Quotas, women's leadership, and grassroots women activists: Bringing women into the party?

Abstract

Research has consistently shown that women are less likely than men to participate in political parties as members and activists; this participation gender gap has persisted despite narrowing gender gaps in education, employment, and in other types of political participation. Yet while the gaps are widespread, their size varies greatly by country as well as by party. To what extent do party organizational factors help explain these disparities? More pointedly, are there any lessons to be learned from past experiences about party mechanisms which might help to reduce these gaps? To answer these questions, this study investigates grassroots partisan participation in 68 parties in 12 parliamentary democracies, considering whether factors that have been shown to boost the number of women candidates and legislators are also associated with changing the traditionally male dominance of grassroots party politics. We find evidence of links between some party mechanisms and higher women's intra‐party participation; however, because the same relationship holds for men's participation, they do not alter the participation gender gap. Only greater participation of women in parties’ parliamentary delegations is associated with smaller grassroots gender gaps. We conclude that parties which wish to close grassroots gender gaps should not rely solely on efforts aimed at remedying gender gaps at the elite level. This article is protected by copyright. All rights reserved

Full text

European Journal of Political Research rr
:rr
–rr
, 2020 1
doi: 10.1111/1475-6765.12382
Quotas, women’s leadership and grassroots women activists: Bringing women
into the party?
ALDO F. PONCE,1SUSAN E. SCARROW2& SUSAN ACHURY3
1Department of Political Studies, Centro de Investigación y Docencia Económicas, Mexico City,Mexico;
2Department of Political Science, University of Houston, Houston, Texas, USA; 3Department of Political
Science, Miami University, Oxford, Ohio, USA
Abstract. Research has consistently shown that women are less likely than men to participate in political
parties as members and activists; this participation gender gap has persisted despite narrowing gender gaps
in education, employment and in other types of political participation. Yet while the gaps are widespread,
their size varies greatly by country as well as by party. To what extent do party organizational factors
help explain these disparities? More pointedly, are there any lessons to be learned from past experiences
about party mechanisms which might help to reduce these gaps? To answer these questions, this study
investigates grassroots partisan participation in 68 parties in 12 parliamentary democracies, considering
whether factors that have been shown to boost the number of women candidates and legislators are also
associated with changing the traditionally male dominance of grassroots party politics. We nd evidence of
links between some party mechanisms and higher women’s intra-party participation; however, because the
same relationship holds for men’s participation, they do not alter the participation gender gap. Only greater
participation of women in parties’ parliamentary delegations is associated with smaller grassroots gender
gaps. We conclude that parties which wish to close grassroots gender gaps should not rely solely on efforts
aimed at remedying gender gaps at the elite level.
Keywords: party membership; political participation; political parties; women and politics
Introduction
Women have been persistently and severely under-represented within the memberships of
political parties in established democracies.One study of 15 Western European democracies
in the late 1980s concluded that in party memberships, ‘male predominance, often of
overwhelming magnitude, is the rule’ (Widfeldt 1995: 147). More recently, surveys of party
members in 10 parliamentary democracies have conrmed the persistence of this gender
imbalance (van Haute & Gauja 2015: 194). This participation gender gap continues despite
narrowing gender gaps in education and employment and in other types of political
participation.
These persisting disparities within their memberships potentially undermine parties’
abilities to represent and respond to the interests of their female supporters (Widfeldt
1995). For instance, half the population may be conspicuously under-represented in the
many parties which now ask their members to directly select candidates, choose leaders or
decide on policies (Cross & Pilet 20 15; Sandri et al. 20 15). Moreover, if women are reluctant
to participate in parties’ lower level activities, parties may have difculties identifying
appropriate female candidates,especially those parties which have traditionally chosen their
C2020 European Consortium for Political Research

2ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
candidates from existing membership pools. Thus, the under-representation of women in
grassroots party politics poses a challenge for party representation in both descriptive and
substantive terms.Despite this, and even though scholars and policy makers have given much
attention to the under-representation of women as candidates and legislators, few studies
focus on the causes of – or remedies for – parties’ grassroots gender gaps.
Such investigation is warranted, not least because parties differ signicantly in the
magnitude of this effect. For instance, women members outnumbered men in 3 out of the
37 parties in the 1980s study cited above,and there were within-country differences in party
gaps. Thus, gender disparities within political party memberships are by no means uniform
or inevitable. Moreover, unlike efforts to boost women’s share of public and party ofces,
there is no inevitable gender trade-off when it comes to grassroots participation levels,
because participation opportunities are not zero-sum. This leads us to ask: Are there party
organizational features which seemingly boost women’s intra-party participation, and, why
are some parties more successful than others in closing the grassroots gender gap?
This study seeks to answer these questions by examining partisan participation patterns
in 12 parliamentary democracies using data from the European Social Survey (2010). In
this survey, women were much less likely than men to self-report party membership or
party activism.1Yet, as we will show, these differences are not uniformly distributed across
countries or parties.
In investigating the sources of these differences, we specically consider the impact of
factors associated with higher women’s participation at parties’ elite levels, including rules
which guarantee leadership roles for women. We want to know whether factors which
promote female participation at the elite level have a trickle-down effect at the grassroots.
As we report below, we nd that factors that promote women’s leadership are indeed
associated with higher women’s grassroots party participation, but most have little effect
in closing the gender gap. In particular, women’s seat share in a party’s national legislative
delegation fosters women’s participation at the grassroots. Thus, if parties want to close
gender gaps in grassroots participation, they may need to do more than just improve the
opportunities for women to obtain political careers: instead,they need to tackle the problem
head on.
Explaining the gender gap in intra-party participation: Competing explanations
Previous research has found signicant and persistent gender differences in propensities to
participate in political activities (Campbell et al. 1964; Norris 2002; Schlozman et al. 1995;
Verba & Nie 1972; Verba et al. 1997; Welch 1977). These differences are evident in both
low-intensity and high-intensity activities, including party membership (Seyd & Whiteley
2004). They appear in countries at various levels of economic and educational development
(Burns, Schlozman, & Verba 2001). Some recent studies have found reductions in the
political participation gender gap (Barnes & Burchard 2013; Desposato & Norrander 2009;
Kittilson & Schwindt-Bayer 20 12; Liu & Banaszak 2017), or have demonstrated that women
over-participate in certain (more individualistic) political activities such as voting, signing a
petition, or raising funds (Coffé & Bolzendahl 2010). Nevertheless, in terms of traditional
modes of political participation, including intra-party activities, this gap is remarkably
durable and widespread.Even so,such individual-level tendencies do not necessarily explain
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 3
the large cross-national and within-country differences in the size of participatory gender
gaps.
