Content uploaded by Raul Gomez
Author content
All content in this area was uploaded by Raul Gomez on Jun 13, 2017
Content may be subject to copyright.
1
Why does Alejandro know more about politics than Catalina? Explaining
the Latin American Gender Gap in Political Knowledge
Marta Fraile (CSIC and EUI)
Raul Gomez (University of Derby)
This is the accepted version of an article published in the British Journal of Political Science.
DOI: https://doi.org/10.1017/S0007123414000532
Please refer to the published article for citation purposes
Acknowledgements:
This work was supported by a grant from the Spanish Ministry of Economy (Reference:
CSO2012-32009).
2
Abstract
This article tests contextual and individual-level explanations about the gender gap in
knowledge in Latin American Countries.We suggest that the gender gap in knowledge is
impacted by political and economic settings through two interrelated mechanisms: gender
accessibility (i.e. the extent to which the context provides opportunities for women to
influence the political agenda) and gender-bias signaling (i.e. the extent to which women
actually play important roles in the public sphere). Analyzing data from the 2008
AmericasBarometer's round of surveys, we show that the gender gap in knowledge is smaller
among highly educated citizens, in rural areas (where both men and women know little about
politics) and in bigger cities (where women's levels of political knowledge are higher). More
importantly, the magnitude of the gender gap varies to a great extent across countries. Gender
differences in income, party-system institutionalization and the representation of women in
national parliaments are all found to play a particularly important role in explaining the
magnitude of the gender gap in political knowledge across Latin America.
Research on public opinion and voting behavior has frequently shown that the average
citizen’s overall level of information, knowledge and understanding of politics is relatively
poor. In addition, political knowledge appears to be unevenly distributed. From all the many
sources of knowledge inequalities identified in previous literature (such as resource-based,
3
motivational-base, race-base), this article deals with perhaps the most puzzling: gender
differences in political knowledge.
Gender differences in knowledge have proven to be remarkably persistent.
Notwithstanding the tremendous increase in the degree of gender equality in political power
and resources in industrialized democracies, women appear to know and to participate less on
politics than men.
1
Although some researchers have noted that the difference is often small in
comparison to other inequalities such as education or social class
2
, gender differences in
knowledge do not only affect more than half of the world's population but also appear to be
the most enduring and strong. The uneven distribution of political knowledge between men
and women raises a number of normative concerns. Political knowledge translates into
political power: governments are more responsive to citizens’ demands when political
knowledge is evenly distributed across social groups.
3
Moreover, governments have more
incentives to be responsive when they can be held accountable, but citizens can only be able
to hold governments accountable for their actions when they know what governments are
actually doing. Consequently, if women systematically have less political knowledge than
men, they may be less well-represented in the democratic system. This would imply a clear
disadvantage in women’s capacity to voice their political needs and wants, and to influence
the political decision-making process. In short, gender constitutes a dimension of political
under-representation both historically and currently, and this fact merits the attention of
researchers.
1
Burns 2007; Delli Carpini and Keeter 1996 and 2000; Norirs 2000.
2
Burns 2007, Norris 2002.
3
Delli Carpini and Keeter 1996.
4
Gender differences in knowledge are particularly worthy of attention in the case of
Latin America where women have been historically prevented from political power and
participation. In spite of very fast changes during the past decade, there are still very large
variations in the levels of women's presence and empowerment across countries in the
region.
4
While the gender gap in knowledge is well documented in the context of advanced
industrial democracies such as the United States, Canada, and Western Europe, we know very
little about the extent of the gender gap in knowledge and its potential determinants in the
case of Latin America.
5
If the gender gap in knowledge persists in countries where the degree
of gender equality in political power and resources is higher, what should be expected from
the case of young or re-emerging democracies facing relevant socio-economic challenges?
The political histories of Latin American countries are plagued with many examples of
political exclusion, persecution and repression of women’s full equality of citizenship,
together with very traditional female-unfriendly economic institutions. Does this additional
disadvantage translate into great levels of gender inequalities in political knowledge? Or have
recent increasing opportunities for women to participate in the political sphere in these
countries helped to close the gender gap in knowledge?
4
Htun and Piscopo 2012.
5
To the best of our knowledge, there are no previous studies about the existence and the
extent of the gender gap in knowledge in Latin America. Moreover there is a complete lack
of studies evaluating the relative impact of contextual versus individual level factors in
explaining the magnitude of the gender gap in knowledge both in advanced industrial
democracies and in Latin American countries. Ours is the first study of this kind. The
following articles however examine the gender gap in political participation and political
values but not in knowledge as we do here: Desposato and Norrander 2009; Morgan and
Buice 2013; Walker and Kehoe 2013.
5
Common explanations of gender inequality in political knowledge point to traditional
social norms that identify women as those responsible for parenting and other caring
activities, as well as to the socioeconomic disadvantages suffered by women, such as lower
levels of socioeconomic and cognitive resources.
6
However, these studies focus only on
explanations at the individual level. In contrast, this article aims to test contextual
explanations about the gender gap in knowledge while controlling for the traditional accounts
at the individual level. We suggest that gender gap in knowledge may also be impacted by
political and economic settings through two interrelated mechanisms: gender accessibility
(i.e. the extent to which the context provides opportunities for women to influence the
political agenda) and gender-bias signaling (i.e. the extent to which women actually play
important roles in the public sphere). Confirming previous studies, findings show that the
gender gap in knowledge is smaller among highly educated citizens, in rural areas (where
both men and women know little about politics) and in bigger cities (where women's levels of
political knowledge are higher). More importantly, findings also show that the gap is
significantly higher in countries that send signals about the assumed marginal role of women
in the public sphere, whereas it is smaller in contexts facilitating the introduction of female
topics in the political agenda.
INDIVIDUAL AND CONTEXTUAL DETERMINANTS OF THE GENDER GAP IN
KNOWLEDGE
Gender differences in political knowledge might arise from three causes: sex differences in
resources, differential effects of the antecedents of knowledge for men and women, and
6
Delli Carpini and Keeter 1996 and 2000; Verba, Burns and Schlozman 1997.
6
differential contexts. To begin with, males and females have differential access to the main
antecedents of knowledge (abilities, resources, and opportunities). For example, where
women have fewer educational opportunities, the lack of these resources might reduce their
aggregate levels of knowledge compared to men. Secondly, men and women may respond
differently to the same factors; education may provide more returns to women than to men, or
vice versa. Thirdly, gender differentials may reflect broader cultural/economic/political
contexts. For instance, Inglehart and Norris argue that the implication of women in politics is
discouraged in those societies where traditional sex roles are dominant (such as agrarian
societies).
7
As Desposato and Norrander have pointed out,
8
contextual effects cannot be measured
in single country studies. Instead, we need a multi-country, two-level interactive model where
knowledge may vary as a function of individual factors, as an interaction of such factors and
gender, and as a function of the interaction of context and gender. If only individual variables
predict knowledge, then the gender gap is caused by sex inequality of access to resources and
opportunities (such as education, income, time). If individual factors interact with gender,
then the gender gap reflects an inequality in the magnitude of the effects of each of the
antecedents of knowledge for men and women. For example, poverty or lack of time might
dampen women’s knowledge more than men’s, and education might promote women’s
knowledge more than men. Finally, if the gender gap varies across countries, then it might
also depend on some contextual variables. Our study includes a diverse set of countries that
allow us to test for each type of mechanism.
