Distinctive Voices: Political Speech, Rhetoric, and the Substantive Representation of Women in European Parliaments

• Published date: 01 November 2023
• Author: Jens Wäckerle,Bruno Castanho Silva
• Date: 01 November 2023
• DOI: http://doi.org/10.1111/lsq.12410

797

LEGISLATIVE STUDIES QUARTERLY, 48, 4, November 2023

DOI: 10.1111/lsq.12410

JENS WÄCKERLE

University of Cologne

BRUNO CASTANHO SILVA

University of Cologne

Distinctive Voices: Political Speech,

Rhetoric, and the Substantive

Representation of Women in

European Parliaments

As the share of women in parliaments rises, increased attention is paid

to how they substantively represent women. Meanwhile, the availability of

parliamentary speech data has enabled researchers to dissect politicians’ rhe-

torical patterns. We combine these two literatures to ask whether rhetorical

differences between men and women in parliament are connected to style,

policy, and preferences of women voters. We apply machine- learning models

to speeches from five West European parliaments (2000– 18) to measure the

femininity of the rhetoric used in each speech. Results show that women and

men talk differently in parliament, and that this distinctiveness is due to both

style and substance. Combining these results with public opinion surveys, we

find that women MPs have the most distinctively “feminine” discourse on is-

sues that are most salient to women in society. These findings showcase the di-

rect connection between descriptive and substantive representation of women

in contemporary democracies.

Europe has seen a general but uneven increase in the share

of women parliamentarians in recent decades. In some countries,

such as Sweden, more than 40% of members of parliament (MPs)

are women while others, like Ireland, rarely have women making

up more than 15% of the legislature. Equality in gender repre-

sentation is a desirable goal in politics as a matter of justice in

itself, but also because we know that, among the public, women

have systematically different preferences and behavior than men

© 2023 The Authors. Legislative Studies Quarterly published by Wiley

Periodicals LLC on behalf of Washington University in St. Louis.

This is an open access article under the terms of the Creative Commons

Attribution License, which permits use, distribution and reproduction in

any medium, provided the original work is properly cited.

798 Jens Wäckerle and Bruno Castanho Silva

(e.g., Chattopadhyay and Duflo 2004; Gerrity, Osborn, and

Mendez 2007; Homola 2019). The presence of fewer women in

politics than their share in the population means that their voices

and interests are underrepresented, marking an essential flaw in

representative democratic systems.

It is necessary to carefully study how having more women in

parliaments (descriptive representation) translates into more dis-

tinctively feminine voices in the political debate, who will also act

for women (substantive representation). As Jane Mansbridge and

others have argued, the presence of women has an important but far

from straightforward impact on how well women are substantively

represented in politics (Mansbridge1999, 2003; Phillips1995). We

argue that in order to get a comprehensive picture of how women

act for women in parliament, we need to look at parliamentary

rhetoric as an expression of political action, focusing thus on how

women’s priorities are represented in parliament. This allows us to

get a step closer to understanding how substantive representation

works in practice.

Research on political rhetoric has shown that men and

women speak differently once elected (Childs2004; Dietrich,

Hayes, and O’Brien 2019; Hargrave and Langengen 2020;

Kathlene1994). While these differences tend to diminish over

time (Hargrave and Blumenau2022), they suggest that politi-

cal debate is changed by the presence of women. In this article,

we ask: is this difference connected to the substantive repre-

sentation of women? Recently, several studies have looked at

whether the share of women in parliament affects policy out-

comes towards assumed women’s preferences (e.g., Clayton and

Zetterberg 2018; Funk and Philips 2019; Mavisakalyan and

Tarverdi2019; Svaleryd 2009; Weeks and Masala 2022), but

we argue that even besides changing policy outcomes, women

speaking distinctively for women in parliament entails substan-

tive representation.

More specifically, we analyze the gendered nature of parlia-

mentary debate using advanced techniques of natural language

processing, applied to a corpus of speeches from five national

parliaments (Germany, Ireland, Netherlands, Spain, and Sweden),

over almost two decades. We adapt a recent insight by Peterson

and Spirling(2018) and use machine- learning models on parlia-

mentary speeches to predict speakers’ gender: the easier it is for a

model to tell whether a speech is given by a man or a woman, the

more distinctively women are speaking in relation to men. We then

799Distinctive Voices

proceed to investigate the sources of that distinctiveness: is it be-

cause women have a different style from men (how they speak), be-

cause women speak about different issues (what they talk about),

or both? Additionally, we use survey data on the national level to

establish which policy areas were important to women at the time

of speeches to see whether women MPs’ rhetoric is connected to

the salience of issues for women in society.

Our contribution is two- fold. First, we connect the study of

rhetoric to policy preferences of women. We not only show that

women speak differently from men, but that they do so particu-

larly when addressing policy areas in which, according to public

opinion surveys, women care the most in relation to men. This al-

lows us to establish a direct connection from women’s priorities to

women MPs’ parliamentary rhetoric. Second, we use an innovative

method, machine learning, to do so. This allows us to approach

the issue without a set preconception of how women’s speech

should be different from men’s. This method can be extended in

the future to study how gendered discourse by MPs is connected

to various other political phenomena. Taken together, these find-

ings provide strong evidence that women’s presence in parliament

is directly linked to substantive representation of women and that

the distinctive way in which women communicate in parliament

reflects societal differences with potential policymaking impacts.

Women and Political Rhetoric

What women MPs express in their speeches, and how, is

something that can mark the difference to words and actions by

men in parliament. The question then is, do women parliamen-

tarians speak differently from men? In a highly formalized setting

such as parliamentary politics, most other social group symbols

are somewhat restricted: conduct and dress codes are standard-

ized with little space for individual expression. With the prevalence

of speech as a centerpiece of political action for individuals to

disclose themselves to others (Arendt1998), parliamentary per-

formance through language is the primary way in which women

politicians could stand out from men and represent women voters

through their actions.

Analyzing the rhetoric of women and men in parliament has fo-

cused on two broad categories. First, analyses of grammar and style

focus on women having a more “conciliatory” or “relational” style,

often attributed to and expected from them (Coates 2015; Park et