Studies of political participation suggest plausible sources of within-country differences,
stemming from both supply-side and demand-side factors (Scarrow 2014; Gauja 2015). On
the supply side are individual resource disparities affecting the likelihood that political
supporters will join. Differences in parties’ gender gaps might stem from systematic resource
differences among their supporters. For instance, trade union membership might be a strong
stimulus for membership in specic parties; if trade union membership is predominantly
male, this would potentially magnify gender disparities in these parties compared with
parties that lack such trade union links. On the demand-side – the extent to which parties
welcome members –participation disparities might be affected by differences in party
ideology and party rules. In trying to explain parties’ differential success in mobilizing
women’s political participation we take account of supply-side (individual-level) factors.
However, we focus on demand side variables, particularly rules of internal organization,
trying to discern the impact of variables over which parties have a great deal of control.
Our investigation considers two types of outcomes: (1) levels of women’s intra-party
participation; and 2) the gender gap in intra-party participation. We separate these effects
because there is not a xed minimum or maximum number of supporters who can be
members or activists (unlike with slots for party candidates or legislators). Thus, there is
no necessary relationship between the share of women and men party supporters who
chose to be active within a party. Participatory gender gaps could equally well disappear
because women become more engaged or because men become less engaged (Wolak 2015).
Similarly, women’s grassroots party participation levels could increase without affecting the
participation gender gap, if men and women respond identically to increased participation
incentives. If there are complementary effects between men’s or women’s participation,
these processes could increase women’s grassroots participation without reducing the
gender gap. For instance, new women members might encourage their male friends and
relatives to also join. This effect could be even greater in those parties which make it easy to
join. But, if both types of growth in participation – women’s and men’s participation – are
unrelated or orthogonal to each other, an increase in women’s participation would reduce
the gender gap.
Gendered candidate selection rules, particularly quotas, can powerfully affect levels
of female partisan participation at the elite level, including boosting the number of
women elected to public ofce (Schwindt-Bayer 2009). These effects have been widely
demonstrated, though they vary across electoral systems, and are also conditional to the
structure of candidate selection rules (Tremblay 2008; Valdini 2012). They could similarly
affect grassroots participation. On the supply side, quotas could spur women to join and
become active in these parties as means for personal political advancement or to support
women candidates.On the demand side,parties with quotas for female candidates may work
harder to recruit female party members. As Davidson-Schmich (2016) shows for Germany,
when quotas extend to municipal and party leadership positions, parties may need large
membership pools just to nd enough female supporters who are willing to be nominated
to all available posts (36).
Whether quotas actually have any of these wider effects is still unresolved. Some cross-
national studies nd that candidate gender quotas stimulate women’s political participation
C2020 European Consortium for Political Research

4ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
at lower organizational levels, including voter turnout (de Paola et al. 2014). However, other
studies nd little or no such effect (Beauregard 2017; Clayton 2015; Davidson-Schmich
2016; Karp & Banducci 2008; Zetterberg 2009). Moreover, even if quota rules do stimulate
women’s grassroots participation, their effects on the gender gap depend on whether they
have a similar impact on men’s and women’s participation. To disentangle these relations,
we thus offer two hypotheses to assess how party-level gender quotas in candidate selection
affect women’s grassroots participation.
Hypothesis 1A: Party regulations promoting women candidates boost women’s intra-
party participation.
Hypothesis 1B: Party regulations promoting women candidates narrow gender gaps
in intra-party participation.
Partisan participation by women might also be affected by differences in women’s status
within party organizations.Variations include whether parties provide sub-organizations for
women partisans and the extent to which party rules facilitate women’s access to leadership
positions. First, many parties promote – or at least differentiate – women’s participation
through separate party sub-organizations. Such bodies could boost women’s partisan activity
by providing additional outlets for intra-party engagement and because their recruiting
targets women. As women gain experience and assume roles of leadership within the
sub-organization, these institutional spaces might also promote women’s ascension within
party organizations,thus making their participation more attractive.There is little evidence
concerning the impact of women’s sub-organizations in these regards. One cross-national
study found no evidence that party women’s organizations boost the number of women
in party leadership positions (Childs & Kittilson 2016: 604), but this did not examine
their impact on women’s grassroots partisan participation. However, given that studies of
parties’ youth organizations show that party sub-organizations can serve as fertile ground
for recruiting and training future party activists (Hooghe et al. 2004), it is conceivable that
women’s organizations might have a similar effect.
Second, rules boosting women’s participation in leadership bodies might encourage
more women to join and become active in the party by signalling that the party offers
women access to inuential positions. These rules might encourage women members to
become more active within party organizations to further their policy interests and personal
aspirations. By the same logic, such signals could potentially dampen men’s interest in
participation, if they feel that such rules reduce men’s chances of attaining leadership
positions.Either or both mechanisms could reduce the gender gap.These arguments lead to
the following hypotheses:
Hypothesis 2A: Sub-organizations and/or rules promoting women’s intra-party
representation boost women’s intra-party participation.
Hypothesis 2B: Sub-organizations and/or rules promoting women’s intra-party
representation narrow gender gaps in intra-party participation.
Women’s intra-party status is not only a matter of formal structures: it is also signalled
by the extent to which women hold top party positions and public ofces. This has been
found to positively affect the attitudes and behaviour of female citizens, including levels of
political knowledge and interest (Hernson et al. 2003; Huddy & Terkildsen 1993; Kittilson
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 5
2006, 20 08; Leeper 1991; Reingold & Harrell 2010; Schwindt-Bayer 2010; Smith & Fox 20 01;
Verba et al. 1997; Witt et al. 1994). More specically, some studies have found that greater
representation of women in legislatures is associated with higher women’s political activity
and a reduced participation gender gap (Barnes & Burchard 2013; Kittilson & Schwindt-
Bayer 2012; Liu & Banaszak 2017;Wolbrecht & Campbell 2007). Such effects could be due to
women legislators paying more attention to concerns that match their own (i.e.,they provide
greater substantive representation) (Schwindt-Bayer 2009). They might also reect a role
model effect changing the supply-side calculus,as women nd it easier to imagine their own
partisan involvement. This also could be due to a demand-side shift, with women in leading
positions boosting women’s grassroots participation by recruiting women into the party to
help with their re-selection or favouring women’s appointment to intra-party posts.