7
Inglehart and Norris 2003.
8
Despostao and Norrander 2009, p 146.
7
At the individual level, previous research has identified numerous factors that
contribute to explaining the gender gap in knowledge. Traditional explanations are based on
socialization theory, which states that conventional social norms define men as citizens who
are in charge of public life and women as being in charge of the domestic or private domain,
as they are seen as more committed to childrearing, caring, and family life more generally.
9
These responsibilities involve an additional cost in the decision to become informed about
politics. Part of this same argument is that men and women are situated differently in the
social structure and have different levels of material and cognitive resources, distinct burdens
of work and responsibilities, and consequently different amounts of time available to dedicate
to getting informed about politics.
10
An additional explanation for the gender differences in knowledge is based on the
possibility that the determinants of what people know or ignore about politics (that is,
abilities, resources, and motivation) are themselves gendered. Women are less likely than
men to possess the precursors of knowledge. But when they attain a greater level of resources
(such as education or income), they might be especially encouraged and therefore receive
significantly larger returns for their political knowledge than their male counterparts. Some
studies argue that even when the socioeconomic differences between men and women are
slight (or even disappear), their effect on what people know or ignore about politics might be
different for men and women. In particular, there is evidence from the US that demonstrates
that the positive effect of education on knowledge is lower for women than it is for men.
11
In
9
Delli Carpini and Keeter 1996.
10
Fraile 2014.
11
Dow 2009.
8
contrast, a recent study of 27 European countries shows that the magnitude of the effect of
education on political knowledge is higher for women than for their male counterpart.
12
In addition, some studies have speculated that the gender gap in political interest and
knowledge could disappear over time due to both changes in socialization and generational
replacement.
13
Previous generations were socialized under (either democratic or
undemocratic) political systems where women were explicitly or implicitly excluded from the
public sphere, not only in politics but also in the labor market. As a consequence, the gender
gap is argued to be greater for older citizens in comparison to their younger counterparts.
However these individual-level explanations account for only one part of the gender
disparity. An alternative approach to explaining the enduring gender gap in knowledge is to
take into account the contextual factors where citizens live. This is what we do in this article:
to focus on contextual factors while controlling for the individual-level factors previously
used in the literature and just mentioned. Knowledge gender gaps have rarely been discussed
in a cross-national context (with the sole exception of Kittilson and Schwindt-Bayer 2012:
52-56). In fact, only studies focusing on gender differences in political participation and
attitudes have adopted this strategy.
14
Here, we suggest that there are at least two mechanisms through which the contextual
setting may influence gender differences in political knowledge. The first of these
mechanisms is gender-bias signaling. We argue that settings which send strong signals that
women only play a marginal role in the public sphere (gender bias) may discourage women's
12
Fraile 2014
13
Bennett and Bennett 1989; Inglehart and Norris 2003.
14
See, for instance Barnes and Burchard 2013; Desposato and Norrander 2009; Karp and
Banducci 2008; Kittilson and Schwindt-Bayer 2010; Morgan and Buice 2013.
9
interest in and access to political information relative to men's. So, for example, perceiving
that most women tend to play a relatively unimportant position in the labor market, or that
politics is mainly a men’s game, may lead to women not paying as much attention to political
issues as men.
The second mechanism, gender accessibility, operates by creating the conditions for
female participation in politics. Women tend to be one of the most disadvantaged groups in
terms of political capital. Therefore, contexts that facilitate women's political participation
and the introduction of female-related issues into the political agenda should also help to
foster their interest, thereby reducing the gender gap in knowledge.
Previous studies about the influence of contextual factors on the gender gap in
political engagement have claimed that elected women constitute symbols of inclusion in the
political system. Thus, the presence of women in political institutions incentivizes the interest
of women (especially young women) to get involved in politics and to be exposed to political
information, thereby raising their knowledge of politics.
15
The logic behind this argument is
linked to the idea of descriptive representation, which refers to the similarity (in terms of
race, social class, or in this case gender) between representatives and citizens.
16
According to
this approach, increases in the descriptive representation of particular groups become
especially relevant when those groups (in this case, women) have been historically excluded
from politics.
17
In addition, the descriptive representation of women may facilitate the
defense of women’s specific interests that are best defended by women themselves, therefore
15
Burns, Schlozman and Verba 2001.
16
Pitkin1967.
17
Phillips 1995.
10
increasing substantive representation as well.
18
We contend that greater representation of
women in political institutions can be interpreted as a signal that politics is not only a men's
game and could therefore help to reduce the gender gap in knowledge. Therefore, this factor
is considered here as an indicator of gender-bias signaling in the political dimension.
Findings in the literature regarding the effects of descriptive representation are mixed,
with a clear predominance of studies focusing on the US. Some of them find that females
living in US states with women in charge of prominent political offices are significantly more
likely to be politically informed, interested, and involved.
19
However, others have found little
support for the role of women’s descriptive representation in encouraging political
engagement.
20
Studies seeking to explore the link between women’s representation and
political engagement (and /or attitudes) across countries also offer mixed evidence. Whereas
some scholars find support for the descriptive representation hypothesis in Latin America,
21
18
Pitkin 1967.
19
Atkeson 2003; Atkeson and Carrillo 2007; Burns, Scholzman and Verba 2001; Campbell
and Wolbrecht 2006; Koch 1997; Reingold and Harrell 2010; Verba, Burns and Scholzman
1997; Wolbrecht and Campbell 2007.
20
Dolan 2006; Lawless 2004; Zetterberg 2009.
21
Desposato and Norrander 2009.
11
Europe,
22
and Sub-Saharan Africa,
23
other studies that have analyzed a wider geographical
area
24
find little support in favor of this hypothesis.
25
It appears then that the inconclusive findings of this literature call for additional tests.
This is what we do here, but accounting for gender-bias signaling not only in the political
dimension but also in the socioeconomic dimension. Although ignored by previous literature,
we argue that the degree of male predominance in the labor market is a relevant economic
dimension of gender-bias signaling. Contexts where women hardly earn the same income as
men may further the idea that female's role in the public sphere is only secondary to and less
important than their traditionally preeminent role in the private sphere. On the contrary,
gender income equality may foster the interest of women in public and political issues and,
therefore, help to reduce the gender gap in knowledge. The gender gap in earnings is acute in
Latin America, and differences between countries have grown very visibly during the 2000s,
with declines in twelve countries and increases in another six.
26
Moving on to the second mechanism, gender accessibility, previous studies have
showed that institutions influence citizens’ political behavior and opinions through a
22
Wolbrecht and Campbell 2007.
23
Barnes and Burchard 2013.
24
Both Karp and Banducci 2008 and Kittilson and Schwindt-Bayer 2010 analyze a set of 29
developing and 34 developed democracies around the world respectively.
25
Karp and Banducci 2008; Kittilson and Schwindt-Bayer 2010.
26
World Bank 2012.
12
psychological mechanism.
27
Institutions are symbols that represent ideals of the way in which
democracy works, and provide cues for citizens on such ideals. Thus, citizens learn more
about political institutions on each occasion they are involved in politics. And what they learn
in turn shapes their political attitudes, opinion and behaviors.
28
A previous study has showed
that gender differences in political engagement tend to be smaller in more proportional
electoral systems, arguably because proportionality opens the political structure up to the
representation and inclusion of diverse social groups that tend to be underrepresented in
majoritarian systems. This may not only help alternative issues to make their way into the
political agenda, but also encourage women to engage themselves in politics.
