ArticlePDF Available

Abstract and Figures

How does the symbolic power of a female president affect female parliamentary behavior? Whereas female descriptive representation has increased around the world, women parliamentarians still face significant discrimination and stereotyping, inhibiting their ability to have a real voice and offer “thick” representation to women voters. We leverage the case of Malawi, a case where the presidency changed hands from a man to a woman through a truly exogenous shock, to study the effect of a female president on female parliamentary behavior. Drawing on unique parliamentary transcripts data, we argue and show that women MPs under a female president become empowered and less confined to stereotypical gendered issue ownership patterns, leading to a significant increase in female MP speech making. Our results speak directly to theories of symbolic representation by focusing particularly on intra-elite role-model effects.
Content may be subject to copyright.
From Thin to Thick Representation- How a Female President
Shapes Female Parliamentary Behavior
Michael Wahman
Assistant Professor
Michigan State University, Department of Political Science
ORCID ID: 0000-0002-7334-5793
Nikolaos Frantzeskakis
PhD Candidate
Michigan State University, Department of Political Science
ORCID ID: 0000-0001-7794-1739
Tevfik Murat Yildirim
Associate Professor
University of Stavanger, Department of Media and Social Sciences
ORCID ID: 0000-0001-7120-6020
Acknowledgement: Previous versions of this paper were presented at the 2018 Annual Meeting of the American
Political Science Association, the 2018 University of Aarhus Challenges to Democratization Workshop, the 2019
Gender, Political, Representation and Development in Africa Conference at University of Ghana, the 2019
Comparative Agendas Project Conference in Budapest, and at the Michigan State University Comparative Politics
Seminar. We are grateful for invaluable comments and support provided by Frank Baumgartner, Sarah Brierley, Ruth
Carlitz, Nicholas Cheeseman, Aaron Erlich, Manolis Frantzeskakis, Hamida Harrison, Andrew Kerner, Staffan
Lindberg, Anna Lührmann, Shahryar Minhas, Merete Bech Seeberg, and Peter VonDoepp. We are also hugely
indebted to the Chief Parliamentarian Librarian in Lilongwe, Maxwell Banda and the Malawi Clerk of Parliament,
Fiona Kalemba. Invaluable research assistance was provided in Malawi by Felix Chauluka and Fannie Nthakomwa.
We are also grateful to two dedicated teams of undergraduate research assistants: one team at University of Missouri,
including Andrew Gilstrap, Trent Hall, Katie Mechlin, Seamus Saunders, Tricia Swartz, Momoko Tamamura, Mandy
Trevor, Hunter Windholtz, and Hanna Wimberly and one team at Michigan State University including Layla Brooks,
Isaac Cinzori, and Anthnoy Luongo.
Corresponding author
How does the symbolic power of a female president affect female parliamentary behavior?
Whereas female descriptive representation has increased around the world, women
parliamentarians still face significant discrimination and stereotyping, inhibiting their ability to
have a real voice and offer “thick” representation to women voters. We leverage the case of
Malawi, a case where the presidency changed hands from a man to a woman through a truly
exogenous shock, to study the effect of a female president on female parliamentary behavior.
Drawing on unique parliamentary transcripts data, we argue and show that women MPs under a
female president become empowered and less confined to stereotypical gendered issue ownership
patterns, leading to a significant increase in female MP speech making. Our results speak directly
to theories of symbolic representation by focusing particularly on intra-elite role-model effects.
“There’s a proverb in Malawi that says, ‘a female cow does not pull the cart, the female cow is
kept for milking’”
-Joyce Banda, Malawi’s first female president
“My sisters, my daughters, everywhere, find your voices!”
-Ellen Johnson Sirleaf, Liberia’s first female president
Across the globe, female heads of government remain rare (Jalalzai 2008; Thames and Williams
2013; Lawless 2015; O’Brien et al. 2015).
Lack of female presidents and prime ministers is a
symptom of wider gender inequalities both social and economic (Stockemer and Byrne 2011).
But, as the literature on symbolic representation reminds us, a dearth of female political leadership
may in itself perpetuate an image of appropriate female roles in public life (Wolbrecht and
Campbell 2007; Simien 2015; Alexander and Jalazai 2020).
Building on the work by Franceshet et al. (2012), Bauer (2016, 224) defines symbolic
representation as: “altering gendered ideas about the role of women and men in politics, raising
awareness of what women can do as political actors and legitimizing them as political actors, or
encourage women to become involved themselves in politics as voters, activists, candidates and
leaders.” Whereas most research on symbolic representation has concentrated on the way in which
female political role models may shape attitudes and behaviors at the mass-level (e.g. Zetterberg
According to the Thames and Williams (2013) data, only about 3% of the world’s executives in the period 1945-
2006 were women.
2009; Morgan and Bruise 2013; Barnes and Taylor-Robinson 2017; Liu and Banaszak 2017; Liu
2018), this paper focuses on the symbolic effect of a female president at the elite-level. More
precisely, we study how a female president may empower female members of parliament (MPs)
to assert more parliamentary leadership and change their parliamentary behavior.
Women have gained increasing numerical parliamentary representation around the world
(e.g. Krook 2010; Stockemer 2011). Still, some authors have questioned the extent to which
increased female parliamentary representation has led to real female parliamentary leadership
(Weldon 2002; Beckwith and Cowell-Meyers 2007; Childs and Krook 2008). Hassim (2006, 173),
writing about the role of women in African parliaments, notes that female MPs remain
marginalized in parliamentary affairs through the subtle (or sometimes not so subtle) gendered
hierarchies that still persist within political institutions. She argues that analysis should distinguish
between “thin” and “thick” representation. Whereas thin representation relates to the mere
presence of women in parliament, thick representation refers to a form of representation where
women are granted real voice and power in legislative assemblies.
We argue that the presence of a female president has an important intra-elite symbolic
effect and enhances female thick parliamentary representation. Specifically, we argue that the
presence of a female president serves to normalize female political power, re-define gendered
norms about appropriate female political behavior and competences, and create momentum for
more assertiveness among female MPs. In term, these mechanisms will lead to increased thick
female representation.
Empirically, we focus on the effect of Malawi’s first female president, Joyce Banda, on
female thick representation. Malawi presents a unique opportunity to study how symbolic
representation of a female president affects female parliamentary behavior. The most serious
challenge to studying the effects of symbolic representation is that women’s election into office is
commonly endogenous to the social and political context that simultaneously shape gendered
perceptions of female leadership. For instance, Thames and Williams (2013) find that cross-
nationally the probability of having a female executive is highly correlated with female legislative
representation and greater history of female political participation. However, Joyce Banda was
never elected to become president, but came to power by the exogenous event of her predecessor’s
natural death. Moreover, her weak political position when taking office makes Malawi a suitable
case for generalizations, even to cases with politically-stronger female executives.
Following recent research on parliamentary behavior (Bäck et al. 2014; Clayton et al. 2014;
Wang 2014; Bäck and Debus 2019; Blumenau 2019), we proxy thick representation by measuring
gender differences in the frequency of parliamentary speech-making. We make use of an original
dataset of Malawi parliamentary speeches in the period 1999-2014 (covering close to 110,000
speeches) created using innovative machine learning techniques. In our analysis, we model how
the exact same set of parliamentarians changed their legislative behavior from the period before
Malawi’s first female president (2009-2012) to the period after Malawi’s first female president
In accordance with our theory, our analysis shows that women MPs speak significantly
more after the inauguration of Malawi’s first female president. Moreover, we separately analyze
women’s participation in debates concerning the economy. The economy is perceived as the most
important issue for Malawian voters (Afrobarometer 2014) but has been stereotypically perceived
as a “male” topic (Krook and O’Brien 2012; Bäck et al. 2014). Due to the electoral salience of the
economy, female presidents have worked actively to redefine gendered notions of women’s
inability to lead on this issue. Our analysis shows that women MPs participated significantly more
often in debates on the economy after a woman became president. Finally, to assess other potential
alternative explanations, we collect and analyze additional data from earlier Malawian
parliamentary sessions. However, none of the alternative explanations find any empirical support.
The paper contributes to the literature on gender and politics in several ways. First, while
most research on symbolic representation has focused on symbolic effects at the mass-level, we
theorize intra-elite consequences of symbolic representation. Second, the paper mitigates the
somewhat gloomy conclusions about the limited impact that female heads of government may
have in otherwise fundamentally patriarchal societies (e.g. O’Brien et al. 2015; Chikapa 2016;
Verge and Pastor 2018). Third, the paper joins a growing literature (e.g. Clayton 2015; Wang
2014) using new data and innovative research methods to place cases from new democracies
within the mainstream of the study of women in politics. All in all, our study shows that a singular
focus on enhancing the descriptive representation of women in parliament is inadequate (Childs
and Krook 2009) and that more emphasis is needed on women’s exclusion from positions at the
very top of the political hierarchy (Jalalzai 2013; Liu and Banaszak 2017).
Beyond Thin Parliamentary Representation
Research on women and legislative politics has virtually exploded in the last few decades.
Whereas most of the early literature studied the role of women in local and national legislatures in
advanced democracies (e.g. Thomas 1994; Dolan and Ford 1997; Lawless 2015; Levendusky
2005), newer research has increasingly focused on the prospects for increased equality in political
representation in less established democracies in regions such as South-East Asia (Liu 2018), Sub-
Saharan Africa (Bauer and Britton 2006), the Middle East (Shalaby and Elimam 2020), and Latin
America (Schwindt-Bayer 2010). Most research on women in legislative politics falls into two
broader categories: women’s descriptive or substantive representation (Wängnerud 2009). The
literature on descriptive representation has highlighted the prevailing numeric under-
representation of women in legislatures globally, but also particularly focused on how certain
institutional solutions, such as gender-quotas, may enhance the number of women in representative
institutions (Tripp and Kang 2008). The literature on substantive representation, on the other hand,
has focused more on how the interests of women citizens are furthered by increased female
political representation (Mansbridge 1999; Chiweza 2016; Clayton et al. 2019; Nwankwor 2019).
However, as Hassim reminds us, enhanced female parliamentary representation may do
little to promote women’s political interests if they remain marginalized within representative
institutions. Despite increased descriptive representation, she concludes that women have had
severe difficulties in being “taken seriously within institutions that are historically and culturally
male” (2006, 173). While women may offer substantive representation and speak more on the
issues of crucial importance to women citizens, their voices may still be quelled within the
patriarchal culture of national legislatures.