On the other hand, other studies have failed to nd such effects when controlling for
other factors (Beauregard 2018; Karp & Banducci 2008) or have found that they depend
on the institutional setting (Kunovich & Paxton 2005). In addition to producing somewhat
inconsistent results, these cross-national studies rarely examine the effects of cross-party
differences in leadership patterns. They thus leave open questions about whether and
when women’s success at elite levels of politics has spill over effects on grassroots partisan
engagement. Given these debates, we propose the following hypotheses:
Hypothesis 3A: Higher proportions of women leaders boost women’s intra-party
participation.
Hypothesis 3B: Higher proportions of women leaders narrow gender gaps in intra-
party participation.
Data and methods
To test our hypotheses, we employ two data sources, the 2010 European Social Survey
(ESS Round 5) and the Political Party Database Round 1 (PPDB) (Poguntke et al. 2016).
First, we employ survey data from the 2010 ESS, which is the latest edition of this survey
incorporating questions on party membership and activism.2These questions allow us to
differentiate between two distinct groups of active partisans: all those who self-identify
as party members, and the sub-set of self-identied members who are active within their
party. We are interested in both groups because some of the mechanisms outlined above
involve a shift in demand for active party members (e.g.,future candidates), not just members
per se. Moreover, while overall a large portion of party members are not active (Ponce
& Scarrow 2016), the gender gap is generally larger for active members as compared to
passive members. Thus, there is reason to suspect that some different factors may be at
play. If changes in quotas or leadership primarily affect women who join because they
see new opportunities to inuence policies and have careers, and/or because party elites
are recruiting them to campaign or run for ofce, this could affect the gender gap among
active members without necessarily affecting relations among passive members. Testing our
models with both groups allows us to pick up on such potential differences in impact.
Our main party-level variables describe conditions in parties from 2010 to 2014,and thus
are roughly contemporaneous with the ESS individual-level data.3We also include 2010 data
on the composition of parliamentary delegations and party leadership.Our universe consists
C2020 European Consortium for Political Research

6ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
of those countries and parties for which we have overlapping data from ESS and PPDB (68
parties from 12 countries).4Our analyses ask how characteristics of specic parties affect
the likelihood that their supporters will take an active role in partisan politics. Because we
need to match each individual with a preferred party, we include only respondents who
name a PPDB party which they support, feel close to or prefer.5We look at expressed
partisanship (where available) or at expressed vote choice for those individuals who do not
state a partisan afliation. Even if voters do not feel close to the party, they are willing to do
something for the party (namely, vote for it and/or say that they did). We use these criteria
to match individuals to the characteristics of the specic party that they would seem most
likely to join if they joined any party. This strategy allows us to maximize the number of
observations.
To test our hypotheses,we use a multilevel logistic model.Since we aim to study the effect
of certain determinants on the likelihood of observing a particular condition (whether or not
the citizen is a member, and whether or not the citizen is a member and also works for the
party), we employ logistic regressions.The model includes two levels of analysis:respondents
and parties. Such a model takes account of the possible lack of statistical independence
across observations, especially among observations within contextual units (in this case,
among respondents supporting particular parties). Institutional or cultural contexts might
vary across parties due to unobserved characteristics, such as party life or socialization,
and these variations could invalidate the assumption of statistical independence across
observations (Raudenbusch & Bryk 2002).6To account for these variations across nested
units,the variance is divided into a between-party component (the variance of the party-level
residuals) and a within-party component (the variance of the respondents-level residuals).
Failure to cluster this type of data may result in biased (underestimated) standard errors,
and, consequently, errors in the estimation of our inference analysis (Steenbergen & Jones
2002).7
To test our hypotheses regarding rules about gendered candidate selection, we use
two independent variables. The rst indicates whether a party has voluntary quotas.8The
second indicates whether a party has legally mandated quotas. We consider voluntary and
mandatory quotas separately because voluntary quotas might be a stronger signal of a
particular party’s determination to favour women’s political careers. Both variables are
constructed using data from the PPDB. To evaluate Hypothesis 2, we consider two PPDB
variables concerning how party statutes construct and promote women’s intra-party roles.
The rst indicates whether the party has a sub-organization for women.The second indicates
whether party statutes require representation of women on the party’s highest executive
body. To evaluate Hypothesis 3, we consider women’s seat share in each parties’ delegation
in the lower house of the national legislature prior to the surveyed ESS data.9Table 1
summarizes the dependent variables and the key independent variables employed in our
empirical analyses.
While our hypotheses focus on the effects of the party-level (demand-side) determinants
of participation, our analysis also controls for supply-side effects determined by individual
differences. Most studies of political participation nd that individual resources and
characteristics shape participation within parties (Ponce & Scarrow 2016). We include this
control because gender differences in the distribution of these resources could account for
lower rates of women’s party membership.Our models thus include the degree of education
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 7
Tabl e 1. Summary of the dependent and key independent variables by hypothesis
Dependent variables
and hypotheses Description of the Variables
Dependent variable 1 Members
Dependent variable 2 Active members: Members who work for the party.
Hypotheses 1A and 1B Voluntary quota: Party statutes or rulebook specify representation of
gender in selecting national legislative candidates even though this is not
required by law.
Mandatory quota: Party operates in country which requires parties to take
account of gender in selecting national legislative candidates.
Hypotheses 2A and 2B Women’s sub-organizations: Party has women’s sub-organizations with
individual memberships.
Executive seat rules: Party statutes have rules about representation of
women on the party’s highest executive body.
Hypotheses 3A and 3B Legislative share: Women’s seat-share in party’s delegation in the lower
house of the national legislature.
and income as individual-level control variables, as well as four other variables associated in
past studies with higher rates of party membership: degree of religious attendance,age,trade
union membership and ideological extremeness10 of the respondent (Coffé & Bolzendahl
2010; Dalton 1996; Marien et al. 2010; Verba et al. 1978; Whiteley 2011).
As a party-level control variable, we include a binary variable to indicate whether
the party is a leftist one.11 In parliamentary democracies, both old and new left parties
have been more likely to adopt gender quotas for candidates (Caul 2001), and they have
traditionally had higher levels of female legislators at both national and local levels (Caul
1999; Kenworthy & Malami 1999; Reynolds 1999; Sundström & Stockemer 2015). Although
we test for quotas and legislative seat share, we also want to account for parties’ ideological
traits that might encourage women’s participation in general, something that could affect
both supply and demand-side incentives for women’s participation. At the country-level,
we include a measure of income per inhabitant to account for cross-national differences in
citizens’ average economic resources.