Other power-sharing institutions such as federalism may have similar effects as they
multiply access points for citizens to interact with institutional politics and to gather political
information.
29
Moreover, studies have found that women tend to know more about local
issues,
30
which leads to the gender gap in knowledge of national politics disappearing when
local politics are under study.
31
For these reasons, the power-sharing effects of federalism
may help reduce gender differences when it comes to political knowledge.
Although previously overlooked in the literature, we contend that there is an
additional and equally important factor which refers to the accessibility of the political
27
Kittilson and Schwindt-Bayer, 2010.
28
Kittilson and Schwindt-Bayer 2010, p. 992.
29
Kittilson and Schwindt-Bayer 2010.
30
Verba, Burns, and Schlozman 1997.
31
See for instance, Shaker 2012 and Rapelli 2014.
13
context: the institutionalization of the party system. There are at least two mechanisms
through which stable party systems may facilitate women's access to politics. On the one
hand, stability is needed in order for stronger links with parties to emerge,
32
which might
potentially be more beneficial for those, like women, who have traditionally less access to
politics. Parties are primary vehicles of political information,
33
and so contexts of party
system stability may help foster women's political knowledge. On the other hand, stability
may also provide women with better access to politics and to the political agenda. Unstable,
weakly institutionalized party systems in Latin America are characterized by competition
based on personal appeals and short-term populist policies
34
with women-friendly policies
and discourses being completely contingent on the will of charismatic leaders.
35
Several
studies have found weak institutionalization to provide a poor environment for women's
engagement in politics. Weakly institutionalized parties tend to be biased towards members
with strong political capital and external support, which are resources that women (in
comparison to men) tend to lack.
36
Studies also indicate that women activists' influence is
stronger in more stable, institutionalized parties, where the rules of the game are clear and
widely known to all and there is less need to rely on the "boys' networks" in order to gain
access to the agenda.
37
Arguably, women's achievements will be more likely to remain and
have an impact on women's political engagement in scenarios of greater institutionalization.
32
Mainwaring and Torcal 2006.
33
Römmele 2003.
34
Mainwaring 1998, 1999.
35
See Kampwirth 2010.
36
Caul 1999.
37
See, for example, Bruhn 2003, Lovenduski 1993, and Waylen 2000, 2003.
14
There is huge variation in the degree of institutionalization of political parties across
Latin America. However, there is a tendency to find relatively unstable, personalistic and
informal party systems in many countries. This context does not seem to be the best place for
women to engage with party politics and to influence the political agenda, which might
reflect in lower levels of political information. In short, we contend that, apart from power-
sharing institutions, there is an additional key factor regarding the accessibility of the political
context: party-system institutionalization. And this needs to be considered in explaining the
gender gap in political knowledge.
To recapitulate, we test here the extent to which certain contextual factors contribute
to increasing or decreasing the gender gap in political knowledge in Latin American countries
while controlling for the typical antecedents of knowledge at the individual level. We suggest
that there are two mechanisms through which the contextual setting may influence the size of
the differences in knowledge between men and women: gender-bias signaling and gender
accessibility. We test the extent to which these mechanisms operate in the case of Latin
America while also controlling for other contextual factors that have been considered by
previous studies: the degree of political freedom
38
and socioeconomic development.
39
38
The potential effect of political freedom on citizens’ political knowledge is not obvious: it
might promote women’s political involvement and knowledge but also decrease it. Women
have often played a relevant part in organizing and developing protest movements under
authoritarian rule (Desposato and Norreder 2009) and in presenting demands to the state for
the provision of public services and the recognition of human rights (Safa 1990)
39
The hypothesis is that economic development leads to a general shift in attitudes, including
(among other things) increasing awareness of the need for more gender equality in the
political arena (Inglehart and Norris 2003). This shift should in turn create more opportunities
15
RESEARCH DESIGN: DATA AND VARIABLES
One reason explaining the lack of attention given to context in relation to the gender gap in
knowledge might be connected with the scarcity of data. There is indeed little comparative
data containing enough information to construct valid indexes of political knowledge that are
comparable across countries.
40
To this respect, the 2008 AmericasBarometer’s round of
surveys offers a unique opportunity to use a comparable set of measures about citizens’
political knowledge across countries.
41
Accordingly, we use data from 18 Latin American
for women's engagement and for female-related issues to enter the political agenda. Empirical
evidence in this regard is, however, not conclusive. Whereas Inglehart and Norris (2003) find
evidence in favour of the hypothesis, the only existing study focusing in Latin American
countries find no support for it (see Desposato and Norrander 2009)
40
To date, only two comparative surveys contain questions on political knowledge: the 2009
European Election Studies Voter Study (EES) and the Comparative Study of Electoral
Systems (CSES). The first three modules of the CSES (from 1996-2011) contain three
questions about democratic institutions, leading politicians and the national parties but these
questions are not standardised across countries, making cross-country comparison
problematic. In contrast, the EES includes up to seven questions about the functioning of EU
institutions and national political actors that are common to all countries. The EES data have
been used, for example, by Fraile (2013 and 2014) to study the contextual determinants of
political knowledge in Europe and the gender differences in knowledge in Europe
respectively.
41
We thank the Latin American Public Opinion Project (LAPOP) and its major supporters
(the United States Agency for International Development, the United Nations Development
16
democracies: Argentina, Bolivia, Brazil, Chile, Colombia,
42
Costa Rica, Dominican Republic,
Ecuador, Guatemala, El Salvador, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru,
Uruguay, Venezuela. All of these are presidential democracies that, despite sharing similar
cultural characteristics and similar histories of (de-)colonization, show variation in our main
variables of interest.
In contrast with the reduced number of items of more recent surveys, the 2008 wave
of the AmericasBarometer contains five questions that relate to various aspects of citizens’
political knowledge. The exact wording of these questions is contained in the On line
Appendix (Table A.2). After testing for the internal consistency of these five items,
43
we
constructed an index of factual political knowledge in the conventional way –that is, creating
an additive measure of correct answers (coded as 1) against incorrect and DK responses
(coded as 0). Thus, the index varies from 0 correct answers to 5 correct answers. The average
number of correct answers across the 18 countries is 2.62, but there is plenty of variation, as
Figure 1 shows. Uruguay, with 3.5 correct answers on average, ranks first in our index, very
Program, the Inter-American Development Bank, and Vanderbilt University) for making the
data available on www.lapopsurveys.org.
42
In the case of Colombia, we replaced the 2008 survey by another conducted in 2009 by the
LAPOP and containing exactly the same questions. We did this because the coding of some
variables in 2008 differed from that in other countries (in Colombia, incorrect, "don't know"
answers and rejection were grouped together under the same category). Needless to say,
Colombian aggregate variables also correspond to the year 2009.
43
The average value of Chronbach’s α is 0.66, with only one country scoring below 0.60
(Uruguay = 0.56) and therefore no country scoring under the minimum benchmark of 0.50.
17
closely followed by Honduras (3.41) and Argentina (3.37); while Nicaragua, with 1.75
correct answers, is found at the other extreme.