Several accounts from around the globe have illustrated the marginalization of women
within parliaments, all pointing towards the lack of thick female representation. For instance,
several studies of parliamentary cultures in cases as diverse as the United Kingdom, Sweden, and
Namibia have attested to frequent harassment and discrimination against women (Levendusky
2005; Bauer 2006; Erikson and Josefsson 2019). Others have focused on women MPs’ slow
career advancement and their absence in cabinets and high-prestige committees (Reynolds 1999;
Heath and Taylor Robinson 2005; Krook and O’Brien 2012; Shalaby and Elimam 2020).
An area in which the lack of thick representation has been particularly apparent is
parliamentary debates. Whereas a number of studies from around the world have highlighted
how female MPs have suffered harassment and ridicule when taking to the floor (e.g. Tørassen
2019), quantitative research has illustrated that women tend to speak significantly less than men
in parliament (Bäck and Debus 2019; Blumenau 2019).
Much of the earlier research on women in parliament has done the important work of
descriptively documenting the lack of thick female representation. Still, we know little about the
ways in which women’s voices within legislative assemblies can be amplified. One existing,
compelling, argument in the literature relates the lack of thick representation to the general lack
of descriptive representation. According to the often-cited critical mass theory, women are
inhibited in their roles as legislators by their minority status. Only when female representation
grows numerically can women collectively advance their position to demand greater space in the
legislative process (Dahlerup 1988). While there is research to support that increased female
representation within parliaments or specific party groups may result in women taking to the
floor more frequently (Yoon 2011; Bäck and Debus 2019), fostering thick representation by
enhancing descriptive representation is, at best, an incomplete strategy. Research from countries
like Sweden, where women are represented in parliament at almost the same rate as men (but
where a woman has still never been prime minister), has shown that women MPs still speak
significantly less in parliamentary debates than men (Bäck et al. 2014)
Fostering Thick Representation Through Symbolic Representation
Leaving the critical-mass theory aside, this paper investigates intra-elite symbolic representation
as an alternative pathway towards enhanced thick representation. Bauer (2019) has argued that
symbolic representation is the least explored, but potentially most powerful path towards
enhanced female political representation. Most existing research on symbolic representation has
not focused on intra-elite effects, but instead on the effect of female political representation on
mass-level public opinion and political engagement. Research has suggested several
consequences of increased female parliamentary representation, such as increased female
political engagement (Atkeson 2003; Wolbrecht and Campbell 2007; Barnes and Burchard
2013), increased likelihood in women running for office (Gilardi 2015), enhanced public support
for female leadership, (Burnet 2011) and generally greater satisfaction with democracy
(Schwindt-Bayer 2010). Liu and Banaszak (2017) and Alexander and Jalazai (2020) have argued
that the symbolic effects of women’s political leadership should be particularly pronounced
when women take up highly visible leadership positions.
These studies on mass-level effects of symbolic representation prompt the question as to
whether symbolic effects may also exist at the intra-elite level. We argue that role model effects
would be particularly pronounced among women of similar social status. As argued by Liu
(2018), when asymmetries are large between the power obtained by women in political
institutions and the power obtained by ordinary women in social structures, role model effects
may be limited. However, even in highly patriarchal societies, women within political elites are
likely to identify strongly with each other and model the behavior of more successful women
within the same political hierarchies. When one of their fellow elite women manages to break the
ultimate glass ceiling to become the head of government, this would have a particularly strong
symbolic effect for parliamentary women.
We argue that the installation of a female president
has several important intra-elite symbolic effects that will promote women’s assertiveness and
shape female MPs’ propensity to speak in parliament.
First, a female president has the ability to normalize the presence of women in political
power. Women remain highly restricted in their political activities; women MPs are often
perceived as representatives of their gender and advocates for female “special interests” (Gilardi
2015). However, the presidency represents the very embodiment of male political power. When a
woman takes up a position of such magnitude, it challenges the traditional understandings of
female leadership being a remarkable deviation from the norm. This reduced salience of gender
for female political elites should transfer from the president to other female elites, such as
parliamentarians. For instance, writing about the early years of Angela Merkel’s leadership in
Germany, Ferree (2006, 106) notes: “She is inevitably going to contribute to changing the
symbolic association of gender and politics [...] Paradoxically, one of the most powerful
evidences that such a change has happened already is the extent to which her gender had become
unremarkable as she goes about the work of exercising political authority.”
Second, a female president can alter stereotypes about appropriate behavior for male and
female MPs and challenge gendered preconceptions about suitable leadership qualities. Past
scholarship has shown that men and women in leadership positions are often held to different
standards when competence and leadership are evaluated. The traits that make women in
leadership positions ‘unlikable’ are the same characteristics that are often deemed necessary for
leadership competence (Heilman and Okimoto 2007). This is an important dilemma. As
We might expect that the symbolic effect of a female president is larger than for a female prime minister.
According to Jalalzai (2013: 3), women are more inhibited by gender stereotypes in presidential than parliamentary
regimes: “Perceptions of women’s negotiation and collaboration skills limit them less than their supposed inability
to act unilaterally, aggressively and decisively- all necessary presidential traits.”
Amanatullah and Tinsley (2013, 110) succinctly put it, “those who act agentically are seen as
competent but unlikable; those who act communally are viewed as likable but incompetent.”
However, the presence of a female president may challenge narrow preconceptions of gender-
appropriate behavior. Such changes would encourage women to participate more frequently in
parliamentary debates. Furthermore, showcasing of women leadership can alter the sort of personal
qualities that are associated with competence. In extension, female perspectives may be perceived
as more valuable in democratic deliberation. Women in power have often drawn on maternal or
matriarchic imagery to legitimize their leadership (Franceschet et al. 2016). Such strategies were
apparent in Joyce Banda’s rhetoric, stating for instance: “where I come from it is the woman who
shoulders the biggest responsibility of supporting the family, through her contributions of labor,
time, emotions and energy” (Banda 2013).
Third, the promotion of a female president signifies a general momentum for women in
politics, allowing female MPs the opportunity to advance their positions and assert leadership.
Such effects have been noted in the general literature on symbolic representation, arguing that the
historic marginalization of women has made women and men alike susceptible to the view that
women are inferior in governing (Alexander 2012; Fox and Lawless 2004). After Joyce Banda
became the president, several MPs used the installation of a female president as a declaration of
victory, a call for more respect, and as a demand for further advancement. For instance, Jean
Kalilani, an MP of Malawi’s Dowa Central constituency (Malawi Parliamentary Hansard
06/20/2012), declared: “The question is no longer whether a woman can be president of a country
or not, but rather what she can deliver. Malawi must get more and more women in decision making
positions.” It should, however, be noted that women’s political momentum may be a double-edged
sword. On the one hand, it will embolden female MPs. On the other hand, it may also inspire
backlash among male MPs, who increasingly regard women as a threat to their political careers
(Krook 2015). We still, on balance, expect that the presence of a female president would lead to
women taking more assertive leadership roles, but will revert back to this possible backlash effect
in the empirical section. Based on this discussion, we formulate the following hypothesis:
Frequency Hypothesis: When a woman becomes head of government, female MPs will speak
more frequently in parliament than before.
A possible objection to our hypothesis is that a female president would empower not only female
MPs, but also their male colleagues. At the mass-level, Schwindt-Bayer (2010) showed that higher
female representation in Latin American parliaments was related to increased satisfaction with
democracy among women as well as men. A woman president could increase the sense of inclusion
among MPs of both genders. However, speaking time in parliament is finite. Although speaking
time it is not a complete zero-sum game, where an increase in women’s speech frequency will
result in the corresponding decline in men’s speech frequency, men and women cannot increase
their number of speeches indefinitely (due to limitations in time and sessions). It is likely that
women MPs will be more affected by the mechanisms proposed above than their male colleagues.
For this reason, our hypothesis relates to female MPs specifically.
A female president may affect not only the frequency with which women MPs speak, but
also the topics on which women parliamentarians access leadership in. Earlier research has shown
that female MPs tend to be particularly marginalized on issues that have traditionally been
perceived as “masculine” or “hard” and have been foregone for appointments to cabinet and
committees relating to such topics (e.g. Reynolds 1999; Krook and O’Brien 2012; Shalaby and
Elimam 2020). Similarly, Bäck et al. (2014) show that women in the Swedish parliament are more
likely to speak on issues that the authors characterize as soft”, and less likely to speak on those
issues that can be considered “hard”.
Above all, one traditional “masculine” topic that stands out is the economy. Previous
research has found women to be conspicuously absent on issues related to the national economy
(Bäck et al. 2014). This is vital for two reasons. First, voters tend to rank the economy consistently
as one of the most important political issues. This is particularly true in developing economies.
For instance, Clayton et al. (2019) showed that African voters, men and women alike, rank the
economy as far more important than any other issue. Women’s particularly thin representation on
issues that have been considered “masculine” can be a consequence both of choice and legislative
marginalization. Women may prioritize other issues than men, particularly those that tend to
overwhelmingly affect the livelihood of fellow women (Wängerud 2009). However, given the
weight that women voters assign to the economy, we cannot explain women MPs relative absence
on this issue as a reflection on women voters’ preferences. Secondly, given the electoral
importance of this issue, women’s absence in economic debates are likely to severely hurt women
MPs career advancement.
Research from a variety of contexts have suggested that gender stereotypes have led voters
to perceive male political candidates as more competent on economic issues (e.g. Hayes and
Lawless 2015). Aware of such negative stereotypes, female executives have often actively
challenged them, trying to redefine the economic issue in more feminine terms. Most famously,
Margaret Thatcher in the United Kingdom made the comparison between the national economy
and a household budget to underscore her competence to deal with a struggling economy (Ponton
2010, 207). In other cases, particularly in the developing world, women leaders such as Joyce
Banda and Ellen Johnson Sirleaf, have actively pushed a more inclusive economic agenda, where
female economic empowerment was placed at the core of the national development mission
(Spiker 2019, ch 2). Broad economic leadership from a female president and her active attempt to
redefine gender stereotypes should also empower women MPs to take more active leadership on
the economic issue. We, thus formulate the following hypothesis:
The economy hypothesis: When a woman becomes head of government, female MPs will speak
more frequently in parliament on the economy than before.
Case Selection and Research Design
Empirically, we leverage the case of Malawi to study the symbolic effect of a female president on
thick female parliamentary representation. Joyce Banda became Malawi’s first and Africa’s
second female president (after Liberia’s Ellen Johnson Sirleaf), when she entered office in April
2012. Banda was, however, never elected to the presidency. She had been the vice president since
2009, but assumed office after her predecessor, Bingu Mutharika, died in office. At the time Banda
took office, she had already fallen out with the late president and had broken away from the
governing Democratic Progressive Party (DPP) to start her own opposition party, the Peoples’
Party (PP). The rift between Banda and the DPP was so wide that top officials within the DPP,
centering around the late president’s brother Peter Mutharika, unconstitutionally tried to prevent
Banda from assuming office (Patel and Wahman 2015).