Analysis 1: Boosting women’s participation?
We start by testing our ‘A’ hypotheses related to the impact of party rules and party
leadership on levels of women’s party participation, using a sample that includes only
women. This allows us to discern which, if any, of our independent variables is related to
levels of women’s participation. Table 2 displays our results for our ‘A’ hypotheses. Our
models assess the three hypotheses together.
Our models displayed in Table 2 show no evidence for voluntary or mandatory party
quotas as promoters of women’s intra-party participation, as we stated in Hypothesis 1A.
Quotas do not boost women’s membership,at least not directly (applying a 0.05 signicance
threshold throughout this discussion).
C2020 European Consortium for Political Research

8ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
Tabl e 2. Multilevel logistic regression: explaining women’s grassroots party participation
Party members Party members who work for the party
Model 1 Model 2
Hypothesis 1A
Voluntary quota .18 .31
(.21) (.26)
Mandatory quota .45*.49
(.24) (.32)
Hypothesis 2A
Women’s sub-organizations .81*** .33
(.19) (.24)
Executive seat rules −.30 .05
(.19) (.24)
Hypothesis 3A
Legislative share .01** .02**
(.01) (.01)
Control variables
Education .02*.08***
(.01) (.02)
Ln (age) .87*** .55*
(.21) (.33)
Trade union member .22 .11
(.15) (.23)
Ideological extremeness .38*** .36***
(.05) (.07)
Personal income .004 .03
(.03) (.04)
Religious attendance .16*** .19***
(.04) (.07)
Leftist parties .21 −.36
(.21) (.29)
Income per inhabitant .03*** .02***
(.01) (.01)
Intercept −10.67*** −10.70***
(.99) (1.50)
Number of observations 6,185 6,185
Number of countries 12 12
Number of parties 68 68
Log-likelihood −953.03 −442.76
Notes: ***statistically signicant at the 1% level; **statistically signicant at the 5% level; *statistically
signicant at the 10% level.
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 9
.005 .01 .015 .02 .025 .03
Predicted Mean of Active Women Membership
0 10 20 30 40 50 60
Legislative Share
Figure 1. The effects of women legislators on active women members.
Note: The calculation of the probability of active women membership is based on Model 2 (Table 2).
Predictive margins with 95% condence intervals.
In regard to the effect of women’s sub-organizations on intra-party participation, Model
1 in Table 2 conrms Hypothesis 2A, showing that parties with this type of organization
tend to have more women members. However, Model 2 shows no relationship between
women’s organizations and women members becoming more active within their parties.
These organizations may help attract women to join the party, but seem to be less effective
in increasing their political involvement; further research on these organizations’ rules and
activities might shed light on this contrast. Moreover, party rules guaranteeing women’s
seats on national party executives (H2A) are not related to higher overall levels of women’s
membership or party activism.
Finally, we nd evidence for the effects outlined in Hypothesis 3A related to women’s
leadership positions.In particular, having more women legislators is associated with greater
women’s participation as party members (Model 1) and as active party members (Model 2).
In other words,having women’s organizations promoting women’s involvement (Hypothesis
2A) and women gaining key legislative jobs (Hypothesis 3A) are both associated with
greater grassroots female participation. The substantive effect of this relationship (H3A)
is displayed in Figure 1, showing how changes in women’s legislative seat shares for each
party are associated with different predicted probabilities of women’s active participation
in grassroots activities (women members who work for the party). For instance, women
supporters are almost twice as likely to work for a party as active members if there is gender
parity in its legislative delegation, compared to a party in which women make up only 10%
of its legislators.12
Our control variables work largely as we expect: the most likely participants are women
who are comparatively radical in an ideological sense, who are more educated, more
religiously devout, and older. Contrary to expectations, our models show that being part
of a trade union, and personal income are not predictors of women’s participation. At the
party level, the binary variable indicating the ideological stand of the party (leftist) is not
signicant. At the country level, we nd that higher levels of national wealth are associated
with higher levels of women’s membership.
C2020 European Consortium for Political Research

10 ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
Analysis 2: Narrowing the gender gap?
Even though we nd that some party rules and practices can help to boost women’s
participation that does not necessarily mean that they also reduce the gender gap. As a
reminder, this is because for grassroots participation levels (unlike with legislative seat
share), it is not a zero-sum game. Thus, to identify whether our key independent variables
help reduce the gender gap, we now analyse the full sample of respondents. We use the
same models, but now include a cross-level interaction term for gender to see if women are
impacted in a distinct way by each of our key independent variables.
As Table 3 shows, these interaction terms provide no evidence that the institutional
factors associated with higher women’s participation narrow the participation gender
gap (Hypotheses 1B and 2B). In particular, we note that women’s organizations do not
have a distinct impact on women supporters in terms of boosting either membership or
activism. In contrast, we nd evidence supporting Hypothesis 3B.The cross-level interaction
between female respondents and the party’s percentage of women legislators is positive and
statistically signicant in the party activism model. Grassroots gender gaps are narrower in
parties with a greater share of female MPs. Parties with more women in the legislatures tend
to have memberships which are more active overall, and also to have smaller participation
gender gaps within them. As our online appendices show, these results are robust and hold
under a variety of model specications.13
To appreciate the magnitude of these effects, Figure 2 shows how the predicted
probabilities of being an active party member vary by gender as the share of female MPs
grows.14 By the point where the gender composition reaches parity, the condence intervals
for men and women supporters overlap, meaning that the gender gap tends to vanish; it is
likely to disappear over this amount. As this gure makes clear, in this case, the gender gap
is closing mainly due to the increased activity among women supporters,not because men’s
activity declines. That is good news for parties seeking to attract more, and more active,
women supporters.15
Additional considerations
Our concern in this paper has been to establish whether the institutional and other forces
which are increasing women’s elite-level partisan participation are having similar effects of
boosting women’s grassroots partisan participation. We have thus posited our hypotheses
as if party-level (demand-side) forces have a leading role in explaining why parties display
very different levels of partisan participation gender gaps. Yet of course it is possible
that some of these relationships might have run in the opposition direction, with active
female participation spurring parties to adopt candidate quotas and/or elect women leaders.
Cross-sectional data have limitations for establishing the direction of these relationships.
Nevertheless, theoretical arguments and some aspects of the data, and the ndings, help
clarify the causal direction.