FIGURE 1 Index of Political Knowledge by Country
0 1 2 3 4 5
Nicaragua
Dominican Rep
Brazil
Mexico
Guatemala
Colombia
Chile
Peru
Paraguay
El Salvador
Costa Rica
Venezuela
Ecuador
Bolivia
Panama
Argentina
Honduras
Uruguay
Source: Our elaboration on LAPOP, 2008.
At the individual level, this study employs indicators of the standard antecedents of
political knowledge, i.e. individual differences in motivation, resources and ability.
44
It is
expected that the gender gap will be smaller among younger, more educated, less religious,
white, working women with fewer children at home, higher standards of living and higher
political interest. It is also expected that the gender gap will be smaller for citizens living in
44
Althaus 2003; Batista Pereira 2001; Delli Carpini and Keeter 1996; Díaz Dominguez 2011;
Luskin1990.
18
big cities (where women are expected to have more political knowledge) and in rural areas
(as a result of both men and women having very low levels of political knowledge).
Once the individual-level antecedents of political knowledge are controlled for, this
study explores the effect on the gender gap of the contextual dimensions discussed earlier.
Two variables account for gender-bias signaling. First, to gauge gender differences in
representative institutions, a measure of the percentage of women in the country’s parliament
is used.
45
Second, gender differences in the labor market are accounted for by the ratio of
estimated female to men earned income,
46
an estimate that ranges from 0 (complete
inequality of earned income) to 1 (complete equality of earned income). We expect a smaller
gender gap in countries with a greater percentage of women in parliament and in countries
with more equality of earned income.
Gender accessibility, on the other hand, is captured through three variables. First, the
average age of the main political parties (Thorsten et al 2001)
47
as an indicator of party-
system institutionalization. The gender gap in knowledge is expected to be lower in countries
with older party systems. Second, Gallagher’s index of disproportionality of electoral
outcomes.
48
The index ranges from 0 to 100, with higher values reflecting less
45
UNDP 2007; UNDP 2009.
46
UNDP 2007; UNDP 2009
47
This is measured by the average of the ages of the 1st government party, 2nd government
party, and 1st opposition party.
48
We calculated these values for Colombia, Dominican Republic, Ecuador, Guatemala and
Venezuela. The rest were retrieved from Gallagher’s website
(http://www.tcd.ie/Political_Science/staff/michael_gallagher/ElSystems/). Note that
aggregated figures for the 2007 parliamentary election in Argentina could not be calculated
19
proportionality. It is expected that the gender gap in knowledge will increase as
disproportionality increases. And third, Watts' index of federalism, which distinguishes
between unitary (0), hybrid unitary (0.5) and federal countries (1).
49
The gender gap in
knowledge should be smaller in less unitary countries.
Finally, the study controls for other two potential contextual determinants of
knowledge. Economic development is measured by GDP per capita in US dollars, as
provided by the World Bank’s World Development Indicators. It is expected that the gender
gap in knowledge will decrease as economic development increases. Political Freedom is
measured through the index of political rights provided by Freedom House,
50
which ranges
from 1 (the most free) to 7 (least free), to measure the quality of democracy. The gender gap
in knowledge is expected to be higher in countries with weaker political rights (although, as
explained above, in the case of Latin America it might also be the contrary, see footnote 38).
Self-reported exposure to media news (a summative index of exposure to radio, TV,
newspaper and internet news) is also controlled for, as this variable may impact on levels of
political knowledge in general. Descriptive statistics of the aggregate- and individual-level
variables employed can be found in the On line Appendix, Table A.1. The measurement of
each of the independent variables is provided at the bottom of Table A.3.
given that alliances between parties differed greatly across constituencies. Thus, we used the
value from the 2005 parliamentary election instead.
49
Data on federalism come from Democracy Time-series Data Release 3.0, January 2009.
http://www.hks.harvard.edu/fs/pnorris/Data/Data.htm.
50
Freedom House 2009 and 2010.
20
Our empirical strategy is to estimate political knowledge as a function of all the
standard antecedents at the individual level plus all the contextual variables using a multilevel
estimation with two levels (individuals and countries). By introducing a random intercept at
the country level, we relax the assumption of independence of errors between respondents
living in the same country and are able to introduce country-level variables with appropriate
standard errors. Since we are interested in differences between men and women, we need to
include not only the main effect of each of the individual and contextual variables, but also an
interactive term of each individual and contextual variable with gender.
FINDINGS
We begin by presenting the size of the gender gap in political knowledge by country in
Figure 2, which shows the average number of correct answers among men and women. As
expected, political knowledge is significantly higher among men in all 18 nations,
51
confirming previous findings analyzing data from industrialized democracies. There is also
clear variation across countries. Argentina presents the smallest gender gap, with men
correctly answering 0.28 more questions than women. At the other extreme, however, is the
Dominican Republic, where men correctly answered almost one more question (0.92) than
women. The gender differences in political knowledge presented in Figure 2 are considerable
if we take into account that the index ranges from 0 to 5 and that the average number of
correct answers across Latin America is, as mentioned earlier, 2.62. Actually, the number of
correct answers given by women across countries is 22% lower on average than the number
of correct answers given by men. This gender gap in political knowledge and its variation
across countries do clearly deserve an explanation.
51
All differences are significant at p<0.01.
21
FIGURE 2. The Gender Gap in Political Knowledge Across Latin America
Mexico
Guatemala
El Salvador
Honduras
Nicaragua
Costa Rica
Panama
Colombia
Ecuador
Bolivia
Peru
Paraguay
Chile
Uruguay
Brazil
Venezuela
Argentina
Dominican Rep
1 2 3 4 5
Women Men
95% Confidence Intervals
Source: Our elaboration on LAPOP, 2008.
To incorporate both individual and contextual information, we employ a hierarchical
model with two different levels (individuals and countries). Table 1 shows the results of the
estimation of the individual determinants of citizens’ political knowledge. Model 1 shows the
magnitude of the gender gap without considering any respondents’ characteristics. On
average, men provide a bit less than one additional correct answer than women (see the
coefficient for the variable gender: 0.76). Model 2 adds all individual predictors under the
expectation that men and women have different access to resources and opportunities, and
that these differences contribute to explaining the gender gap in knowledge. Finally, Model 3
adds the significant interaction terms of all the relevant individual-level independent
variables and gender under the expectation that men and women may respond differently to
22
the same factors. From all the possible interactions, only three turned out to be statistically
significant: education, urbanization and ethnicity.
Results in Model 2 in Table 1 demonstrate the impact of resource differences on the
levels of political knowledge of Latin American citizens. The number of correct answers
increases with education, age, and (subjective) standard of living. In addition to men,
knowledge is also greater among those citizens who are white, married, employed, declare
themselves to be interested in politics and are intensively exposed to media news.
Conversely, the level of knowledge appears to be smaller for those having more children at
home and for those living in a rural area.
Results also suggest that, even when individual differences in motivation, resources
and ability are all taken into account, political knowledge continues to be significantly higher
among men than it is among women. The magnitude of the effect of gender is, however,
reduced once all the individual determinants of knowledge are considered (the reduction is
from 0.760 to 0.515 = 0.245). Thus, it appears that compositional differences between men
and women on the antecedents of knowledge contribute to explaining only part of the gender
gap.
52
Put another way, part of the gender gap is explained by the fact that men in Latin
American still have on average more socioeconomic and cognitive resources than women.