Figure 1 positions Malawi in a global and African comparison on two vital variables: the
V-Dem Female Political Empowerment Index
and female parliamentary representation
(Coppedge et al. 2020).
Figure 1: Female Political Empowerment and Female Parliamentary Representation
Note: Data from 2012. Figure only includes countries classified as electoral democracies or
liberal democracies by the V-Dem Regimes of the World Index. Dashed lines represent African
averages and solid lines represent global averages.
The index ranges from 0-1 and is obtained by averaging three different V-dem indices: the women’s civil liberties
index, women’s civil society participation index, and women’s political participation index.
Malawi scores below both the African and global averages on the political empowerment index.
This is not to say that women have been completely sidelined economically, culturally, and
politically in Malawi. Indeed, much of the literature on cultural gender norms in Malawi has
emphasized the importance of the country’s mostly matrilineal culture for granting political and
economic access for women both contemporarily and historically in a way that has not been the
case in many other African countries (Muriaas et al. 2019; Robinson and Gottlieb Forthcoming).
In terms of female political representation, Malawi in 2012 is slightly above the African and
slightly below the global average. From democratization in 1994 to the 2009 election, female
parliamentary representation increased significantly from 7% to 22% (see Table 1). The Malawian
public had also become increasingly aware of gender disparities in political representation through
the 50:50 campaign, a campaign launched in the 2009 election to enhance female political
representation (Kayuni and Muriaas 2014). Nevertheless, despite acknowledging progress,
Amundsen and Kayuni (2016: 1) conclude: “Much of the traditional role of women still prevail in
Malawi. She is the caretaker; her role is largely limited to the private domain, and much social and
cultural prejudice against her participation in politics persists.”
Table 1: Female Parliamentary Representation in Malawi 1994-2014
Parliamentary Term
Female Representation
Source: Chimunthu Banda (2017:176)
It is not clear whether a country with relatively high or relatively low levels of gender
equality could be expected to experience the largest symbolic effect of a female president on thick
parliamentary representation. On the one hand, one might argue that the largest effect would be
observed in the most gender-unequal societies (often found in the developing world) where a
female president would break the strongest with traditional gender roles. On the other hand, it
might be that political gender roles would be more amendable in otherwise relatively gender equal
societies (often found within OECD countries). Findings from studies on symbolic representation
in various settings do not provide clear priors.
Although we remain agnostic on whether more or
less gender-equal cases would make for a least-likely case, we believe that the Malawian case has
one critical benefit in terms of empirical generalization. While much of the literature on female
For examples of studies showing positive symbolic effects in advanced industrialized countries see e.g. Wolbrecht
and Campbell (2007) and Karp and Banducci (2008); showing positive symbolic effects in developing countries see
Beaman et al. (2009) and Barnes and Burchard (2013); showing no symbolic effects in advanced industrialized
countries see Dolan (2006) and Wolak (2019); showing no symbolic effects in developing countries see Zetterberg
(2009); Clayton (2015); Liu (2018).
executives has focused on politically strong women with considerable electoral mandates, strong
political parties and significant political clout, such as Angela Merkel, Michelle Bachelet or Ellen
Johnson Sirleaf (e.g. Ferree 2006; Thomas and Adams 2010), Banda was politically weak, her
political future was unpredictable, and her party was unorganized. She also suffered from the sort
of misogynistic public attacks that women in leadership positions across the world have often
endured (Lora-Kayambazinthu and Shame 2016; Chikaipa 2019). For instance, when first
becoming president, former first lady, Callista Mutharika, questioned her ability to rule, dismissing
her as a ‘simple market woman’. Banda’s popularity also vanished during her term in office due
to a serious economic crisis, inherited by her predecessor, and the revelation of a systemic
corruption scandal, popularly referred to as “Cashgate (Dulani and Chunga 2015). Although the
corruption scandal involved both the Banda and the Mutharika regimes, and although Banda was
never proven to be personally implicated, voters placed the Cashgate scandal squarely at the feet
of Banda (Zimmerman 2015). It is, hence, possible that sexism directed towards Banda and her
vanishing popularity may have reduced the positive symbolic effect and lead to backlash against
female MPs (Krook 2015). We argue that if a politically relatively weak president like Banda could
change the behavior of female parliamentarians, we would expect the same effect in cases with
politically stronger executives.
The Malawi case also offers a unique opportunity in terms of internal validity. The unusual
circumstances surrounding President Banda’s rise to power make Banda’s presidency exogenous
to changes in perceptions concerning women’s leadership. Indeed, the endogenous nature of
female political representation represents one of the greatest challenges for causal analysis of
symbolic representation. Moreover, since there was no parliamentary election in this period, the
same MPs were in parliament throughout the period June 2009-April 2012 (under President
Mutharika) and April 2012-April 2014 (under President Banda) and there was no change in the
number of female MPs.
One possible limitation with our research design is that although Banda’s appointment was
exogenous, we cannot rule out the possibility that other political events happening after
Mutharika’s death shaped the legislative behavior of men and women differently. However, to
mitigate this risk, our analysis will also study other alternative explanations that could account for
changes in female legislative behavior, using data from additional parliamentary terms.
Data and Methods
In order to study the legislative behavior of MPs, we rely on parliamentary transcripts, an
underutilized resource in the African context. With the notable exception of the innovative work
by Clayton, Josefsson, and Wang (Clayton et al. 2014; Clayton et al. 2017; Wang 2014) from the
Ugandan parliament, most quantitative content analysis based on parliamentary transcripts come
from the North American or European context (e.g. Proksch and Slapin 2012; Fernandes et al.
We introduce a new dataset on speeches in the Malawian parliament for the period 1999-
2014. These transcripts had not previously been made readily available to the public but were
obtained through intense fieldwork and direct communication with the librarian of the parliament
in Lilongwe. Our collection of transcripts covers 32 parliamentary sessions. Each session
represents one legislative sitting in parliament and varies in length. In total, the transcripts cover
almost 110,000 speeches held by the 463 MPs in our dataset.
Given the large volume of speeches, we relied on machine learning to code their content.
Machine learning has proven to be a valuable tool for social scientists, particularly for text analysis
(Lucas et al. 2015) and text classification (Hopkins and King 2007; Clayton et al. 2016).
Specifically, we used a supervised learning approach. We employed transfer learning on the pre-
trained network BERT (Devlin 2018) and trained it to code the speeches based on the widely used
coding scheme of the Comparative Agendas Project (Baumgartner et al. 2006). The training set
consisted of 2,500 hand-coded speeches from original transcripts selected at random. We discuss
our approach in more detail in the online appendix.
In our analysis, we use two dependent variables: the number of speeches made by an MP
and the speeches on the economy. To create the second dependent variable, we recoded as speeches
on the economy all speeches that, based on the CAP scheme, were coded as domestic
macroeconomics, labor and employment, foreign trade, or banking, finance, and domestic
Independent Variables
Our main independent variable is the gender of the MP. We obtain the gender of each MP from
Ott and Kanyongolo (2010, 412ff.) and code whether an MP is female. We are primarily interested
in the interactive effect of gender and whether a speech was held in the Banda term. Our data
contains eleven parliamentary sessions for the period June 2009 April 2014; sessions 1-7 occur
before the Banda presidency and sessions 8-11 during the Banda term.
We also include a number of control variables. Malawi is known for having high MP
turnover, particularly for female MPs. Since we may expect that experienced MPs are more
engaged in debates than less experienced MPs, we control for whether an MP is a newcomer. We
code this variable by scanning electoral results from previous elections. We also control for
important offices held by MPs. Specifically, if an MP was serving as minister, deputy minister,
committee chairperson, or as part of their party’s legislative leadership
they are coded as Senior
MPs. We expect that women would be less likely to have ministerial portfolios (Arriola and
Johnson 2014) and that MPs holding such offices are more likely to appear in parliament. In
models measuring the number of speeches on the economy, we control for membership in the
Budget and Finance Committee. In addition, we control for whether an MP belongs to either of
the two ruling parties: DPP (in the Mutharika period) and PP (in the Banda period), the main
opposition party MCP, or if they were independent.
Finally, in order to account for time trends,
we include a time variable for the number of months the government has been in office, as well as
its squared term. Descriptive statistics for all variables are provided in Table 2.
Table 2: Descriptive Statistics
Dependent Variables
Standard Deviation
Total Number of
Speeches on the
Independent Variables
Senior MP
Finance Committee
Legislative leadership roles include being Leader of the House, Government Chief Whip, Deputy Government
Chief Whip, Leader of the Opposition, Party Whip, and Party Deputy Whip.
Using information from the transcripts
Model Estimation
For our main analyses, we use MP-month as our unit of analysis and counts of speeches as our
dependent variable. Looking at the distribution of our dependent variables, we find that the
standard deviation is much larger than the mean, and for a large percentage of the observations the
dependent variable takes the value 0. Taking into consideration the data distribution, with a large
number of structural 0s, we conduct our analysis using a zero-inflated negative binomial model
with standard errors clustered on the individual MP (to account for heteroscedasticity across MPs)
to compare women's and men's speech counts during the Mutharika period and the Banda period.
A possible alternative to the model-based approach we opt for here would be a difference-in-
difference design, however the structure of the data would make such an approach problematic.
During the Banda period, we expect that female MPs are "treated" by the fact that the
president is a fellow woman. On the other hand, we do not expect this “treatment” to affect male
MPs. We thus focus on the change in the number of speeches made by female MPs in the two
periods and on how they compare to the number of speeches made by male MPs during the same
two periods. In order to investigate these relationships, we include an interaction term between the
variables Female, denoting the gender of the MP, and Banda, denoting under which government
the speech took place. Our main model is specified as follows with i denoting the individual and t
denoting time:
Number of Speechesi,t=b0+b1*Bandai,t+b2*Femalei,t+b3*(Bandai,t*Femalei,t)+b4*SeniorMPi,t
The benefit of a difference-in-difference design is maximized when the model used is linear. Even if standard
difference-in-differences assumptions hold in a non-linear model (including the parallel trends assumption),
nonlinear models cannot exploit them and often violates them (Lechner 2010). Given that our dependent variable is
a count variable with high overdispersion and a large number of structural 0s, the use of a linear model would not be
appropriate. Instead, we use a count model and follow the advice by Ai and Norton (2003) on estimating effects
from nonlinear models, using simulations keeping confounding variables at substantively meaningful values.