Above all, the timing suggests that the observed participation relations are owing
from party level to individual behaviour, not the other way around. Most parties in our
sample adopted gender-related candidate selection rules at least a decade before our survey
data (Caul 2001; Krook 2006). Moreover, studies concerning the approval of gendered
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 11
Tabl e 3. Multilevel logistic regression: explaining the gender gap in grassroots party participation
Party members
Party members who
work for the party
Model 1 Model 2
Hypothesis 1B
Voluntary quota .28 .41
(.25) (.26)
Voluntary quota * female −.18 −.26
(.20) (.32)
Mandatory quota .45*.91***
(.26) (.26)
Mandatory quota * female −.27 −.54
(.23) (.35)
Hypothesis 2B
Women’s sub-organizations .44** .33
(.21) (.21)
Women’s sub-organizations * female .21 .06
(.18) (.27)
Executive seat rules −.07 .12
(.21) (.22)
Executive seat rules * female −.02 .03
(.18) (.28)
Hypothesis 3B
Legislative share −0.001 −.004
(0.01) (.01)
Legislative share * female 0.01*.02**
(0.01) (.01)
Control variables
Female −0.85*** −1.22***
(0.26) (.41)
Education 0.02*** .05***
(0.01) (.01)
Ln (age) 0.73*** .25
(0.13) (.18)
Trade union member 0.39*** .49***
(0.09) (.14)
Ideological extremeness 0.32*** .27***
(0.03) (.04)
Personal income 0.05*** .07***
(0.02) (.03)
Religious attendance 0.17*** .19***
(0.03) (.04)
(Continued)
C2020 European Consortium for Political Research

12 ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
Tabl e 3. Continued
Party members
Party members who
work for the party
Model 1 Model 2
Leftist parties 0.16 .12
(0.18) (.20)
Income per inhabitant 0.03*** .02***
(0.01) (.01)
Intercept −8.97*** −8.13***
(0.66) (.85)
Number of observations 12,337 12,337
Number of countries 12 12
Number of parties 68 68
Log-likelihood −2,371.16 −1,282.93
Notes: ***statistically signicant at the 1% level; **statistically signicant at the 5% level; *statistically
signicant at the 10% level.
0.02 .04 .06
Predicted Mean of Active Membership
0 10 20 30 40 50 60
Legislative Share
Male Female
Figure 2. The effects of women legislators on the grassroots gender gap.
Note: The calculation of the probability of active women membership is based on Model 2 (Table 3).
Predictive margins with 95% condence intervals.
candidate quotas have not identied women’s membership share as a factor in the adoption
of these rules; they have pointed to other factors, including the number of women within the
highest ranks of the party and the adoption of these quotas by other countries (Caul 2001;
Krook 2006). Similarly, while party women’s organizations may have become instruments
through which (women) party leaders seek party policy change and promote women in
parliament (Childs & Kittilson 2016), most of these organizations were established decades
ago, not by current members (elites or otherwise). Similarly, most parties adopted rules
about representation of women on the party’s highest executive body long ago. Thus, we do
know that membership levels in our 2010 data are not an immediate cause of these party-
level structures.Finally, previous studies have also concluded that the election of women in
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 13
legislatures worldwide depends mostly on the existence of gender quotas; levels of women
members have not been posited as a cause (Jones 2009; Schwindt-Bayer 2009, 2010). All
these theoretical and empirical considerations reduce the likelihood of endogeneity between
these elite party variables and membership participation.
For those readers unconvinced by the idea of a single causal direction from the election of
women in legislatures to party membership, we also perform several checks to conrm that
this is a one-way relationship.Most importantly, we implement a two-step probit regression
selecting as an instrument the logarithm of the inverted value of the Gallagher index. This
variable is highly correlated (0.52) with the women’s seat share in the party’s delegation
in the lower house and uncorrelated with the residuals of the estimation (0.028). Through
different tests, we verify the exogeneity of the instrument.16
In addition, we can get additional insight into the possible mechanisms at work by
investigating the impact of women’s elite representation on grassroots supporters who
do not choose to become party members: are women supporters in general more likely
to become more active in politics as parties’ top levels became more gender-balanced
(a role model effect)? In fact, we nd no impact of our key independent variables on
women’s participation level or on the gender gap for self-identied supporters who had
not enrolled in their party. This nding suggests that party membership is an important
institutional condition in boosting women’s participation at the grassroots and closing the
participation gender gap. Thus, it supports the idea that the most important mechanisms at
work are demand-side forces (e.g., top-down relationships such as women elites recruiting
other women members to support their ascension within party organizations or make them
candidates in the future), and perhaps career-related factors (supply side), rather than more
generalized role model forces that encourage women to take a more positive view of partisan
political participation. This top-down relationship also reduces the likelihood that our other
results are affected by endogeneity.
Conclusions
This study offers several lessons for those who are interested in decreasing gender gaps
in grassroots party participation. First, while previous studies have shown that rules about
gender in candidate selection can increase women’s presence in legislatures, we nd that
some of the same mechanisms are also associated with increases in women’s grassroots
partisan participation. Party women’s organizations, parties’ pro-women candidate rules
and the women legislators they help to elect, are all associated with increases in women’s
party membership and party activism. Because these measures seemingly enliven parties’
membership organizations,they should make it easier for parties to nd high-quality women
candidates within their memberships. On the other hand, because many of these measures
increase engagement by all supporters,not just women, they do not uniformly work to close
participation gender gaps. Even so, the good news is that we nd no evidence that women-
promoting measures create a participation backlash, under which gender gaps might close
because men’s participation declines.
Overall, this study suggests that quotas and women sub-organizations have, at best, an
indirect role to play in closing participation gender gaps at the grassroots. To the extent that
these mechanisms promote women’s election to public ofce, they may boost the grassroots
C2020 European Consortium for Political Research

14 ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
mobilization efforts led by women ofce seekers. Thus, our ndings suggest that if parties
want to take steps to eliminate this gap, they probably need to adopt policies targeted
explicitly towards this goal, rather than just waiting for legislative delegations to achieve
gender parity. For instance, parties could try to enlist volunteers in a broader range of local
campaign activities, in recognition that women and men may have different participation
preferences (Coffé & Bolzendahl 2010). One way to get a better idea of what strategies might
work would be to establish whether men and women partisans have differing perceptions of
the costs or benets of party membership and party activism (putting a gendered focus on
studies along the lines of Poletti et al.(2019) or the chapters in van Haute & Gauja (2015)).
But even if more research is needed to identify plausible specic remedies, what does seem
clear from the results presented here is that parties should not count on closing the grassroots
participation gap merely by using the tools that have promoted women’s access to elite levels
of the party. Further measures are needed.