To test the extent to which the same antecedents of knowledge might differently
affect men and women, Model 3 in Table 1 shows the results of an additional estimation
52
Looking at individual variables in Model 2, the impact of compositional differences on the
gender appears to be negligible. For example, if women had the same average level of
education as men (i.e. 9.3 years instead of 8.8 years) they would be able to give 0.06 more
correct answers. If the proportion of employed women was similar to that of men (72% rather
than 36%), the gender gap would be reduced by only 0.01 more correct answers.
23
where we introduce interactions of all the relevant individual-level variables with gender.
Surprisingly, most of the variables that have been identified in the literature to reduce the
gender gap do not seem to be relevant in Latin America, as their interaction with gender is
not statistically significant.
53
Thus, the impact of gender on political knowledge does not
seem to be conditioned by the number of children at home, labor market position, church
attendance or income. Moreover the gender gap in knowledge in Latin America does not
appear to decrease among the youngest generations in comparison to their older counterparts,
since the interaction term of age and gender did not reach statistical significance.
54
These
results challenge the recurrent speculation in the literature that the gender gap in politics will
disappear over time due to changes in socialization and accompanying generational
replacement.
55
Model 3 in Table 1 shows the results of the estimation including only the significant
individual-level interactions. Only three variables contribute to significantly reducing the
gender gap: education, urbanization and ethnicity. Although education tends to increase the
levels of political knowledge for both men and women, the effect is smaller for men as can be
53
We previously estimated different models introducing the interaction term of each single
individual variable with gender separately in order to avoid multicollinearity. As mentioned
in the main text, only three interactions turned out to be statistically different from zero.
54
Following Thomas Brambor, Clark, and Golder 2005, we calculated the marginal effects of
the difference between men and women for all the values of age in order to explore the
possibility that the gender gap is significant for some particular cohorts but not for others.
However, this was not the case. The gap is significantly persistent across all the range of
age’s values.
55
Bennett and Bennett 1989; Inglehart and Norris 2003.
24
appreciated by the negative sign of the interaction term. Comparing people with no formal
education to those who have spent the longest amount of time in formal education (18 years)
results in a reduction of 28% in the magnitude of the gender gap.
56
56
On average, females without formal education give 0.60 less correct answers than their
male counterparts. Among people with 18 years of education this effect is reduced to 0.43
less correct answers.
25
TABLE 1 Individual-level Determinants of the Gender Gap in Political Knowledge
(1)
Gender only
(2)
Individual-level
model
(3)
Individual-level
interactions
GENDER (1 = male)
0.760***
0.515***
0.735***
(0.020)
(0.020)
(0.048)
Education
0.116***
0.121***
(0.002)
(0.003)
Rural area
-0.127***
-0.055
(0.023)
(0.030)
Big city
0.037
0.100***
(0.021)
(0.027)
Ethnicity (1 = white)
0.088***
0.125***
(0.021)
(0.026)
GENDER * Education
-0.010**
(0.004)
GENDER * Rural
-0.161***
(0.044)
GENDER * Big city
-0.145***
(0.041)
GENDER * Ethnicity
-0.095**
(0.037)
Married/cohabiting
0.150***
0.147***
(0.021)
(0.021)
Church attendance
-0.012
-0.012
(0.007)
(0.007)
Subj. standard of living
-0.119***
-0.119***
(0.011)
(0.011)
Children at home
-0.021***
-0.021***
(0.006)
(0.006)
Employed
0.037
0.033
(0.020)
(0.020)
Age
0.040***
0.040***
(0.003)
(0.003)
Age²
-0.000***
-0.000***
(0.000)
(0.000)
Political interest
0.131***
0.131***
(0.009)
(0.009)
News media exposure
0.091***
0.091***
(0.004)
(0.004)
Intercept
2.248***
-0.082
-0.182
(0.116)
(0.128)
(0.130)
Var (ind)
1.86***
1.32***
1.318***
(0.019)
(0.013)
(0.013)
Var (country)
0.241***
(0.081)
0.169***
(0.057)
0.169***
(0.057)
N Level 1
19,311
19,311
19,311
N Level 2
18
18
18
-2LL
-33404
-30123
-30109
Source: Our elaboration on LAPOP, 2008. Standard errors in parentheses *** p<0.01, ** p<0.05
The measurement of all independent variables is provided at the bottom of Table A.3 in the On line
Appendix
26
The effect of urbanization turned out to be as expected. In Latin American rural areas,
where men also have lower levels of political knowledge, the gender gap is greatly reduced,
as demonstrated by the negative sign of the interaction term (rural*male). On average, the
gender gap is 16% lower in rural areas than it is in urban areas.
57
In urban areas, on the other
hand, the gender gap is significantly weaker in bigger cities (including capital and
metropolitan areas), where women have easier access to political resources and therefore also
have more political information than women in other areas. The average number of correct
answers is 0.63 less for women in smaller cities and towns than it is for men, but this
difference decreases to 0.48 in bigger cities (that is, a decrease in the gender gap of 21%).
58
Finally the difference between males and females appears to be smaller among white
people. While non-white men give 0.54 more answers than non-white women, white men
give 0.45 more correct answers than their female counterparts. Nevertheless, it is worth
noting that, as mentioned later on, the effect of ethnicity on the gender gap ceases to be
significant when contextual-level variables are controlled for (see Table 3).
We next introduce interactions between gender and each of our aggregate variables.
Results are shown in Table 2 (where the other independent variables at individual level are
not shown but are included in the estimation; see Table A.3 in the On line Appendix where
we list all the independent variables). As expected, the gender gap is smaller in countries with
older political parties (see the corresponding column of Model 4) a greater percentage of
women in Parliaments (see the corresponding column of Model 7), a smaller sex difference in
57
Men in rural areas give 0.41 more correct answers than women, which compares to 0.57
more correct answers on average in urban areas.
58
The effect of gender is not significantly different in rural areas compared to bigger cities.
27
earned income (see the corresponding column of Model 8), and in federal countries (see the
corresponding column of Model 5). Smaller gender gaps are also found in countries with
higher levels of economic development (measured as GDP per capita; see the corresponding
column of Model 9) and more political freedom (see the corresponding column of Model 10,
recall that this variable goes from more to less degree of freedom), our two control variables
at the aggregate level. However, contrary to our expectation this study does not find evidence
that sex differences in political knowledge in Latin America depend on a country’s electoral
disproportionality (see the corresponding column of Model 6).
28
TABLE 2 Contextual Determinants of the Gender Gap in Political Knowledge
(4)
(5)
(6)
(7)
(8)
(9)
(10)
VARIABLES
PARTY
AGE
FEDERAL
ISM
PROPORTI
ONALITY
WOMEN
IN
PARL.