Research on parliamentary behavior around the world has found that women speak less in
parliament than their male colleagues (Bäck et al. 2018). This is also true in Malawi. Looking at
the 2009-2014 parliament, we find that the average female MP spoke only 69% as often as the
average male MP. We also find the sort of gendered differences in speech topics, observed from
other countries. Figure A1 in the appendix shows speech frequency by gender broken up by
speech topic. Gender differences are particularly noteworthy on economic issues. In the 2009-
2014 parliament, the average female MP only made 49% as many economy speeches as the
average male MP.
The noteworthy discrepancy in speech making between the genders is troubling. Not only
is parliament dominated by men descriptively, but female MPs are also less active in
parliamentary debates. Low female activity in parliamentary debates, combined with low
descriptive representation means that out of the 51,981 parliamentary speeches that were made in
the Malawian parliament in the period 2009-2014, 84.3% were made by male MPs.
However, the gendered patterns among the MPs with modal characteristics
seem to
change when Malawi gets its first female president, Joyce Banda. In Figure 2 we plot a kernel-
weighted local polynomial of the female-to-male MP speech ratio on the months that the
parliament was in session. This preliminary analysis shows a clear break between the Mutharika
presidency, where female MPs spoke about 60% as often as male MPs, and the Banda
Newcomer, party-affiliated (not independent), male and female MPs that do not hold ministerial office or are a part
of parliamentary leadership.
presidency, where they spoke about 85% as often as their male counterparts. Figure 2 also fills a
second important function in observing general time-trends, independent of the break offered by
the Banda presidency. This is important to rule out the possibility that the “Banda effect” is
simply due to a generally positive time trend. The 2009 election was a breakthrough for female
parliamentary representation in Malawi (Kayuni and Wang 2014). Furthermore, members of the
women’s caucus in parliament were offered various training opportunities to become more
effective legislators (Chiweza et al. 2016; Adams and Wiley 2020). These factors could have
resulted in gradual and increased activity of women during the entire parliamentary term.
However, figure 2 shows no general positive time-trend in the data. The female/male MP speech
ratio is relatively consistent within the Mutharika and Banda periods respectively.
Figure 2: Female to male MP speech ratio over time
To further explore the potential intra-elite effect of symbolic representation on thick female
representation, we present a number of zero-inflated negative binomial regression models. We
report our findings using simulations for our two main hypotheses: the frequency hypothesis and
the economy hypothesis. In linear-additive regression models, the effect and statistical significance
of independent variables can be interpreted directly from the table of results based on a single
coefficient. However, this is not true for multiplicative interaction models, like the one we present
here. The most important basis for statistical inference is not the p-value of the interaction effect
itself, substantive effects are better assessed through substantively meaningful simulations (Ai and
Norton 2003; Brambor et al. 2005). As we are mostly interested in the behavior of the typical MP,
we set all variables at their mode. As a result, our estimates describe newcomer, party-affiliated
(not independent), male and female MPs that do not hold ministerial office or are a part of
parliamentary leadership. The estimations for the frequency hypothesis are presented in the top
row of graphs in Figure 3.
Figure 3: Simulations Main Analysis
The top left panel of Figure 3 shows the expected number of speeches by male and female
MPs during the Mutharika and Banda periods, respectively. Before we move on to interpret these
results, it is worth reiterating that our hypothesis states that women will speak significantly more
Full results table in Table A1 (appendix)
2 4 6 8
Predicted Number Of Events
Male Female
Predictive margins of Banda over Female
for all speeches
-1 0 1 2 3 4
Contrasts of Predicted Number Of Events
First Differences between governments
for all speeches
-4 -2 0 2
Contrasts of Predicted Number Of Events
First Differences between genders
for all speeches
.4 .6 .8 1 1.2 1.4
Predicted Number Of Events
Male Female
Predictive margins of Banda over Female
for speeches on the economy
0.2 .4 .6 .8
Contrasts of Predicted Number Of Events
First Differences between governments
for speeches on the economy
-.6 -.4 -.2 0 .2
Contrasts of Predicted Number Of Events
First Differences between genders
for speeches on the economy
Note: The 95% confidence intervals are included
in the Banda-term. However, if we see the same kind of increase for men, this would suggest other
possible substantive explanations, not directly related to gender. It could also suggest that a
woman president empowers both men and women to speak more frequently in parliament. We did
not explicitly hypothesize that the gender gap between men and women would disappear (although
this could potentially be the case), nor did we hypothesize that women will speak significantly
more than men in the Banda period.
During the Mutharika period, male MPs are projected to speak 5.7 times/month, while
female MPs are only expected to speak 3.6 times. On the other hand, during the Banda period,
male MPs were expected to speak 6.1 times and female MPs 5.9 times/month. In the middle top-
row panel, we show how the expected number of speeches for men and women changes in the
Banda period, compared to the Mutharika period. We find that the increase of 2.2 speeches/month
for women under Banda (observed in figure 3) is statistically significant (p=0.027). On the other
hand, men do not speak significantly more in the Banda period compared to the Mutharika period.
Finally, in the top right panel, we contrast the expected number of speeches made by male and
female MPs over the two periods. We find that women spoke significantly less than men during
the Mutharika period, but not during the Banda period. Taken together, these graphs provide sound
support for our frequency hypothesis. Our main models use the number of speeches to
operationalize thick representation. In the appendix we also re-run the analysis using the number
of words as our dependent variable (Figure A2). This robustness test does not yield substantially
different results.
Economy Hypothesis
The bottom left panel of Figure 3 shows the expected number of speeches on the economy by
male and female MPs during the Mutharika and Banda periods, respectively. In both periods,
male MPs are projected to speak more often on the economy than female MPs. In the Mutharika
period, the expected number of speeches per month on the economy was 0.86 for men, compared
to 0.52 for women. In the Banda period, however, the gender gap is reduced, where the expected
number of speeches per month on the economy was 1.06 and 0.89 for men and women,
To directly test our hypothesis, the middle graph of the bottom row contrasts the predicted
number of economy speeches/month for each gender, comparing the Banda period to the
Mutharika period. We find that while women increase their number of economy speeches
significantly (p = 0.045) in the Banda period, there is no statistically significant difference in the
number of economy speeches for men between the Mutharika and Banda period.
Finally, the bottom right panel of Figure 3 contrasts the expected number of speeches on
the economy made by male and female MPs over the two periods. We find that women spoke
significantly less than men on the economy during the Mutharika period, but not during the Banda
period. As a result, the analysis provides strong support for our economy hypothesis and lends
further support to the theory on the symbolic importance of a female head of government. The
results are not substantially different if we measure our dependent variable with number of words
rather than number of speeches (Figure A2).
To further investigate changes in gendered patterns on various issues, figures A3-A5 in the
appendix replicate the same analysis for the six other topics that are most frequently discussed in
the Malawi parliament. This additional analysis shows that the economy is a topic where female
participation is substantially increased due to the intra-elite symbolic representation effect.
However, we do also find interesting variations between the Mutharika and Banda period on two
other topics that have traditionally been perceived as “male” topics: government operations and
crime. We show that women spoke significantly more about government operations under Banda,
we also show that there is no longer a statistically significant difference between male and female
MPs in their expected number of speeches on crime in the Banda period.
Alternative Explanations
Our empirical strategy offers many advantages, but there are still limitations. Specifically, some
of the effects found in the main analysis could have been related to idiosyncratic factors in the
Mutharika period or other possible factors affecting women and men differently after the
installation of Banda. To assess the most important alternative explanations, we collected and
coded additional transcripts for the period 1999-2009 and engaged in further analysis.
A Mutharika Effect Rather than a Banda Effect?
First, we evaluate the possibility that the observed effect was associated particularly with the
Mutharika presidency rather than the female presidency. The 2009 parliament was Mutharika’s
second term after re-election in the 2009 presidential election. To evaluate whether Mutharika is
an exceptional male president, we compare male and female parliamentary behavior during
Mutharika’s first term in office (2004-2009) with male and female parliamentary behavior during
the last term of Mutharika’s male predecessor, Bakili Muluzi. If it turns out that women spoke
This represents all transcripts available through the Parliamentary Library in Lilongwe.
more during Muluzi than they did under Mutharika, the observed increase in female speech under
Banda could be more a consequence of Mutharika losing office than Banda gaining it.
Figure 4: Difference between the Muluzi and Mutharika Periods
Figure 4 shows the results of our simulations, using the same model specification as in our main
models (full results in Table A3 in the appendix). In support of our hypothesis and contrary to the
alternative explanation, it does not seem like the Mutharika regime was particularly hostile towards
women parliamentarians. On the contrary, the middle panel on the top row of Figure 4 shows that
women spoke significantly more under Mutharika than they did under Muluzi. Furthermore, the
middle panel on the bottom row shows that women spoke significantly more on the economy under
Mutharika than they did under Muluzi.
2 4 6 8 10
Predicted Number Of Events
Male Female
Predictive margins of Mutharika over Female
for all speeches
Contrasts of Predicted Number Of Events
First Differences between governments
for all speeches
-4 -2 0 2 4
Contrasts of Predicted Number Of Events
First Differences between genders
for all speeches
0.2 .4 .6 .8 1
Predicted Number Of Events
Male Female
Predictive margins of Mutharika over Female
for speeches on the economy
0.2 .4 .6 .8
Contrasts of Predicted Number Of Events
First Differences between governments
for speeches on the economy
-.6 -.4 -.2 0 .2 .4
Contrasts of Predicted Number Of Events
First Differences between genders
for speeches on the economy
Note: The 95% confidence intervals are included
Do Women Generally Speak More Later in the Parliamentary Term?
One potentially problematic aspect of our research design is that we compare women early on in
the parliamentary term with the same women later in the same term. Although we control for
earlier parliamentary experience, it may be that women without experience are more reluctant to
speak early on in the term than men without experience. Research on gender difference in
communication styles has argued that women improve their ability to navigate gendered
communicational expectations as they gain in experience (Pfafman and McEwan 2014).