Acknowledgements
This research was supported in part by a grant from the National Science Foundation (NSF
1419401). We would like to thank Miki Kittilson, Mónica Lara, Mariano Sánchez-Talanquer,
Leslie Schwindt-Bayer, Zeynep Somer-Topcu, Daniel Weitzel and the anonymous reviewers
for their helpful comments and suggestions.
Online Appendix
Additional supporting information may be found in the Online Appendix section at the end
of the article.
Table A1: Operationalizations used in the analysis and sources
Table A2: Descriptive Statistics
Table A3: Number of Women Members and Women Supporters in our Sample
Table B1: Explaining Women’s Grassroots Party Participation
Table B2: Explaining the Gap in Women’s Participation
Table C1: Explaining Women’s Participation
Table C2: Explaining the Gap in Women’s Participation
Table D1: Rightist Parties, Explaining Women’s Participation
Table D2: Rightist Parties, Explaining the Gap in Women’s Participation
Table D3: Party Families, Explaining Women’s Participation
Table D4: Party Families, Explaining the Gap in Women’s Participation
Table D5: Women’s Position in the Party, Explaining Women’s Participation
Table D6: Women’s Position in the Party, Explaining the Gap in Women’s Participation
Table E1: Explaining Women’s Participation
Table E2: Explaining the Gap in Women’s Participation
Table F1: Multilevel Logistic Regression: Explaining Women’s Participation (Excluding
National-Level Covariates)
Table F2: Multilevel Logistic Regression: Explaining the gap in Women’s Participation
(Excluding National-Level Covariates)
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 15
Table G1: Multicollinearity Test (VIF)
Table H1: Explaining Women’s Participation
Table H2: Explaining the Gap in Women’s Participation
Figure I1: Proportion of women party members among those showing preference for a party,
by Country & Party
Notes
1. 5.5% of male respondents self-reported as party members compared with 3.2% of female respondents.
2.5% of male respondents self-reported as active members,compared to 1.1% of female respondents.
2. It includes a question about whether the respondent has “worked in a political party or action group in
last 12 months.”
3. Table A3 in the Appendix of the Supporting Information (last column) displays the available years in
the PPDB Database on party rules for each of the parties included in this study. For more details on the
PPDB,see Poguntke et al. (2016). To access the data, see Scarrow (2017).
4. The 68 parties of our sample represent most or all parties with parliamentary representation in the 12
surveyed countries (except for France,with only two included parties). This makes it unlikely that there
is sample selection bias at the party level, though of course our results are valid only for the group of
countries included in our analyses. Table A3 in the Appendix in the Supporting Information displays
parties’ names.
5. When joining both databases, we assign surveyed individuals to specic parties based on the following
criteria: (1) for those who are members: the party to which they belong; 2) for non-members: the party
to which they feel close, or, if they do not express such a preference, the party for which they voted (in
that order).
6. We employ random intercepts at the party level in our multilevel model.
7. We use a two-level model due to the limited number of parties by country. However, as a robustness
check we replicate the models displayed in Tables 2 and 3 employing three levels in our multilevel
models: respondents, parties and countries (see Tables B1 and B2 in the Appendix in the Supporting
Information). The results from this more conservative test are similar to those reported in the body of the
paper.
8. The variable takes the value of 1 only when there are party rules AND national requirements do not
exist. According to the information in the International IDEA database of gender quotas, none of the
parties in our sample required proportionality levels that were higher than those mandated by law.
9. Tables A1 and A2 in the Appendix (in the Supporting Information) display detailed descriptions and
statistics of variables used in the analyses. For each party in our sample Table A3 in the Appendix
(in the Supporting Information) shows the number of supporters and members, and the percentage of
members (among supporters), with total gures and for women only. Figure I1 in the Appendix (in the
Supporting Information) displays the range of within-country differences in the proportion of women
party members among those showing preference for a party in the 12 countries in our sample.
10. This variable transforms the usual 10-point scale measuring the ideological continuum left-right into
one of only ve points,in which the extreme values (1 and 10) take the value of 5 (two and nine take the
value of 4, and so on).
11. This data comes from the classication made by the 2010 ESS. The classication is available at the
following web page: https://www.europeansocialsurvey.org/docs/round5/survey/ESS5_appendix_a3_e04_
0.pdf
12. We use Model 2 in Table 2 to calculate the predicted probabilities of participating as an active member
as women’s share of party´s seats in the lower house varies.
13. As additional robustness checks, we replicate models displayed in Tables 2 and 3 under various
circumstances.First, to evaluate whether relatively small parties are driving the results,we run the models
excluding parties with fewer than 31 supporters in the ESS, then excluding those with fewer than 61
supporters (respective 7 and 10 parties out of our sample of 68). In both cases, our results are robust to
C2020 European Consortium for Political Research

16 ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
this reduction in the sample (Tables C1 and C2 in the Appendix in the Supporting Information). Second,
we test different party family effects rst by adding an additional party-level control variable identifying
rightist parties, then adding binary variables for multiple party families (Christian Parties, Conservative,
Social Democrat, and Greens) (Tables D1–D4 in the Appendix in the Supporting Information). Third,
we test our hypotheses employing alternative measures of our independent variables:for H1, party rules
that take account of gender in selecting party congress delegates;for H3, the share of women in the party
executive, and a dichotomous variable for a woman party leader (Tables D5 and D6 in the Appendix
in the Supporting Information). Fourth, we consider whether gender disparities in attitudes affect the
supply of women’s partisan participation, including interest in politics, satisfaction with democracy, and
perceptions about whether women and men are equally entitled to a job when jobs are scarce (Tables F1
and F2 in the Appendix in the Supporting Information). Finally,we exclude all national-level variables to
see whether they distort our results (Tables G1 and G2 in the Appendix in the Supporting Information).
None of these alternative model specications alter our initial ndings, or show additional statistically
signicant relationships.The validity of our key ndings holds even after employing country xed effects.
Finally, we test whether or not multicollinearity poses a concern in our empirical analysis. The Variance
Ination Factor (VIF) analysis indicates that this potential problem does not seem to distort our results
signicantly. The mean VIF equals 2.82 (for the second regression in Table 3) and VIF does not exceed
10 for any independent variable.(Any value greater than 10 would suggest further investigation) (Table
H1 in the Appendix in the Supporting Information ). Finally, we also run our models separating the
variables mandatory quota and voluntary quota. Our key ndings hold and these variables do not seem
to play a role as previously stated.