GENDER
INCOME
EQUAL
GDP
POL
RIGHTS
GENDER (MALE = 1)
0.580***
0.546***
0.573***
0.654***
0.821***
0.632***
0.395***
(0.029)
(0.022)
(0.037)
(0.044)
(0.090)
(0.035)
(0.047)
GDP p/c
-0.000
(0.000)
Gender * GDP
-0.000***
(0.000)
Pol Rights
-0.100
(0.102)
Gender * Pol Rights
0.051***
(0.018)
Party Age
0.005
(0.003)
Gender * Party Age
-0.002***
(0.001)
Federalism
-0.232
(0.224)
Gender * Federalism
-0.155***
(0.044)
Disproportionality
-0.027
(0.029)
Gender * Disproportionality
-0.009
(0.005)
Parl women (%)
0.013
(0.011)
Gender *Parliawomen
-0.007***
(0.002)
Income Equality
0.611
(0.978)
Gender * Income Equality
-0.609***
(0.174)
Intercept
-0.271
-0.016
0.086
-0.333
-0.387
0.013
0.157
(0.180)
(0.136)
(0.216)
(0.246)
(0.498)
(0.216)
(0.269)
Var (indiv)
1.319***
1.318***
1.319***
1.319***
1.32***
1.318***
1.319***
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
(0.013)
Var (country)
0.156***
0.154***
0.158***
0.162***
0.168***
0.164***
0.164***
(0.052)
(0.052)
(0.053)
(0.054)
(0.057)
(0.055)
(0.055)
N Level 1
19,311
19,311
19,311
19,311
19,311
19,311
19,311
N Level 2
18
18
18
18
18
18
18
-2LL
-30118
-30117
-30121
-30117
-30117
-30115
-30119
Source: Our elaboration on LAPOP, 2008. Standard errors in parentheses
*** p<0.01, ** p<0.05
The following control variables are not shown (but are included in the estimation): Education (years), Urbanization,
Married/cohabiting (ref: Single), Religiosity (service attendance), Subjective standard of living, Number of children at
home, Labor market position: Employed, Inactive (ref: homemaker), Age, Age squared, Political interest, Exposure to
media news (summative index based on radio, TV, newspapers and the internet news exposure). Table A.3 in the On line
Appendix shows the control variables at the individual level (and the measurement of all independent variables).
29
After specifying each of the country-level variables separately, an additional equation
was estimated (Model 11 in Table 3) where all significant interaction terms were specified
simultaneously: party age, percentage of women in Parliaments, income equality, political
freedom and GDP, together with education, ethnicity and degree of urbanization of the place
of residence. This strategy is useful to test which variables appear more relevant in
contributing to increasing or decreasing the gender gap in knowledge in Latin America.
Again, other independent variables at the individual level are not showed in Table 3 but they
are included in the estimation (see Table A.4 in the On line Appendix where all the
independent variables are shown).
As can be seen in Table 3, when all these variables are simultaneously specified, the
effect of federalism, economic development and political freedom all cease to be statistically
significant, even though coefficients retain the right sign. Ethnicity, which interacted
significantly with gender in Model 3, ceases to be significant when contextual variables are
introduced. However, all the other interactions continue to reach statistical significance in the
final model.
Findings are then in line with our expectation that the contextual setting influence
gender differences in political knowledge through two specific mechanisms: gender-bias
signaling and gender accessibility. Regarding gender-bias signaling, the magnitude of the
gender gap in knowledge significantly decreases as the percentage of women in Parliaments
(the political gender-bias signaling dimension) increases and as income equality between
female and male (the economic gender-bias signaling dimension) increases.
The gender accessibility mechanism appears to also work in the case of Latin
American countries but less through the effect of power-sharing institutions and more
through the effect of the institutionalization of the party system. Institutionalized party
systems provide policy stability to parties and better opportunities for women's activists to
30
access party leadership and to influence the political agenda contributing to a decrease in the
magnitude of the gender gap in knowledge.
31
TABLE 3 Contextual Determinants of the Gender Gap in Political Knowledge. Full models.
(11)
(12)
All Aggregate
Final Model
GENDER (1 = male)
1.261***
1.221***
(0.168)
(0.107)
Education
0.121***
0.121***
(0.003)
(0.003)
Rural Area
-0.046
-0.057
(0.030)
(0.030)
Big City
0.098***
0.103***
(0.028)
(0.028)
Ethnicity (1 = white)
0.102***
0.086***
(0.026)
(0.021)
Cross-level interactions
GENDER * Party Age
-0.003***
-0.003***
(0.001)
(0.001)
GENDER * Women in Parliament
-0.009***
-0.009***
(0.002)
(0.002)
GENDER * Income Equality
-0.379**
-0.475***
(0.182)
(0.179)
GENDER * Political Rights
0.003
(0.029)
GENDER * Federalism
-0.106
(0.072)
GENDER * GDP
-0.000
(0.000)
Individual-level interactions
GENDER * Education
-0.010***
-0.009**
(0.004)
(0.004)
GENDER * Rural
-0.182***
-0.156***
(0.044)
(0.044)
GENDER * Big City
-0.139***
-0.149***
(0.041)
(0.041)
GENDER * Ethnicity
-0.036
(0.039)
Contextual variables
Party Age
-0.003***
0.008**
(0.001)
(0.004)
Women in Parliaments
-0.009***
0.022
(0.002)
(0.011)
Income Equality
-0.385**
0.096
(0.182)
(0.913)
Political Rights
0.001
(0.029)
Federalism
-0.106
(0.072)
GDP p/c
-0.000
(0.000)
Intercept
-0.891
-0.907
(0.779)
(0.490)
Var (indiv)
1.314***
1.315***
(0.013)
(0.013)
32
Var (country)
0.123***
0.137***
(0.041)
(0.046)
N Level 1
19,311
19,311
N Level 2
18
18
-2LL
-30081
-30090
Source: Our elaboration on LAPOP, 2008. Standard errors in parentheses
*** p<0.01, ** p<0.05
Control variables at the individual level are not shown (but are included in the estimation). Table A.4 shows the control
variables at the individual level whereas Table A3 shows the measurement of all independent variables (See the On line
Appendix)
We rely on a graphical presentation of the results to assess more clearly the magnitude
of the impact of the aggregate variables. To this end, and following the recommendations of
Kam and Franzese (2007), the marginal effects on gender differences in political knowledge
have been computed (ie. the difference between the marginal effect of each variable on men’s
political knowledge minus its effect on women’s political knowledge), along with their
associated standard errors. Figure 3 shows how differences between men and women (that is
to say, the gender gap in knowledge) are clearly reduced in countries with older party
systems, which provide stable settings for women to access the agenda and leadership of
political parties. In our sample, the effect ranges from a difference of 0.60 questions when
party age is 5 (minimum) to a difference of 0.35 when party age is 95 (maximum). This
implies an overall reduction in the magnitude of the gender gap in knowledge of about 42%.
If we consider that the average gender gap in knowledge amounts to 0.66 questions, this is
clearly an important effect.
33
FIGURE 3 Effect of Party Age on the Gender Gap in Political Knowledge
(with 95% confidence intervals)
Source: Our elaboration based on the estimations presented in Model 12 of Table 3.
A very similar reduction can be seen in countries with a higher proportion of women
in Parliaments (Figure 4). On average, men provide 0.61 more correct answers in countries
with the minimum percentage of women in Congress (9%), a figure that is reduced to 0.33
when the maximum (39%) is reached. This involves an overall reduction of almost half
(46%) in the magnitude of the gender gap in political knowledge. The presence of more
women in the legislature signals that women play a relevant role in the political sphere,
thereby activating the motivation and interest of women to get involved and informed about
politics, and consequently reducing the knowledge differences between men and women.