Figure 5: Difference early and late parliamentary terms in Muluzi and Mutharika I
First Half
Second Half
0 5 10 15
Predicted Number Of Events
Male Female
Predictive margins of Second half over Female
for all speeches
-12 -10 -8 -6 -4
Contrasts of Predicted Number Of Events
First Differences between governments
for all speeches
First Half
Second Half
-2 0246
Contrasts of Predicted Number Of Events
First Differences between genders
for all speeches
First Half
Second Half
0.5 1 1.5
Predicted Number Of Events
Male Female
Predictive margins of Second half over Female
for speeches on the economy
-1 -.8 -.6 -.4 -.2 0
Contrasts of Predicted Number Of Events
First Differences between governments
for speeches on the economy
First Half
Second Half
-.4 -.2 0 .2 .4
Contrasts of Predicted Number Of Events
First Differences between genders
for speeches on the economy
Note: The 95% confidence intervals are included
To evaluate this potential alternative explanation, we used additional data from the two earlier
parliamentary terms in our dataset (1999-2004 and 2004-2009). We divided the terms into two
periods, breaking the terms at the half-point of the term. We thus compare women and men in the
early months of the 1999 and 2004 parliaments, with the same women and men in the late months
of the 1999 and 2004 parliaments. Results are displayed in Figure 5 (full table in the appendix
Table A4). Contrary to this alternative explanation, we do not find that women generally speak
more later in the parliamentary term. In fact, the opposite is true. Looking at the middle panel of
the top-row, we find that in the earlier Malawian parliaments both women and men actually spoke
significantly less in the later part of the parliamentary term than they did in the earlier part of the
same term. Looking at the middle panel of the lower row, we find that the same is true if we focus
particularly on the economy.
A Party Leader Rather than a Presidential Symbolic Effect?
President Banda was not only the president, but also the leader of a major party in parliament: the
People’s Party (PP). No MPs were elected on a PP ticket; the party was founded by Joyce Banda
in July 2011, two years after the 2009 election and only 10 months before the start of the Banda
presidency. Although a few early MPs joined the party before April 2012, most PP members joined
the party after Banda became president (Svåsand 2015). It is commonplace in Malawi politics that
opposition MPs and independents join the ruling party to gain access to executive resources
(Young 2014).
Although empirical literature on female party leaders has not unequivocally found that
female party leaders are more inclined to promote the careers of female politicians (O’Brien et al
2015), some research has argued that gender dynamics within parliamentary groups may affect
female legislative behavior. For instance, Bäck and Debus (2019) hypothesize that women MPs in
parties with higher female representation speak more frequently than their female colleagues in
more male-dominated parties. It is possible that the Banda effect can be attributed to women within
PP being empowered by their female party leader and that the increased speech frequency of
women MPs is related to the PP gaining in parliamentary numeric strength after Banda became
To investigate this possibility, we re-ran the analysis excluding all PP members of
parliament from the analysis. The results are displayed in Figure 6. The results of our analysis
remain substantially unchanged even when excluding PP MPs from the analysis.
Figure 6: Difference Between Banda and Mutharika Periods, Excluding PP MPs
To further probe the importance of political parties, Figure 7 shows the change in the female/male
speech ratios for PP’s two main party rivals, DPP and MCP, in the Mutharika and Banda periods,
respectively. It is possible that opposition to women’s participation would increase particularly in
the DPP as misogynistic opposition to the female president grew stronger. However, this is not
what we find. On the contrary, the female/male speech ratio increased in both DPP and MCP in
the Banda period. Looking at non-senior, newcomer MPs, the group which makes up the vast
majority of parliament, the average DPP female MP made only 67% as many speeches as the
average male DPP MP in the Mutharika period, compared to 92% in the Banda period. The
corresponding averages was 71% for MCP female MPs in the Mutharika period and 91% in the
2 4 6 8 10
Predicted Number Of Events
Male Female
Predictive margins of Banda over Female
for all speeches
-2 0 2 4 6
Contrasts of Predicted Number Of Events
First Differences between governments
for all speeches
-4 -2 0 2 4
Contrasts of Predicted Number Of Events
First Differences between genders
for all speeches
.5 1 1.5
Predicted Number Of Events
Male Female
Predictive margins of Banda over Female
for speeches on the economy
0.5 1
Contrasts of Predicted Number Of Events
First Differences between governments
for speeches on the economy
-.5 0 .5
Contrasts of Predicted Number Of Events
First Differences between genders
for speeches on the economy
Note: The 95% confidence intervals are included
Banda period. All in all, this additional analysis confirms that the symbolic effect of the female
president is not confined to her own political party, but runs across parties.
Figure 7: Female/Male Speech Ratio for DPP and MCP in Mutharika and Banda Periods
Note: Bars represent male/female speech ratio for non-senior, newcomer MPs.
Do Women Speak Less During Democratic Erosion?
The second Mutharika term was characterized by political instability and creeping
authoritarianism. As the executive was put under intense pressure, the regime resorted to political
repression and strengthened executive control. The extent of Malawi’s democratic backsliding was
most acutely felt at the time of the July 2011 demonstrations, at which point Malawi police shot
and killed 11 pro-democracy protesters (Cammack 2012). After Mutharika’s death, Malawi’s
democracy normalized under the reign of President Banda. Banda implemented important reforms
to scale back many of the draconian laws put in place by the Mutharika regime and restore the
freedom of the press (Chinsinga 2015). Could it be that women were particularly reluctant to speak
in parliament during Mutharika due to eroding democracy? To investigate this alternative
explanation, we study an earlier episode of democratic erosion in Malawian history, the time of
President Muluzi’s third-term presidential bid. This period lasted roughly between January 2001
and March 2003
and has many similarities with the backsliding period under Mutharika
(VonDoepp 2005). In order for Muluzi to pave the way for a constitutional amendment to allow
for a third term in office, the government increased repression on opposition, journalists, and civil
society (Cammack 2012). In the period, Malawi was downgraded on Freedom House’s Freedom
scale from a rating of 3 to a rating of 4 on both Civil Liberties and Political Rights (higher values
represent lower levels of freedom). If we find that women speak less frequently in the period of
democratic backsliding under Muluzi, it could be that the effect observed in our main analysis is a
consequence of the shrinking democratic space in Mutharika’s second term. Figure 8 compares
the frequency of speeches between men and women in President Muluzi’s second term, comparing
the period of “democratic backsliding” (January 2001-March 2003) with that of “democratic
stability” (June 1999- January 2001 and March 2003-May 2004).
Starting with the first serious floating of the third-term agenda and ending with the final recognition that Muluzi
would not run for re-election (VonDoepp 2005).
Full results in Table A5
Figure 8: Difference between the democratic stability and democratic erosion period in the Muluzi
As with the other alternative explanations, we do not find any support. The middle upper-row
panel shows that while men spoke significantly less in the backsliding period there was no
difference for women MPs. Similarly, the right-hand panel on the upper-row shows that while
women spoke significantly less than men in the democratic stability period, there was no
statistically significant difference in the backsliding period. Looking particularly at speeches on
the economy, the middle panel in the bottom row of Figure 8 shows that while men spoke
significantly less in the backsliding period than they did in the stability period, there is no
statistically significant difference between the periods for women. All in all, these models show
Democratic Stability
2 4 6 8 10
Predicted Number Of Events
Male Female
Predictive margins of Backsliding over Female
for all speeches
-6 -4 -2 0 2
Contrasts of Predicted Number Of Events
First Differences between governments
for all speeches
Democratic Stability
-6 -4 -2 0 2
Contrasts of Predicted Number Of Events
First Differences between genders
for all speeches
Democratic Stability
0.5 1 1.5
Predicted Number Of Events
Male Female
Predictive margins of Backsliding over Female
for speeches on the economy
-1 -.5 0 .5
Contrasts of Predicted Number Of Events
First Differences between governments
for speeches on the economy
Democratic Stability
-1 -.5 0
Contrasts of Predicted Number Of Events
First Differences between genders
for speeches on the economy
Note: The 95% confidence intervals are included
that, in fact, men’s advantage in speech making vis-á-vis women is greater in the stability period
than it is in the backsliding period.
A Banda Backlash?
While a woman president challenges gendered stereotypes, female political advancements may
also enhance the degree to which male MPs perceive their female colleagues as a threat to their
own political advancement. For this reason, we might experience a backlash against female MPs
(Krook 2015). Although we have not observed a general backlash, it is possible that the symbolic
effect of a female president is highly related to the popularity of the president. In cases where the
female president is perceived as performing poorly, women politicians might collectively be
blamed and gender stereotypes reinforced. Over time, Banda’s popularity vanished and prominent
scholars of Malawi politics such as Tiyesere Mercy Chikapa (2016) have argued that Banda’s
perceived failure had negative repercussions for other female politicians in their effort to gain re-
election. Is it the case that the symbolic effect disappears as Banda’s popularity diminished?
Although we do not have monthly data on President Banda’s approval ratings,
Afrobarometer data collected shortly after Banda’s installation and just before the 2014 election
make clear that Banda’s popularity was reduced significantly during her term in office (Dulani and
Chunga 2015). It is hard to pinpoint exactly when Banda’s popularity drops, but it is widely
believed that the Cashgate corruption scandal revealed in September 2013 was a decisive moment
associated with Banda’s fall from grace (Zimmerman 2015). In Figure 9 we break the Banda
presidency into a pre-Cashgate and a post-Cashgate period. Although women speak less post-
Cashgate, we see an even larger reduction in the number of speeches for men. In fact, the predicted
difference in the number of speeches between men and women is smaller in the post-Cashgate
period than it is in the pre-Cashgate period (the difference between men and women remains
statistically insignificant in both periods). In other words, the intra-elite symbolic effect persists
despite the female president’s reduced popularity.
Figure 9: Difference between the Pre- and Post-Cashgate period in the Banda regime
Although women’s descriptive representation in parliament has increased globally, women are still
lacking in thick representation (Hassim 2006). Much important work has been conducted to
Before Cash Gate
After Cash Gate
2 4 6 8 10 12
Predicted Number Of Events
Male Female
Predictive margins of Cash Gate over Female
for all speeches
-8 -6 -4 -2
Contrasts of Predicted Number Of Events
First Differences between periods
for all speeches
Before Cash Gate
After Cash Gate
-6 -4 -2 0 2
Contrasts of Predicted Number Of Events
First Differences between genders
for all speeches
Before Cash Gate
After Cash Gate
0.5 1 1.5 2
Predicted Number Of Events
Male Female
Predictive margins of Cash Gate over Female
for speeches on the economy
-2 -1.5 -1 -.5
Contrasts of Predicted Number Of Events
First Differences between periods
for speeches on the economy
Before Cash Gate
After Cash Gate
-1 -.5 0 .5
Contrasts of Predicted Number Of Events
First Differences between genders
for speeches on the economy
Note: The 95% confidence intervals are included
describe the form of obstacles faced by female legislators around the world, but we still need more
research that identifies paths towards enhanced thick female representation. As argued by previous
research, increasing female descriptive representation is insufficient (Weldon 2002; Childs and
Krook 2009). Women legislators, even in practically gender-balanced parliaments, still possess
less voice than their male colleagues in vital parliamentary debates (Bäck et al. 2014). This paper
has offered a new path towards enhanced thick parliamentary representation: intra-elite symbolic
We argue that elite women will find inspiration from other more successful women within
the same political hierarchies. The presence of successful women at high positions in political
hierarchies have the ability to redefine the role of women in politics, challenge perceptions of male
and female leadership, and create a general momentum for women in politics. Looking particularly
at the importance of Malawi’s first-ever female president, Joyce Banda, we showed that having a
woman president led to women taking up more space in parliamentary debates and being less
confined in their parliamentary roles.
