
Percentage of municipal population 49 to 59 years old
Percentage of municipal population 60 to 69 years old
Percentage of municipal population 70 to 79 years old
Dependent Variables
Gender ratio - Average income in terms of minimum salaries
Gender ratio - Number of workers earning up to 0.5 minimum salaries
Gender ratio - Number of workers earning between 0.51 and 1 minimum salary
Gender ratio - Number of workers earning between 1.01 and 1.5 minimum salaries
Gender ratio - Number of workers earning between 10.01 and 15 minimum salaries
Gender ratio - Number of workers earning between 15.01 and 20 minimum salaries
Gender ratio - Number of workers earning over 20 minimum salaries



















Table 2B
Descriptive statistics and t-tests by elected.

Margin of victory Between -5% and 5%

Std. Dev.




Std. Dev.


Mayors who completed primary education but not high school
Mayors who completed high school but not higher education
Mayors who completed higher education
Worker´s Party (PT)
Brazilian Social Democratic Party (PSDB)
Front Liberal Party (PFL)
Percentage of total population with access to main water supply
Percentage of women in total population
Gender ratio of hours worked – Formal Workers
Gender ratio of age – Formal Workers
Percentage of municipal population less than 1 year old
Percentage of municipal population 1 to 4 years old
Percentage of municipal population 49 to 59 years old
Percentage of municipal population 60 to 69 years old
Percentage of municipal population 70 to 79 years old













Dependent Variables
Gender ratio - Average income in terms of minimum salaries
Gender ratio - Number of workers earning up to 0.5 minimum salaries
Gender ratio - Number of workers earning between 0.51 and 1 minimum salary
Gender ratio - Number of workers earning between 1.01 and 1.5 minimum salaries
Gender ratio - Number of workers earning between 10.01 and 15 minimum salaries
Gender ratio - Number of workers earning between 15.01 and 20 minimum salaries
Gender ratio - Number of workers earning over 20 minimum salaries
Note: * p < 0.01, ** p < 0.05, * p < 0.1.



The Leadership Quarterly xxx (xxxx) xxx–xxx

P.R. Arvate et al.

McCrary test


Density Estimate




Absolute frequency





-.6 -.5 -.4 -.3 -.2 -.1 0

.1 .2 .3 .4 .5 .6



Percentage margin of votes







Percentage margin of votes


Fig. 1. Frequency and McCrary test.

determining which characteristics differ between male and female
candidates (see Table 3B).
Between the examined characteristics of elected mayors, which include education level and party affiliation (i.e., Worker's Party-PT,
Brazilian Social Democratic Party-PSDB, and Liberal Front Party-PFL),
we find no significant differences. Parties may have either a favorable
strategy for electing women candidates or a bias against women candidates (Kittilson, 2006; Krook, 2006).
Imbens and Lemieux (2008) and Eggers et al. (2015) did not establish the importance of covariate balance in the individual characteristics of same-gender candidates (i.e., women/winners compared
with women/losers and men/losers compared with men/winners) in
the sharp regression discontinuity design. In this study, we compare
same-gender characteristics. We adopt this procedure to check whether
the women/men elected differ from other candidates of the same sex.
The premise of this verification is to exclude the bias that may be attributed to “super” women/men winning (a woman with characteristics
that are superior to those of other women in her group – between
winners and losers – which could explain the electoral result; the same
is valid for men) (see Table 3C).
The individual characteristics of same-gender candidates appear
well-balanced and the candidates appear to be very similar. We find a
small discrepancy between male candidates. Among male candidates,
two of the six characteristics are not balanced at the 10% level. We have
fewer winning men than losing men in the PT (Worker's Party) and
fewer winning men who completed higher education. We have two
other parties (PFL and PSDB) and two other education levels (i.e.,
completed primary education and completed high school) that are balanced between the candidates. There is no strong evidence of “super”
candidates, and this residual difference between male candidates does
not affect the internal validity of the quasi-experiment.4

election than lose it. Potential manipulation of the selection mechanism
would occur if there were a jump in frequency of the running variable
(percentage margin of votes) close to the cut-off point. Based on the
procedure proposed by McCrary (2008), we test the null hypothesis of
continuity of the density of the running variable (i.e., margin of votes)
against the hypothesis of an interruption at the cut-off point. Our results
indicate that the null hypothesis cannot be rejected (t = −0.1556),
indicating no discontinuity around the cut-off point. Therefore, there is
no discontinuity around the cut-off point.
We used a second procedure to investigate whether municipal
characteristics affected the electoral result. We also use here the
method proposed by Calonico et al. (2014). The main idea is that there
is no “jump” in these variables in elections with a zero margin of votes
(i.e., close to the cut-off point). If there is a jump, then municipal
characteristics might explain the electoral result (the environment). For
instance, a higher percentage of women as a proportion of the total
population when a woman wins a close-run election could be the factor
explaining the victory of a woman over a man. We need, therefore,
results without a “jump” in these variables. In other words, we need
balanced results for the covariates (see Table 3A).
The municipal characteristics are well-balanced; that is, we observe
no significant differences between the treatment and the control group on
municipal covariates. The variables we used included: (a) percentage of
the total population with access to the public drinking water supply
system;, (b) the total population of the municipality; (c) women as a
percentage of the total population; (d) the ratio of hours worked in formal
organizations, by gender; (e) the ratio of ages in formal organizations, by
gender; (f) the percentage of the population < 1 year old; (g) the percentage of the municipal population 1 to 4 years old (children can affect
women's progress in the labor market); (h) the percentage of the municipal population 49 to 59 years old, the percentage of the municipal population 60 to 69 years old; and (i) the percentage of the municipal population 70 to 79 years old (the retirement age of women and men may
differ, given that women live longer than men in Brazil, IBGE).
We also examine the differences in the observable characteristics of
candidates for mayor. This assessment provides a means for

According to Guala (2012), internal validity is threatened in an experiment when
confounding variables compete with the “jump” in the zero margin of victory to explain
the dependent variables. The non-dominant difference (two out of seven) between male
candidates does not explain the victory of women over men.


Related documents

1 s2 0 s1048984317305179 main
gender analysis april 2012
googles ideological echo chamber
10 1023 a 1010243318426
2251 s15 qp 13
three hidden ways education contributes to discrimination

Related keywords