34
FIGURE 4 Effect of % of Women in Parliament on the Gender Gap in Political
Knowledge (with 95% confidence intervals)
Source: Our elaboration based on estimations presented in Model 12 of Table 3
Likewise, the economic gender-bias signaling dimension appears to work in the Latin
American case. In fact, more equality of income between men and women is associated with
a smaller gender gap in political knowledge (Figure 5). To be more precise, the difference
between men and women is 0.61 correct questions in the most unequal context in our sample
(female-to-male earned income ratio= 0.3). The gap, however, decreases significantly in
countries with greater levels of income equality between both genders, reaching 0.42 when
the maximum level of equality in the sample is reached (female-to-male earned income
ratio=0.7). This involves an overall reduction of almost one third (31%) in the magnitude of
the gender gap in political knowledge.
35
FIGURE 5 Effect of Income Equality on the Gender Gap in Political Knowledge
(with 95% confidence intervals)
Source: Our elaboration based on estimations presented in Model 12 of Table 3
Lastly, the effects of federalism, political rights and economic development are much
smaller than those of the other aggregate variables and do not appear to be conditioned on
gender (since the interaction term of all of them with gender did not reach statistical
significance at p<0.05). The implications of these findings both for the study of political
knowledge and for the functioning of representative democracy are discussed in the last
section
DISCUSION AND CONCLUSIONS
This study analyses the relationship between gender and political knowledge in 18
Latin American Countries with data from the 2008 AmericasBarometer’s round of surveys.
36
Findings show that, after controlling for individual differences in motivation, resources and
ability, political knowledge continues to be significantly higher among men than it is among
women across all countries analyzed here confirming thereby previous findings in Europe,
the United States, and Canada. However the magnitude of the gender gap varies to a great
extent across Latin American countries.
A substantive relevant finding of this article is that sex differences decrease with
citizens’ level of education. This suggests that a considerable increase in the general level of
education in Latin America might contribute to reducing the gender gap in politics to about
one fourth its actual size. In addition, gender differences appear smaller in rural areas (where
both men and women have low levels of political knowledge) and in bigger cities (where
women's levels of political knowledge is higher). Surprisingly, evidence relating to
differential effects of other individual factors such as the number of children, or the economic
status was not found. Furthermore, the size of the gender gap in knowledge does not appear
to depend on age, thereby challenging previous expectations from the literature about a
decrease in the size of the gender gap in politics over time due to changes in socialization and
accompanying generational replacement.
59
These findings are relatively pessimistic about the
potential of closing the gender gap in the near future and suggest that Latin America
countries probably need more radical changes in socialization that still lie ahead.
But perhaps more importantly, this study demonstrates that context also matters when
it comes to explaining cross-country differences in the magnitude of the gender gap in
knowledge in Latin America. It is argued that this may take place through two interrelated
pathways: gender-bias signaling and gender accessibility. In terms of gender-bias signaling,
we find that as women’s descriptive representation increases, the gender gap in knowledge
59
Bennett and Bennett 1989; Inglehart and Norris 2003.
37
decreases: for example, if Costa Rica (the country with the highest level of women’s
representation in parliament) is compared with Brazil (the country with the lowest level) on
average the size of the gender gap in knowledge in the former is one third smaller than in the
latter. This study also demonstrates that an ignored factor in previous studies —income
equality between sexes— appears to be of great relevance in contributing to a reduction in the
magnitude of the gender gap in knowledge. If we compare Nicaragua (the most unequal
country in our sample) to Colombia (the most equal polity) on average the size of the gender
gap in knowledge in the former is 37% smaller than in the latter. These finding suggests, first,
the need to incorporate socioeconomic explanations about the gender gap in politics not only
at the individual level (measured through citizens’ socioeconomic resources) but also at the
contextual level. And second, that the presence of acute gender biases in different aspects of
the public sphere (not only political but also economic) may have important symbolic (de-)
mobilization effects and contribute to enlarging the gender gap in knowledge.
The study also shows that the greater the consolidation of party systems, the more
gender differences in knowledge diminish. On average, gender differences decrease from
0.60 in El Salvador (the country with the lowest level of party system institutionalization) to
0.35 in Uruguay (the country with the highest level). This implies a reduction of 42% in the
size of the gender gap. We interpret this finding as evidence in favor of the gender
accessibility mechanism, according to which stable institutions that facilitate women's
involvement in politics and sustained access to the political agenda may also help to reduce
the gender gap in political knowledge. This further speaks to the harmful effect that political
instability can have on enlarging sex differences in political knowledge in Latin America.
The findings presented here also suggest the need to conduct further research on the
topic in a larger geographical area and possibly across a long period of time. We urge
researchers to continue pushing for the inclusion of a range of comparable knowledge
38
measures that result in valid representations of what people know about politics across
countries. To this respect the inclusion of four knowledge items standardised across countries
in the fourth module of the CSES appears to us promising.
60
Other international survey
programs could also adopt this strategy.
A last but very important implication of these findings is that what people know about
politics is not only a function of their capabilities and motivation but also of the opportunities
that the context where they live offers to them to become informed about politics, confirming
also previous findings for the European case (Fraile 2013). Moreover, socioeconomic and
institutional settings providing opportunities for women to influence the political agenda and
to play a central role in the public sphere significantly contribute to reducing the magnitude
of the gender gap in knowledge. Consequently, this gap does not necessarily need to
perpetuate itself over time. It all depends on the capacity of politicians and citizens to create
the opportunities for females to increase their level of knowledge. This way, women might
know what governments are actually doing to the same extent than men, holding
governments accountable for their actions while in power and creating additional incentives
to such governments to be responsive to all of the citizens.
60
Only ten countries are currently included in the fourth module of the CSES (at the time we
checked it, in November 2014). These are not enough to replicate the analyses conducted
here. Nevertheless the number of countries is expected to expand significantly. As a
consequence, replication of our analyses with a different set of countries will hopefully be
possible in the future.
39
REFERENCES
Althaus, Scott. 2003. Collective Preferences in Democratic Politics. Cambridge: Cambridge
University Press
Atkeson, Lonna Rae. 2003. Not All Cues Are Created Equal: The Conditional Impact of
Female Candidates on Political Engagement. Journal of Politics 65 (4): 1040–1061.
Atkeson, Lonna Rae, and Nancy Carrillo. 2007. More Is Better: The Influence of Collective
Female Descriptive Representation on External Efficacy. Politics and Gender 3 (1): 79-101.
Batista Pereira, Federico. 2011. Political Knowledge Levels Across the Urban-Rural Divide
in Latin American and the Caribbean. AmericansBarometer Insights Series 68.
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.
Bennett, Linda and Stephen Earl Bennett. 1989. Enduring Gender Differences in Political
Interest: The Impact of Socialization and Political Dispositions. American Politics Quarterly
17(1): 105-122
Brambor, Thomas, Williams Roberts Clark, and Matt Golder. 2006. Understanding
Interaction Models: Improving Empirical Analysis. Political Analysis 14 (1): 63-82.
Bruhn, Kathleen 2003. Whores and Lesbians: Political Activism, Party Strategies, and
Gender Quotas in Mexico. Electoral Studies 22 (1): 101-119.
Burns, Nancy. 2007. Gender in the Aggregate, Gender in the Individual, Gender and Political
Action. Politics and Gender 3(1): 104-124.
40
Burns, Nancy, Kay Lehman Schlozman, and Sidney Verba. 2001. The Private Roots of
Public Action. Cambridge, MA: Harvard University Press
Campbell, David E and Christina Wolbrecht. 2006. See Jane Run: Women Politicians as Role
Models for Adolescents. Journal of Politics 68 (2): 233–247.
Caul, Miki. 1999. Women’s Representation in Parliament. The Role of Political Parties.