To be sure, our study still leaves many questions about intra-elite symbolic representation
to be answered by future work. Most importantly, further explorations of causal mechanisms
would certainly help in establishing more precise links between female leadership and women’s
empowerment within political institutions. Much of this work is likely to be at the micro-level and
should also consider studying the effect of female leadership on male and female elites respectively
(E.g. SchwindtBayer 2010). More work is particularly needed on the ways in which male elites
may respond negatively to female political empowerment (Krook 2015).
Secondly, future work should also consider similar intra-elite symbolic effects at other
levels of political hierarchies and within institutions other than national parliaments. For instance,
research has pondered the ways in which female chiefs (Bauer 2016) or female justices (Dawuni
and Kang 2015) may inspire female leadership within other political institutions.
Thirdly, looking more specifically at legislative politics, more work is also needed to study
the ways in which formal institutions may mitigate positive intra-elite symbolic effects.
Particularly, one may hypothesize that such effects may be affected by the presence of gender
quotas (Clayton et al. 2014) or be quelled by highly structured parliamentary speech-making
procedures giving significant power to male gatekeepers (Proksch and Slapin 2012).
Fourth, this paper studies the effects of a first-ever female president. The novelty of a first-
ever female president is likely to have the greatest short-term impact on female parliamentary
behavior (Schwindt-Bayer and Reyes-Housholder 2017). Nevertheless, future work from countries
where female executive power is more normalized through repeated and/or long spells of female
executive power may study both short-term effects of having a female executive in office and long-
term effects of accumulated experience of female executive power (see also Beauregard 2018).
All in all, we believe that this study has wide implications for debates on women in politics.
Our findings stress the interconnected nature of political representation at various levels of
government and highlights the need to study these institutions in tandem. Previous research has
questioned the role of women in executive positions in enhancing women’s descriptive
representation in legislatures and cabinets (O’Brien et al. 2015; Chikapa 2016). Our argument is
considerably more optimistic about the role of female executives in enhancing women’s political
representation. Focusing on thick, rather than thin, representation, we show how a female president
may enhance the political role of women by changing the nature of political representation for
women already present within male-dominated political institutions. The concept of intra-elite
symbolic representation was applied here to the highest levels of government, but there is little
reason to believe that such effects are limited to this level. Our results confirm that real political
empowerment diffuses between institutional levels and that a singular focus on increasing
women’s descriptive representation in parliament may not be enough for reaching political gender
equality within contemporary democracies.
Human Subjects
The authors affirm this research did not involve human subjects.
Ethnics and Conflicts of Interest
The authors declare no ethnical issues or conflicts of interest in the research
Data Transparency
Research documentation and data that supports the findings of this study are openly available at
APSR Dataverse at Wahman, Michael; Frantzeskakis, Nikolaos; Tevfik Murat Yildirim, 2021,
“From Thin to Thick Representation: How a Female President Shapes Female Parliamentary
Behavior”,, Harvad Dataverse
Adams, Melinda, and Kristin Wylie. 2020. “Transnational Organizing, the Boosting
Effect, and Women’s Legislative Caucuses in Africa.” Politics, Groups, and Identities, 8(3):
615-626. Women in Politics in Malawi: An Introduction. in Amundsen, Inge and Happy Kayuni
(Eds.) Women in Politics in Malawi. Bergen and Zomba: CMI and PAS.
Ai, Chunrong, and Edward C. Norton. 2003. “Interaction Terms in Logit and Probit
Models.” Economics Letters, 80(1):12329.
Alexander, Amy C. 2012. “Change in women's descriptive representation and the belief
in women's ability to govern: A virtuous cycle”. Politics & Gender, 8(4), 437-464.
Alexander, Amy C., and Farida Jalalzai. 2020. “Symbolic Empowerment and Female
Heads of States and Goverments: A Global Multilevel Analysis”. Politics, Groups, and
Identities, 8(1): 24-43.
Amanatullah, Emily T. and Catherine H. Tinsley. 2013. “Punishing female negotiators
for asserting too much. . .or not enough: Exploring why advocacy moderates backlash against
assertive female negotiators.” Organizational Behavior and Human Decision Processes, 120:
Amundsen, Inge and Happy Kayuni. 2016. “Introduction.” In Amundsen, Inge and
Happy Kayuni (Eds.) Women in Politics in Malawi. Bergen and Zomba: CMI and PAS.
Arriola, Leonardo R., and Martha C. Johnson. 2014. “Ethnic Politics and Women's
Empowerment in Africa: Ministerial Appointments to Executive Cabinets”. American Journal of
Political Science, 58(2), 495-510.
Atkeson, Lonna Rae. 2003. “Not all cues are created equal: The conditional impact of
female candidates on political engagement”. The Journal of Politics, 65(4), 1040-1061.
Banda. Joyce. 2013. “Address by Her Excellency Dr. Joyce Banda”. International
Labour Organization.
Barnes, Tiffany D., and Stephanie M. Burchard. 2013. “’Engendering’ politics: The
impact of descriptive representation on women’s political engagement in Sub-Saharan
Africa”. Comparative Political Studies, 46(7), 767-790.
Barnes, Tiffany D., and Michelle M. Taylor-Robinson. 2017. “Women Cabinet Ministers
in Highly Visible Posts and Empowerment of Women: Are the Two Relates?” in Alexander,
A.C., Bolzendahl, C., & Jalalzai J. (Eds.) Measuring Women’s Empowerment Across the Globe.
Switzerland: Palgrave.
Bauer, Gretchen. 2006. Namibia: Losing Ground Without Mandatory Quotas. in Bauer,
Gretchen and Hannah E. Britton (Eds.) Women in African Parliaments. Boulder: Lynne Rienner
Bauer, Gretchen. 2016. “What is Wrong with a Woman Being Chief? Women Chiefs and
Symbolic and Substantive Representation in Botswana”. Journal of Asian and African Studies,
51(2), 222-237.
Bauer, Gretchen. 2019. “Women in African Parliament” Progress and Prospects.” in
Olajumoke, Yacob-Haliso and Toyin Falola (Eds.) The Palgrave Handbook of African Women’s
Studies, New York: Palgrave.
Bauer, Gretchen, and Hannah Evelyn Britton. 2006. “Women in African Parliaments: A
Continental Shift?” in Bauer, Gretchen and Hannah E. Britton (Eds.) Women in African
Parliaments. Boulder: Lynne Rienner Publishers.
Baumgartner, Frank R., Christoffer Green-Pedersen, and Bryan D. Jones. 2006.
“Comparative studies of policy agendas.” Journal of European Public Policy, 13(7), 959-974.
Beaman, Lori, Raghabendra Chattopadhyay, Esther Duflo, Rohini Bande, and Petia
Topalova. 2009. “Powerful Women: Does Exposure Reduce Bias?” The Quarterly Journal of
Economics, 124(4): 1497-1540.
Beauregard, Katrine. 2018. “Women’s Representation and Gender Gaps in Political
Participation: Do Time and Success Matter in a Cross-national Perspective?” Politics, Groups,
and Identities, 6(2): 237-263.
Beckwith, Karen, and Kimberly Cowell-Meyers. 2007. “Sheer Numbers: Critical
Representation Thresholds and Women’s Political Representation.” Perspectives on Politics,
5(3), 553-565.
Blumenau, Jack. 2019. “The Effects of Female Leadership on Women's Voice in Political
Debate.” British Journal of Political Science, pp.1-22.
Bäck, Hanna, Marc Debus, and Jochen Müller 2014. “Who Takes the Parliamentary
Floor? The Role of Gender in Speech-making in the Swedish Riksdag.” Political Research
Quarterly, 67(3), 504-518.
Bäck, Hanna., and Marc Debus. 2019. “When do Women speak? A Comparative
Analysis of the Role of Gender in Legislative Debates.” Political Studies, 67(3): 576-596.
Brambor, Thomas, William Roberts Clark, and Matt Golder. 2006. “Understanding
Interaction Models: Improving Empirical Analyses.” Political Analysis, 14(1), 6382.
Burnet, Jennie E. 2011. “Women have found respect: Gender quotas, symbolic
representation, and female empowerment in Rwanda.” Politics & Gender, 7(3), 303-334.
Cammack, Diana. 2012. “Malawi in Crisis, 2011-2012". Review of African Political
Economy, 39(132), 375-388.
Childs, Sarah and Lena Krook Mona. 2008. “Critical Mass Theory and Women’s
Political Representation”. Political Studies, 56: 725-736.
Childs, Sarah, and Mona Lena Krook. 2009. “Analysing Women’s Substantive
Representation: From Critical Mass to Critical Actors.” Government and Opposition, 44(2): 125-
Chimunthu Banda, Henry. 2017. Malawi Parliament: Origins, Reforms and Practice.
Lilongwe: Pan-African Publishers.
Chinsinga, Blessings. 2015. The Political Economy Context. in Patel, Nandini, and
Wahman, Michael (Eds.) The Malawi 2014 Tripartite Elections: Is Democracy Maturing?.
National Imitative for Civic Education Press.
Chikapa, Tiyesere Mercy. 2016. “The ‘Joyce Banda Effect’: Explaining the Discrepancy
Between Public Opinion and Voter Behaviour. In Amundsen, Inge and Happy Kayuni (Eds.)
Women in Politics in Malawi. Bergen and Zomba: CMI and PAS.
Chikaipa,Victor. 2019. “Caring Mother or Weak Politician? A Semiotic Analysis of
Editorial Cartoon Representations of President Joyce Banda in Malawi Newspapers.” Critical
Arts, 33(2): 14-28.
Chiweza, Asiyati Lorraine. 2016. “Women in Local Government: A Potential Arena for
Women’s Substantive Representation.” CMI Briefing, 15:13.
Chiweza, Asiyati Lorraine, Vibeke Wang, and Ann Maganga. 2016. “The Women’s
Parliamentary Caucus: Promoting Cross-Party Substantive Representation.” In Amundsen, Inge
and Happy Kayuni (Eds.) Women in Politics in Malawi. Bergen and Zomba: CMI and PAS.
Clayton, Amanda. 2015. “Electoral Gender Quotas, Female Leadership, and Women’s
Political Engagement: Evidence from a Randomized Experiment.” Comparative Political
Studies, 48(3), 333-69.