14. We employ the estimations made from Model 2 in Table 3 to calculate the predicted probabilities of
active party membership for men and women party supporters, depending on the share of women
legislators.
15. In addition, we implement a partial hypothesis test, based on Model 2 from Table 3, to determine whether
the variables ‘female’, ‘Women’s legislative seat share: party specic’, and the interaction between them
are (jointly) statistically signicant. The test’s outcome indicates that these covariates are statistically
signicant at the .004 level (p-value). Another partial hypothesis test to determine whether the variables
‘female’ and the interaction between the variables ‘female’ and ‘Women’s legislative seat share: party
specic’ reveals that these variables are (jointly) statistically signicant at the .004 level (p-value). The
goal of this exercise is to conrm the statistical signicance of the combined effect of these variables as
a whole.
16. Even at a conservative .10 signicance threshold, the Wald Test of exogeneity supports accepting
the assumed exogeneity for our instrument. Section H´s tables in the Appendix (in the Supporting
Information) show these results. We also nd support for the Hypothesis 3 through this empirical
approach. Again, we show the effect of the variable women’s seat-share in party’s delegation in the lower
house is robust regardless of the approach adopted. We employ the command ivprovit in Stata for this
estimation. Furthermore, we implement a global test of endogeneity of the regressors. Specically, we run
a test of endogeneity after a GMM estimation as an additional robustness test to evaluate the presence
of endogeneity in the set of all regressors (Hausman 1978; Wooldridge’s 1995; Wu 1974). We fail to reject
the null hypothesis of exogenous regressors applying a conservative .10 signicance threshold. Finally,
we implement mediation analysis employing the variables used in those regressions displayed in Section
H’s tables to examine whether our instrument is able to affect our dependent variables (membership and
active membership) directly or indirectly (through the variable ‘women’s seat-share in party’s delegation
in the lower house’). While the instrument could also affect the dependent variables indirectly (the
effect of the instrument on the variable women’s seat-share in party’s delegation in the lower house is
statistically signicant), it did not directly.
References
Barnes, T. & Burchard, S. (2013). Engendering politics: The impact of descriptive representation on women’s
political engagement in sub-Saharan Africa. Comparative Political Studies 46(7): 767–790.
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 17
Beauregard, K. (2017). Quotas and gender gaps in political participation among established industrial
European democracies: Distinguishing within-and across-country effects. Political Research Quarterly
70(3): 657–672.
Beauregard, K. (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.
Burns, N., Schlozman, K. & Verba, S. (200 1). The Private Roots of Public Action: Gender, Equality, and
Political Participation. Cambridge, MA: Harvard University Press.
Campbell, A., Converse, P.E., Miller, W.E. & Stokes, D.E. (1964). The American voter. New York: Wiley.
Caul, M.(1999). Women’s representation in parliament: The role of political parties. Party Politics 5(1):79–98.
Caul, M. (2001). Political parties and the adoption of candidate gender quotas: A cross-national analysis.
Journal of Politics 63(4): 1214–1229.
Childs, S.& Kittilson, M.C. (2016). Feminizing political parties: Women’s party member organizations within
European parliamentary parties. Party Politics 22(5): 598–608.
Clayton, A. (2015). Women’s political engagement under quota-mandated female representation: Evidence
from a randomized policy experiment. Comparative Political Studies 48(3): 333–369.
Coffé, H. & Bolzendahl, C. (2010). Same game, different rules? Gender differences in political participation.
Sex Roles 62(5): 318–333.
Cross, W.P. & Pilet, J-B. (eds.). (2015). The politics of party leadership: A cross-national perspective. Oxford:
Oxford University Press.
Dalton, R.J. (1996). Citizen politics. Chatham, NJ: Chatham House.
Davidson-Schmich, L. (2016). Gender quotas and democratic participation: Recruiting candidates for elective
ofces in Germany. Ann Arbor: University of Michigan Press.
De Paola, M, Scoppa, V. & De Benedetto, M.A. (2014). The impact of gender quotas on electoral
participation: Evidence from Italian municipalities. European Journal of Political Economy 35: 141–157.
Desposato, S. & Norrander,B. (2009). The gender gap in Latin America: Contextual and individual inuences
on gender and political participation. British Journal of Political Science 39(1): 141–162.
ESS Round 5: European Social Survey Round 5 Data (2010). Data le edition 3.4. NSD - Norwegian Centre
for Research Data, Norway,Data Archive and distributor of ESS data for ESS ERIC.
Gauja, A. (2015). The construction of party membership. European Journal of Political Research 54(2):
232–248.
Hausman, J. (1978). Specication tests in econometrics. Econometrica 46: 1251–1271.
Hernson, P., Lay, J.C. & Stokes, A.K. (2003). Women running “as women”: Candidate gender, campaign
issues, and voter-targeting strategies. The Journal of Politics 65(1): 244–255.
Hooghe, M., Stolle, D. & Stouthuysen, P. (2004). Head start in politics: The recruitment function of youth
organizations of political parties in Belgium (Flanders). Party Politics 10(2): 193–212.
Huddy, L & Terkildsen, N. (1993). Gender stereotypes and the perception of male and female candidates.
American Journal of Political Science 37: 119–147.
International IDEA. Gender Quotas Database. www.idea.int/data-tools/data/gender-quotas/database.
Jones, M. (2009). Gender quotas, electoral laws, and the election of women. Evidence from the Latin
American vanguard. Comparative Political Studies 42(1): 56–81.
Karp, J. A. & Banducci, S.A. (2008). When politics is not just a man’s game: Women’s representation and
political engagement. Electoral Studies 27(1): 105–115.
Kenworthy, L. & Malami, M. (1999). Gender inequality in political representation: A worldwide comparative
analysis. Social Forces 78(1): 235–268.
Kittilson, M.C. (2006). Challenging parties, changing parliaments: Women and elected ofce in contemporary
Western Europe. Columbus: Ohio State University Press.
Kittilson, M.C. (2008). Representing women: The adoption of family leave in comparative perspective. The
Journal of Politics 70(2): 323–334.
Kittilson, M.C. & Schwindt-Bayer, L.A. (2012). The gendered effects of electoral institutions: Political
engagement and participation. Oxford: Oxford University Press.
Krook, M. (2006). Reforming representation: The diffusion of candidate gender quotas worldwide. Politics
& Gender 2: 303–327.