Party Politics 5(1): 79-98.
Delli Carpini, Michael X. and Scott Keeter. 1996. What Americans Know about Politics and
Why it Matters. New Haven: Yale University Press.
Delli Carpini, Michael, and Scott Keeter. 2000. Gender and Political Knowledge. In Gender
and American Politics: Women, Men, and the Political Process, edited by Sue Tolleson
Rinehart and Jyl Josephson, 21-52. Armonk, NY: M.E. Sharpe
Desposato, Scott and Barbara Norrander. 2009. The Gender Gap in Latin America:
Contextual and Individual Influences on Gender and Political Participation. British Journal of
Political Science 39 (1): 141–162.
Díaz Dominguez, Alejandro. 2011. Political Knowledge and Religious Channels of
Socialization in Latin America. AmericansBarometer Insights Series 55.
Dolan, Kathleen. 2006. Symbolic Mobilization? The Impact of Candidate Sex in American
Elections. American Politics Research 34 (6): 687–704.
Dow, Jay K. 2009. Gender Differences in Political Knowledge: Distinguishing
Characteristics-based and Returns-based Differences. Political Behavior 31 (1): 117–136.
Fraile, Marta. 2013. Do Information Rich Contexts Reduce Knowledge Inequalities? The
Contextual Determinants of Political Knowledge in Europe. Acta Politica 48(2): 119-143.
41
Fraile, Marta. 2014. Do women know less about politics than men? The Gender Gap in
Political Knowledge in Europe. Social Politics 21(2): 261-289.
Freedom House. 2009. Freedom in the World: Setbacks and Resilience. Washington, DC:
Freedom House.
Freedom House. 2010. Freedom in the World: Global Erosion of Freedom. Washington, DC:
Freedom House.
Htun, Mala, and Jennifer Piscopo. 2012. Presence Without Empowerment? Women in
Politics in Latin America and the Caribbean. Social Science Research Council, Brooklyn.
Inglehart, Ronal and Pippa Norris. 2003. Rising Tide. Gender Equality and Cultural Change
around the World. Cambridge. Cambridge University Press.
Kam, Cindy D. And Robert J. Franzese. 2007. Modeling and Interpreting Interactive
Hypothesis in Regression Analysis. Ann Arbor: The University of Michigan Press.
Kampwirth, Karen. 2010. Introduction. In Gender and Populism in Latin America:
Passionate Politics edited by Karen Kampwirth. University Park: Penn State University Press
Karp, Jeffrey A and Susan A Banducci. 2008. When Politics is not Just a Man’s Game:
Women’s Representation and Political Engagement. Electoral Studies 27(1): 105–115.
Kittilson, Miki Caul and Leslie Schwindt-Bayer. 2010. Engaging Citizens: The Role of
Power-Sharing Institutions. Journal of Politics 72(4): 990-1002.
Kittilson, Miki Caul and Leslie A. Schwindt-Bayer. 2012. The Gendered Effects of Electoral
Institutions. Political Engagement and Participation. Oxford: Oxford University Press.
Koch, Jeffrey. 1997. Candidate Gender and Women’s Psychological Engagement in Politics.
American Politics Research 25(1): 118–133.
42
Lawless, Jennifer. 2004. Politics of Presence? Congress Women and Symbolic
Representation. Political Research Quarterly 57(1): 81–99.
Lovenduski, Joni 1993Introduction: The Dinamics of Gender and Party’, in Gender and
Party Politics, edited by Joni Lovenduski and Pippa Norris, 1-15. London: Sage.
Luskin, Robert. 1990. Explaining Political Sophistication. Political Behaviour 12(4): 331-
361.
Mainwaring, Scott. 1998. Rethinking Party Systems Theory in the Third Wave of
Democratization: The Importance of Party System Institutionalization. Kellogg Institute
Working Paper #260, University of Notre Dame
Mainwaring, Scott.1999. Rethinking Party Systems in the Third Wave of Democratization:
The Case of Brazil. Stanford, CA: Stanford University Press.
Mainwaring, Scott, and Mariano Torcal. 2006. Party System Institutionalization and Party
System Theory After the Third Wave of Democratization, in Handbook of Party Politics,
edited by Richard Katz and William Crotty, 204-227. London: Sage.
Morgan, Jana and Melisa Buice. 2013. Latin American Attitudes Towards Women in
Politics: The Influence of Elite Cues, Female Advancement, and Individual Characteristics.
American Political Science Review 107(4): 644-662
Norris, Pippa 2002. Democratic Phoenix: Reinventing Political Activism. Cambridge:
Cambridge University Press.
Norris, Pippa. 2009. Democracy Time-series Data Release 3.0, January 2009.
http://www.hks.harvard.edu/fs/pnorris/Data/Data.htm
43
Phillips, Anne. 1995. The Politics of Presence: The Political Representation of Gender,
Ethnicity, and Race. Oxford: Clarendon Press.
Pitkin, Hanna. 1967. The Concept of Representation. Berkeley: University of California
Press.
Rapelli, Lauri. 2014. Comparing Local, National and EU Knowledge: The Ignorant Public
Reassessed. Scandinavian Political Studies. Doi: 10.1111/1467-9477.12032
Reingold, Beth and Jessica Harrell. 2010. The Impact of Descriptive Representation on
Women’s Political Engagement: Does Party Matter? Political Research Quarterly 63(2):
280–294.
Römmele, Andrea. 2003. Political Parties, Party Communication and New Information and
Communication Technologies. Party Politics 9(1): 7-20.
Safa, Helen Icken. 1990. Women's Social Movements in Latin America. Gender and Society
4 (3): 354-369.
Shaker, Lee. 2012. Local Political Knowledge and Assessment of Citizen Competence.
Public Opinion Quarterly 76(3): 525-537.
Thorsten Beck, George Clarke, Alberto Groff, Philip Keefer, and Patrick Walsh. 2001. New
Tools in Comparative Political Economy: The Database of Political Institutions. The World
Bank Economic Review 15(1): 165-176.
United Nations Development Programme (UNDP). 2007. Human Development Report
2007/2008. New York: Palgrave Macmillan.
United Nations Development Programme (UNDP). 2009. Human Development Report 2009.
New York: Palgrave Macmillan.
44
Verba, Sidney, Nancy Burns and Kay Lehman Schlozman. 1997. Knowing and Caring about
Politics: Gender and Political Engagement. Journal of Politics 59 (4): 1051–1072.
Wayler, Georgina 2000. Gender and Democratic Politics: a Comparative Analysis of
Consolidation in Argentina and Chile. Journal of Latin American Studies 32 (3): 765-793.
Waylen, Georgina 2003. Gender and Transitions: What do We Know? Democratization
10(1): 157 –178.
Watts, Ronald L. 1998. Federalism, Federal Political Systems, and Federations. Annual
Review of Political Science 1(1): 117–137.
Walker, Lee Demetrius and Gennevieve Kehoe. 2013. Regime Transition and Attitudes
toward Regimen: The Latin American Gender Gap in Support for Democracy. Comparative
Politics 45(2): 187-204
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): 921–
939.
World Bank. 2012. The Effect of Women's Economic Power in Latin America and the
Caribbean. Washington, DC: World Bank.
Zetterberg, Par. 2009. Do Gender Quotas Foster Women’s Political Engagement? Political
Research Quarterly 62(4): 715-730.