Clayton, Amanda, Cecilia Josefsson, and Vibeke Wang. 2014. “Present without
presence? Gender, quotas and debate recognition in the Ugandan parliament.” Representation,
50(3), 379-392.
Clayton, Amanda, Cecilia Josefsson, and Vibeke Wang. 2017. “Quotas and women's
substantive representation: Evidence from a content analysis of Ugandan plenary
debates.” Politics & Gender, 13(2), 276-304.
Clayton, Amanda, Cecilia Josefsson, Robert Mattes, and Shaheen Mozaffar. 2019. “In
Whose Interest? Gender and MassElite Priority Congruence in Sub-Saharan
Africa.” Comparative Political Studies, 52(1), 69-101.
Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Jan Teorell,
David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Anna
L¨uhrmann, Kyle L. Marquardt, Kelly McMann, Pamela Paxton, Daniel Pemstein, Brigitte Seim,
Rachel Sigman, Svend-Erik Skaaning, Jeffrey Staton, Steven Wilson, Agnes Cornell, Nazifa
Alizada, Lisa Gastaldi, Haakon Gjerløw, Garry Hindle, Nina Ilchenko, Laura Maxwell, Valeriya
Mechkova, Juraj Medzihorsky, Johannes von R¨omer, Aksel Sundström, Eitan Tzelgov, Yi-ting
Wang, Tore Wig, and Daniel Ziblatt. 2020. ”V-Dem Dataset v10” Varieties of Democracy (V-
Dem) Project.
Dahlerup, Drude. 1988. “From a small to a large minority: women in Scandinavian
politics.” Scandinavian Political Studies, 11(4), 275-298.
Dawuni, Josephine, and Alice Kang. 2015. “Her Ladyship Chief Justice: The Rise of
Female Leasers of the Judiciary in Africa.” Africa Today, 62(2): 45-69.
Dolan, Kathleen, and Lynne E. Ford. 1997. “Change and Continuity Among Women
State Legislators: Evidence From Three Decades.” Political Research Quarterly, 50(1): 137-151.
Dolan, Kathleen. 2006. “Symbolic Mobilization? The Impact of Candidate Sex in
American Elections.” American Political Research, 34(6): 687-704.
Dulani, Boniface, and Joseph Chunga. 2015. When is Incumbency No Longer an
Advantage? In Patel, N., & Wahman, M. (Eds.) Is Democracy Maturing? The Malawi 2014
Tripartite Elections. Lilongwe: The National Initiative for Civic Education.
Erikson, Josefina, and Cecilia Josefsson. 2019. “The legislature as a gendered workplace:
Exploring members of parliament’s experiences of working in the Swedish
parliament”. International Political Science Review, 0192512117735952.
Fernandes, Jorge M., Cristina Leston-Bandeira, and Carsten Schwemmer. 2018.
“Election proximity and representation focus in party-constrained environments”. Party
Politics, 24(6), 674-685.
Ferree, Myra Marx. 2006. “Angela Merkel: What Does it Mean to Run as a Woman?”
German Politics and Society, 78(24), 93-107.
Franceschet, Susan., Mona Lena Krook, and Jennifer M. Piscopo. 2012. The Impact of
Gender Quotas. Oxford: Oxford University Press.
Franceschet, Susan, Jennifer M. Piscopo, and Gwynn Thomas. 2016. “Supermadres,
Maternal Legacies and Women’s Political Participation in Contemporary Latin America.”
Journal of Latin American Studies, 48(1), 1-32.
Gilardi, Fabrizio. 2015. “The temporary importance of role models for women's political
representation.” American Journal of Political Science, 59(4), 957-970.
Hayes, Danny, and Jennifer L. Lawless. 2015. “A Non-Gendered Lens? Media, Voters,
and Female Candidates in Contemporary Congressional Elections.” Perspectives on Politics,
13(1): 95-118.
Hassim, Shireen. 2006. “The Virtuous Circle of Representation: Women in African
Parliaments.” in Bauer, Gretchen and Hannah E. Britton (Eds.) Women in African Parliaments.
Boulder: Lynne Rienner Publishers.
Heilman, Madeline E., and Tyler G. Okimoto. 2007. “Why are women penalized for
success at male tasks? The implied communality deficit.” Journal of applied psychology, 92(1),
Hopkins, Daniel, and Gary King. 2007. “Extracting systematic social science meaning
from text.” Manuscript available at http://gking. harvard. edu/files/words. pdf 20.07.
Jalalzai, Farida. 2008. “Women rule: Shattering the executive glass ceiling.” Politics &
Gender, 4(2), 205-231.
Jalalzai, Farida. 2013. Shattered, Cracked, or Firmly Intact? Women and the Executive
Glass Ceiling World Wide. Oxford: Oxford University of Press.
Karp, Jeffrey A. and Susan A. Banducci. 2008. “When Politics is not Just a Man’s Game:
Women’s Reperesentation and Political Engagement.” Electoral Studies, 27(1): 105-115.
Kayuni, Happy M., and Ragnhild L. Muriaas. 2014. “Alternatives to Gender Quotas:
Electoral Financing of Women Candidates in Malawi.” Representation, 50(3): 393-404.
Krook, Mona Lena. 2010. Quotas for women in politics: Gender and candidate selection
reform worldwide. Oxford University Press.
Krook, Mona Lena. 2015. “Empowerment versus Backlash? Gender Quotas and Critical
Mass Theory”. Politics, Groups and Identities, 3(1), 184-188.
Krook, Mona Lena, and Diana Z. O’Brien. 2012. “All the President’s Men? The
Appointment of Female Cabinet Ministers Worldwide”. The Journal of Politics, 74(3), 840-855.
Lawless, Jennifer. 2015. “Female Candidates and Legislators.” Annual Review of
Political Science, 18: 349-366.
Lechner, Michael. 2010. “The Estimation of Causal Effects by Difference-in-Difference
Methods Estimation of Spatial Panels.” Foundations and Trends® in Econometrics 4(3):165
Liu, Shan-Jan Sarah. 2018. “Are Female Political Leaders Role Models? Lessons from
Asia.” Political Research Quarterly, 71(2), 255-269.
Liu, Shan-Jan Sarah, and Lee Ann Banaszak. 2017. “Do Government Positions Held by
Women Matter? A Cross-National Examination of Female Ministers’ Impacts of Women’s
Political Participation.” Politics & Gender, 13, 132-162.
Lora-Kayambazinthu, Edrinnie, and Edith Kalilombe Shame. 2016. A Different
Yardstick: Gender and Leadership in Political Discourse. In Amundsen, Inge, and Happy
Kayuni. (Eds.) Women in Politics in Malawi. Bergen and Zomba: CMI and PAS.
Lucas, Christopher, Richard A. Nielsen, Margaret E. Roberts, Brandon M. Stewart, Alex
Storer, and Dustin Tingley. 2015. “Computer-assisted text analysis for comparative
politics”. Political Analysis, 23(2), 254-277.
Malawi Parliamentary Hansard. 06/20/2012. Malawi Parliament
Mansbridge, Jane. 1999. “Should Blacks Represent Blacks and Women Represent
Women? A contingent ‘yes’.” Journal of Politics, 61(3), 628-657.
Morgan, Jana, and Melissa Buice. 2013. “Latin American attitudes toward women in
politics: The influence of elite cues, female advancement, and individual
characteristics.” American Political Science Review, 107(4), 644-662.
Muriaas, Ragnhild L., Vibeke Wang, Lindsay Benstead, Boniface Dulani, and Lise
Rakner. 2019. “Why the Gender of Traditional Authorities Matter: Intersectionality and
Women’s Rights Advocacy in Malawi.” Comparative Political Studies, 52(12): 1881-1924.
Nwankwor, Chiedo. 2019. “Women Cabinet Ministers’ Substantive Representation in
Africa.” Social Politics, 0(0): 1-24.
O'Brien, Diana Z., Matthew Mendez, Jordan Carr Peterson, and Jihyun Shin. 2015.
“Letting Down the Ladder or Shutting the Door: Female Prime Ministers, Party Leaders, and
Cabinet Ministers.” Politics & Gender, 11(4), 689-717.
Ott, Martin, and Fidelis Edge Kanyongolo. 2010. Democracy in Progress: Malawi’s
2009 Parliamentary and Presidential Elections. Zomba: Kachere Books.
Patel, Nandini, and Michael Wahman. 2015. “The presidential, parliamentary and local
elections in Malawi, May 2014.” Africa Spectrum, 50(1), 79-92.
Pfafman, Tessa M., and Bree McEwan. 2014. “Polite Women at Work: Negotiating
Professional Identity Through Strategic Assertiveness.” Women’s Studies in Communication,
37(2), 202-219.
Ponton, Douglas. 2010. “The Female Political Leader: A Study of Gender Identity in the
Case of Margaret Thatcher.” Journal of Language and Politics, 9(2), 195-218.
Proksch, Sven‐Oliver, and Jonathan B. Slapin. 2012. “Institutional foundations of
legislative speech.” American Journal of Political Science, 56(3), 520-537.
Reynolds, Andrew. 1999. “Women in the legislatures and executives of the world:
Knocking at the highest glass ceiling.” World Politics, 51(4), 547-572.
Robinson, Amanda, and Jessica Gottlieb. Forthcoming. “How to Close the Gender Gap in
Political Participation: Lessons from Matrilineal Societies in Africa.” British Journal of Political
Science, Pages: TBA.
SchwindtBayer, Leslie A. 2010. Political Power and Women’s Representation in Latin
America. Oxford: Oxford University Press.
SchwindtBayer, Leslie A, and Catherine Reyes-Housholder. 2017. “Citizen Response to
Female Executives: is it Sex, Novelty or Both?” Politics, Groups, and Identity, 5(3): 378-398.
Shalaby, Marwa M., and Laila Elimam. 2020. “Women in Legislative Committees in
Arab Parliaments.” Comparative Politics, 53(1): 139-167.
Simien, Evelyn M. 2015. Historic firsts: How symbolic empowerment changes US
politics. Oxford University Press.
Spiker, Julia. 2019. Empowering Women: Global Voices of Rhetorical Influence. London:
Lexington Books.
Stockemer, Daniel. 2011. “Women's parliamentary representation in Africa: the impact of
democracy and corruption on the number of female deputies in national parliaments.” Political
Studies, 59(3), 693-712.