C2020 European Consortium for Political Research

18 ALDO F. PONCE, SUSAN E.SCARROW & SUSAN ACHURY
Kunovich,S. & Paxton, P. (2005). Pathways to power: the role of political parties in women’s national political
representation. American Journal of Sociology 111(2): 505–552.
Leeper, M. S.(1991). The impact of prejudice on female candidates: An experimental look at voter inference.
American Politics Quarterly 19: 248–261.
Liu, S-J.S. & Banaszak, L.A. (2017). Do government positions held by women matter? A cross-national
examination of female ministers’ impacts on women’s political participation. Politics & Gender 13(1):
132–162.
Marien, S., Hooghe, M. & Quintelier, E. (2010). Inequalities in non-institutionalised forms of political
participation: A multi-level analysis of 25 countries. Political Studies 58(1): 187–213.
Norris, P. (2002). Democratic phoenix: Reinventing political activism. Cambridge: Cambridge University
Press.
Poguntke, T., Scarrow, S.E., Webb, P.D., Allern, E.H., Aylott, N., van Biezen, I., Calossi, E., Costa Lobo, M.,
Cross, W.P., Deschouwer, K., Enyedi, Z., Fabre, E., Farrell, D.M., Gauja, A., Pizzimenti, E., Kopecký, P.,
Koole, R., Müller, W.C., Kosiara-Pedersen, K.,Rahat, G., Szczerbiak, A., van Haute, E. & Verge, T.(20 16).
Party rules, party resources and the politics of parliamentary democracies: How parties organize in the
21st century. Party Politics 22(6): 661–678.
Poletti, M., Webb,P.D. & Bale, T.(2019). Why do only some people who support parties actually join them?
Evidence from Britain. West European Politics 42(1): 156–172.
Ponce, A. & Scarrow, S.E.(20 16). Which members? Using cross-national surveys to study party membership.
Party Politics 22(6): 679–690.
Raudenbusch, S. & Bryk, A. (2002). Hierarchical linear models: Applications and data analysis methods,2nd
edn. Newbury Park, CA: Sage.
Reingold, B. & Harrell,J. (2010). The impact of descriptive representation on women’s political engagement:
Does party matter? Political Research Quarterly 63(2): 280–294.
Reynolds, A. (1999). Women in the legislatures and executives of the world: Knocking at the highest glass
ceiling. World Politics 51(4): 547–572.
Sandri, G., Seddone, A. & Venturino, F. (eds.). (2015). Party primaries in comparative perspective. Farnham,
UK: Ashgate Publishing.
Scarrow, S.E.(20 14). Beyond party members:Changing approaches to partisan mobilization. Oxford:Oxford
University Press.
Scarrow, S.E. (2017). Political Party Database Version 1a.2 (2017 Update), https://doi.org/
10.7910/DVN/UJDXNE, Harvard Dataverse, V1, UNF:6:3zpYpiffBYoUuVeP02Vn2w ==[leUNF
(2017).
Schlozman, K.L., Burns, N., Verba, S. & Donahue, J. (1995). Gender and citizen participation: Is there a
different voice? American Journal of Political Science 39(2): 267–293.
Schwindt-Bayer, L.A. (2009). Making quotas work: the effect of gender quota laws on the election of women.
Legislative Studies Quarterly 34(1): 5–28.
Schwindt-Bayer, L.A. (2010). Political power and women’s representation in Latin America. Oxford: Oxford
University Press.
Seyd, P. & Whiteley, P. (2004). British party members. An overview. Party Politics 10(4): 355–366.
Smith, E. & Fox, R. (2001). The electoral fortunes of women candidates for Congress. Political Research
Quarterly 54(1): 205–221.
Steenbergen, M. & Jones, B. (2002). Modeling multilevel data structures. American Journal of Political
Science 46(1): 218–237.
Sundström, A. & Stockemer, D. (2015). What determines women’s political representation at the local level?
A ne-grained analysis of the European regions.International Journal of Comparative Sociology 56(3–4):
254–274.
Tremblay, M. (ed.). (2008). Women and legislative representation. Basingstoke, UK: Palgrave Macmillan.
Valdini, M.E. (2012). A deterrent to diversity: The conditional effect of electoral rules on the nomination of
women candidates. Electoral Studies 31(4): 740–749.
van Haute, E. & Gauja, A. (eds.) (2015). Party members and activists. Abingdon, UK: Routledge.
Verba, S. & Nie, N. (1972). Participation in America: Political democracy and social equality. New York:
Harper & Row.
C2020 European Consortium for Political Research

QUOTAS, WOMEN’S LEADERSHIP AND GRASSROOTS WOMEN ACTIVISTS 19
Verba, S., Burns, N & Schlozman, K.L. (1997). Knowing and caring about politics: Gender and political
engagement. Journal of Politics 59(4): 1051–1072.
Verba, S., Nie, N & Kim, J. (1978).Participation and political equality: A seven nation comparison. New York:
Cambridge University Press.
Welch, S. (1977). Women as political animals? A test of some explanations for male-female political
participation differences. American Journal of Political Science 21(4): 711–730.
Whiteley, P. (2011). Is the party over? The decline of party activism and membership across the democratic
world. Party Politics 17(1): 21–44.
Widfeldt, A. (1995). Party membership and party representativeness. In D. Fuchs & H.D. Klingemann (eds.),
Citizens and the state (pp. 134–182). Oxford: Oxford University Press.
Witt, L., Paget, K & Matthews, G. (1994). Running as a woman: Gender and power in American politics.New
York: Free Press.
Wolak, J. (2015). Candidate gender and the political engagement of women and men. American Politics
Research 43(5): 872–896.
Wolbrecht, C. & Campbell, D. (2007). Leading by example: Female members of parliament as political role
models. American Journal of Political Science 51(4): 921–939.
Wooldridge, J. (1995). Score diagnostics for linear models estimated by two stage least squares. In G.S.
Maddala, T.N. Srinivasan & P. Phillips (eds.), Advances in econometrics and quantitative economics:
Essays in honor of Professor C. R. Rao (pp. 66–87). Oxford: Blackwell.
Wu, D-M. (1974). Alternative tests of independence between stochastic regressors and disturbances: Finite
sample results. Econometrica 42: 529–546.
Zetterberg, P. (20 09). Do gender quotas foster women’s political engagement? Lessons from Latin America.
Political Research Quarterly 62(4): 715–773.
Address for correspondence: Susan E. Scarrow, Department of Political Science, University of Houston,
Houston, TX 77204-3011, USA. Email: sscarrow@uh.edu
C2020 European Consortium for Political Research