Stockemer, Daniel, and Maeve Byrne. 2011. “Women's representation around the world:
the importance of women's participation in the workforce.” Parliamentary Affairs, 65(4), 802-
Svåsand, Lars. 2015. Political Parties: Fragmentation and Consolidation, Change and
Stability. In Patel, Nandini, & Wahman, Michael (Eds.) Is Democracy Maturing? The Malawi
2014 Tripartite Elections. Lilongwe: The National Initiative for Civic Education.
Thames, Frank C, and Margaret S. Williams. 2013. Contagious Representation: Women’s
Political Representation in Democracies Around the World. New York: New York University
Thomas, Sue. 1994. How Women Legislate. Oxford: Oxford University Press.
Thomas, Gwynn, and Melinda Adams. 2010. “Breaking the Final Glass Ceiling: The
Influence of Gender in the Elections of Ellen Johnson-Sirleaf and Michelle Bachelet.” Journal of
Women, Politics, ad Policy, 31(2): 105-131.
Tripp, Aili Mari, and Alice Kang. 2008. “The global impact of quotas: On the fast track
to increased female legislative representation.” Comparative Political Studies, 41(3), 338-361.
Tripp, Aili Mari. 2015. Women and power in post-conflict Africa. Cambridge University
Tøraasen, Marianne. 2019. “Gender Parity and the Symbolic Representation of Women
in Senegal.” The Journal of Modern African Studies, 57(3), 459-481.
Verge, Tània, and Raquel Pastor. 2018. “Women’s Political Firsts and Symbolic
Representation.” Journal of Women, Politics & Policy, 39(1), 26-50.
VonDoepp, Peter. 2005. “Party Cohesion and Fractionalization in New African
Democracies: Lessons from Struggles over Third-Term Amendments”. Studies in Comparative
International Development, 40(3), 65-87.
Wang, Vibeke. 2014. “Tracing Gender Differences in Parliamentary Debates: A Growth
Curve Analysis of Ugandan MPs’ Activity Levels in Parliamentary Sessions, 1998-2008.”
Representation, 50(3), 365-377.
Weldon, S. Laurel. 2002. “Beyond bodies: Institutional sources of representation for
women in democratic policymaking.” The Journal of Politics, 64(4), pp.1153-1174.
Wolak, Jennifer. 2020. “Descriptive Representation and the Political Engagement of
Women.” Politics & Gender, 16(2): 339-362.
Wolbrecht, Christina, and David E. Campbell. 2007. “Leading by example: Female
members of parliament as political role models.” American Journal of Political Science, 51(4),
Wängnerud, Lena. 2009. “Women in parliaments: Descriptive and substantive
representation.” Annual Review of Political Science, 12, 51-69.
Yoon, Mi Yung. 2011. “More Women in the Tanzanian Legislature: Do Numbers
Matter?” Journal of Contemporary African Studies, 29(1): 83-98.
Young, Daniel. 2014. “An Initial Look into Party Switching in Africa: Evidence from
Malawi.” Party Politics, 20(1): 105-115.
Zetterberg, Pär. 2009. “Do Gender Quota’s Foster Women’s Political Engagement?
Lessons from Latin America.” Political Research Quarterly, 62(4): 715-730.
Zimmerman, Brigitte. 2015. Voter Response to Scandal: Cashgate. In Patel, Nandini, and
Wahman, Michael. (Eds.) Is Democracy Maturing? The Malawi 2014 Tripartite Elections.
Lilongwe: The National Initiative for Civic Education.
Affective polarization is pervasive in modern US politics, and can be intensified by strategic messaging from members of Congress. But there are gaps in our knowledge of the dynamics of polarizing appeals from elected representatives on social media. We explore the usage of polarizing rhetoric by members of Congress on Twitter using the 4.9 million tweets sent by members of Congress from 2009 to 2020, coded for the presence of polarizing rhetoric via a novel and highly accurate application of supervised machine learning methods. Fitting with our expectations, we find that more ideologically extreme members, those from safer districts, and those who are not in the president’s party are more likely to send polarizing tweets, and that polarizing tweets garner more engagement, increasing campaign funding for more polarizing members.
Full-text available
Much of the vast literature on the substantive representation of women takes as its point of departure important a priori assumptions about the nature of women as a group. Calling for rethinking of many of those assumptions, a recent body of work recommends an inductive approach to defining women's interests. In line with this view, I draw on a recently constructed dataset that codes nearly a million Americans' policy priorities over the past 75 years to explore what constitutes women's interests and whether gender differences in priorities cut across partisan and racial divisions. Results suggest consistent gender gaps across a large number of policy categories, with women showing particular concern for policy areas traditionally associated with 'women's interests'. While in many policy areas women were more likely to share policy priorities with other women than with their male counterparts of the same race or partisan background, results also document considerable heterogeneity among women in various policy areas, which has major policy implications that speak to the representation of women's interests.
Full-text available
Following the adoption of the Senegalese Law on Parity and the subsequent influx of women to the National Assembly, we saw a strong gendered polarisation of attitudes towards women as political leaders among the Senegalese population. This study explores whether similar attitudinal changes are found among political elites in the Senegalese National Assembly. Theory suggests that an increase in the number of women elected to legislatures (‘descriptive representation’) will challenge people's perceptions of politics as a male domain and strengthen belief in women's ability to govern (‘symbolic representation’). Although the effects of gender quotas on women's representation has received considerable scholarly attention, the field of symbolic representation remains under-studied. A case study of the effects of the Senegalese parity law addresses this knowledge gap, contributing with new empirical insights. This study also develops indicators that can help measure potential developments in the symbolic representation of women. Parity appears to have contributed to slightly more acceptance towards women as political leaders within the National Assembly. The findings are discussed in the last section.
Full-text available
This article analyses media representations of Joyce Banda, Malawi's first female president, by newspaper editorial cartoons. A total of 50 cartoons were collected, comprising Point of Order and Puludzu, from The Nation and The Daily Times newspapers, spanning the period 2012 to 2014. Using theories of feminism and representation, this paper reveals how the editorials, as unique visual forms of communication, offer an opportunity for deep reflection on attitudes concerning female, presidential leadership. The findings suggest that both cartoon series depict President Banda as an indecisive, inefficient, clueless and weak politician based on ingrained gender stereotypes of women, rather than on her actions as a president governing the country. This study regards cartoonists as social critics who have brandished Banda's motherhood characteristics at every occasion to represent female presidency as a failure in Malawi. It is shown that the graphic artists were largely absorbed in the ideologies that trivialise women as being only suitable for domestic roles. Their relentless scrutiny of "gender" eccentricities, infirmities and frailties regarding Banda effectively destroyed her power consciousness and fuelled misogynistic perceptions that, by extension, no woman can ever be suitable for electoral leadership positions, especially not as Malawi President.
Research on women executives’ substantive representation is currently drawing serious attention given women’s increasing inclusion in cabinets globally. However, these studies rarely question whether women ministers across Africa represent women’s interests. Using primary semi-structured interview data conducted with forty-nine interviewees, including seventeen ministers in Nigeria and South Africa, this article explores whether women politicians advocate for women’s interests. It centers a critical understanding of the cabinet as an embedded institution, embraces a context-specific understanding of substantive representation, and demonstrates that women ministers’ representation of women’s interests occurs mostly in nonlegislative ways. It presents an expansive three-level conception of substantive representation.
Extant studies have predominantly focused on women’s numerical presence in the Middle East and North Africa (MENA)’s legislatures, yet, research examining the role played by female politicians continues to be limited. To bridge this gap, we study one of the most important, albeit overlooked, bodies within these assemblies: legislative committees. Using an original dataset on committee memberships (n=4580), our data show that females are significantly marginalized from influential committees and tend to be sidelined to social issues and women’s committees. To explain this, we develop a theory of provisional gender stereotyping. We argue that the duration of quota implementation shapes women’s access to influential committees. We focus on two mechanisms to support our argument: a redistribution of power dynamics within legislative bodies and women’s political expertise.
The share of women holding legislative office in Africa more than doubled between 1998 and 2019, reaching almost 24%. Despite this increase in women's descriptive representation, women members of parliaments (MPs) continue to encounter exclusionary masculinized environments. One way to facilitate women’s work in these male-dominated parliaments is through the creation of Women’s Legislative Caucuses (WLCs). This article explores the diffusion of WLCs across Africa and examines the factors that influence caucus functioning. The establishment of WLCs in Africa began in 1993 with the creation of a women’s caucus in Ghana. Caucus formation sped up in the 2000s, peaking in 2007. This trend has touched all regions of Africa, with 38 countries having established WLCs as of 2015. We identify critical actors at international, regional, and national levels involved in WLC establishment and examine how their collaboration generates a boosting effect that amplifies the voices and gender equity work of women MPs and feminist civil society. We argue that transnational collaboration can boost domestic advocacy but only when it leverages global and regional resources in ways that bring domestically situated civil society organizations into the halls of formal political power.
Do female leaders amplify the voices of other women in politics? The author addresses this question by examining parliamentary debates in the UK House of Commons. In the context of a difference-in-differences design that exploits over-time variation in the gender of cabinet ministers, the article demonstrates that female ministers substantially increase the participation of other female MPs in relevant debates, compared to when the minister is male. It also uses a measure of debate influence, based on the degree to which words used by one legislator are adopted by other members, to show that female ministers also increase the influence of female backbenchers. To explore the mechanisms behind these results, the author introduces a new metric of ministerial responsiveness and shows that female ministers are significantly more responsive than their male counterparts to the speeches of female backbenchers.
While gender gaps in political participation are pervasive, especially in developing countries, this study provides systematic evidence of one cultural practice that closes this gap. Using data from across Africa, this article shows that matrilineality – tracing kinship through the female line – is robustly associated with closing the gender gap in political participation. It then uses this practice as a lens through which to draw more general inferences. Exploiting quantitative and qualitative data from Malawi, the authors demonstrate that matrilineality's success in improving outcomes for women lies in its ability to sustain more progressive norms about the role of women in society. It sets individual expectations about the gendered beliefs and behaviors of other households in the community, and in a predictable way through the intergenerational transmission of the practice. The study tests and finds evidence against two competing explanations: that matrilineality works through its conferral of material resources alone, or by increasing education for girls.
When women are represented on the campaign trail and in elected office, women in the electorate have been shown to report greater engagement in politics. However, most evidence of the effects of descriptive representation on women's empowerment is drawn from surveys from the 1980s and 1990s. I update these studies to consider how women candidates and officeholders affect the political knowledge, interest, and participation of other women in the electorate. Using responses from the Cooperative Congressional Election Study from 2006 to 2014, I find that both men and women are more politically knowledgeable when represented by women in Congress and in state government. Considering political engagement, I find little evidence that women are more politically interested or participatory when residing in places with more female officeholders or candidates. Women's political presence as candidates and officeholders does not uniquely encourage other women to engage in political life.